■ .■.'. ■.. ; ' —;> . 
 
 Ww» 
 
®mnrt\ WLnivmity |f itatg 
 
 BOUGHT WITH THE INCOME) 
 FROM THE 
 
 SAGE ENDOWMENT FUND 
 
 THE GIFT OF 
 
 Henrg m. Sage 
 
 xSgx 
 
 AM.3..5-JM . & 
 
 uLajk.. 
 
Cornell University Library 
 
 arY685 
 
 v.1-2 
 
 Practical electrical engineering 
 
 3 1924 032 174 892 
 ohn.anx 
 
The original of this book is in 
 the Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/details/cu31924032174892 
 
PRACTICAL 
 
 ELECTRICAL ENGINEERING. 
 
'!* %iW 
 
 !!%!' mm I*l ;| i 
 
 M 1)L ' , m. 3D>J i . t : Q). $ I : ' 3 
 
PRACTICAL 
 ELECTRICAL ENGINEERING 
 
 A complete treatise on the Construction and Management of 
 
 Electrical Apparatus as used in Electric Lighting and 
 
 the Electric Transmission of Power. 
 
 Illustrated with many Hundreds of Illustrations. 
 
 W. W. BEAUMONT, M.I.C.E. 
 
 C. H. W. BIGGS, M.I.E.E. (Editor). 
 
 C. CAPITO, M.I.E.E. 
 
 G. KAPP, M.I.E.E. 
 
 BY 
 
 M. H. KILGOUR, M.I.E.E. 
 A. RECKENZAUN, M.I.E.E. 
 P. SELLON, M.I.E.E. 
 J SWINBURNE, M.I.E.E. 
 
 H. SWAN, MJ.E.E. 
 
 London 
 Published for the Proprietors by 
 BIUCS & CO., 139-140, SALISBURY COURT, FLEET STREET, E.C 
 
INTBODTJCTIOJN 1 . 
 
 ,N these modern days, when the dissemination 
 of information is so easy, the time required 
 for the development of an industry is short 
 compared with what it was before the era 
 of the technical newspaper. No sooner 
 does a new idea seem practical than hun- 
 dreds of minds are directed towards it, each attempting 
 some modification to render it cheaper and more prac- 
 tical. Till the time of Davy, Morse, Cooke, and 
 Wheatstone, between the years 1830 and 1840, there 
 was no practical application of electricity. These men 
 have the credit between them of introducing electric tele- 
 graphy. Towards the latter part of the same period 
 Spencer, Jacobi, Smee, and the Blkingtons introduced 
 the art of electro-deposition. Also during or about the 
 same time Faraday, Sturgeon, and Henry obtained the 
 germs and constructed the first rough apparatus of 
 what has developed into the dynamo machine. Far 
 and away before the other names mentioned must that 
 of Faraday be placed. He was the Homer and the 
 Shakespeare of investigators. More recently still 
 Hughes applied with wonderful success his knowledge 
 of mechanism to the wants of telegraphy. Sir W. 
 Thomson, too, invented the apparatus that made ocean 
 telegraphy practicable. Nearer still as to time Holmes, 
 Ladd, Siemens, Pacinotti, and Gramme have given us 
 the machines ; Brush, Crompton, Brockie, and others 
 the arc lamps; Swan and Edison the incandescent 
 lamps, that furnish the apparatus for electric lighting 
 purposes and for the transmission of power. 
 
 The purpose of this treatise is to consolidate the 
 knowledge that applies more directly to the production 
 and the use of apparatus connected with electric lighting, 
 and the transmission of power. The use of electricity 
 in the transmission of power is destined to become far 
 more important than its use as a lighting agjent, 
 
 The plan adopted is, in the first place, to consider 
 the requirements of a central station. Supposing the 
 building finished, the prime motors must be considered. 
 Usually these will be steam-engines — hence it is neces- 
 sary for the electrical engineer to consider generally 
 the relation between this part of his apparatus and the 
 other parts. There will be no attempt to deal ex- 
 haustively with engines and boilers ; but the general 
 practical details necessary to be followed in con- 
 structing these in order that they may comply with 
 the special requirements in ordinary working for 
 regulation and maintenance in electrical installations, 
 will be amply described. Gearing and lubrication 
 must also be considered. The main object of the 
 book, however, is the electrical part. Owing to the 
 subject having never till recently been considered 
 from an engineer's point of view, there is absolutely 
 no literature upon the subject. There is no book to 
 inform the practical man how to construct a dynamo. 
 Books with a mass of mathematical formulae there 
 are, though only one of them, that of Dr. Silvanus 
 Thompson, is worth reading— hence plain information 
 given by practical men to practical men should supply 
 a vacant place as regards this class of literature. To 
 Gisbert Kapp and Dr. Hopkinson we owe the investiga- 
 tions which rendered possible the specification of a 
 machine which, when completed, should have the 
 output for which it is designed. Till the publication 
 of their papers, machines were made very much by rule 
 of thumb. The notation and nomenclature of elec- 
 trical literature was altogether against the engineer. 
 Up to the introduction of the dynamo electricians had 
 had to deal with minute quantities, when all at once 
 the consideration of the subject involved quantities 
 thousands and millions of times greater than had been 
 considered before. Engineers have done something to 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 assist them in this part of the subject, but they cannot 
 as yet prevail against the pure theorist. Dr. Hopkin- 
 son introduced, in his papers to the Institution of 
 Mechanical Engineers, a graphic notation which has 
 proved of inestimable value. Mr. Kapp in his papers 
 has suggested for one practical unit a value six 
 thousand times greater than that used in laboratory 
 practice, and this suggestion falls in with the require- 
 ments of engineers, as it takes " revolutions per 
 minute" instead of "revolutions per second." It is 
 comparatively easy to count " revolutions per minute," 
 but far from easy to count them " per" second." 
 
 The materials of which the apparatus is constructed 
 will be discussed, the best methods of using them, and 
 the right proportions described. Dynamos are required 
 for various purposes, and the design of each machine 
 should, be prepared with a due consideration of the 
 work it will have to do. Thus, large constant-current 
 and alternate-current machines are required for large 
 central stations, smaller ones for smaller installations 
 for isolated plants, and for ship lighting. Then the 
 machines differ in construction if required to produce 
 high pressure, from those required to produce a lower 
 pressure, and the slow-speed machines from the high 
 speed. 
 
 These requirements will all be discussed, and as 
 far as possible plain, workable instructions given for 
 the design of any class of, machine. 
 
 Before leaving the interior of the central station the 
 various methods of control must be described, also the 
 instruments and methods of measurement. Outside 
 the station the various systems of distribution will 
 claim attention, as will the transformers, lamps, 
 whether arc or incandescent, methods of wiring, safety 
 devices and fittings. 
 
 Another important part of the work is the utilisation of 
 electricity for obtaining mechanical power. In hundreds, 
 if not thousands of cases the use of electricity is not 
 only less costly, but far more convenient, and as the pos- 
 sibility of obtaining the necessary current becomes more 
 and more common, there will be an immense demand for 
 small and for large motors. These may be termed re- 
 versed dynamos, for the latter absorb mechanical energy, 
 and produce pressure which gives rise to current, while 
 the former absorb electrical energy and give out 
 mechanical energy. Motors, again, will be largely used 
 in tramway work, in haulage work in mines, for lifts, 
 for lifting, for travelling cranes, and probably for the 
 training of big guns. 
 
 The principles of electrical science involved in these 
 applications will be given as briefly as possible, and 
 those desirous of knowing more of these principles will 
 have to study special books written to elucidate par- 
 ticular points. The general principles of the science 
 as applied in engineering work are neither difficult to 
 understand, nor do they involve a knowledge of higher 
 mathematics. Formulae, it is true, must be extensively 
 used, but the greater number of these formulas when 
 used in practice involves a knowledge only of the rules 
 of simple arithmetic, and by using these aids the 
 
 matter can be presented in a far simpler and more 
 concrete form than by restricting description to words 
 only. There is every indication, too, that as our know- 
 ledge of the subject advances, the mathematics of the 
 subject will be considerably simplified. Hitherto, 
 mathematicians have made their mathematics fit the 
 knowledge, and this they can always do by using more 
 or less complex formulae, many of which tax the highest 
 mathematical skill to expound, and are altogether un- 
 translatable to the non-mathematician. Considerable 
 use is made of the graphic notation and its develop- 
 ments since its first introduction. 
 
 No doubt in many things this book will be found to 
 wander considerably from the popular notions now 
 prevailing. But this is a necessary sign of progress, 
 and especially may this be thought to be the case in 
 taking the stand that all electrical apparatus as used 
 generates electrical pressure, or creates a difference of 
 electrical pressure only. The term electrical pressure, 
 or difference of pressure, will be used instead of the 
 commoner terms electromotive force or difference of 
 potential, terms which do not appeal to the engineer, 
 and which the foremost thinkers have, since the 
 general introduction of dynamos desired to see replaced 
 by " pressure." The view that the apparatus employed 
 generates pressure only is unorthodox to this extent, 
 that the writers of school books have not yet seen that 
 it is involved in the view which now generally prevails 
 in the front ranks that " electricity," whatever it is, is 
 a constant quantity in the universe. It can neither be 
 destroyed nor created, and if this be so, then it is far 
 simpler and more in accordance with such views to 
 consider the apparatus as altering the condition " of 
 what already exists," than as creating equal quantities 
 of something in two different conditions, a creation 
 which complicates the whole consideration of the sub- 
 ject. The views thus put forward in this book are 
 merely the advanced views held by the men who 
 devote their working lives almost exclusively to the 
 investigation of the subject, whose opinions unortho- 
 dox to-day become orthodox to-morrow. 
 
 Beaders should make themselves thoroughly ac- 
 quainted with the elementary principles which form 
 the introduction to the practical part of the work. 
 No attempt has been made therein to apply the prin- 
 ciples, except perhaps in the solution of a few questions, 
 done more to show how to use the formula? than to 
 directly apply them, or in a casual remark as to the 
 particular branch where the principle involved is 
 prominent. 
 
 The use of such a knowledge of principles to handi- 
 craftsmen is obvious; it gives every man a chance, 
 according to his natural talents, of becoming an im- 
 prover of the art he works at, and even a discoverer 
 in the science connected with it. He is daily handling 
 the tools and materials with which new experiments 
 are to be made and daily watching the operations 
 so that a knowledge of the principles will frequently 
 suggest a more or less slight modification of the appa- 
 ratus, tending to improvement. All opportunities of 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 making experiments must be improved, all strange 
 appearances observed, and in this way the apparatus 
 in use perfected. It must never be forgotten, however, 
 that perfection does not mean development in one 
 direction only. A good machine is always a com- 
 promise, and the problem to be solved in designing 
 a first-class machine is the best relations that all the 
 principles involved shall bear to each other. Eor 
 example, a dynamo is a mechanical as well as an elec- 
 trical machine, hence the electrical requirements can- 
 not be pushed to extremes without considering the 
 requirements of the mechanism. Then, again, the 
 dynamo is only one part of a system — a system in- 
 volving boilers, engines, and distributing arrange- 
 ments. The dynamo therefore should always be 
 designed with a view to its position in this system — 
 that is, with a view to make it fit its work as perfectly 
 as possible. After all, in one way the introduction of a 
 dynamo or any other piece of mechanism into a circuit 
 must be looked upon as a loss. The prime considera- 
 tion is the energy involved in the combustion of fuel, , 
 and the amount of that energy we can apply to 
 humanity's requirements. Every piece of apparatus 
 inserted between the burning fuel and the point 
 where the energy developed is utilised causes a loss 
 of some of the energy, so that if the energy of com- 
 bustion could be directly utilised as electrical energy, 
 an immense step would be gained. According to the 
 present state of our knowledge the intervention of 
 apparatus is necessary, and all improvement in such 
 apparatus tends towards less loss between the point 
 where the energy is developed and where it is utilised. 
 A further consideration of this subject will show that 
 if a piece of apparatus, however good and perfect in 
 itself, involves the use of other apparatus, it may prove 
 that this latter necessity completely vitiates the value 
 of the perfect apparatus. 
 
 Very few discoveries have been made by chance or 
 by ignorant persons — much fewer than is generally 
 supposed. It is commonly told of the steam-engine that 
 an idle boy, being employed to stop and open a valve, 
 saw that he could save himself trouble by fixing it to a 
 moving part of the engine. This is possible, no doubt, 
 yet the tale is a little doubtful ; but improvements 
 
 of value are seldom so easily found out, and hardly 
 another instance can be named of important discoveries 
 being so accidental. 
 
 So far as electrical apparatus is concerned it owes 
 little to accident. Faraday, was a born investigator 
 into nature's secrets, and to his labours we owe the germ 
 of the modern dynamo. Sturgeon and Henry devoted 
 almost the whole of their lives to the study of electrical 
 phenomena — hence, the electromagnet is by no means 
 an accidental discovery. Pacinotti made a great step 
 in advance, but he did not realise the value of his 
 work. Thirty years or so after Faraday's discovery, 
 Gramme really fashioned a machine that was to 
 create an industry. The work of Siemens or of 
 Wilde must not be overlooked, but sufficient names 
 have been mentioned to show that electrical progress 
 is due to persevering study, and not to accidental dis- 
 covery. But, in so far as chance has anything to do 
 with discovery, surely it is worth the while of those 
 who are constantly working in a particular employ- 
 ment to obtain the knowledge required, because their 
 chances are greater than other people's of being able to 
 apply that knowledge to new and useful ideas ; they are 
 always in the way of perceiving what is wanting, and 
 this in itself is half-way towards obtaining what is 
 wanting. 
 
 One of the most important features in modern engi- 
 neering of all kinds is the accuracy of the methods of 
 measurement employed. It is essentially important in 
 all that concerns electrical engineering to make careful 
 and accurate measurements. The instruments em- 
 ployed are simple and extremely elegant in construc- 
 tion, nor does their use involve difficulties. Perhaps 
 the least successful instrument at the time of writing is 
 the " meter." It is desirable that purchasers of elec- 
 trical energy should be able to measure as easily the 
 energy they consume as they measure the gas that 
 passes through the gas-meter. 
 
 These then are the subjects to be treated, and when 
 it is remembered that the electrical industry will 
 become, nay must become, one in which an exceed- 
 ingly large sum of money is invested, it will be allowed 
 that it is high time the existing principles and practice 
 were rigidly defined. 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 CHAPTEE 1. 
 
 GENEKAL PEINCIPLES. 
 
 t HAT electricity is we do not know. Certain 
 strange phenomena have been noticed 
 through many centuries, and to these has 
 been given the name electrical. It is 
 these phenomena that have been studied 
 and investigated, and laws relating to their 
 action have been deduced. From time to 
 time the deductions made have had to be 
 remodelled, as new experiments threw new light upon 
 the subject, and not even now can we say for certain 
 that this or that deduction is absolutely correct. All 
 we can say is that it fits best the results of all known 
 experiments and phenomena. In electrical matters, 
 then, the student must keep his mind open, and suffer 
 no dogmatism to rule supreme. An effort will be made 
 in this introductory sketch to present as clearly and 
 succinctly as possible the principles of electrical science 
 which in the state of our present knowledge must guide 
 the engineer in his work. Just as the subject of chemistry 
 has grown so large that no one man can hope to be- 
 come thoroughly proficient in all its branches, and 
 restricts his studies to one branch, thus becoming a 
 specialist, so must the student of electricity restrict his 
 energies to one branch if he desires to become proficient 
 therein. 
 
 The work of the electrical engineer is a special branch 
 of mechanical engineering, and the better the mechani- 
 cal engineer the better the electrical engineer. The 
 builder of steam-engines should have some knowledge 
 of the theory of steam, the strength and the properties 
 of iron and steel ; so also the dynamo builder and 
 the dynamo user should have some knowledge of the 
 theory of electricity. The operations which come 
 within the scope of this work are those connected with 
 electric lighting, with isolated or central station light- 
 ing, the machinery and apparatus used therein and in 
 connection therewith; also with the electrical trans- 
 mission of power, whether by means of isolated or self- 
 contained plants or from central stations. 
 
 Without entering to any great length upon the 
 various theories which have, from time to time, been 
 put forward as explanatory of electrical phenomena, we 
 shall, for the purposes of this book, speak of electricity 
 as an entity, as a something which under electrical 
 
 pressure is moved from point to point along well 
 defined paths. The phenomena due to electricity can 
 be controlled by controlling the pressure under whose 
 influence the movements are made. 
 
 Constant in Quantity. — The idea that electricity, so 
 far as this mundane sphere is concerned, is constant in 
 quantity is by no means new, although it is only during 
 recent years that it has been formulated. It is a con- 
 venient euphonism sometimes to speak about the 
 generation of electricity, but it is inaccurate. Unfortu- 
 nately, the majority of text-books, following each other 
 like sheep do the bell-wether, insert paragraphs of ante- 
 diluvian origin about the generation of electricity — 
 hence the popular notion is that the apparatus em- 
 ployed generates electricity, whereas it produces only 
 electrical pressure. Just as we get no flow of water 
 without a difference of water level, so we get no elec- 
 trical manifestations without a difference of electrical 
 levels or pressure. 
 
 All the apparatus constructed and employed is for 
 the purpose of producing this difference of pressure, for 
 measurements connected therewith, or for placing the 
 electricity at the point where it is intended it shall be 
 utilised. Again, just as in the distribution of water we 
 have to provide duly suitable channels, so in the dis- 
 tribution of electricity it is necessary to provide suit- 
 able channels, paths, or, as they are technically termed, 
 circuits. 
 
 Circuits. — Electricity manifests itself in two diffe- 
 rent ways, and for these different manifestations two 
 different kinds of channels, paths, or circuits have to 
 be provided. These two kinds of circuits will be dis- 
 cussed under the respective heads of the Conductive 
 Circuit and the Magnetic Circuit. 
 
 All the phenomena due to electricity can be best 
 studied by a careful examination of these circuits, and 
 we are fain to believe in a much better and much 
 simpler manner than by any other method. It may be 
 thought necessary to introduce a subsidiary circuit, to 
 be called an inductive circuit, as an aid to the reader, 
 but the idea of "action at a distance," which seems in- 
 volved in the ordinary notions of induction, will find 
 no place here. It is assumed that no action originates 
 at one point and its effects perceived at another 
 point without some physical connection between the 
 two points. The phenomena of the circuits are many, 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 and the interactions of the circuits important, inasmuch 
 as it is these interactions that can be modified and 
 moulded to suit the requirements of the engineer. The 
 latter troubles himself little about the immense mass 
 of investigations which delight the worker in the 
 domain of pure science, and has but one simple ques- 
 tion about everything, "What is the use of it?" If 
 there is any direct application the engineer is glad to 
 know all about it, and sooner or later he will probably 
 
 harness it in another direction, forming a newer and 
 more perfect combination. Still it is requisite to respect 
 the work of pure science, as at any moment a new ex- 
 periment may give results that can be usefully applied 
 in practical work. Thus there may prove to be pheno- 
 mena connected with the circuits not discussed, and if 
 so, it may be taken for granted that our knowledge of 
 them is so meagre that their place in applied science 
 cannot at present be determined. 
 
 CHAPTER II. 
 
 THE CONDUCTIVE CIRCUIT. 
 
 SSUME that we have a source of electrical 
 prt'ssure, which source, as will be seen 
 further on, can be obtained in many different 
 forms, it is only necessary in addition to 
 provide the channel or circuit through 
 which the electricity is to pass. This circuit 
 must be a closed path or a series of closed 
 paths, and if a suitable circuit is obtained, it is only 
 Decessary to insert the source at some point in the 
 circuit and put it in action, when it appears as if 
 the electricity naturally existing in the circuit is set 
 into greater or lesser motion according to the pressure. 
 The path or circuit may be long or short, symmetrical 
 or unsymmetrical, of the same or of different but 
 suitable materials; the essential point about it is that 
 from whatever point it starts, to that point it must 
 again return. We have been careful to mention suit- 
 able materials, for all materials are not suitable, and 
 some are more so than others ; in fact, it seems as if 
 a circuit of one material presented more difficulties to 
 the action of the electricity than a circuit of another 
 material. The difficulty placed by the material of the 
 circuit in the way of electricity is termed resistance. 
 Thus metals usually offer very little resistance, while 
 dry air, glass, indiarubber, etc., offer very great resist- 
 ance — hence it is customary to divide materials into 
 two classes, conductors and non-conductors or insulators, 
 the former being those of least resist- 
 ance, the latter those of the greatest 
 resistance. We may indicate a simple 
 circuit diagrammatically as in Fig. 1, 
 where S indicates the source of the 
 electrical pressure, shortly called the 
 source, and K the circuit outside of 
 the source, the interior of the source completing the 
 entire circuit. 
 
 Conductors and Insulators. — Conductors are used to 
 convey the electricity to the point where it is to be used, 
 insulators are used to stop it from going where it is not 
 wanted. The best conductors present some obstacles 
 to the motion of the electricity, generally called the 
 
 Fio. l. 
 
 current, and the best insulators are only those which 
 present an enormously greater resistance to the current. 
 So far as investigation has yet gone, we have neither 
 perfect conductors nor perfect insulators, nor can a line 
 of demarcation be drawn to say where conduction ceases 
 and insulation begins. Thus the difference between 
 conductors and insulators is merely one of degree. The 
 following list may be of some use, as showing the best 
 conductors and the best insulators. The best conduc- 
 tors are put at the top of the list of conductors, and the 
 best insulators — that is, the worst conductors — at the 
 top of the list of insulators. 
 
 Conductors. 
 
 All metals. 
 
 Well-burned charcoal. 
 Plumbago. 
 Acid solutions. 
 Saline solutions. 
 Metallic ores. 
 Animal fluids. 
 Living vegetable sub- 
 stances. 
 Moist earth. 
 Water. 
 
 Insulators (Non-conductors). 
 
 Dry air. 
 
 Shellac. 
 
 Paraffin. 
 
 Amber. 
 
 Eesins. 
 
 Sulphur. 
 
 Wax. 
 
 Jet. 
 
 Glass. 
 
 Mica. 
 
 Ebonite. 
 
 Guttapercha. 
 
 Indiarubber. 
 
 Silk. 
 
 Dry paper. 
 
 Parchment. 
 
 Dry leather. 
 
 Porcelain. 
 
 Oils. 
 
 Phenomena of the Circuit.— The phenomena to be 
 considered in connection with the circuit are three : 
 
 1. Eesistance, for which the symbol E will be 
 
 used. 
 
 2. Electrical Pressure, or as it is often called 
 
 electromotive force, for which the symbol 
 E will be used. 3 
 
 3. Electricity in Motion, usually called current 
 
 for which the symbol C will be used. 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 Resistance. — The phenomena appearing in one cir- 
 cuit are comparable with the phenomena appearing in 
 other circuits, and in order to be able to make these 
 comparisons with facility, certain units have been 
 adopted for each case. For the conductive circuit, 
 a series of units suitable for laboratory work and pure 
 science have been adopted, as well as the practical 
 units required in our workshops. Unfortunately, as 
 yet only the former kind of units have been adopted 
 for the magnetic circuit, hence calculations for the 
 latter are more cumbrous than is the case for the 
 conductive circuit. It will be unnecessary to enter 
 further than possible into the domain of pure science, 
 so we shall simply give the names and definitions of 
 the practical units as they are required. It may, 
 however, be stated that these practical units are 
 derived from the absolute units, based on the centi- 
 metre-gramme-second (C.Gr.S.) system, so called 
 because the " centimetre" is the unit of length, the 
 " gramme" the unit of mass, and the " second" the unit 
 of time. This absolute system is very well adapted to 
 the wants of pure science, but not to practical work. The 
 unit of resistance for practical work is called the ohm, 
 after Prof. Ohm, to whose labours the present position 
 of our knowledge of the conductive circuit is greatly 
 due. The ohm, according to the latest definition, is the 
 resistance of a column of pure mercury 106 centimetres 
 (41"73 inches) long, 1 square millimetre ('00155 square 
 inch) in sectional area, at a temperature of deg. C. It 
 may safely be said that the ideal ohm, according to this 
 definition, never has and never will be made. Instru- 
 ment makers construct their resistances of german 
 silver or platinum silver in terms of standard resistances 
 first obtained by a committee of the British Association, 
 and known as the B.A. unit. The above definition is 
 that known as the legal ohm. 
 
 1 B.A. unit = 0-9889 legal ohm. 
 1 legal ohm = 1-0112 B.A. units. 
 Instrument makers construct boxes containing various 
 multiples and sub-multiples of the ohm, arranging them 
 as far as possible in the manner best suited for the 
 measurements for which they are to be used. 
 
 Equivalent Conductors. — The resistance of any con- 
 ductor varies directly as its length and inversely as its 
 sectional area. The increasing the length increases 
 the path or circuit, and it seems almost axiomatic that 
 if a conductor of a given length has a certain resistance, 
 and is increased to double that length, it will have 
 double resistance. Increasing the area of a conductor 
 is similar to adding another conductor to the circuit, 
 increasing the size of the path over which the elec- 
 tricity has to travel, rendering it less difficult. If R is 
 the resistance of a conductor, I its length, and s its 
 
 sectional area, 
 
 I 
 Itcc- 
 
 If we call the power of a body for conducting elec- 
 tricity its conductivity, and represent this by e, recol- 
 lecting that conductivity is the inverse of resistance, 
 we may write 
 
 I I 
 
 E = — and c = — t>. 
 s c s a 
 
 It is easy to compare the conductivities of various, say 
 two, substances. Taking them of the same length and 
 section, and let c x and c 2 represent the conductivities, 
 B x and B 2 the respective resistances, then 
 
 • c • • .i ■ J_ 
 
 or the dimensions being the same, the conductivities 
 are inversely as the resistances. 
 
 It is sometimes convenient to be able to replace one 
 conductor by another conductor of different materials, 
 at the same time not convenient to increase or decrease 
 the total resistance of the circuit. A conductor so used 
 to replace another without altering the total resistance 
 is termed an equivalent conductor. All that is necessary 
 to obtain equivalent conductors is to measure or calcu- 
 late that the resistance or conductivity of the conductor 
 required is equal to the resistance or conductivity of 
 that to be replaced. In a section further on the method 
 of measuring resistances will be described. Meanwhile, 
 the sectional area, s, is known, if we know the length , 
 weight, and specific gravity of the substance. Let w be 
 the weight, and a- the specific gravity, then the volume 
 v = Z s; and the volume, I s, multiplied by the specific 
 gravity, a-, is the weight, w ; or 
 w = Isa- 
 
 I a- 
 
 Putting this value of s in the formula 
 
 I 
 
 c- 
 
 we get 
 
 c= 
 
 Per 
 
 Then, with two conductors, C, C v of length I, l v of con- 
 ductivities c, c v and sectional areas s, s v we should get 
 the same resistance, and one might be substituted for 
 the other, when 
 
 J Zj_ 
 
 c s ~ c l s 1 
 
 Another way of attaining the same result is by starting 
 as before with 
 
 I 
 Eqc — 
 
 s 
 
 Knowing that different conductors have, for equal 
 lengths and sections, different resistances, and taking 
 unit lengths and unit sections of some conductor, such 
 as pure silver, as our standard, the resistances of 
 similar pieces of every other conductor can be expressed 
 in terms of the standard. Such expression may be called 
 the specific resistance of the material. Thus, if the con- 
 ductivity of silver is 1, and of copper - 99, the specific 
 
 resistance of silver will be 1, and of copper :Kq = -qq-> 
 
 the specific resistance being the reciprocal of conduc- 
 tivity. The reciprocal of a number is unity divided by 
 that number. A table of conductivities then gives also 
 
8 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 a table of specific resistances. Using specific resistance, 
 a, the above formula becomes 
 
 r =4 
 
 If we have two wires whose lengths are I, l u their 
 sectional areas s, s v their specific resistances a, a u 
 then their actual resistances, E and B^ may be found. 
 
 I L 
 
 R=a— and E, = a,—. 
 
 s 1 1 s 1 
 
 Dividing the second by the first we get 
 
 K x a x /j s 
 
 Example. 
 
 Assume that the specific resistance of iron is seven 
 times that of copper. How thick must an iron wire be 
 which for the same length shall offer the same resis- 
 tance as a copper wire '5sq. in. in section? In the 
 formula let the symbols E lf a u Zj, s lt refer to iron, and 
 E, a, I, s, to copper. 
 
 Kj <jj Zj s 
 
 E ~~ a I s x ' 
 
 but Ej = E and l± = I, a x = l, a =1, s = '5, therefore 
 the formula becomes 
 
 1=TX 1 X ~ 
 1 Sj 
 
 s 1 = 7 x -5 = 3-5, 
 
 which is simply saying that iron to have the same 
 length and resistance must have a sectional area of 
 3'5in., as against the sectional area of copper 'Sin., 
 which of course, in so simple a case, would be got in 
 practice by saying that as iron had seven times greater 
 resistance than copper, the same lengths would require 
 seven times greater sectional area in the iron than in 
 the copper to have wires of same resistance. 
 
 For the purposes of the British Association Com- 
 mittee on the unit of resistance, Prof. Matthiessen 
 determined the conductivities of various metals at 
 different temperatures. These experiments are classic, 
 and the results are given in the following table. Prof. 
 Matthiessen put the conductivity of pure silver at 
 Odeg. C.-100. 
 
 
 Conductivities. 
 
 Metals. 
 
 At deg. C. 
 „ 32 deg. F. 
 
 At 100 deg. C. 
 „ 212 deg. F. 
 
 Silver, hard 
 
 100 
 
 99-95 
 
 77-96 
 
 29-02 
 
 23-72 
 
 18-00 
 
 16-80 
 
 12-36 
 
 8-32 
 
 4-76 
 
 4-62 
 
 1-60 
 
 1-245 
 
 71-56 
 
 Copper, hard 
 
 70-27 
 
 Gold, hard 
 
 55-90 
 
 Zinc, pressed 
 
 20-67 
 
 Cadmium 
 
 16-77 
 
 Platinum, soft 
 
 
 
 
 Tin 
 
 8-67 
 
 Lead 
 
 5-86 
 
 Arsenic 
 
 333 
 
 Antimony 
 
 3-26 
 
 Mercury, pure 
 
 
 
 0-878 
 
 The conductivities of the metals are very high, and 
 consequently the resistances very low compared with 
 other substances. Thus, if we take a piece of copper 
 to have a resistance equal 1, a similar length and sec- 
 tional area of distilled water will have, according to 
 Culley, a resistance of 6,754 million times as great. 
 
 Influence of Temperature. — The resistance of most 
 conductors increases with rise of temperature, and this 
 peculiarity has to be allowed for in the construction of 
 dynamos and motors, as these machines, and especially 
 dynamos, have frequently to work at temperatures, not 
 only considerably higher than that of the air outside 
 the dynamo-room, but much higher than the surround- 
 ing air. 
 
 Arnasten, Matthiessen, and Siemens made very 
 elaborate experiments to determine the relations be- 
 tween resistance and temperature, and the result of 
 their work comes practically to this, that for every 
 degree Fahrenheit copper increases in resistance about 
 
 2 12 
 
 two-tenths of one per cent., ^ of tkx or TnnTj or '002. 
 
 A length of copper having a resistance of 1 ohm at 
 32deg. F. would therefore have a resistance of 1 + 
 (-002 x 50), or l'l at 82deg. F. Put in another way, a 
 length of wire that had a resistance of 10 ohms at 
 32deg. F. would have a resistance of 11 ohms at 82deg. 
 F. We have adopted the Fahrenheit scale because that 
 scale is the most frequently used in England. If the 
 centigrade scale be used the increase of resistance will 
 be about - 0036 for every ldeg. C, for 
 
 100 deg. C. = 180 deg. F., or 1 deg. C. = 9 ° F ', 
 
 and 
 
 ~ of -002 = -0036. 
 5 
 
 According to the experiments above referred to, the 
 conductivities of all pure metals in the solid state de- 
 crease, or the resistances increase, in the same ratio, 
 except in the metals iron and thallium. The conduc- 
 tivity of iron between ldeg. C. and 100 deg. C, or 32 deg. 
 F. and 212deg. F., varies to the extent of 39-2 per cent., 
 while that of thallium between the same temperatures 
 varies 31'4 per cent., other metals varying only 29-3 
 per cent. This result we attribute rather to the possi- 
 bility, although the greatest care might be taken, that 
 the iron and the thallium experimented with were not 
 pure. 
 
 The following table shows the amount of resistance 
 of a few substances used for various electrical purposes 
 by which 1 ohm is increased by a rise of temperature 
 ldeg. F., or ldeg. C. 
 
 Eise of E. of 1 Eise of E. of 1 
 
 Material, ohm when heated ohm when heated 
 ldeg. F. ldeg. C. 
 
 Platinoid '00013 '00021 
 
 Platinum-Silver.. '00018 '00031 
 
 German Silver . . . '00024 " -00044 
 
 Gold, Silver '00036 ... -00065 
 
 Cast Iron '00044 '00080 
 
 Copper -00215 -00388 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 The resistance of true alloys — that is, alloys in which 
 there is no chemical action — depends upon the pro- 
 portional volumes of the metals entering into their 
 composition. Alloys of lead, tin, cadmium, and zinc 
 are of this class, while the alloys of bismuth, antimony, 
 platinum, palladium, iron, aluminium, sodium, gold, 
 silver, and copper have greater resistance than would 
 be calculated from the properties of metals forming the 
 alloy. 
 
 Annealing. — The degree of hardness or softness of a 
 metal or alloy affects its resistance. That of a hard 
 drawn wire is not the same as when the wire has been 
 made hot and let cool again. Eesistance is lessened by 
 annealing. Matthiessen gives the following results for 
 copper and silver : 
 
 Metals Temn Conductivity. 
 
 Metals. ±emp. Hard Annealed. 
 
 Copper ... 11° 95-31 9783 
 
 Silver 14-6° 95-36 103-33 
 
 The comparison being made with pure silver at 
 lOOdeg. C. 
 
 Dr. Siemens, who compared the conductivities of 
 copper, silver, and brass with pure mercury at Odeg. C, 
 gives the following results : 
 
 Metal. Hard. Annealed. 
 
 Copper 52-207 55253 
 
 Silver 56-252 64"380 
 
 Brass 11-439 13"502 
 
 These results will show the necessity for all practical 
 purposes of employing soft annealed metals as con- 
 ductors. Further, as a rule, hard drawn wires do not 
 have a permanent resistance. The question of altera- 
 tion in resistance is unsettled, but it is believed that 
 there is a molecular change going on in conductors 
 long, for years, after manufacture, which slightly 
 affects their resistance. This change is ordinarily so 
 inappreciable that for constructional purposes it is 
 never considered. 
 
 Ohm's Law. — This is by far the most important law 
 that holds in the domain of electrical science. It was 
 first promulgated by Prof. S. J. Ohm in his work, 
 "Die Galvanische Kette, Mathematisch Bearbeitet" 
 (The Galvanic Circuit Mathematically Considered) in 
 1827, in which was given the general differential equa- 
 tion for the propagation of electricity through pris- 
 matic bodies, and from this equation he deduced the 
 simple law which justly bears his name. This law gives 
 us the relationship existing between the three funda- 
 mental phenomena of the conductive circuit — viz., 
 electrical pressure, resistance, and current — and it may 
 
 be written 
 
 Electrical Pressure 
 Current = 
 
 orC 
 
 Eesistance 
 E 
 
 when B stands for electrical pressure (electromotive 
 force), C for current and B for resistance. The 
 formula can be written 
 
 E 
 E=CEorK = g. 
 
 volts 
 amperes = -r — ; 
 r ohms 
 
 and gives the means of ascertaining the third quantity 
 when the other two quantities are known. We have 
 already referred to the practical unit of resistance 
 adopted, we now give the others. 
 
 The unit of resistance is called the ohm. 
 „ „ electrical pressure „ volt. 
 „ „ current „ ampere. 
 
 And in these units the above formula become 
 
 volts 
 
 volts = amperes x ohms ; ohms = -. 
 
 r ' amperes 
 
 Thus practical calculations become exceedingly simple. 
 Examples of Calculations. 
 In Fig. 1 let s be the source of pressure, and give an 
 effective pressure, = 100 volts, through the resistance, 
 B, = 2 ohms, to find the current, C. 
 
 n E volts 
 
 G= E' am P eres = ohm"s 
 _100 
 ~ 2 
 The number of amperes through B therefore is 50. 
 
 "What pressure is required to give 50 amperes through 
 a resistance of 2 ohms ? 
 
 Pressure = amperes x ohms 
 = 50 x 2 
 = 100 Answer, 100 volts. 
 
 What resistance is required to obtain a current of 50 
 amperes when the pressure is ] 00 volts ? 
 
 = 50. 
 
 Eesistance = 
 
 pressure 100 
 
 = 50 = 2 ' 
 
 current 
 
 Answer, 2 ohms. 
 
 But all circuits are not so simple as this, and we may 
 now consider 
 
 Divided Circuits. — If the simple circuit is represented 
 in Fig. 2, where S is the source and B the resistance, 
 this circuit can be modified by introducing a second 
 
 5 
 
 Fig. 2. 
 
 Fig. 3. 
 
 path, as in Figs. 3 and 4. The current due to the pressure 
 of the source, S, has now the choice of two parts, or to 
 divide itself between both. It does the latter and 
 divides itself inversely as the resistances, or rather the 
 original current is added to because the total resistance 
 is lessened, the total current now acting dividing itself 
 as stated. There is no alteration of current in the 
 original circuit. Under the new conditions it is the 
 division of the total current that has to be considered. 
 Let B and B t represent the resistances, and C and G t 
 the currents. 
 
 C x E = C x x Rj, 
 
 or the product of current x resistance is constant for 
 
10 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 both branches. This can be written 
 
 C 
 
 
 which shows the currents are inversely as the resist- 
 ances. From this equation, if any three of the 
 quantities are known, the fourth can be found — thus : 
 
 C = Cl TT 
 
 1 Bi 
 
 C 
 
 R 1 = 
 
 CE 
 
 In the case of the double circuit just described, one 
 circuit is said to be a shunt to the other, or the circuits 
 are in multiple arc or in parallel. The latter terms, 
 however, are more used when there are several circuits. 
 The use of parallel circuits is very extensive. It will 
 be seen further on that electromagnets are made by 
 coiling a conductive circuit around a core of soft iron. 
 Now if we have two circuits in parallel, there are two 
 ways of winding around two cores. Let A and B 
 represent sections of iron cores, they can be encircled 
 by only one branch, as in Kg. 5, or they may be en- 
 
 Fia. 5. 
 
 Fro. 6. 
 
 circled by both branches, as in Fig. 6. In the former 
 case they are said to be shunt-wound, in the latter case 
 compound-wound, so that shunt-winding means that 
 one branch of the circuit only encircles the cores 
 compound-winding means that both branches of the 
 circuit are wound round the cores. 
 
 We have already stated that the adding a shunt to 
 the circuit or any number of branches within certain 
 limits does not interfere with the current going through 
 the original branch or branches. The analogy of 
 water will help to make this clear. If we have a cistern 
 filled with water, and so arranged as to be kept full, or 
 the head of water in it kept constant, and fix an 
 ordinary pipe to the cistern, we can draw off a certain 
 quantity of water per minute through this pipe, and 
 the quantity we can so draw off is not affected by 
 fixing another pipe to the cistern and drawing water 
 through that. As long as the head of water in the cistern 
 is constant, the pipes will each give its full complement of 
 water proportional to that head. If the second pipe is 
 in all respects similar to the first, we get from the two 
 pipes just double the quantity of water in a given time 
 that we should get from either pipe alone— that is, the 
 quantity of water discharged is proportionally increased 
 by the addition of the second pipe. The same holds 
 with electricity. Keeping the electrical pressure con- 
 stant, the addition of a branch circuit means the motion 
 through it of a current proportional to the pressure 
 
 and its own resistance. There is practically no inter- 
 ference with the original branch. This, indeed, might 
 be seen from a study of Ohm's law. The addition of a 
 branch decreases the total resistance of the circuit, 
 and as the current is inversely as the resistance of the 
 latter is decreased, the former is increased. In order 
 to know the increase of current, or to know the total 
 current, it is necessary to know the total resistance of 
 the circuit. 
 
 Total or Joint Resistance. — If conductors are 
 arranged one after the other as in Fig. 7 or Fig. 8, 
 they are said to be in series, and the total resistance 
 
 A 
 
 B 
 
 C 
 
 Ri 
 
 Bo 
 
 Fig. 7. 
 
 B, 
 
 Fig. 8. 
 
 is the sum of their separate resistances. Thus if A, B, 
 and C be three conductors arranged in series, having 
 respectively the resistances B 1; B 2 , B 3 , while B repre- 
 sents the total resistance, then 
 
 E=B 1 + B 2 + B 3 . 
 
 In the simple circuit referred to in Fig. 9, the total 
 resistance of the circuit is the sum of two resistances, 
 for there is the resistance of the in- 
 ternal parts of the source, and also the 
 resistance of the conductor external 
 to the source, and generally in all 
 circuits these two sets of resistances 
 have to be considered. Both internal 
 and external resistances may be in 
 parallel. It is customary in Ohm's formula to dis- 
 tinguish the external from the internal resistance hy 
 writing B for the total external, and r for the total 
 
 E 
 
 internal resistance, or C = ^ 
 
 B + r 
 
 "We will now find the total or joint resistance of two 
 branches. 
 
 Unless the contrary is stated, the electrical pres- 
 sure referred to is that available in the external circuit. 
 
 In Figs. 10 and 11 let Bj, 
 B 2 represent the resist- 
 ances of the two branch 
 circuits, and B the total 
 resistance of the external 
 circuit. The total cur- 
 rent in the external 
 circuit will be the sum 
 of the separate currents 
 . x , in each branch. Let C 
 
 be the total current, and C„ C 2 the currents in 
 branches Bj, B 2 respectively. Then C = C +C • but 
 BE ^ E ' " 
 
 * 3 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 11 
 
 Let E-l, then C = 
 
 1_ 
 E 
 
 1 
 
 '%' andC 3 = E; 
 
 i 
 
 or — = — _). - 
 
 R Rj Rn 
 
 _R2 +Rj 
 
 R^lg 
 = r7+r7 
 
 whence R = 
 
 that is, a current C would flow through the circuit 
 
 having a resistance — — ^~, 
 j^n + E 2 
 
 joint resistance of E x and E 2 . 
 
 therefore 
 
 EjE, 
 
 Ej + E 
 
 |- is the 
 
 -6- 
 
 "j 
 
 «3 
 
 Similarly it can be shown that the joint 
 resistance of three branches, Eig. 12, having 
 respectively the resistances E x , E 3 , E 3 is 
 
 W 
 
 E = 
 
 EjE;>E 3 
 
 EjEj + E 1 E 3 + E 2 E;j 
 
 Fro. 12. 
 
 When the branch resistances are equal this formula 
 is very simple, and becomes 
 
 where 
 
 IV 
 
 x n 
 
 E, 
 
 the resistance of one branch. 
 n = the number of branches. 
 
 Kirchhoffs Laws. — These laws or propositions are 
 two, one of which may almost be said to be axiomatic 
 or self-evident ; the other is merely an amplification of 
 Ohm's law. It is sometimes convenient to use these 
 propositions instead of the above methods in calcula- 
 tions. The first of these propositions is that 
 
 The sum of the currents in all the wires which meet in 
 
 a point is equal to nothing. 
 Which is another way of saying all currents in a con- 
 ductive circuit going to a point must also go from it, 
 because the point does not act as a reservoir. Let 
 
 Fio. 13. 
 
 C . C a , C s convey currents to the point 0, Kg. 13, and 
 c c 2 , c 3 , c 4 convey currents away, then C x + C 2 + C 3 = 
 c^+c 2 + c 3 + c 4 ,or Cj + C, + C, - (cj + c, + c 3 + c«) = 0, 
 
 The second proposition is that 
 
 The sum of all the products of the currents and 
 resistances in all the branches forming a closed circuit 
 is equal to the sum of all the electrical pressures in the 
 same circuit. 
 
 Which is another way of saying that when E = E x + 
 E t + E 3 , etc., and C = C x + C 2 + C 3 , etc., and E is the 
 total resistance of Ej, E 2 , E : „ etc., then 
 
 E x + E 2 + E 3 + etc. = C^Rj + C 2 R 2 + C 3 R 3 + etc. 
 
 that is E = C R. 
 
 The following table of resistance will perhaps give a 
 better idea of the relative values of the metals as con- 
 ductors than the table of conductivity previously given : 
 
 Metal. 
 
 R in ohms of 1 
 
 yard of wire 30 
 
 mils diam. 
 
 Metal. 
 
 R in ohms of 1 
 
 yard of wire 30 
 
 mils diam. 
 
 w 
 
 a v 
 
 Cu 
 
 Fig. 14. 
 
 Silver "0359 Nickel "2526 
 
 Copper -0324 Tin -2678 
 
 Gold -0417 Lead "3980 
 
 Aluminium "0590 Antimony '72 
 
 Zinc -1274 Bismuth 2"73 
 
 Platinum "1836 Mercury 2" 
 
 Iron "1970 German silver ... '4244 
 
 A mil is one-thousandth of an inch, or 1,000 mils 
 make one inch. 
 
 Current. — If we take a source of electrical pressure, 
 such as the combination, Eig. 14, consisting of a glass 
 jar three parts filled with dilute sulphuric 
 acid (ten parts water to one of acid) , with 
 strips of copper (Cu) and zinc (Zn), placed 
 as shown, and joined externally, that is, 
 outside the liquid, by the conductor, W, 
 we get certain electrical phenomena, said 
 to be due to a current of electricity passing 
 along the conductor. It may be well to 
 state here that opinions differ somewhat 
 as to the exact part played by the conduc- 
 tor in the electric circuit. It is usually 
 surrounded by air, or some other non-conductor, and 
 it is very probable that the non-conductor plays 
 as important, if not a more important, part in the 
 phenomena as does the conductor. After due con- 
 sideration we shall assume that both the conductor 
 and the surrounding dielectric, or non-conductor, 
 have to play their respective parts in the observed 
 phenomena. It is not for us, in a practical work of this 
 kind, to enter into long discussions upon the tendencies 
 of modern scientific thought and investigations, and this 
 reference has been made solely with a view to say that 
 in every case the most advanced, as well as the older, 
 theories have been reviewed, and the views put forward 
 are those deemed most fitted to appeal to workers. 
 
 As soon as the external wire in the above combina- 
 tion is connected, if its resistance be not too great, 
 small bubbles will be seen to form upon the copper 
 plate in the fluid, gradually enlarging, and finally 
 bubbling to the surface; also if a small magnet be 
 brought near the wire it is visibly affected. Then, 
 again, if the temperature of the wire was noted in its 
 state before connecting, and again noted in its state 
 
12 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 after connecting, it would be found that the latter 
 temperature was higher than the former. These 
 phenomena are said to be due to the current. We have 
 assumed that electrical pressure has been set up at apoint 
 in the combination, or that a difference of pressure has 
 been made between two points, and under the influence 
 of this pressure electricity is forced around the circuit. 
 In years gone by, when our knowledge of the subject 
 was less, and when all scientific men spoke of the 
 apparatus as generating electricity, the simplest way to 
 explain certain phenomena was to imagine the genera- 
 tion of two kinds of electricity. There is, as a matter 
 of fact, no reason to imagine two kinds of electricity, or 
 even to imagine one kind. Electrical phenomena may 
 be due, as light phenomena are due, to certain actions 
 upon the ether. However, we have preferred to con- 
 sider the phenomena as due to action upon electricity. 
 A pressure is exerted upon it, and a path is arranged 
 for its progress ; hence the motion of electricity through 
 this path, giving us the phenomena of the current. The 
 pressure generated must be looked upon as directive — 
 that is, acting in a certain direction, or causing action 
 in a certain direction. 
 
 Analogy with Flow of Water.— Prof. Perry and Dr. 
 Wormell have both tabulated the analogies between 
 flow of electricity, or current, and flow of water. We 
 quote the following sentences from Dr. Wormell: It 
 will be known as a simple principle of dynamics that 
 to lift a weight a certain height, work must be done, the 
 measure of which is the product of the weight and 
 height to which it is raised, and when a body falls from 
 a given height it acquires an active energy, measured by 
 the product of the weight of the body and the height 
 through which it falls. If, for instance, 1,0001b. fall 
 through a height of 10ft., the energy accumulated at 
 the end of the fall is 10,000 foot-pounds. If the 1,0001b. 
 are then arrested, and no other result follows, this 
 10,000 foot-pounds of energy is turned into heat. Now 
 Joule has taught us that 772 foot-pounds of dynamical 
 energy are equivalent to the heat that will raise the 
 temperature of lib. of water 1 deg. F. Hence the heat 
 produced in the instance we have taken for illustration 
 is 10,000-^772 units of heat. These principles enable 
 us to point out certain analogies that exist between the 
 work done by a flow of water and that done by a current 
 of electricity. 
 
 Water. 
 
 1. Suppose a quantity of 
 water to flow from a higher 
 level to a lower, the energy 
 liberated will be measured 
 by the product of the quan- 
 tity of water, Q, and the 
 difference of the two levels, 
 D. If the water flows 
 away without doing work, 
 and leaves the lower level 
 with the same speed as it 
 enters the upper, then the 
 energy due to the fall has 
 all been converted into heat 
 
 Electricity. 
 
 1. Suppose a quantity of 
 electricity to flow between 
 points that have a differ- 
 ence of potential, the en- 
 ergy set free will be mea- 
 sured by the product of 
 the quantity, Q, and the 
 difference of pressure of 
 the two points, D. If the 
 electricity flows without 
 doing work, all the elec- 
 trical energy is converted 
 into heat, because of the 
 electrical resistance of 
 
 by the frictional resistance 
 of the pipes through which 
 it flows. If H be the quan- 
 tity of heat produced in the 
 pipes, H = QD-h 772. 
 
 2. If we let the water do 
 work on the way, turn a 
 mill, for instance, then the 
 backward pressure of the 
 machine tends to stop the 
 flow of water, and if it is 
 to issue at the lower level 
 with the same velocity as 
 before, the resistance of 
 the pipes must be dimi- 
 nished, and less heat will 
 be produced. If W is the 
 measure of the work done, 
 then H = (Q D— W) di- 
 vided by 772. 
 
 3. Water may be brought 
 from any distance, and be 
 made to give out nearly all 
 its energy to work a ma- 
 chine. To secure this re- 
 sult there must be a great 
 head of water, and only a 
 small quantity allowed to 
 flow per second. 
 
 From the first of these analogies we conclude that if 
 C is the quantity of electricity flowing per second, and 
 E the difference of pressure between two points of the 
 circuit, then the amount of energy expended between 
 these points is C E. But if E be the resistance, then, 
 by Ohm's law, E = C E, and the amount of energy 
 liberated is C 2 E. 
 
 Phenomena Connected with a Wire Carrying a 
 Current— -If we arrange a circuit, as in Fig. 15, with a 
 portion of the conductor vertical, and pass this part of 
 
 wire. If H be the mea- 
 sure of the heat produced 
 in the wire,H = Q D divided 
 
 by 772. 
 
 2 . If we let the electricity 
 work a machine in passing 
 between the two points, 
 the motion of the machine 
 produces a backward pres- 
 sure which diminishes the 
 flow of electricity, and in 
 order that the current may 
 remain the same, the re- 
 sistance must be di- 
 minished, and then less 
 heat will be produced. If 
 W is the measure of the 
 work done, then H = (Q D 
 — W) divided by 772. 
 
 3. Electricity may be 
 brought from any distance, 
 and be made to expend it- 
 self in working a machine. 
 To secure this result there 
 must be a great difference 
 of pressure, and a small 
 current. 
 
 Fig. 15. 
 
 the wire through a piece of cardboard, we shall find 
 when a current is passing that if some iron filings are 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 Large Electrolier for 1. Lights m the Venetian Room, Holborn Restaurant, London-Wrought Brass and Lacquered in Imitation of Matted Gold. 
 
 MESSll*. 1!. VEKITl' AND soxV ELEL'TIUC LIGHT FlTTISUs.. 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 B11USH -UX LAJ11'. 
 
PRACTICAL ELECTRICAL ENGINEERIttd. 
 
 13 
 
 sprinkled upon the cardboard the filings will arrange 
 themselves in circular form, as in Fig. 16. From this 
 
 Fig. 16. 
 
 and other experiments it is found that a current through 
 a conductor has an influence external to the conductor, 
 and the whole of the space through which such influence 
 extends is called the field of the conductor. The resul 
 upon the iron filings is similar to that of a directive 
 force or forces starting from the conductor and acting 
 in a direction tangential to the axis of the filings when 
 arranged. 
 
 Again, if the terminal wires of a source, instead of 
 being connected together, are placed in a liquid solution 
 which is a conductor, the circuit is completed just as if 
 it were wholly composed of wire. In many solutions, 
 as soon as a current is started in the circuit a chemical 
 action takes place, and the solution is decomposed. 
 Such decomposition is called electrolysis. This action 
 seems to take place in the direct fine of the conductor, 
 hence we may say the phenomena connected with the 
 current are twofold : 
 
 1. Those within the conductor. 
 
 2. Those external to the conductor. 
 
 Heating Effect of Current. — Of the phenomena con- 
 nected with the effects of the current within the con- 
 ductor the two most important are the heating effect 
 and the chemical effect." It is to the heating effect 
 that the value of electricity as a lighting agent is due. 
 Looking, then, at the production of the electric light 
 as one of the most important branches of electrical 
 engineering, this part of the subject to be discussed 
 may be briefly stated. Accepting the modern theory 
 that light is due to vibrations of the ether caused by 
 highly heated bodies, in order to obtain artificial light 
 the highly heated bodies must be placed at the points 
 where the light is desired, and assuming that every 
 reader is fully alive to the impossibility of each indi- 
 vidual member of the community arranging for inde- 
 ' pendent sources of generation of fight at every point 
 required, we come to the problem of how best to 
 distribute from central points the means of providing 
 artificial lights over large areas, and at the various 
 points in those areas where the light is required. The 
 two great systems in vogue are the production of gas 
 at certain points, and distributing that gas over large 
 areas; and the generation of electrical pressure at 
 certain central points, and distributing an electric 
 current over the required area. Gas is an inflammable 
 material and its burning, or its combustion, with the 
 
 oxygen in the air gives an artificial nluminant. The 
 value of electricity for this purpose, on the other hand, 
 depends upon its heating effect on the conductor pro- 
 vided to carry the current. Usually, as will be seen 
 further on, the prime motors in electric light instal- 
 lations are steam engines, or, rather, the prime motor 
 of all is the burning fuel in the firebox of the boiler. 
 Then the economical question for all engineers becomes, 
 given a certain amount of heat generated in the fire- 
 box of the boiler, how best to get the largest percentage 
 of heat so generated to the points requiring illumina- 
 tion. 
 
 It is found that the heat generated in a conductor by 
 a current is proportional to the product of the square 
 of the current into the resistance of the conductor. 
 
 Heat generated = C 2 E. 
 
 Now as the current flowing across each section of the 
 circuit is the same, and as at some parts of the circuit 
 the heat is not wanted, whilst at other parts it is 
 wanted, the natural course is pursued of inserting a 
 conductor of high resistance at the point where the 
 heat is required. The heat thus obtained raises the 
 temperature of the conductor till we obtain brilliantly 
 incandescent conductors giving light where wanted. 
 Trace the various changes from the burning fuel to the 
 light, and it will be seen how many opportunities there 
 are for loss on the way. 
 
 The burning fuel is used to generate steam. 
 
 Expansion of steam used to move the piston of the 
 engine. 
 
 Belt from engine drives dynamo. 
 
 Cmrrent due to electrical pressure from dynamo circu- 
 lates around circuit. 
 
 Or, 
 
 Heat energy is transformed into mechanical energy. 
 Mechanical energy is transformed into electrical 
 
 energy. 
 Electrical energy is transformed back again into heat 
 
 energy. 
 
 In every transformation there is a loss, and the loss 
 can be measured and calculated. By energy we mean 
 the capacity for doing work. A force does work when 
 its point of application moves in the direction of the 
 force, and the work done is measured by the product of 
 the magnitude of the force and the distance through 
 which the point of application moves in the direction 
 of the force. The unit of work used by English 
 engineers is the foot-pound. It is the work done in 
 raising one pound through the space of one foot 
 against gravity. This old-fashioned unit did not com- 
 mend itself to the electricians who set to work in the 
 interests of telegraphy to introduce units. They 
 preferred the metric system, hence the attempt to 
 introduce either the grammetre or the Mlogrammetre, 
 which gives as units either a gramme raised a metre, 
 or a kilogram raised a metre. The kilogram = 2'21b., 
 and the metre 3'28ft., so a kilogrammetre is 22 x 3"28 
 or 7 - 2 foot-pounds. 
 
14 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 1 kilogrammetre 
 1 foot-pound 
 
 7 - 2 foot-pounds. 
 1 
 
 As Y0 s * — £ = horse-power, the unit of v\ orl< 
 
 7-2 
 
 = 0138 kilogrammetre. is _ yolt x ampere; and taB been calle d the watt. 
 
 But not only have engineers to consider the quantity 
 of work, they must know the rate of working ; hence 
 time has to be considered. The unit here adopted is 
 the horse-power, equivalent to the raising of 
 33,0001b. one foot in one minute. 
 
 Unit of work = foot-pound. 
 
 Unit of rate of work = horse-power = 33,000 foot- 
 pounds per minute. 
 
 It has long been understood that heat and work are 
 convertible, but it was not till the classic experiments 
 of Joule that the exact equivalents were determined. 
 He proved that lib. falling 772'43ft. would raise the 
 temperature of lib. of water one degree Fahrenheit. 
 This is called the mechanical equivalent of heat, or, 
 shortly, Joule's equivalent, and is taken as 772 foot- 
 pounds. 
 
 The British thermal unit then is 772 foot-pounds, 
 and is designated by J. 
 
 33 000 
 1 horse-power = -},W- = 42-75 thermal units 
 
 (British). 
 
 If the centigrade scale is used, then the heat 
 equivalent is = 1,390 — that is, the work done in raising 
 lib. of water through 1 deg. C. is equivalent to the fall- 
 ing of lib. through a height of 1,390ft., for 1 deg. C = 
 
 \ deg. F., and 772 x % = 1,389-6. 
 o 5 
 
 According to the French system, the thermal unit is 
 
 called a calorie. 
 
 As a foot-pound = 0-1383 kilogrammetre, 
 the calorie = 1390 x 0-1383 = 192 kilogram- 
 metres. 
 The calorie = 1,390 foot-pounds. 
 
 It has been stated that electrical energy is con- 
 vertible into heat energy, and as heat energy is 
 convertible into mechanical energy, we see that elec- 
 trical energy is convertible into mechanical energy, 
 and mechanical energy is convertible into electrical 
 energy. 
 
 If the whole of the electrical energy passing through 
 a circuit in t minutes is converted into heat, then 
 Heat = C 2 K*. 
 
 It is customary to speak of electrical energy as 
 
 equivalent to so many horse-power hours, and using 
 
 tbe units ampere, ohm, volt, we bave to use the 
 
 C 2 B 
 
 -^ = horse-power ; or, as E = C B, this may 
 
 C F 
 
 be written — — P horse-power. 
 74( 
 
 If, then, a current of 100 amperes flows through a 
 circuit having a resistance of 10 ohms continuously for 
 
 an hour, we get 1Q 0' * 10 = 134 horse-power (about) 
 
 formula 
 
 746 
 
 for an hour. 
 
 1 watt = 1 volt x 1 ampere. 
 746 watts = 1 horse-power. 
 
 It will be seen that we have here a simple way of 
 ascertaining the loss by conversion between engine and 
 dynamo. By taking engine diagrams we get the horse- 
 power developed by the engine, and by measuring the 
 current and resistance in the external circuit, or 
 current and pressure, we get the horse-power available 
 for use, the difference between the two is so much 
 loss. 
 
 Further, we can ascertain exactly the horse-power 
 put into the dynamo, and the horse-power given out as 
 available electrical energy, and so ascertain the 
 efficiency of the dynamo as a machine. We have said 
 that the heating properties of the current are required 
 at certain parts of the circuit, while the heat developed 
 in ether parts of the circuit is so much waste. The 
 resistance of the ordinary conductors increases with 
 temperature, and as our available energy depends upon 
 current as well as upon pressure, the resistance should 
 be kept as low as possible. If the temperature is 
 allowed to rise, it may lead to the fusing of wires, to 
 the destruction of insulation, to the causing of fires to 
 the buildings in which the conductors are placed. So 
 long as a current is passing through a conductor there 
 will be heat generated according to the formula C 2 B, 
 and this heat continually increases the temperature of 
 a conductor unless it is radiated from the surface of the 
 conductor or got rid of in some other way. If the 
 temperature of the conductor rises to a certain point 
 and then remains constant, it means that the heat 
 generated is equal to the heat radiated or conducted 
 away. It will at once be seen that there is less diffi- 
 culty in getting rid of the heat generated in a straight, 
 or rather in a single, conductor uan when the con- 
 ductor is coiled. 
 
 Professor George Forbes has paid particular atten- 
 tion to this problem, and in his paper before the 
 Institution of Electrical Engineers, and from the 
 results of his experiments, the following practical con- 
 clusions are derived: "An insulated wire carries a 
 greater current without overheating than a bare wire, 
 if the diameter be not very great." This would 
 naturally have been expected, as the heat is dissipated 
 from the conductor by conduction through the insu- 
 lator, and radiation from the surface of the insulator. ' 
 "Assuming," says Prof. Forbes, "the diameter of the 
 cable to be twice that of the conductor, a greater 
 current can be carried in insulated cables than in bare 
 wires up to 1-9 (nearly two) inches diameter of con- 
 ductor. But if the insulated cable have a diameter 
 four times that of the conductor, this is the case up to 
 1-lin. diameter of conductor." 
 
 Where the thickness of insulation is made very great 
 the limiting size of conductor which favours the insu- 
 lated wire is given in the table. 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 15 
 
 Dia. of insulation 
 
 Dia. 
 
 of conductor. 
 
 2 
 4 
 6 
 8 
 10 
 100 
 
 Limiting dia. of conductor 
 which favours insulation. 
 
 1'9 inches. 
 11 
 
 •986 
 
 •859 
 
 •786 
 
 •393 
 
 Prof. Forbes calculates a table for wires under the 
 conditions given at the head of the table, viz. : 
 
 Table. — Subaqueous and Aerial Cables (insulated). 
 Diameter of cable _ . 
 Diameter of conductor 
 Temperature of air = 20 deg. C. 
 t = excess of temperature of conductor over air. 
 
 
 
 Current in Amperes. 
 
 
 Diameter in 
 
 
 
 
 
 
 centimetres and 
 
 
 
 
 
 
 
 
 
 
 
 mills. 
 
 t=\°. 
 
 «=9°. 
 
 <=25°. 
 
 i=49°. 
 
 C = 81°. 
 
 Cm. 
 
 Mills. 
 
 
 
 
 
 
 •1 
 
 40 
 
 3 7 
 
 11 
 
 17-8 
 
 24-0 
 
 29-5 
 
 •2 
 
 80 
 
 91 
 
 27-0 
 
 43-8 
 
 59 
 
 72:5 
 
 •3 
 
 120 
 
 150 
 
 44-4 
 
 72-1 
 
 97-3 
 
 119 
 
 •4r 
 
 160 
 
 21-2 
 
 62-5 
 
 102 
 
 137 
 
 168 
 
 •5 
 
 200 
 
 27-4 
 
 81 
 
 131 
 
 177 
 
 218 
 
 •6 
 
 240 
 
 33 7 
 
 100 
 
 164 
 
 219 
 
 268 
 
 ■7 
 
 280 
 
 40-1 
 
 119 
 
 192 
 
 259 
 
 319 
 
 •8 
 
 310 
 
 46 4 
 
 137 
 
 223 
 
 301 
 
 369 
 
 •9 
 
 350 
 
 529 
 
 157 
 
 253 
 
 342 
 
 420 
 
 1-0 
 
 390 
 
 59 3 
 
 175 
 
 285 
 
 384 
 
 472 
 
 2-0 
 
 780 
 
 124 
 
 367 
 
 595 
 
 803 
 
 988 
 
 3 
 
 1180 
 
 189 
 
 559 
 
 908 
 
 1225 
 
 1503 
 
 40 
 
 1570 
 
 254 
 
 753 
 
 1221 
 
 1646 
 
 2021 
 
 5 
 
 1970 
 
 319 
 
 945 
 
 1534 
 
 2068 
 
 2523 
 
 6 
 
 2360 
 
 385 
 
 1138 
 
 1846 
 
 2491 
 
 3058 
 
 7-0 
 
 2760 
 
 450 
 
 1330 
 
 2158 
 
 2846 
 
 3575 
 
 8-0 
 
 3150 
 
 514 
 
 1525 
 
 2472 
 
 3335 
 
 4094 
 
 9-0 
 
 3540 
 
 580 
 
 1716 
 
 2785 
 
 3755 
 
 4611 
 
 10-0 
 
 3940 
 
 645 
 
 1909 
 
 3097 
 
 4178 
 
 5130 
 
 Computed from the formula, 
 
 °=v( 
 
 •48 K 
 
 ,t x 
 
 3D 2 I 
 
 10 + 3D 2 log. o gH 
 
 K = thermal conductivity of insulator, = -0-9048 for 
 
 guttapercha ; 
 E = coefficient of cooling, = '0003. 
 
 The difficulties surrounding the case of buried con- 
 ductors are too great to give any practical guidance, 
 but the prevailing practice will be fully discussed in the 
 section dealing with distribution. 
 
 A most important practical question, however, is 
 that relating to coils, because as a rule we require to 
 lose as little energy in heating the coils as we possibly 
 can, and especially so in the armature coils of dynamos. 
 Prof. Forbes says : "It becomes an easy thing in 
 any case to calculate the heating of a coil when we 
 know its cooling surface and its resistance. 
 
 Let p = the resistance of a coil in ohms at the 
 permissible temperature. 
 S = the surface exposed to the air measured 
 in centimetres (1 square cm. = 155 
 square inch. 1 square inch = 615 
 square cm). 
 
 Let t = the rise in temperature, centigrade scale. 
 C = the current in amperes. 
 •24 C 2 P = heat generated = E t S, 
 where E is McFarlane's constant varying from -0002 
 to "0003. The latter value may be taken. If 50 deg. C. 
 be the permissible rise in temperature 
 
 °=V 
 
 0003 x 50 x S 
 
 = -25 
 
 \ 
 
 r\ 
 
 ■24 x p ■>■ p 
 
 N.B. — It must be remembered in practice that the 
 resistance (cold) must be increased by A of its value to 
 give P . 
 
 Example. — The resistance of the field-magnets of a 
 dynamo is 1'5 ohms cold, and the surface exposed to 
 the air is 1 metre ; find the current to heat it not more 
 than 50 deg. C. Here S = 10,000, P = P8 ohms, and 
 
 C = •a.sy iO.OOO = 33-5 amperes. 
 
 Those who are accustomed to handling dynamos will 
 know that this is very much what we actually find, 
 and it gives us confidence in the applications of theory." 
 
 According to the most recent practice, Mr. Esson 
 gives the ultimate rise in temperature of coils wound 
 with double cotton-covered wires, varnished on the 
 exterior, to be approximately in degrees centigrade for 
 square centimetres of surface. When W = energy in 
 watts dissipated, and S = cooling surface in square 
 centimetres, 
 
 W 355 
 
 C° = 
 
 S 
 
 If square inches be taken as the unit, then the formula 
 becomes 
 
 W55 
 
 C° = 
 
 S 
 
 when W = energy in waits dissipated, and S = cooling 
 surface in square inches. 
 9, 
 
 -C°, these formulse become on theFahren- 
 
 Asl° F 
 heit scale : 
 
 F° = — for square centimetres of surface ; 
 
 o 
 
 A ■, ™ W 30-5 , . , 
 
 And F = — - — for square inches. 
 
 b 
 
 These formulae being admittedly approximate, we 
 may use the constant 200 instead of 197 without 
 practically affecting the result, and write 
 
 W200 
 
 F° = 
 
 - — for square centimetres. 
 
 The coefficients, says Mr. Esson, have been obtained 
 from practice, but they are liable to some uncertainty. 
 In the cases from which they have been derived the 
 coils were wound on formers of sheet iron, which fitted 
 close to the magnets. These formers were fitted with 
 brass end flanges, and in the spaces between the flanges 
 and the wire, also between the body and the wire, was 
 a thick layer of insulating material. The surface of 
 the end flanges is not included as making up surface S, 
 nor is the inside surface of the former next the maenet. 
 
16 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Though insulated so well thermally, there is certainly 
 a flow of heat towards the interior of the magnets and 
 towards the end flanges, so that, as a matter of fact, 
 the virtual cooling surface is greater than given above. 
 Its value it is, however, difficult to determine, while 
 the above equation is approximately true for machines 
 of medium size, and may be used for all cases in which 
 the coils do not exceed 2£in. or 3in. in thickness. 
 
 For armatures — which in rotating induce a current 
 of air — the equation is different, and the coefficient 
 given above may have a much lower value, according 
 to the velocity and ventilation. For armatures the 
 surface velocity of which is 50ft. per second (3,000ft. 
 per minute), and the exterior diameter of the core 1£ 
 times the interior, the length being about equal to the 
 diameter, the rise is, approximately, 
 
 C = - (for sq. centimetres), 
 
 or 
 
 C° 
 
 s 
 
 W35 
 
 S 
 
 (for sq. inches), 
 
 accordingly as the surface is taken in square centi- 
 ■ metres or square inches. Here the whole of the cooling 
 surface inside, outside, and at the ends is counted, W 
 being the total watts dissipated in both electrical waste 
 and hysteresis. The latter must, under no circum- 
 stances, be neglected, as it may reach 20 per cent, or 
 more of the total. In giving the above equation, it 
 should be stated that the armatures of the machines 
 which furnished the results are closed at the 
 commutator end, except at the periphery, where a 
 draught of air through the centre from the pulley end 
 is expelled. For slow-speed machines the coefficient is 
 necessarily higher, and a larger surface must be 
 allowed, or special means adopted, in the shape of 
 vanes or blades, to force a draught through the arma- 
 ture. Experience can be the only guide here. . . 
 
 In any case, the ultimate temperature of the machine 
 ought not to rise above a certain value, whatever the 
 temperature of the room in which it works. In Mr. 
 Esson's opinion, an ultimate temperature of from 
 70 deg. to 75 deg. C. may be permitted with perfect 
 safety, but this should not be exceeded. 
 
 Mr. C. Hering, when discussing coils on field- 
 magnets, gives formulae for them. These formulas are 
 very diffeient to those of Mr. Esson, whose formulae 
 accord with English practice, while Mr. Hering has 
 derived his formulae from American practice. Mr. 
 Hering gives under his conditions a watt of energy as 
 dissipated for every 223 square inches of surface, when 
 the difference between the temperature of the coil and 
 the surrounding air is 1 deg. F. 
 
 Putting this into a formula, we get 
 
 W = C E = JL T S = 0-004476 T S, 
 
 degrees Fahren- 
 
 where W = watts lost in coil, T 
 heit, and S = square inches. 
 
 From the formula W = C E = C 2 E, — = -i-TS 
 
 2 E 223 
 
 when C E E have their ordinary significations, E, 
 
 however, being the resistance at the temperature to 
 which the coil may be raised, we get some very practical 
 information. Suppose we take 
 
 CE= L s 
 223 
 
 C 
 
 T.S 
 223 E 
 
 this gives the greatest current which can be used in the 
 magnet coils of a shunt machine having a certain 
 pressure, in order that they do not heat above a certain 
 temperature. This current can be calculated before 
 the winding is determined, as it is independent of the 
 number of turns or the resistance, if only the size of 
 the external surface is approximately known. This 
 maximum current should never be exceeded. 
 
 Again, the greatest current for any allowable 
 temperature above that of the surrounding air can 
 easily be divided— say, for example, 50 deg. F. 
 
 C 
 
 50 S -= -224 §-• 
 
 223 E ' E 
 
 its equivalent C E, we get C = . / 
 
 If here we substitute for E 
 
 •224 £ 
 
 E 
 
 If 80 deg. F. is the maximum difference of tempera- 
 ture, 
 
 ° = ®4-= ■ 36 I = ' 60 VI 
 
 The formula can be used for series machines when 
 C is known, for writing 
 
 we get 
 
 E 
 
 TS 
 223 C 2 
 
 With a permissible rise of 50 deg. F., or of 80 deg. F., 
 we have respectively 
 
 E = 
 
 •224 S 
 
 and E = "36 
 
 S 
 
 C 2 ~" C 2 
 
 The surface area of the coil in square inches may be 
 found from 
 
 223 W 223 CE 223 C 2 E 
 
 S = 
 
 T 
 
 T 
 
 For a rise of temperature of 50 deg. F. or 80 de<*. F. 
 respectively, the surface will be 
 
 S = 
 
 223W 
 ~50~ 
 
 = 4-46 W; and S = 
 
 223 W 
 80 
 
 = 2-8 W. 
 
 As the number of watts to be allotted to the magnets 
 is approximately known at the outset, this formula 
 enables one to determine about what the least surface 
 of the coils should be, and, in fact, knowing the cross 
 section of the core, to arrive at an approximate idea of 
 the least length of core. 
 
 When we desire to employ the heating properties of 
 the current for the production of light, it is necessary 
 to employ some conductor having high resistance that 
 does not fuse at the temperature to which it has to be 
 raised. None of the metals are suitable, and hitherto 
 all attempts to find another conductor for this purpose 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 ELEVATION 
 
 ceompton's electkio teavellins okane. 
 
 PLAN 
 
 oeomfton's eleotkic travelling crane. 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 AMERICAN FITTINGS. 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 17 
 
 than carbon have failed. Used as ordinary conductors, 
 however, the metals are seldom raised to a temperature 
 near their melting point. We shall see, however, that 
 metals and alloys are used for protective purposes — that 
 is, they are put in the circuit that should there be by 
 any accident an excess of current, they will melt and 
 break the circuit. The following table of fusibility 
 gives an idea of the utility of the metals for this 
 purpose : 
 
 Table of Fusibility. 
 
 Tin melts at 442 deg. F. 
 
 Lead „ „ 617 „ 
 
 Zinc „ „ 773 „ 
 
 Silver „ „ 1,800 „ 
 
 Copper „ 1,990 „ 
 
 Gold „ ,,2,000 „ 
 
 Cast Iron „ 2,800 „ 
 
 Steel „ ,,4,000 „ (about) 
 
 In papers to the Eoyal Society, Mr. W. H. Preece, 
 F.B.S., described a long series of experiments made to 
 determine the heating effects of currents upon wires, 
 from which he deduced the formula for the fusion 
 current. 
 
 C = a d? 
 and in his third paper he calculated two tables from 
 the final value of the constant a, as determined by the 
 experiments. The metals experimented with were 
 copper, aluminium, platinum, German silver, platinoid, 
 iron, tin, alloy (2 of lead to 1 of tin), lead. The value 
 of a for these different metals is given as follows : 
 
 Inches. Centimetres. Millimetres. 
 
 Copper 10,244 2,530 80-0 
 
 Aluminium 7,585 1,873 59"2 
 
 Platinum 5,172 1,277 40"4 
 
 German silver 5,230 1,292 40"8 
 
 Platinoid 4,750 1,173 37'1 
 
 Iron 3,148 7774 24"6 
 
 T m 1,642 4055 128 
 
 Alloy (lead & tin 2 to 1) 1,318 325"5 10"3 
 
 Lead 1.379 340-6 10-8 
 
 We give a part of the further tables, the part relating 
 more directly to the metals used extensively in electrical 
 engineering. (See Tables I. and II. on next page.) 
 
 The Chemical Action of the Current.— The chemical 
 action of a current is best seen in the art of electro- 
 deposition, which to be thoroughly understood requires 
 a considerable knowledge of chemistry. In fact, the art 
 depending upon this chemical action is rapidly being 
 divided into three distinct branches— which may be 
 termed electro-chemical, electro-metallurgical, and 
 electro-depositing. 
 
 The chemist tells us that all substances m nature 
 can be broadly divided into two classes : 
 
 1. Elementary substances, which the art of the 
 
 chemist is not yet able to divide into anything 
 dissimilar to themselves. 
 
 2. Compound substances, which can be divided up 
 
 into their component elements, 
 
 Chemical action is either the combination of ele- 
 mentary substances, or elements with compounds, or 
 compounds with compounds, or the reverse action of 
 the separation of compounds or elements from com- 
 pounds. Chemical action may be altogether indepen- 
 dent of electricity or it maybe assisted thereby. Thus, 
 if the clean blade of a knife be dipped into a solution of 
 sulphate of copper, it will become coated with the 
 copper. This we should term simply a chemical action, 
 although copper is deposited upon the steel. If we 
 take an ordinary Darnell's battery, we find that the 
 copper from the sulphate of copper is deposited upon 
 the copper plate, and this only when a current is 
 circulating through the circuit. This deposition, then, 
 is due to electrical action. The full consideration of 
 this subject does not come within the scope of this 
 treatise, but we may state the more important principles 
 connected with electrolysis. It was Faraday who gave 
 us the nomenclature relating to electrolysis. He called 
 the compound to be decomposed the Electrolyte, and 
 the process Electrolysis. The plates or poles of the 
 battery he called Electrodes. The plate where the 
 greatest pressure exists he called the Anode, and the 
 other pole the Cathode. The products of decomposi- 
 tion he called Ions. 
 
 Elementary substances, inasmuch as they cannot be 
 decomposed, are not electrolytes. As a rule electrolytes 
 are electrolysed only in the fluid state, because only in 
 that state are they conductors. The components of an 
 electrolyte are resolved during electrolysis into two 
 groups which, so to speak, travel through the electrolyte 
 in opposite directions, one going towards one electrode 
 and one going towards the other. Those bodies only 
 are electrolytes that are composed of a conductor and a 
 non-conductor. It is to be carefully noticed that : 
 
 1. The amount of chemical action is the same at 
 
 whatever part of the circuit it occurs. 
 
 2. The amount of an element liberated at an electrode 
 
 during a given time is proportional to the 
 current. 
 
 This latter law permits another way to obtain " unit 
 current." Very careful experiments have been made 
 by Lord Eayleigh, and he finds that a current of one 
 ampere will deposit 0-017253 grain, or 0-00111815 
 gramme, of silver per second on one of the plates of a 
 silver voltameter, the liquid employed being a solution 
 of silver nitrate containing from 15 to 20 per cent, of 
 the salt. 
 
 The weight of hydrogen similarly set free is "00001035 
 gramme — that is, one ampere per second sets free 
 •00001035 gramme of hydrogen from water in that time. 
 
 Knowing the amount of hydrogen thus set free, and 
 certain other chemical facts, we can calculate exactly 
 what weight of other substances will be set free or 
 deposited in a given time by a given current. 
 
 Thus the current that liberates 1 gramme of 
 hydrogen will liberate 8 grammes of oxygen, or 108 
 grammes of silver, the numbers 8 and 108 being the 
 chemical equivalents for oxygen and silver respectively. 
 
18 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 FUSE TABLES, BY W. H. PEEEOE, F.K.S. 
 
 I. 
 Giving the Diameters of various Wires which will be Fused by a given Current, from the formula 
 
 d 
 
 CM 
 
 a = 1642 for Tin = 1379 for Lead = 10244 for Copper = 3148 for Iron. 
 
 
 Tin Wire. 
 
 Lead Wire, 
 
 Copper Wire. 
 
 Iron Wire. 
 
 Current 
 
 
 
 
 
 
 in 
 
 Amperes. 
 
 Diameter 
 
 Approx. 
 
 Diameter 
 
 Approx. 
 
 Diameter 
 
 Approx. 
 
 Diameter 
 
 Approx. 
 
 
 Inches. 
 
 S.W.G. 
 
 Inches. 
 
 S.W.G. 
 
 Inches. 
 
 S.W.G. 
 
 Inches. 
 
 S.W.G. 
 
 1 
 
 0-0072 
 
 36 
 
 0-0081 
 
 35 
 
 0-0021 
 
 47 
 
 0-0047 
 
 40 
 
 2 
 
 0-0113 
 
 31 
 
 00128 
 
 30 
 
 0-0034 
 
 43 
 
 0-0074 
 
 36 
 
 3 
 
 0-0149 
 
 28 
 
 0-0168 
 
 27 
 
 0-0044 
 
 41 
 
 0-0097 
 
 33 
 
 4 
 
 0-0181 
 
 26 
 
 0-0203 
 
 25 
 
 0-0053 
 
 39 
 
 0-0117 
 
 31 
 
 5 
 
 0-0210 
 
 25 
 
 0-0236 
 
 23 
 
 0-0062 
 
 38 
 
 0-0136 
 
 29 
 
 10 
 
 0-0334 
 
 21 
 
 0-0375 
 
 20 
 
 0-0098 
 
 33 
 
 0-0216 
 
 24 
 
 15 
 
 0-0437 
 
 19 
 
 0-0491 
 
 18 
 
 0-0129 
 
 30 
 
 0-0283 
 
 22 
 
 20 
 
 0-0529 
 
 17 
 
 0-0595 
 
 17 
 
 0-0156 
 
 28 
 
 0-0343 
 
 20-5 
 
 25 
 
 0-0614 
 
 16 
 
 0-0690 
 
 15 
 
 0-0181 
 
 26 
 
 0398 
 
 19 
 
 30 
 
 00694 
 
 15 
 
 00779 
 
 14 
 
 0-0205 
 
 25 
 
 0-0450 
 
 18-5 
 
 35 
 
 0-0769 
 
 14-5 
 
 0-0864 
 
 13-5 
 
 0-0227 
 
 24 
 
 0-0498 
 
 18 
 
 40 
 
 0-0840 
 
 13-5 
 
 0-0944 
 
 13 
 
 0-0248 
 
 23 
 
 0-0545 
 
 17 
 
 45 
 
 0-0909 
 
 13 
 
 0-1021 
 
 12 
 
 0-0268 
 
 22 
 
 0-0589 
 
 16-5 
 
 50 
 
 0-0975 
 
 12-5 
 
 0-1095 
 
 11-5 
 
 0-0288 
 
 22 
 
 0-0632 
 
 16 
 
 60 
 
 0-1101 
 
 11 
 
 0-1237 
 
 10 
 
 0-0325 
 
 21 
 
 0-0714 
 
 15 
 
 70 
 
 0-1220 
 
 10 
 
 0-1371 
 
 9-5 
 
 0-0360 
 
 20 
 
 0-0791 
 
 14 
 
 80 
 
 01334 
 
 9-5 
 
 0-1499 
 
 8-5 
 
 0-0394 
 
 19 
 
 0-0864 
 
 13-5 
 
 90 
 
 0-1443 
 
 9 
 
 0-1621 
 
 8 
 
 0-0426 
 
 18-5 
 
 0-0935 
 
 13 
 
 100 
 
 0-1548 
 
 8-5 
 
 0-1739 
 
 7 
 
 0-0457 
 
 18 
 
 0-1003 
 
 12 
 
 120 
 
 0-1748 
 
 7 
 
 01964 
 
 6 
 
 0-0516 
 
 17-5 
 
 0-1133 
 
 11 
 
 140 
 
 0-1937 
 
 6 
 
 0-2176 
 
 5 
 
 0-0572 
 
 17 
 
 0-1255 
 
 10 
 
 160 
 
 0-2118 
 
 5 
 
 0-2379 
 
 4 
 
 0-0625 
 
 16 
 
 0-1372 
 
 9-5 
 
 180 
 
 0-2291 
 
 4 
 
 0-2573 
 
 3 
 
 0-0676 
 
 16 
 
 0-1484 
 
 9 
 
 200 
 
 0-2457 
 
 3-5 
 
 0-2760 
 
 2 
 
 0725 
 
 15 
 
 0-1592 
 
 8 
 
 250 
 
 0-2851 
 
 1-5 
 
 0-3203 
 
 
 
 0-0841 
 
 13-5 
 
 0.1848 
 
 6-5 
 
 300 
 
 0-3220 
 
 
 
 0-3617 
 
 00-5 
 
 0-0950 
 
 12-5 
 
 0-2086 
 
 5 
 
 II. 
 
 Giving Current in . 
 
 imperes required to Fuse Wires according to the formula C = 
 
 = ad* 
 
 No. 
 
 Diameter. 
 
 s 
 
 Tin. 
 
 Lead. 
 
 Copper. 
 
 Iron. 
 
 S.W.G. 
 
 Inches. 
 
 a = 1642. 
 
 a = 1379. 
 
 a = 10244. 
 
 a = 3148. 
 
 14 
 
 0-080 
 
 0-022627 
 
 37-15 
 
 31-20 
 
 231-8 
 
 71-22 
 
 16 
 
 0-064 
 
 0-016191 
 
 26-58 
 
 22-32 
 
 165-8 
 
 50-96 
 
 18 
 
 0-048 
 
 0-010516 
 
 17-27 
 
 14-50 
 
 107-7 
 
 33-10 
 
 20 
 
 0-036 
 
 0-006831 
 
 11-22 
 
 9-419 
 
 69-97 
 
 21-50 
 
 22 
 
 0-028 
 
 0-004685 
 
 7-692 
 
 6-461 
 
 48-00 
 
 14-75 
 
 24 
 
 0-022 
 
 0-003263 
 
 5-357 
 
 4-499 
 
 33-43 
 
 10-27 
 
 26 
 
 0-018 
 
 0-002415 
 
 3-965 
 
 3-330 
 
 24-74 
 
 7-602 
 
 28 
 
 0-0148 
 
 0-001801 
 
 2-956 
 
 2-483 
 
 18-44 
 
 5-667 
 
 30 
 
 0-0124 
 
 0-001381 
 
 2-267 
 
 1-904 
 
 14-15 
 
 4-347 
 
 32 
 
 0-0108 
 
 0-001122 
 
 1-843 
 
 1-548 
 
 11-50 
 
 3-533 
 
 To find the weight of silver deposited by a given 
 current in a given time, we find the weight of hydrogen 
 liberated by the given current in the given time, and 
 multiply by the chemical equivalent, 
 
 Examples. 
 Find weight of silver deposited in 10 seconds by a 
 current of 10 amperes : 
 
 Weight of silver = weight of hydrogen liberated per 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 19 
 
 Table III. 
 
 Elements. 
 
 Electro-Positive. 
 
 Hydrogen 
 
 Potassium 
 
 Sodium 
 
 Aluminum 
 
 Magnesium 
 
 Gold 
 
 Silver 
 
 Copper (Cupric) 
 
 „ (Cuprous) 
 
 Mercury (Mercuric) .. 
 
 „ (Mercurous). 
 Tin (Stannic) 
 
 „ (Stannous) 
 
 Iron (Ferric) 
 
 „ (Ferrous) 
 
 Nickel 
 
 Zinc 
 
 Lead 
 
 Electro-Negative. 
 
 Oxygen 
 
 Chlorine 
 
 Iodine , 
 
 Bromine , 
 
 Nitrogen 
 
 Valency. * 
 
 H 1 
 K 1 
 
 Na 1 
 
 AF 
 
 Ms; 3 
 
 Au 3 
 
 Ag 1 
 
 Cu 2 
 
 Cu 1 
 
 Hg 1 
 
 Hg 1 
 
 Su* 
 
 Su= 
 
 Fe* 
 
 Fe 2 
 
 Ni 2 
 
 Zu 2 
 
 Pb 2 
 
 O 2 
 
 CI 1 
 
 I 1 
 
 Br 1 
 
 X 3 
 
 Atomic 
 Weighty 
 
 Chemical 
 Equivalent 
 
 1- 
 
 3904 
 
 2299 
 
 273 
 
 2394: 
 
 196-2 
 
 10766 
 
 63- 
 
 63- 
 
 199-8 
 
 199-8 
 
 117S 
 
 117-8 
 
 55-9 
 
 55-9 
 
 58-6 
 
 64-9 
 
 206-4 
 
 15-96 
 35-37 
 126-53 
 79-75 
 1401 
 
 Electro - Chemical 
 
 Equivalent 
 
 (Milligrammes 
 
 per Coulomb). 
 
 1- 
 
 3904 
 
 22-99 
 91 
 
 11-97 
 
 65-4 
 107 66 
 
 31-5 
 
 63- 
 
 99-9 
 199-8 
 
 29-45 
 
 58-9 
 
 18-64+ 
 
 27-95 
 
 29-3 
 
 3245 
 103-2 
 
 7-9 5 
 
 35-37 
 
 126-53 
 
 79-75 
 
 4-67 
 
 •010384 
 
 •40539 
 
 •23S73 
 
 ■09449 
 
 •12430 
 
 ■67911 
 
 1-11800 
 ■32709 
 ■65419 
 
 1-03740 
 
 207470 
 •30581 
 •61162 
 •19356 
 ■29035 
 ■30425 
 •33696 
 
 1-07160 
 
 •08286 
 •36728 
 1-31390 
 ■82812 
 •04849 
 
 Coulombs 
 
 per 
 Gramme. 
 
 96293-00 
 2467-50 
 4188-90 
 1058-30 
 
 80403 
 1473-50 
 
 894-41 
 3058-60 
 1525-30 
 
 963-99 
 
 481-99 
 3270-00 
 163500 
 5166-4 
 3445-50 
 3286-80 
 2967-10 
 
 933-26 
 
 Grammes 
 
 per 
 
 Ampere 
 
 Hour. 
 
 to 
 
 003738 
 1-45950 
 0-85942 
 3-40180 
 447470 
 2-44480 
 4-02500 
 1-17700 
 2-35500 
 3-73450 
 7-46900 
 1-10090 
 2-20180 
 ■69681 
 1-04480 
 1-09530 
 1-21330 
 3-85780 
 
 Loor. 
 
 4-9836353 
 3-3922572 
 3-6221000 
 3 0245939 
 2-9055411 
 31683501 
 2-9515366 
 3-4855227 
 3-1833553 
 29840725 
 2-6830380 
 3-5145478 
 3-2135178 
 3-7131894 
 3-5372523 
 3-5167733 
 3-4723322 
 2-9700027 
 
 Log. 
 
 ^■5727090 
 0-1642041 
 1 -9342055 
 0-5317086 
 0-6507614 
 0-3882433 
 0-6047659 
 0-0707765 
 3719909 
 05722325 
 0-8732625 
 0-0417479 
 0-3427779 
 T -8431131 
 0-0190332 
 00395331 
 00839682 
 0-5863397 
 
 * Valency is the atom fixing or atom replacing power of an element compared with hydrogen, whose valency is unity. 
 
 + Atomic weight is the weight of one atom of each element compared with hydrogen, whose atomic weight is unity. 
 
 t Beequerel's extension of Faraday's law showed that the electro-chemical equivalent of an element is proportional to its chemical 
 equivalent. The latter is equal to its combining weight, and not to atomic weight 4- valency, as defined by Thompson, Hospitalier, and 
 others who have copied their tables. For example, the ferric salt is an exception to Thompson's rule, as are sesqui salts in general. 
 
 second x no seconds x current strength x 108 = 
 ■00001035 x 10 x 10 x 108 = 11178 gramme. 
 
 Find weight of copper deposited in 1 hour by a 
 current of 10 amperes : 
 
 •00001035 x 60 x 60 x 10 x 32 = 11"923 grammes. 
 
 We have seen that 1 ampere per second liberates 
 "00001035 gramme of hydrogen, therefore strength of 
 current in amperes = 
 
 weight in grammes of H. liberated per sec. 
 -00001035 
 
 Or strength of C in amperes = 
 
 weight of silver liberated per sec. 
 -00001035 x 108 
 
 n p weight of element liberated per sec. 
 
 "00001035 x chemical equivalent of element' 
 
 The accompanying Table DX is calculated upon 
 Lord Eayleigh's determination of the electro-chemical 
 equivalent and Eoscoe's atomic weight. 
 
 Voltameter. — The voltameter represented in Pig. 17 
 consists of a vessel, A, containing acidulated water, 
 two platinum plates with terminals to which the 
 battery can be connected, together with tubes and a 
 measuring apparatus to contain the gas when liberated. 
 The measuring tube is filled with water, then inverted 
 over water in the vessel B, When the battery is 
 
 joined up to the terminals the water is decomposed. 
 The gases thus obtained pass through the tube and 
 into the tube C, forcing down the water The volume 
 
 Fiq. 17. 
 
 of mixed gases, hydrogen and oxygen, produced pei 
 
 second by a current of one ampere, is in cubic inches = 
 
 •03 x 76(273 + 0°) , _. , A . a 
 
 - where O = degrees centigrade and 
 
 Ax 273 
 h= pressure in inches of mercury 
 
20 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 If the Fahrenheit scale is used the formula becomes 
 
 • 01058x30{491 + (F°-32)} 
 h x 491 
 The phenomena connected with the interior of the 
 conductor, thus briefly described, are exceedingly- 
 important, though not more so than the phenomena 
 connected with the exterior of the conductor. 
 
 Phenomena connected with the Exterior of a Conduc- 
 tor carrying a Current. — We have seen that the current 
 in a wire exercises a directive influence upon iron 
 filings placed around the wire. This directive influence 
 is found at whatever point in the circuit the experiment 
 is tried, and we may assume that a force emanating 
 from the conductor and due to the current in the con- 
 ductor acts upon these small particles of iron in a 
 certain and in a definite direction. If the lines in Pig. 
 18 represent the concentric circles into which the 
 filings arrange themselves under the action of the cur- 
 rent, a number of circles one above the other, Fig. 19, 
 
 I'm. 18. 
 
 Fio. 19. 
 
 would indicate how filings would arrange themselves all 
 along the wire. The influence of the current can also 
 be examined by bringing a small magnet near the wire, 
 or a part of the conductor belonging to another con- 
 ductive circuit. 
 
 It is found that the influence of the current gets less 
 and less the further we get away from the conductor. 
 The whole space through which the influence is felt is 
 termed the field of the conductor. Many term it the 
 magnetic field of the conductor. For most practical 
 purposes we may assume that the force or forces 
 acting in the interior of the conductor act in directions 
 parallel to the axis of the conductor, while those in the 
 field act at right angles to the axis of the conductor. 
 It probably will be found that this assumption is not 
 quite accurate, there being something analogous to a 
 spiral in the direction of the action of the forces. 
 
 Faraday, to whose discoveries we owe the present 
 position of electrical science, suggested an excellent 
 
 and convenient method for discussing questions relating 
 to the field, in supposing the effects on filings and 
 magnets due to " lines of force " coming from the con- 
 ductor and acting upon the filings. It must not be 
 understood that such lines of force actually exist, but 
 it is very convenient to imagine lines in the direction 
 of which a force or forces act, and " number of lines " 
 or "length of lines " to indicate the magnitude of the 
 force or forces. In this case we use " number of lines " 
 to indicate magnitude of forces acting. It is seen that 
 the iron filings arrange themselves in concentric circles 
 —that is, we can assume that the forces under which 
 they so arrange themselves may be represented by 
 " closed curves," or instead of " lines " we may use the 
 term " loops " of force. The forces all seem to cause 
 polarity, or the arrangement of the ends of the particles 
 acted upon in definite directions. 
 
 It will be well to recapitulate the conventional as- 
 sumptions made with regard to the loops of force con- 
 nected with the conductive circuit for the purposes of 
 this book. 
 
 1. That the lines or loops of force in the conductor 
 
 are parallel to the axis of the conductor. 
 
 2. That the loops of force external to the conductor 
 
 are proportional in number to the current in the 
 conductor — that is, a definite current generates 
 a definite number of loops of force. These may 
 be stated as the strength of field is proportional 
 to the current. 
 
 3. That the radii of the loops of force are at right 
 
 angles to the axis of the conductor, as shown in 
 Fig. 20. 
 
 MX 
 
 X 
 
 > '/ L ----_-* u _ ;_- 1- L 
 
 Fig. 20. 
 
 The interaction between fields and the influence of 
 fields upon conductors is a most important branch of 
 study. The phenomena connected with dynamos, 
 motors, transformers, etc., are connected with this 
 branch of the subject, which, however, cannot be 
 thoroughly explained till after a study of the magnetic 
 circuit. For the present, then, the interactions of the 
 field will be confined to two cases : 
 
 1. The interaction between a conductor carrying a 
 
 current— that is, having a field — and one with 
 no current. 
 
 2. The interaction between two conductors carrying 
 
 currents. 
 
 It is found that if a conductor is brought into the 
 field of another conductor a current is set up in the 
 conductor so brought into the field, the direction of 
 the current in the conductor brought into the field 
 being opposite to that of the conductor causing the field. 
 
Practical electrical engineering. 
 
 2i 
 
 Thus, if A B, Fig. 21, represent a conductor surrounded 
 by its field, the arrows showing the direction of the 
 current, and another conductor, C D, is brought into 
 the field of AB, a current is set up in CD in the 
 direction of the arrows and opposite to that in AB. 
 
 -^= — v- 
 
 -f-^r 
 
 * ' 
 
 When we come to consider the interaction between 
 two fields due to two conductors carrying currents, we 
 have two cases : first, when the currents are in the same 
 direction, and, secondly, when the currents are in oppo- 
 
 cJ- 
 
 4f 
 
 L_ 
 
 Y\ 
 
 /" ^ 
 
 
 W 
 
 v\ 
 
 T~B 
 
 Fig. 21. 
 
 Fig. 23. 
 
 This current so set up is but momentary. If, while 
 C D is in the position indicated, the current in A B is 
 doubled, there will be another momentary current set 
 up in C D, equal and in the same direction as in the 
 former case. This double current in AB may be 
 represented by double the number of loops, as in Fig. 
 22. Or if the conductor CD is brought nearer to 
 
 c, \ i 
 
 -D, 
 
 Fio. 22. 
 
 AB, so that a greater number of the concentric loops 
 go round it, as in the dotted line, C x D v a further 
 momentary current is set up, still in the same direction 
 as the former ones. The action may now be reversed, 
 and as the conductor C 1 D 1 is taken to the second 
 position C D, there will be a momentary current in the 
 reverse direction to the original current — that is, in the 
 same direction as that in A B. On making the current 
 in AB what it was originally, there is again another 
 momentary current in C D in the same direction as in 
 A B ; and finally, on taking C D out of the field, there 
 is again a momentary current in the same direction as 
 that in A B. 
 
 To summarise these actions. On bringing a con- 
 ductor into the field a momentary current is set up 
 opposite in direction to that causing the field, and this 
 momentary current is continued again and again by 
 "looping" more lines of force round the conductor 
 brought into the field. Thus we find that the current 
 set up is proportional to the number of "loops" passed 
 round the conductor. This is a reverse action. 
 The current causes the " field," the field causes a 
 " current." The field is proportional to the current, 
 and the current is proportional to the field. A current 
 caused by the action of a field of force is said to be an 
 induced current, and the current in C D is said to be 
 induced by the current in AB. The latter might be 
 called the inductor current, then we should have the 
 induced current on entering or advancing in the field 
 is opposite in direction to the inductor current, but 
 in the same direction on retiring or leaving the field. 
 
 site directions. Let A B and C D, Fig. 23, be two parallel 
 conductors with cm-rents in the same direction. Sup- 
 pose the currents in each case to be equal, then the 
 forces causing the fields will be equal. Take any pair 
 of loops, which interact, at a point where these loops 
 first come into contact and under the influence of equal 
 and opposite directive forces, and the result is the 
 forces neutralise each other at this point, the effect 
 being as if the inner parts of the loops of force were 
 neutralised, the remaining parts forming one loop 
 around both wires, tending to shorten its diameter, 
 and bringing the conductors AB, CD nearer together. 
 Hence two parallel conductors seem to attract each 
 other. This attraction is due to the interaction of the 
 fields. 
 
 The interaction of two fields due to currents in oppo- 
 site directions has a reverse effect, and the conductors 
 seem to be repelled. Let A B, C D, Fig. 24, represent 
 two conductors with currents in opposite directions. 
 
 ^ 
 
 N \\ 
 
 / 
 
 i 
 
 B 
 
 rr 
 
 r? 
 
 Fig. 24. 
 
 The forces acting in the fields will be in opposite direc- 
 tions, but a little examination of the figure will show 
 that if the fields are brought together the forces at the 
 point of contact are in the same direction, and there is 
 therefore no neutralisation of the field. By extending 
 the ideas involved by lines or loops of forces somewhat, 
 we think the whole of these and other interactive phe- 
 nomena can be made clear. The result of the force is 
 to form a closed curve of polarised particles. If the 
 field is filled with iron filings we should get a series of 
 concentric rings of iron filings around each conductor. 
 No two rings of filings can occupy the same space. All 
 are agreed upon this. Instead of filings, suppose the 
 matter acted upon to be particles of ether or particles 
 of electricity. Surely the same result will be admitted 
 that no two rings can occupy the same space ; carrying 
 
22 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 the idea a little further, we may assume that no two 
 lines or loops of force can occupy the same space. With 
 this assumption many difficulties are cleared away. 
 
 In the preceding case the loops of force partly 
 neutralised and partly merged into each other. In 
 the present case there can be no neutralisation and no 
 merging. The effect, therefore, of bringing the two 
 fields of A B and the field of CD into contact is very 
 similar to the bringing of two faggots into contact. A 
 weak field brought into contact with a weak field will 
 experience a weak resistance ; a strong field brought 
 into contact with a strong field will experience a strong 
 resistance. This resistance, or, if it is preferred, this 
 repulsive force, is simply proportional to the number 
 of concentric circles of polarised particles brought into 
 contact, and to nothing else. The number of loops of 
 force is proportional in each case to the currents, 
 hence this resistance or repulsive effect is also pro- 
 portional to the currents. 
 
 It is to Ampere we owe the discovery of the inter- 
 actions between these fields, and the laws discovered 
 by him may be thus stated : 
 
 1. Two parallel conductors attract one another if 
 the currents in them are flowing in the same direction, 
 and repel one another if the currents flow in oppo- 
 site directions. 
 
 The conductors may be portions of the same or of 
 different circuits. 
 
 2. The parts of conductors crossing one another 
 obliquely attract one another if both the currents are 
 either towards or from the point of crossing, and repel 
 one another if one current is towards and one from that 
 point. Thus currents in conductors as in Fig. 25 
 attract or repel according as shown. 
 
 Attract. 
 
 Attract. 
 
 Attract. 
 
 Fig. 25. 
 
 Repel. 
 
 AB, Fig. 26, represent a straight portion of a con- 
 ductor. The number of loops of force are equal along 
 
 M 
 
 M 
 
 ! S 
 
 Fig. 26. 
 
 equal lengths of the conductor, and the direction of the 
 force or forces carrying the loops is the same. Now 
 
 'W 
 
 ft 
 
 Fig. 27. 
 
 double the wire upon itself as in Fig. 27, and we get on 
 first contact of the fields a repulsive action like we get 
 in the case of unlike currents. On bringing them 
 
 ^ 
 
 ■\ 
 
 K^J 
 
 D 
 
 
 Fig. 28. 
 
 closer and closer together, as in Fig. 28, the respective 
 forces causing " loops " become opposite in direction, 
 and as they are equal they exactly balance or neutralise 
 each other. In Fig. 27 two loops are shown, the 
 arrows pointing in the direction of the forces. It will 
 thus be seen that to produce no effect, the return must, 
 as the proposition states, be close to the outward-going 
 conductor. It does not matter whether the wires be 
 straight or curved — the effect is the same. Thus the 
 field is neutralised by doubling and twisting around, 
 as in Fig. 29. One further remark is necessary here ; 
 it is that the doubling back of the wire upon itself, and 
 
 3. The force exerled between two parallel portions 
 of circuits is proportional to the product of the strength 
 of the two currents, to the length of the portions, and 
 inversely proportional to the distance between them. 
 
 Ampere also demonstrated certain other proposi- 
 tions, the most important of which is that a conductor 
 doubled back upon itself, so that the return path of the 
 current is close to the outward path of the current 
 exerts no force upon external points. Examine this 
 proposition in the light of what has been said. Let 
 
 Fig. 29. 
 
 hereby the destruction of the field, increases the re- 
 sistance of the circuit; indeed, the partial destruc- 
 tion of the field before referred to also increases the 
 resistance of the circuit, and, in fact, any interference 
 of field with field increases the resistance of one or of 
 both circuits. 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 23 
 
 CHAPTER III. 
 
 THE MAGNETIC CIECUIT. 
 
 '. HE popular notion of a magnet is that it is 
 a piece of steel having certain peculiar 
 properties, the most prominent of which 
 are : (1) If free to move horizontally it takes 
 up a certain definite position, one end point- 
 ing nearly to the geographical north, the 
 other end pointing nearly to the geographical south. 
 The end pointing northwards is generally called the 
 north pole of the magnet, the other end being called 
 the south pole. (2) Another property of the magnet is 
 its attraction for iron or steel. (3) It has certain actions 
 upon other magnets. There are other properties which 
 need not detain us now. 
 
 A conductive circuit has all the properties of a 
 magnet. It is a magnet. On this part of the subject 
 Dr. Lodge says : " Coil up a wire conveying a current. 
 . . . The result is it behaves like a magnet; 
 compass needles near it are affected, steel put near it 
 gets magnetised, and iron nails or filings get attracted 
 to it — sucked up into it if the current be strong enough ; 
 in short, it is a magnet, not, of course, a permanent 
 one, but a temporary one, lasting as long as the current 
 flows. It is thus suggested that magnetism may 
 perhaps be simply electricity in motion. Let us work 
 out this idea more fully. 
 
 " First of all, one may notice that everything that 
 can be done with a permanent magnet can be imitated 
 by a coiled wire conveying a current (it would not do 
 altogether to make the converse statement). Float a 
 coil attached to a battery vertically on water, and you 
 have a compass needle ; it sets itself with its axis north 
 and south. Suspend two coils and they will attract or 
 repel or turn each other round just like two magnets." 
 Consider a single turn of a conductive circuit. The 
 loops of force are all around. If now we coil another 
 turn (the conductor must be insulated, otherwise the 
 turns will not be separate) the loops are, as we have 
 previously explained, partly neutralised and partly 
 merged, the result being loops around both turns. 
 If we continue adding turns of conductor to the coil, 
 the loops of the field still pass around the turns, as in 
 Fig. 30. In the consideration of the conductive circuit 
 it was stated that the number of loops around the con- 
 ductor is proportional to the current through the 
 conductor; also that the coiling of a conductor upon 
 itself increased the resistance of the circuit. Hence, 
 with the same pressure, the current would be less when 
 the circuit was made into a coil than if it consisted of 
 only one large turn. The total number of loops, then, 
 threading the core of the coil is less than the number of 
 loops would be if the coil was straightened out, with the 
 same pressure acting in the circuit. In the coil as shown 
 
 the loops form closed curves through the air spaces, 
 one part of the curve being inside the coil, the 
 other part outside the coil. The total space in which 
 "loops of force" are found is called the "field," and 
 we now see why the term " magnetic field " is suitable, 
 for we have seen this coil— or, as it is usually termed, 
 solenoid — is a magnet, and the space through which its 
 influence extends is its "magnetic field." 
 
 Fio. 30. 
 
 In the form of coils shown all the loops thread the 
 core and distribute themselves equally around the 
 outside of the coil, provided the path through which 
 they pass is of one homogeneous material, such as air. 
 
 It may at once be stated that the great utility of 
 the magnetic field depends almost entirely upon the 
 portion of the loops outside the core, and their density. 
 A perfect magnet may be defined as one where 
 there is no external field — that is, the total field is 
 within the surfaces bounding- the material of which 
 the magnet is made. We may assume that in the 
 case described the closed curves leave the coil at 
 one end or pole, and winding round through the air go 
 into the coil at the other pole. The pole by which the 
 loops are supposed to leave the core is conventionally 
 termed the north pole, N, that by which they enter 
 is called the south pole, S. 
 
 Fig. 31. 
 
 If we bend the coil round, as in Fig. 31, so as to 
 make almost a perfect ring of conductor-turns, the 
 
24 
 
 PRACTICAL ELECTRICAL E'NGINEERltiC. 
 
 " loops of force " go through the air space between the 
 end turns almost entirely, as shown in the figure. There 
 are a few leakage loops, but these may for the moment be 
 left out of consideration. If the ring of conductor-turns 
 is completed, as in Fig. 32, then the " loops of force " are 
 
 " loops of force" from the permanent magnet are 
 similar to those from the coil, leaving one pole and 
 entering the other. A portion of the path of these 
 
 Fig. 32. 
 
 entirely within the core, and none are found in the 
 external air space. The magnet as shown in Fig. 32 
 is the perfect magnet, but of no practical value. All 
 the magnets thus composed of conductor-turns are 
 termed " electromagnets," in contradistinction to 
 permanent magnets. The former are temporary mag- 
 nets, lasting as long as there is a current through the 
 coils, the latter being, as their name implies, permanent 
 and are not dependent upon a current through coils. 
 Permanent magnets were known long before electro- 
 magnets, as the particular iron ore which possesses 
 magnetic properties is found in various parts of the 
 world. Natural magnets, however, are of no use in 
 practical work. They are at once cumbersome and of 
 little power. In very ancient times it was known that 
 if steel was rubbed with a magnet it became itself 
 magnetic and retained the properties imparted by the 
 rubbing. It was known also that soft iron similarly 
 obtains the magnetic properties, but does not retain 
 them — hence steel bars were magnetised by rubbing 
 with natural or other artificial magnets. Sturgeon, in 
 1825, made the first electromagnet. Profs. Henry 
 and Moll and Mr. Watkins, between 1825 and 1830, 
 made many experiments with electromagnets, and 
 these names are worthy of record. 
 
 If now we take a permanent bar magnet and inves- 
 tigate its field by means of iron filings, we find the 
 filings arrange themselves as in Fig. 33. A simple 
 method of doing this is to place a piece of cardboard 
 just above the magnet, and sprinkle the filings upon 
 the cardboard. A gentle tapping of the cardboard will 
 assist the experiment. But this and dozens of similar 
 experiments are fully described in every little book on 
 the subject. The filings will be seen to be arranged 
 in curves from one pole of the magnet to the other. 
 This shows the influence of the magnet upon the 
 filings in the plane of the cardboard, and if the card- 
 board were made to describe a complete revolution 
 around the axis of the magnet, the influence would be 
 found to exist at every position. This shows that the 
 
 Fig. 33. 
 
 closed curves is through the steel forming the magnet, 
 the other portion of the path being through air. It is 
 customary in practice to make the length of the path 
 through the air as short as possible by having the 
 magnet bent into the shape of a horse-shoe, bringing 
 the poles near together, as in Fig. 34. Both straight 
 
 Fig. 34. 
 
 and bent permanent magnets are used, though mostly 
 for purposes where no great magnetic strength is 
 required. As engineers generally require greater mag- 
 netic strength, electromagnets are now almost univer- 
 sally used for dynamos and motors, and it is to these 
 magnets we shall more closely direct attention. The 
 number of magnetic loops around a conductor is pro- 
 portional to the current. If a second turn of con- 
 ductor is added, keeping the current constant, the 
 number of magnetic loops is doubled, and similarly 
 each turn of conductor adds its quota to the total. So 
 far, then, the number of magnetic loops around a coil 
 depends upon — 
 
 1. The strength of the current. 
 
 2. The number of conductor-turns. 
 
 The number, however, depends also upon another 
 factor, the resistance of the path traversed by the 
 loops of force. Just as in the conductive circuit the 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 !S«fe 
 
 MATHER AND FLATTs ALTERNATE UUHRENT DYNAMO. 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 LONGITUDINAL 
 TffAVCRSE 
 
 ELEVATION. 
 
 ANDJUSON S ELECTRIC TRAVELLING CEAXF. 
 
 ifif Mlllllllllllllinillllllllllhlllll 
 
 MOTOR 
 
 PLAN. 
 
 asdersox'r'electkic travelling cease. 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 25 
 
 various metals have different conductivities or resis- 
 tances, so in the magnetic circuit various substances 
 have different magnetic conductivities or magnetic 
 resistances. 
 
 Of all known materials, iron has the greatest mag- 
 netic conductivity, or the least magnetic resistance. 
 Commercial iron is far from pure, and almost every 
 sample differs somewhat in its chemical constitution, 
 hence the magnetic conductivity of various irons differs 
 considerably. Experiments have shown that the con- 
 ductivity of iron varies from a little to many thousands 
 of times better than that of air — that is, the magnetic 
 resistance of the iron experimented with has varied to 
 many thousands of times less than that of air. 
 
 Besides iron, nickel and cobalt are almost the only 
 substances which possess magnetic conductivity, 
 though a few others, such as chromium, cerium, and 
 manganese, possess magnetic power to an exceedingly 
 feeble extent. 
 
 If, then, we take a given current passing through a 
 given number of conductor-turns, the number of mag- 
 netic loops will depend upon the resistance of the mag- 
 netic circuit, just as the current with a given pressure 
 in the conductive circuit depends upon the resistance 
 of the circuit. 
 
 B 
 
 In the conductive circuit we have C = 
 
 E' 
 
 In the magnetic circuit we have — 
 No. of loops of force Current x conductor-turns 
 
 or magnetism 
 
 Eesistance of magnetic circuit. 
 
 As the current is measured in amperes — in practice 
 
 we use the expression ampere-turns — which allows us 
 
 to write — 
 
 „ - , Ampere-turns 
 
 Eesistance of magnetic circuit" 
 
 This shows that in the magnetic circuit there rules 
 a law similar to that of Ohm in the conductive circuit. 
 
 In the conductive circuit we have the electromotive 
 force causing pressure or difference of pressure. The 
 pressure determines the current through a given resis- 
 tance, while in the magnetic circuit the ampere-turns 
 determine the magnetism or number of magnetic loops 
 through a given magnetic resistance. 
 
 If N represents the number of loops of force, E m the 
 total magnetic resistance, and A, the ampere-turns, 
 we get 
 
 -Km 
 
 The magnetic pressure due to the ampere-turns 
 
 = -4ttTC 
 
 where T = turns, and C = amperes. This brings the 
 
 formula to 
 
 •4*-TC 
 
 N = - 
 
 E„ 
 
 The total magnetic resistance, like the total conductive 
 resistance, may be and usually is made up of several 
 separate resistances. When the magnetic resistances 
 are in series the total resistance is the sum of the 
 
 separate resistances. Thus, if we consider the magnetic 
 resistance of the field-magnets of a dynamo, we get, first, 
 the magnetic resistance of the core and yoke of the 
 field-magnets; secondly, the magnetic resistance of 
 the air spaces within the armature and between the 
 armature and the field-magnet poles ; and, thirdly, the 
 magnetic resistance of the armature. Thus, if E m = 
 total magnetic resistance and E a ; E A ; E F the magnetic 
 resistances of the air spaces, the armature and the field- 
 magnets respectively 
 
 E m = E a + E A + Ep 
 or the formula can be written 
 
 N=- 
 
 •4ttTC 
 
 E a + E + Ep 
 
 Leakage. — The magnetic field of a conductor may be 
 deemed to arise from leakage from every part of the 
 conductor surface. A perfect insulator would stop 
 this leakage. Every unit area of surface may then be 
 looked upon as the formation of a branch or shunt 
 circuit, and the total resistance of any conductor is the 
 joint resistance of the conductor itself and these 
 shunt circuits. In the magnetic circuit there is similar 
 need of a magnetic insulator or screen. It may be 
 heterodox, but it is nevertheless true that the best 
 conductor is in one sense the best insulator, and the 
 best magnetic conductor is also the best magnetic 
 insulator — that is, if we wish to screen a particular 
 object from the influence of a magnetic field we do so 
 by surrounding the object with a mass of iron. The 
 loss by leakage is simply a question of resistance. If 
 there is a shunt circuit at all, both the conductive 
 (current) effect and the magnetic (loops of force) effect 
 divide inversely as the resistance. In most magnetic 
 circuits there will be leakage, and it should be the aim 
 in practice to reduce leakage to a minimum by making 
 the leakage circuits of very great resistance compared 
 with the direct circuit. 
 
 Effect of Heat on Magnets. — If a permanent steel 
 magnet is heated it gradually loses its magnetism, till 
 when at a bright red it seems to lose it altogether. 
 Nickel also loses its magnetic properties when heated, 
 while cobalt does not. If a piece of iron is rapidly 
 magnetised and demagnetised it becomes hot, as if the 
 magnetisation caused internal friction. Thus it may 
 be said that heating an iron or steel magnet increases 
 its resistance. We have already seen that heating the 
 conductor of a conductive circuit increases its resist- 
 ance, so that this property seems common to both the 
 conductive and the magnetic circuits. 
 
 Interactions of Fields. — The interactions between the 
 magnetic fields of conductors have been described. 
 
 Fig. 35. 
 
 The interactions between the fields of magnets can be 
 considered in the same way, and then the interactions 
 
26 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 between the fields of conductors and the fields of mag- 
 nets. Let N S, Fig. 35, represent a magnet. The 
 direction of action of the loop of force is from N to S ; 
 and if iron filings are put .into the field they will form 
 a polarised chain from N to S. The circuit is through 
 air, and has a certain resistance. If now we bring 
 another bar of magnetic material, such as iron, into 
 the vicinity of the field, as in Fig. 36, the resistance of 
 
 N, 
 
 i) 
 
 it 
 
 Fig. 36. 
 
 the magnetic circuit will be lessened, if the circuit 
 passes as N to S v N x , S. In fact, we make a path of 
 less resistance through the iron, and immediately the 
 loops follow the law, and go through the circuits 
 inversely as their resistance. All the loops will not 
 pass through S x N x , only the number agreeing with the 
 law enunciated. The remainder still go through the 
 shunt circuit presented by the air path. We have 
 previously agreed to the convention that the pole of a 
 magnet, where the loops leave, is the N pole, and where 
 they enter it is the S pole. The loops enter at 8 V 
 owing to the initial directive action of N, and leave at 
 Nj ; therefore, as S x N x now possesses all the properties 
 of a magnet, we say it has been magnetised by the 
 influence of N S. In electrical language, any con- 
 ductor, magnet, or magnetic substance acted upon by 
 the field of any other conductor or magnet, is said to 
 be acted upon by induction. The action is said to be 
 inductive action. 
 
 Thus the bar S x N x is said to be magnetised by 
 induction. A magnet pole acting upon a magnetic 
 substance induces in the latter a pole of opposite 
 polarity to itself. A north pole induces a south, and a 
 south pole induces a north. 
 
 V- ----- A s, 
 
 y- ^ •, 
 
 Fig. 37. 
 
 Let N S, Nj S 1; Fig. 37, represent two bar magnets 
 of about equal strength. Bring them into proximity 
 so that their fields interact. Bach magnet is sur- 
 rounded by its chains of polarised particles, and the 
 attempt to bring the magnets together means the 
 attempt to put the two fields into the same space. If 
 each field corld be indicated by iron filings it would be 
 attempting to put two lots of filings into the same 
 space, which of course is a physical impossibility. If 
 we agree upon the convention that two loops of force 
 cannot occupy the same space no wore than can two 
 
 chains of iron, the whole matter becomes clear. There is 
 a resistance of one field to the introduction of the other. 
 The term repulsion has been given to this resistance, 
 and two similar poles are said to repel each other. 
 This repulsion, then, is of the same nature as the 
 resistance offered to matter being put into the same 
 place as other matter. However near the poles may 
 be, there is no action between magnets except when 
 their fields interact. In many cases, however, it assists 
 calculation to consider this resistance as due to a 
 repulsive force. If the fields of two such magnets are 
 made to interact, the strength of each magnet is 
 weakened, as may easily be shown by the explanations 
 given. If one magnet is considerably more powerful 
 than the other, then the polar chains of the weaker one 
 are reversed by the stronger, and the weaker magnet is 
 magnetised in the opposite direction to what it was 
 orginally, the pair then exhibiting similar phenomena 
 to the pair in Fig. 38. 
 
 s, 
 
 N, 
 
 it 
 
 )) 
 
 ti\ 
 
 Flu. 38. 
 
 Let Nj S 1; N S, Fig. 38, be two magnets of about the 
 same strength brought into proximity so that their fields 
 interact, but with opposite poles towards each other. 
 The action of the pair is to merge the two circuits into 
 one, the one magnetic circuit having a less resistance 
 than either of the two separate. The tendency of all 
 magnetic loops is to make themselves as short as 
 possible, so the tendency of the circuit is now to 
 shorten itself, and this can only be done by the bars 
 getting closer together. Let Aj, Fig. 39, represent a 
 magnet with its chain of polarisation as shown, and B x 
 a similar magnet. If the opposite poles are brought 
 near, it will be seen that the polarised particles can 
 satisfy each other — that is, be so arranged that the 
 + (shaded) pole of one is in contact with the - 
 (unshaded) pole of the next, when the line of polarisa- 
 
 — A 
 
 Fig. 39. 
 
 tion passes direct from + pole of k x to - pole of B , 
 making A x and B x parts of one magnetic system. The 
 
 N, 
 
 rE3--: 0v 
 
 : s — =" *- 
 
 ( * 
 
 
 Fig. 40. 
 
 Fig. 41. 
 
 tendency to shorten the lines tend to bring A and 
 Bj into close contact, or make the system like that in 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 27 
 
 Fig. 40, where N S and N 1; Pig. 41, were the original 
 magnets. This tendency gives rise to the phenomena 
 of attraction, whence, in short, the laws have been 
 promulgated — 
 
 1. That unlike poles attract. 
 
 2. That like poles repel. 
 
 Both these properties are of great service in practice. 
 
 We have now to briefly consider the interaction of 
 conductors and of magnet fields. 
 
 If a conductor is 'brought into a magnetic field, a 
 number of the magnetic loops pass around the con- 
 ductor. It has been stated that the magnetic loops of 
 a conductor are caused by the current; the reverse 
 action also takes place, and magnetic loops passed 
 around a conductor induce a current in that conductor. 
 Further, just as the number of loops is proportioned 
 to the current, so the current is proportioned to the 
 number of loops. If a number of magnetic loops 
 encircle a conductor, a temporary current is set up in 
 the conductor in a certain direction ; if tbe number of 
 encircling loops be diminished an equivalent current is 
 set up, in direction the reverse to the original current. 
 If the whole number of loops be taken away, the 
 reverse current is equal to but opposite in direction to 
 the original current. If the number of loops is added 
 to, the current set up is in the same direction as the 
 original current. A continuous current can only be 
 obtained by continuously varying the number of loops 
 passed around the conductor. A continuous current 
 always in one direction could be obtained if the number 
 of loops passed round the conductor could be con- 
 tinuously increased. It is customary to speak of the 
 current or the pressure that regulates the current as due 
 to the number of lines or loops cut per second by the 
 conductor, and one volt pressure is obtained in a con- 
 ductor when the conductor cuts 100,000,000 loops of 
 force per second. The usual method of obtaining a 
 current is to bring a conductor into as strong a field as 
 possible, then to take it out again. This action gives 
 first a current in one direction, then an equal current, 
 but in the opposite direction. Apparatus, however, is 
 used so that these reversed currents may, if required, 
 appear at the point where they are to be utilised, in the 
 same direction. 
 
 conductor, the action of the field will be either to 
 augment or to decrease the current according to the 
 direction of the original current. If it is in the same 
 direction as that due to the field, we get increase; if in 
 the opposite, we get decrease. 
 
 Lastly, we have the action of a conductor field upon 
 a magnetic substance or upon a magnet. If instead 
 of the open coil, Fig. 31, we make our turns of wire 
 around a piece of iron, Fig. 42, and send a current 
 
 Fig. 42. 
 
 through the wire, the iron core is in the magnetic 
 circuit, and becomes a magnet. If the viewer is look- 
 ing at one pole of the magnet, and the current in the 
 surrounding coil is in the same direction as the hands 
 of a watch travel, he is looking at the south pole. The 
 direction of the magnetic loops is in at S, out at N, the 
 current in the coil being in the direction shown by the 
 arrows. The resistance of the magnetic circuit is 
 lessened by bending the coil till the two poles are 
 near each other, as in Fig. 43. If a core of steel is 
 
 Fig. 43. 
 
 used instead of a core of iron, the steel may be made 
 into a permanent magnet, whereas the iron (and the 
 more completely so the softer it is), when the current 
 is stopped, loses its magnetic properties, though even 
 with the softest iron a little magnetism always remains 
 — termed residual magnetism. 
 
 There is one phenomena connected with field inter- 
 actions of great use in practical work. It is the action 
 of a small magnet in the magnetic field of a conductor. 
 To understand this action, it must be remembered that 
 if a magnet and a conductor are parallel to each other 
 their fields will be at right angles. Suppose D, Fig. 44, 
 
 - - -S-_. 
 
 Fig. 31. 
 
 Fig. 44. 
 
 The action of a magnetic field upon a conductor is to represents a magnet with polarised loops as shown, 
 alter the electrical pressure of the conductor, and so If these polarised loops are brought within the 
 obtain a current. If a current already exists in the influence of another polarising force, the latter will try 
 
28 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 to polarise in its own direction. If the new polarising 
 force be at right angles to the original, the action of the 
 force will be to polarise the particles in a direction at 
 right angles to their original direction. This is what 
 happens with a small magnet in a conductor field, so 
 that if N S, Fig. 45, is brought into the field of C, the 
 
 n 5 
 
 ^ 1---------==* ---^ 
 
 /A<r^ .',' 
 
 Fig. 45. 
 
 stronger field tends to polarise the loops of N S in 
 the direction in which its own loops are polarised. As 
 the magnet N S is itself a chain of polarised particles, 
 its direction is influenced, and the result is that a 
 magnet so brought into a conductor field tends to set 
 itself or to be placed in a position at right angles to 
 the conductor. If we assume that two forces in 
 different directions are acting on material particles, 
 B A, Fig. 46, we know that the particles will tend to a 
 direction which is the resultant of that of the forces, 
 and this direction in the case under consideration will 
 
 bo more and more in the direction of the loops of force 
 of C, the stronger the field of C compared with N S. 
 If the magnet field is very strong compared with that 
 of the conductor, then the conductor will move, or 
 
 
 N 
 
 
 s 
 
 B 
 
 
 
 _ L 
 
 
 
 
 t — -, 
 
 S" \ 
 
 
 
 
 , 
 
 
 
 „ — ■ 
 
 
 <8/j 
 
 
 
 d 
 
 =]0 
 
 Fig. 46. 
 
 tend to move, in accordance with the foregoing 
 considerations. 
 
 The interactions of the greatest use in electrical 
 engineering are : 
 
 1. Those between a magnetic field and a conductor, 
 which give the basis of dynamo action. 
 
 2. Those between conductor fields and conductors, 
 which give the basis of transformers. 
 
 3. Those between conductor field and magnets, 
 which give the basis of many of our measur- 
 ing instruments. 
 
 CHAPTER IV. 
 
 THE INDUCTIVE CIECUIT. 
 
 T is customary to view the inductive circuit as 
 something altogether different from the con- 
 ductive, whereas the one is in reality but a 
 modification of the other. The modification 
 introduces more prominently certain pheno- 
 mena which are important, and have now to 
 be considered. Induction is a term used to denote a 
 condition which in different circumstances seems 
 altogether distinct from what it does under other 
 circumstances, yet a closer examination will show 
 the similarity. There can be no flow of current 
 without difference of electrical pressure, but there 
 can be difference of electrical pressure without cur- 
 rent. The former case, difference of pressure with 
 current, exists in the conductive circuit ; the latter 
 case, difference of pressure but no current, exists in the 
 inductive circuit. We have said electrical apparatus is 
 used for producing electrical pressure. It may aid the 
 imagination in some cases if we look upon the appara- 
 tus as having a double object, raising pressure on the 
 one side and proportionately decreasing it on the other. 
 Of course the assumptions previously made ought to 
 lead to this view, for as there is no introduction of 
 electricity into a circuit, the increase of electrical pres- 
 
 sure in one direction naturally means that if measured 
 in the opposite direction the pressure is below the 
 maximum. Thus we get two conditions the counter- 
 part of each other, a quantity of electricity on the one 
 side under more electrical pressure, an equal quantity 
 on the other side under less electrical pressure. The 
 lifelong object of electricity in these different states is 
 to resume the normal condition, or the condition of a 
 normal pressure when no electrical phenomena are 
 indicated. 
 
 We can now represent the simple inductive circuit, 
 Fig. 47. Suppose a source, S, of electrical pressure, and 
 conductors, A and B, separated by the 
 air space, D. Suppose, further, the 
 pressure to be increased in the direc- 
 tion of A, and diminished in direction 
 of B, then when contact is made with 
 the source we get a momentary rush 
 or current in direction A, and as it 
 were a slight heaping up of the electricity in A 
 and a corresponding diminution in B. This heaping 
 up, accumulation, or charge, as it is called, shows 
 itself principally at the surfaces of A and B, im- 
 mediately opposite, or the electricity in unlike 
 conditions of pressure gets as near together as 
 possible, 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 29 
 
 The quantity so accumulated at the ends of wires is 
 extremely small, but it is easy to arrange the circuit so 
 that quantities measurable by our instruments can be 
 dealt with. First, however, it may be wise to point 
 out that these two electrical conditions are fully recog- 
 nised. 
 
 If the apparatus we use increases the pressure 
 in one direction, this has the same effect as diminish- 
 ing it in another, hence the electricity appears mani- 
 fested to us in two different conditions. In the 
 olden days one electricity was spoken of, the pheno- 
 mena being said to be due to a surplus or to a want 
 of electricity ; subsequently two electricities were 
 named ; still later the conditions were simply termed 
 positive and negative, and symbolised by + and - . 
 These symbols are very suitable and we shall adopt 
 them, using + to symbolise electricity at the point 
 of higher pressure, and - to symbolise it at the point of 
 lower pressure. It will at once be seen that if the 
 pressure is taken away the electricity assumes its 
 normal condition, when it is not manifest to our 
 instruments. Thus only in its abnormal state is elec- 
 tricity manifest to us, or useful for any purpose for which 
 it is desirable to use it. It is convenient sometimes to 
 suppose the apparatus raises the pressure on the one 
 hand, and lowers the pressure on the other to equal 
 degrees, just as pumping water from a cistern both 
 raises and lowers, though only in equal degrees under 
 certain conditions. 
 
 If the normal electrical condition algebraically be 
 equivalent to 0, this is represented by the sum of the 
 positive and negative conditions. If x represents the 
 electrical pressure in any circuit + x is exactly equal to 
 - x, or in another way a pressure of 10 in one direction 
 is equivalent to a diminution of 10 in the opposite direc- 
 tion, as + 10 - 10 = 0, the normal condition of pressure. 
 We cannot by any possible device obtain a condition 
 that more pressure is obtained in one direction than 
 depression in the other. This is the same as saying, 
 like the old books, whenever + electricity is generated 
 an equal quantity of- electricity is generated at the 
 same time. 
 
 The statement, " a heaping up of electricity," is not 
 scientifically accurate. What takes place is the heaping 
 up or accumulation of electrical energy. Just as in 
 the conductive circuit C x E represents electrical 
 energy, so in the inductive circuit the energy may 
 be represented by A x E, where A represents the 
 electricity accumulated under a pressure, E. 
 
 Simple Inductive Circuit. — A simple inductive cir- 
 cuit may be arranged as in Fig. 48. Let S be the 
 source of electrical pressure, A and B wires joining the 
 terminals of the source with the parallel conducting 
 surfaces or plates shown in section P V 1 at a distance 
 from other conductors. The air space between P and 
 V 1 forms the dielectric, non-conductor, or inductive 
 resistance. If the distance between P and Pj be taken 
 as the length of the resistance, and the surface area of 
 the opposing surface of one of the plates as the sectional 
 area of the resistance, then, just as in the conductive 
 
 circuit, the resistance varies directly as the length of 
 the conductor and inversely as its sectional area, so in 
 the inductive circuit the resistance varies directly as 
 the length and inversely as the sectional area of the 
 
 Fig. 48. 
 
 dielectric. In other words, if you double or treble the 
 length of the dielectric, you double or treble its resis- 
 tance, while if you double its sectional area you halve 
 its resistance. This is symbolised by 
 
 L 
 
 Boc 
 
 S 
 
 where B = resistance of dielectric, L its length, and S 
 its sectional area. In both circuits, then, we have 
 similarly the source of electrical pressure and resis- 
 tance of circuit. The result of electrical pressure in 
 the conductive circuit is current, in the inductive 
 circuit it seems to be accumulation on the surfaces of 
 the opposite conductors. We say seems because the 
 probable action is at the surface of the dielectric in 
 contact with the conductors, or in the dielectric, and 
 not on the conductors. In this case, however, the use 
 of the popular expression will do no harm. 
 
 The idea of accumulation involves that of density per 
 unit area, and if the plates are, throughout their surfaces, 
 at the same distance apart the same source accumulates 
 or charges the opposite surfaces to the same density. 
 The quantity accumulated, or the total charge, will 
 evidently be the density per unit area x number of 
 units of surface. It will now be seen that accumulation 
 in the inductive circuit is analogous to current in the 
 conductive circuit, and that the same laws apply to 
 both. 
 
 Current x resistance = electrical pressure. 
 
 Or C x B = E. 
 
 Accumulation x resistance = electrical pressure. 
 
 Or A x B = E. 
 
 Inductive Resistance, — Series. — Multiple Arc. — In- 
 
 Fig. 49. 
 
 ductive resistances are added in series just as conductive 
 resistances. Thus Figs. 49 and 50 show two resistances 
 
30 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 in series ; Figs. 51 and 52 show three resistances in 
 series. In either case the length of dielectric has either 
 been doubled or trebled, and the total resistance 
 
 In the first case B = E x + B 2 . 
 
 Or in the second case B = B x + E 2 + E 3 . 
 
 If Ei = E 2 = E 3 , etc., then E = wE r 
 
 In the conductive circuit the practical effect, as far 
 as total resistance is concerned, of adding conductors in 
 multiple arc is the same as if the sectional area of the 
 conductor is increased. Thus the resistance of the 
 circuit, E, Fig. 53, would be halved either by adding, 
 in multiple arc, another similar con- 
 ductor, or doubling the sectional area of 
 E. Similarly, in the inductive circuit the 
 practical effect of the addition of resis- 
 tances in multiple arc is to increase 
 the sectional area, and correspondingly 
 to decrease the resistance. 
 
 s 
 
 Fig. 53. 
 
 Fig. 54. 
 
 Thus, if, in Fig. 54, P and P t be the opposing 
 conductors and E the inductive resistance, and if 
 P and Pj be increased to double size, as shown by the 
 dotted lines in Fig. 55, or a similar resistance be added 
 
 R, 
 
 p, 
 Fig. 55. 
 
 Fie. 56. 
 as shown in Fig. 56, the total resistance, E, will be 
 K = B +B aS in tlie conductive circuit, for let the 
 respective resistances be E ( and B,, and the pressure be 
 
 unit pressure between the surfaces, and let the joint or 
 total resistance be E, then : 
 
 Accumulation on P or P x = — - 
 
 Ei 
 
 „ „ S or S x = jj- 
 
 Total accumulation on P and S, or P x and S x = 
 
 
 ^- ; but as accumulation is inversely 
 
 BiE, 
 
 E X E 2 
 
 as resistance E = 
 
 1\ + K 2 
 
 Distribution of Charge. — It has been stated that in 
 the conductive circuit the current is increased when a 
 new path is inserted into the circuit, and the total 
 current distributes itself inversely as the resistances; 
 so in the inductive circuit the charge or accumulation 
 is increased by the insertion of a new inductive path, 
 and the total charge distributes itself inversely as the 
 resistances. Thus in Figs. 55 or 56 the addition of plates, 
 S S x , equal to P + P x with E 2 = E } pressure remaining 
 constant, leads simply to doubling the charge because 
 the resistance is halved, the quantity and density on 
 P or P a being equal to quantity and density on S 
 or Si. If, however, E 2 is different from E x -, as in Fig. 57, 
 
 Fig. 57. 
 
 where Q Q 1 represent the new opposing conductors, 
 the total accumulation will be increased because the 
 total resistance is lessened. Suppose in this case E 2 is 
 twice Ej, we may represent 
 
 E 1 = l andE 2 = 2. 
 
 Total resistance E = JMJf = ~ 
 Ei + E 2 3 
 
 that is, the resistance of the original circuit, E^l, is 
 now by the addition of E 2 reduced to -|~, and the charge 
 will therefore be correspondingly increased. And if 10 
 represents the old quantity accumulated, and x the new 
 quantity, 
 
 Ne w resistance _ Old charge 
 Old resistance ~~ New charge 
 2 
 
 ~3~ 10 
 1 ~x 
 2 
 -3-2 = 10 
 
 whence x = 15. 
 
 In symbols, if Q = old quantity and Ql new quantity 
 accumulated, u J 
 
 E i B, ,, 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 31 
 
 This new charge is distributed inversely as the 
 resistances — that is, 10 parts with resistance E x and 
 5 parts with resistance E 2 . It will be seen that no 
 change has taken place as regards the charge on P 
 or P x , and this will be found to hold good whatever 
 branch resistances may be added to the circuit. Simi- 
 larly, in the conductive circuit, it has been seen, the 
 adding of parallel, branch, or shunt circuits with pressure 
 remaining constant, does not affect the current through 
 the original circuit. If the opposed conductive sur- 
 faces have at all points the same inductive resistance 
 between them, the distribution of the charge is 
 uniform over the surfaces. If, however, as is frequently 
 the case, the resistance is not uniform, then the distri- 
 bution is not uniform. A simple diagram will illustrate 
 this. Let S, Fig. 58, represent the source, one pole of 
 
 Fiu. 58. 
 
 which is connected with the conductor, A, the other 
 with the wall of the room, B. The conductor, A, is 
 insulated from the room, and thus can receive a charge 
 from the source, S. Let air be the dielectric. The 
 dielectric resistance between the points C D will be 
 much less than the resistance between the points E F, 
 hence the charge upon unit area C will be correspond- 
 ingly greater than the charge on unit area E. Experi- 
 menters have often wondered why certain results did 
 not agree with theory, and many a young student has 
 forgotten that his body near a piece of apparatus causes 
 a distribution of charge different to what would be if 
 the body were not there. 
 
 Short Circuits. — When the resistance of the con- 
 ductive or of the inductive circuit is decreased to a 
 small quantity, the current or the accumulation is 
 enormous. An examination of the formulae will show 
 that theoretically any amount of current or accumula- 
 tion can be had with any pressure, however small. 
 Anyone who understands division in arithmetic will 
 easily understand our meaning. Thus, suppose we 
 have to divide any number — say 5 — by any other 
 number greater than unity, the answer will be less 
 than 5 ; it is exactly^ when we divide by unity, but when 
 we get on the other side of unity and divide by numbers 
 less than unity the answer is greater than 5. 
 
 Take these examples : 
 
 5 
 5 -^ 2 or -tj- = 2h 
 
 1 or -j-= 5. 
 
 5 -*■ I or :g-= 10. 
 
 5-ior |g= 20. 
 
 F E 
 
 Thus in C = -p or A =^ whatever pressure E may 
 
 indicate, the current, C, or accumulation, A, may be as 
 great as we please provided we make E in each case 
 small enough. 
 
 It is this short-circuiting in practical work that leads, 
 as we have before said, to great dangers. Whether the 
 original circuit be conductive or inductive the result of 
 a short-circuit is the same, the passage of a large 
 current through a small resistance. 
 
 Specific Inductive Besistance. — Faraday found that 
 by substituting another gas for air as the dielectric the 
 resultant charge was unaltered, and he also found for 
 all the solid and liquid dielectrics he experimented with 
 a greater capacity for accumulation than with gases. 
 He thus introduced the term specific inductive capacity. 
 It would now seem preferable to look upon all gases as 
 having the same resistance, and solids and liquids as 
 having different resistances. If the resistance of unit 
 length and unit sectional area of air be taken as the 
 unit, then the resistances of other insulators compared 
 with this might be termed their specific inductive 
 resistances. 
 
 Quantity and Capacity . — The unity of quantity is an 
 ampere per second, and is called a Coulomb. The 
 capacity of a conductor is measured by the number of 
 coulombs of charge that can be given to one conductor 
 with a volt pressure between that conductor and the 
 opposing conductor. Thus, the capacity of P, Fig. 
 59, is the number of coulombs upon it when there is a 
 
 Fig. 59. 
 
 The 
 
 difference of pressure of a volt between P and P lP 
 unit of capacity is called the Farad. 
 
 The capacity of a single pair of plates is small, and 
 to obtain a greater capacity in practice a series of 
 insulated conductors are arranged so that each alternate 
 conductor is connected to its own terminal, as in Fig. 60. 
 
 *>t 
 
 1* 
 
 Fip. 60. 
 
 This apparatus is called a condenser, and is frequently 
 made of sheets of tinfoil, separated by sheets of thin 
 paper coated with paraffin wax. If the conductors are 
 separated by air, we have an air condenser. If A is 
 the surface area of one of the sets of conductors in 
 
32 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 square inches, and d the distance in inches between 
 them — that is, the length of the dielectric — then the 
 capacity in farads is 
 
 F = - 
 
 10 12 x 4-452 x d' 
 
 Condensers are usually made of so many millionths of a 
 farad, or micro-farads, capacity, and the above formula 
 becomes in micro-farads (M.F.). 
 
 A. 
 
 ^•^ = 10 6 x 4-452 x a" 
 If the measurements are in square centimetres and 
 centimetres, these formulas become respectively 
 
 A 
 
 F 
 
 '10 13 x 1-131 x d 
 
 and 
 
 < MJ? -> = 1(F x 1-131 x d - 
 
 If, instead of air, a dielectric of greater specific in- 
 ductive capacity be used, then these formulae must be 
 multiplied by the " specific inductive capacity " to give 
 the capacity of the condensers. Thus, if paraffin wax is 
 used, this having a specific inductive capacity of about 
 2, the formula 
 
 A 
 
 F = 
 
 becomes 
 
 F = 2 
 
 10 12 x 4-452 x d 
 A 
 
 10 12 x 4-452 x d 
 
 Prof. Ayrton gives the following table of specific in- 
 ductive capacities : 
 
 Specific Inductive Capacity. 
 
 Substance. 
 
 Specific Inductive 
 Capacity. 
 
 Authority. 
 
 Vacuum, air at about O'OOl 
 
 millimetre pressure 
 
 Vacuum, air at about 5 milli- 
 
 Hydrogen at about 760 milli- 
 
 Air at about 760 millimetres' 
 
 Carbonic Dioxide at about 760 
 
 defiant Gas at about 760 milli- 
 metres' pressure 
 
 Sulphur Dioxide at about 760 
 
 Paraffin Wax, Milky 
 
 J0-94 about. 
 
 / 0-9985 
 \ 0-99941 
 / 0-9997 
 \ 0-9998 
 
 }• 
 
 / 1-000356 
 1 1-0008 
 
 j 1-00037 
 
 j 1-0037 
 
 rl-92 
 
 1-96 
 
 1 1-977 
 
 U-32 
 
 2-47 
 
 2-34 
 
 2-94 
 
 2-55 
 (2-56 
 J 2-76 
 13-15 
 
 I 2-88 to 3-21 
 13-84 
 
 2-95 to 3-73 
 
 4-2 
 
 5 
 
 6-57 
 
 6-85 
 
 7-4 
 
 10-1 
 
 Ayrton. 
 
 Ayrton. 
 Boltzmann. 
 Boltzmann. 
 Ayrton. 
 Taken as the 
 
 standard 
 Boltzmann. 
 Ayrton. 
 
 Boltzmann. 
 
 Ayrton. 
 
 Schiller. 
 Wiillner. 
 Gibson and 
 
 Barclay. 
 Boltzmann. 
 Schiller. 
 Schiller. 
 Schiller. 
 Boltzmann. 
 Wiillner. 
 Schiller. 
 Boltzmann. 
 Wiillner. 
 Boltzmann. 
 Wiillner. 
 
 „ Light 
 
 „ „ Dense 
 
 „ ,, Double extra dense 
 
 Vj. Hopkinson. 
 
 Leyden Jar. — Condenser. — The original form of the 
 condenser was that known as the Leyden jar so named 
 
 after the town of Leyden, where it was accidentally dis- 
 covered by Muschenbroeck and Cunius in 1746. The 
 Leyden jar is usually a glass jar coated inside and out- 
 side to about three-fourths or rather more of its height 
 
 A 
 
 Fig. 61. 
 
 with tinfoil. Fig. 61 shows a sheet of glass, A B, 
 coated on both sides with tinfoil, A A, while Fig. 62 
 shows the coated glass in the form of a jar, A B form- 
 ing the external and internal coatings of tinfoil. The 
 conductor from B can be connected to one terminal of 
 source through the knob and portion outside the jar. 
 
 The charged condenser contains a store of electrical 
 energy 
 
 F xE 2 
 = — q. 7 foot-pounds. 
 
 Some writers have suggested the storing up of electrical 
 
 energy by means of huge condensers, like gas is stored 
 up in gasometers. Perhaps a little 
 consideration of the matter would 
 show that the idea is not feasible, for 
 it can be shown that one-half the 
 energy expended in charging the 
 condenser will be lost when the 
 charge is obtained from a source, 
 whether battery or dynamo, with the 
 electrical pressure, or potential, con- 
 stant. If the pressure of the charging 
 source could be continuously in- 
 creased in proportion to the decrease 
 of pressure from the condenser in 
 Fio. 62. discharging, we should get minimum 
 
 loss. Besides loss, the space required 
 
 for the condensers would be very large. 
 
 Suppose, for example, a source of E volts be used 
 
 to charge a condenser of F farads, then, as above, the 
 
 store of energy 
 
 FxE 2 
 = 2-7 foot -P oun cls = -37 F E 3 foot-pounds. 
 
 Let Q = charge of condenser in coulombs, 
 Q = FE. 
 The total work done by the source 
 
 = -7375 QE foot-pounds. 
 • = -7375 FE 2 foot-pounds. 
 Hence loss = (-7375 - -37) F E 2 foot-pounds, 
 or one-half the energy expended. 
 
 We have thus briefly indicated the salient features 
 which have to be applied in practice concerning the 
 three circuits. 
 
 Each special portion of the book will expand into 
 greater detail the features to be applied in the specific 
 application under discussion, 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 ~3 
 
 
 I 1 
 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 K 
 
 25 H.P. RECKENZAFN MOTOR — J-SIZE. 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 33 
 
 CHAPTER V. 
 
 ELECTB0-GEAPH1CS. 
 
 *T is essential to the electrical engineer that he 
 should be able to rightly interpret the graphic 
 representation of physical phenomena, as it 
 is for the mechanical engineer to interpret 
 the diagram giYen by a steam-engine. A 
 little curve speaks clearly and plainly to the 
 mechanical engineer, telling him at a glance almost as 
 much as could be written in several pages, and en- 
 abling hundreds to read and to understand what is 
 taking place, although if the information was put 
 before them in words and figures it would remain 
 incomprehensible. The use of graphic methods in elec- 
 trical engineering is increasing, and the introduction 
 of these methods, more particularly to phenomena con- 
 nected with dynamos and motors, has largely led to the 
 astounding increase of our knowledge about such 
 apparatus in a short time. It may be said that Dr. J. 
 HopMnson, F.B.S., in his papers before the Institution 
 of Mechanical Engineers, in April, 1879, and April, 
 1880, first effectively used the system as an instrument 
 of electrical research, and the remarks of Mr. B. E. B. 
 Crompton during the discussion of the second paper 
 conclusively show how soon its value was acknow- 
 ledged in practice. Mr. Crompton said, "If Dr. 
 Hopkinson's diagram was correct, as no doubt it was, 
 the problem was solved which had perplexed him for a 
 long time — namely, to find out the point at which to 
 regulate the lamps in order to get stability of light." 
 
 A full and complete knowledge of graphical methods 
 requires an extended knowledge of mathematics, but 
 very much may be done by men who have little 
 or no mathematical skill. A man who can properly 
 use a pair of compasses and a scale of lengths will 
 frequently be able to solve problems of a certain class, 
 for all practical purposes, as satisfactorily as the more 
 accomplished mathematician. In many cases a plotted 
 curve speaks more plainly and tells more to the un- 
 trained mathematician than the equation to that curve, 
 be it simple or be it complex. Within certain limits a 
 curve can be interpreted by everyone. It must not for 
 a moment be supposed that the non-mathematician can 
 ever attain to the position of the mathematician either 
 in accuracy or extent of interpretation. The former 
 can do much, but the much is very limited compared 
 with that of the latter. As has been stated, the sym- 
 bolisation and the solution of problems arising in the 
 mechanical arts by means of lines and curves — that is, 
 by graphical methods — is very extensive, so that the 
 application of the system to electrical problems was 
 perfectly natural. Although Dr. Hopkinson may be 
 taken as introducing the system to the student so far 
 as dynamos and that class of electrical apparatus is 
 
 concerned, it will be found that the sys.em has long 
 been extensively used in discussing problems connected 
 with cable work. It is only necessary to consult the 
 writings of Latimer Clark, F. C. Webb, H. E. Kempe, 
 and others who have written on submarine cables and 
 their testing to see this. We imagine that the great 
 value of the system is because it pertains to the concrete 
 rather than to the abstract. 
 
 Although the title of this chapter is Electro-Graphics, 
 it is intended to indicate certain directions wherein 
 books on special branches of mathematics should be 
 consulted, rather than to enlarge upon the subject. 
 
 It is impossible to understand the references made 
 by writers on dynamos without an elementary know- 
 ledge of trigonometrical ratios. Take the case of the 
 induced electromotive force in a one-coil armature, 
 which is " proportional to the sine of the angle through 
 which the coil has turned." Simple as this statement 
 is to the initiated, it is incomprehensible to the 
 uninitiated except it be accompanied by a graphic 
 representation of a " curve of sines." Let us examine 
 this simple statement — 
 
 In Fig 63, let A B be a straight line. Upon A B place 
 another straight line, AD, equal to AB in length, the 
 end A of A D resting exactly upon the end A of A B. 
 
 Eotate A D around A, taking up the position shown 
 at AD. 
 
 From D drop a perpendicular line onto A B, cut- 
 ting A B at C. 
 
 The rotation of A D has caused the opening, or angle 
 BAD, between the two lines, A B, A D. 
 
 No matter how long or how short A B, A D, the 
 lines A B, AD, A C, C D, etc., all have a constant pro- 
 portion to each other as long as the angle BAD 
 remains unaltered. 
 
 If, however, A D be rotated further, say, to A D 1; the 
 proportion of these lines to one another has altered. 
 This will easily be seen, for C D becomes C x D lf a much 
 
Missing Page 
 
Missing Page 
 
36 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 But the forces may neither act together nor exactly- 
 opposite, but at any angle. Any elementary book on 
 statics will show that the resultant of any two such forces 
 acting at a point may be represented in magnitude and 
 in direction by the diagonal of the parallelogram, of 
 which the lines representing the two forces form adjacent 
 sides. Let A B, A C, Fig. 69, represent forces of 41b. 
 and 31b. respectively acting in the directions A B, A C. 
 The resultant is found by completing the parallelogram 
 A B C D, by drawing through C the line C D parallel 
 to A B, and through B the line B D parallel to A C. 
 
 Fig. 69. 
 
 Join A D. Then A D represents the resultant of the 
 force A C, A D in direction and in magnitude — that is, 
 the same effect would be obtained if A C and A B were 
 replaced by a force represented in direction and in 
 magnitude by A D. If A B, A C act at right angles to 
 each other, and be drawn to scale, A B = 4 units, and 
 A C = 3 units, then A D will on the same scale measure 
 5 units. For the angle A B D being a right angle, 
 
 AB 2 + BD 2 = AD 2 
 
 4 2 + 3 2 = AD 2 
 
 V25 = 5= VAD 2 = AD. 
 
 The resultant of two forces acting at any other angle 
 than a right angle may be similarly represented, though 
 in many cases the calculation is not so simple. Take 
 as an example : 
 
 Two forces represented by 121b. and 151b. acting 
 upon a point at an angle of 60 deg., required to find the 
 magnitude and direction of the resultant. 
 
 1st. Graphical solution, Fig. 70. 
 
 Fig. 70. 
 
 Draw AB = 15 units of any convenient scale. 
 
 Make the angle B A C = 60 deg. 
 
 Make A C = 12 units of same scale as A B. 
 
 Through C draw CB parallel to AB, and through B 
 draw B E parallel to A C. 
 
 Join A E. Then AE is the resultant of AB, AC, 
 and if measured off on the same scale as was used in 
 
 drawing A B, A C will be found to be 23 2-5 parts of 
 that scale. The angle B A E will be found to 
 measure very nearly 26° 20'. The resultant force then 
 is23 2-51b. 
 
 2nd. Calculation : 
 
 AE 2 = AB 2 + BE 2 - 2 AB, BE, cos. 120 deg. 
 = AB 2 + BE 2 + 2AB, BE, cos. 60 deg. 
 = 15 2 + 12 8 + 2 x 15 x 12 x -5. 
 = 549. 
 A E = V549 = 23-43. 
 
 An answer which differs from the preceding one 
 only by "03, or T £ „. 
 
 If more than two forces act upon a point, their 
 resultant may be found by first finding the resultant of 
 two of the forces, then of two more, then the resultant 
 of these resultants, and so on. 
 
 Moment of Forces, Couples, Torque. — Forces may not 
 act at the same point on a body, Fig. 71. 
 
 Fig. 71. 
 
 Let the two forces P and Q, represented in magni- 
 tude and direction respectively by A B, C D, act respec- 
 tively at the points A and B on a body. The resultant 
 of these forces can be found. Also the point at which 
 it acts, graphically thus— Join A C. Produce D C, B A 
 to meet at the point E. Along E C take ED 1= 'cD 
 and along E A take E B = A B. Complete the parallelo- 
 gram E B y B D x . Join E R. Produce E E, cutting A C 
 in F and make FE 1 = E E. Then F E x represents the 
 resultant of A B t C B, and A B x C D may be replaced by 
 the single force F Ej acting at F. 
 
 From F draw F M, PM, perpendicular to the 
 direction of the forces AB, CD respectively. It is 
 found that the force P represented by A B x F M = 
 the force Q represented by C D x F M., or shortly 
 P x F M = Q x F U v l } ' 
 
 These products are termed the moments of the 
 forces. 
 
 If the body acted upon be reduced to a rioid line \ C 
 A C is called a lever, and the point F its fulcrum. If 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 37 
 
 the force P alone were to act on the body at the point 
 A, or, say, on the lever, it would evidently twist the 
 lever about the fulcrum, and Q, acting alone at C, would 
 twist the lever about the fulcrum in the opposite direc- 
 tion. When two equal and opposite parallel forces act 
 at different points of a rigid body in the same plane, 
 their effect cannot be obtained by a single or resultant 
 force, and their tendency will be to twist the body in 
 the direction of the plane in which they act. Thus P, 
 Fig. 72, acting on A in the direction A P, and P acting 
 on B in the direction B P, will twist the body round 
 in the direction B A P or A B P. 
 
 *P 
 
 Pia. 72. 
 
 The term Couple is applied to such a system of 
 forces. 
 
 The product of one of the forces and the arm of the 
 couple is called the moment of that couple. Thus 
 P x AB = moment of the couple represented in 
 Fig. 72. This applied to a magnet gives the magnetic 
 force at the N pole or at the S pole x the length of 
 the axis of the magnet as the moment of the magnet. 
 
 This moment is a measure of the tendency of the 
 couple to twist the body on which it acts, and it is 
 customary to indicate a couple by its moment. 
 
 A couple tending to twist a body in the same direc- 
 tion as the hands of a watch is termed a positive 
 couple ; if in the opposite direction, a negative couple. 
 The term " torque " is generally used by electrical 
 engineers instead of "moment of couple," "statical- 
 moment," or "turning moment." 
 
 The produet of two quantities ean be represented by an 
 
 area. 
 
 This kind of graphic representation is principally due 
 to Prof. Silvanus Thompson. An example, taken from 
 his well-known book, will show the excellence and 
 the simplicity of his work. It relates to the efficiency 
 
 c 
 
 / 
 
 / 
 
 
 Fia. 73. 
 
 of motors. Let AB, Fig. 73, represent the electro- 
 motive force, E, of the electric supply. On A B con- 
 struct the square ABCD. Draw the diagonal BD. 
 From B along B A measure B F, representing on the 
 same scale the counter electromotive force of the motor. 
 
 Through F draw FGH parallel to B C, and through G 
 draw KGL parallel to AB or D C. The actual electro- 
 motive force producing a current is the difference be- 
 tween the electromotive force of the supply and the 
 counter electromotive force of the motor. Let the 
 latter = e. The difference is E - e, which may be 
 represented by A F, K G, D H, K D, G H, or L C. The 
 electric energy put into the motor per second is C 2 E = 
 
 E -e 
 E C ; and as C = —^ — , where B = total resistance in 
 
 FA 
 the circuit, C is represented in the diagram by -=5-7 
 
 forFA = AB-BF = E-e. The energy expressed 
 
 fv. u •«. E (E-e) 
 thus may be written ^ 
 
 The work converted by the motor is 
 g(E-e) 
 
 but B is constant, so that these values may be written 
 respectively E (E - e) and e (E - e) . 
 
 Now the product E (E - e) is represented by the 
 rectangle A F H D, for A D = A B representing E ; and 
 A F represents E — e. 
 
 While the product e (E - e) is represented by the 
 rectangle G L C H, for G L = F B represents e, and 
 GH = GK=AF represents E - e. 
 
 Fig. 74. 
 
 These areas, then, are proportional to the work ex- 
 pended and recovered, and when shaded, as in Fig. 74, 
 exhibit very clearly these proportions. 
 
 It will be seen that the construction of such diagrams 
 is very simple, but when carefully constructed can be 
 made to give considerable information to those who 
 might be unable to understand analytic treatment. 
 
 The Use of Co-Ordinates. 
 
 The third kind of graphic representations to which 
 we shall briefly allude are those in which it is usual to 
 define the position of a line by referring to its distance 
 from two other lines generally at right angles to one 
 another in the plane of the paper. The branch of 
 mathematics that treats this subject fully is termed co- 
 ordinate geometry. As an instrument of research the 
 co-ordinate theory is unequalled for power and facility. 
 Assume that lines, whether straight or curved, are 
 made up of a series of points. It will be best to con- 
 sider first the position of a point in a plane, then to 
 consider series of points or lines. 
 
38 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Draw two lines at right angles X X x> Y Y l5 Fig. the distance of the point P from the axis of Y, is 
 75. nothing; in other words, that the point lies somewhere 
 
 on OY. Similarly, y = tells us that the point lies 
 
 M 
 
 Fig. 75. 
 
 These are called the co-ordinate axes, or, briefly, co- 
 ordinates. Usually only one quadrant is required, 
 XOY. The point is said to be the origin of the 
 co-ordinates, lines measured along X are named the 
 abscissae, and along Y the ordinates, while X and 
 Y together are called the co-ordinates, XOXj and 
 
 Y Yj are the co-ordinate axes, here rectangular, but 
 not necessarily so. The position of any point P in the 
 same plane as the co-ordinate axes is fixed if we know 
 its distance from the co-ordinates. A further conven- 
 tion is that all ordinates above XOX x are reckoned 
 + , all below — , while all abscissae to the right hand of 
 
 Y Yj are reckoned + , all to the left hand - . If, then, 
 the point P is in the first quadrant, and its abscissa, the 
 line P N, drawn through P parallel to X, and its 
 ordinate, P M, drawn through P parallel to Y, both 
 ordinate and abscissa are positive. If the point P 
 were in the second quadrant the ordinate would be + 
 and the abscissa— ; in the third quadrant both would 
 be — , and in the fourth the ordinate would be - and 
 the abscissa + . Thus the position of P can be rigidly 
 defined. It is usual to describe the abscissa by the 
 letter x and the ordinate by y, so that taking P as in 
 the diagram PN = OM=a;; PM = ON=2/. 
 
 The point whose co-ordinates are x and y is simply 
 denoted by (a;, y) . When a point is given, and its co- 
 ordinates are consequently known, the known quantities 
 are generally represented by a,b, etc. 
 
 Thus, if x = a and y = b, to find the point we measure 
 off M = a divisions of the scale in which a is given, 
 and N = b divisions of the same scale, then drawing 
 MP parallel to OY and NP parallel to OX. The 
 •intersection of the lines denotes the position of the 
 point. The following numerical examples give points 
 in each quadrant (Figs. 76, 77, 78, and 79) : 
 
 1. 
 
 x = 
 
 3 
 
 y = 
 
 4 
 
 2. 
 
 x = 
 
 - 3 
 
 y = 
 
 4 
 
 3. 
 
 X = 
 
 - 3 
 
 y = 
 
 - 4 
 
 4. 
 
 X = 
 
 3 
 
 y = 
 
 - 4 
 
 This representation of the position of a point in a 
 figure drawn to scale is called plotting or constructing 
 the point. We often, however, know something about 
 points, but not enough to determine them. In this 
 case the point can take any one of a series of positions, 
 and yet its co-ordinates satisfy the conditions imposed 
 upon them. Thus let x = 0. This tells us simply that 
 
 p 
 
 2 
 
 
 1 
 
 i ' 
 
 x 
 
 Y, 
 Fig. 77. 
 
 Y, 
 
 Fig. 78. 
 
 X! 
 
 ■i — r 
 4- 
 
 Y! 
 
 Fig. 79. 
 
 somewhere on X. If x = a, it means that the point 
 P is distant a unit from Y, and that it is somewhere 
 in a line drawn parallel to Y, distant a units from it. 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 39 
 
 Again, if x = y, we see that the co-ordinates of P are 
 always equal, and of the same sign. The point to 
 satisfy these conditions is anywhere on the line AB, 
 which bisects the angle XOY, Fig. 80, and no 
 point not on this line can fulfil them. If 
 the point lies on the part of the line in the angle 
 XOY, its co-ordinates are equal and positive ; if on the 
 part in the angle X x Y lf they are equal and negative. 
 
 Fig. 80. 
 
 This series of positions which a point can occupy while 
 its co-ordinates satisfy a given equation, is called the 
 locus of that equation. The word curve is used here to 
 indicate any geometric locus, whether a curve, a 
 straight line, or an isolated point. Consider a little 
 further Fig. 80. Each point in A B is equally 
 distant from O X and O Y u a property which (remem- 
 bering the distinction between positive and negative 
 directions) no other point in the plane possesses. Now, 
 the distances of a point from Y and X are the 
 co-ordinates x and y of that point, and, therefore, the 
 equation x = y expresses algebraically the distinctive 
 property of points on A B, or, as we may say, of the 
 line A B. This equation, therefore, is called the equa- 
 tion o/AB. Hence we define the equation of a curve 
 as : 
 
 The equation of a curve is the expression in an 
 equation of the relation which exists between the co- 
 ordinates of every point of that curve, and of no other 
 points. 
 
 This equation expresses some law which governs the 
 changes of the co-ordinates of P, and controls the part 
 of P in the plane: hence a curve that is drawn at 
 random is not the locus of any equation, for, to repeat, 
 the locus of the equation must be a curve governed by 
 some law. We have then two cases : 
 
 I. A curve governed by some law, which curve has 
 its corresponding equation. 
 
 II. A curve not controlled by any law, which curve 
 has no corresponding equation. 
 
 When the curve is the locus of an equation, if the 
 equation is given, we can construct or plot the curve 
 by determining the position of a number of points in 
 the curve, and joining these points, or we can experi- 
 mentally determine a number of points, and thus plot 
 the curve, and, if any, obtain the corresponding 
 equation. 
 
 Let us plot the curve whose equation is x - y + 2 = 0. 
 
 If we choose any arbitrary value for x, we can find a 
 
 value for y, which, together with the chosen value of x, 
 will satisfy the equation. These values — viz., the 
 arbitrarily chosen one and the one found — are therefore 
 the co-ordinates of one point of the locus of the equa- 
 tion. A second chosen value for x enables us to find 
 another value of y, and these are the co-ordinates of 
 another point of the locus of the equation, and so on. 
 If we choose x = 1, then y = 3, in order that x - y + 2 = 0, 
 therefore (1, 3) is a point of the required locus. Again, 
 if x = 2, then y = 4, and (2, 4) is a point of the required 
 locus. Similarly (4, 6), (8, 10), (0,2), (-1, 1), (-2, 0), 
 (-3,-1), (-5,-3), (—8,-6), are points of the 
 locus. Plotting these points, we see, Fig. 81, they all lie 
 in a straight line, and from this we infer that the equa- 
 tion x — y + 2 = represents a straight line, which 
 cuts O Y two units above O, and O X two units to the 
 left of O. It saves a good deal of time and trouble if 
 ruled paper, especially prepared for this purpose, is 
 used for plotting work. 
 
 Fig. 81. 
 
 Let us plot the curve whose equation is y 2 = ix. In 
 this equation 
 
 if x = y = 
 
 * = i V = ± 1 
 
 * = | J=t s / 2=+ 1-41 
 x = I y = ± 2 
 x = 2 y = ± 2 Ji = + 2-82 
 x = 3 y = ± 2 s'3 = + 3-46 
 x = 4 y = + 4 
 3 = 5 y = ± 2 Jb = + 4-48 
 x = 6 y = ± 2 JQ = ± 4-90 
 
 and so on. 
 
 If we choose a negative value for x, we obtain an 
 
 imaginary value for y, which leads to the conclusion 
 
 that x has no negative value — that is, that there are no 
 
 points on this locus with negative abscissae, no points, 
 
 to the left of O Y. Again, each abscissa has 
 
40 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 two ordinates, each of which satisfies the equation, from 
 which we infer that from every point above X there 
 is a point symmetrically situated below X. Further, 
 it is evident that as x increases so also y increases. 
 There can he no limit, therefore, to the extent of the 
 curve to the right of Y. Fig. 82 shows this curve. 
 
 Fig. 82. 
 
 We have seen that straight lines can he used to 
 represent quantities — that if a line lin. long represents 
 1 bushel of corn a line 10in. long will represent 10 
 bushels of corn ; so if a line 1 unit (any unit on any 
 selected scale) represent 1 ohm resistance, then a line 
 n units will represent n ohms resistance ; or if 1 unit 
 represents 1 volt, or 1 ampere, n units will represent n 
 volts or n amperes. Thus, lengths upon the co-ordi- 
 nates are made to represent ohms, volts, amperes, 
 revolutions of armature, ampere-turns, resulting mag- 
 netism, or whatever is required for the investigation in 
 hand. The consideration of a few simple examples will 
 enable the reader to see how useful this kind of repre- 
 sentation can be made. Let X, Fig. 83, represent 
 
 Fig. 83. 
 
 resistance, and Y electrical pressure or electromotive 
 force. Let the end, X, of the resistance, X, be con- 
 nected to earth, so that at the point X we get zero 
 pressure. Join Y X. In X take any point, A, and 
 through A draw A B parallel to Y ; then A B will 
 represent the electrical pressure at the point A. It can 
 be shown that Y B bears the same proportion to B X 
 as A does to X ; so that if we know the fall of 
 electrical pressure between Y B we know the resistance 
 A, if we know that of X. Suppose Y = 50 volts 
 and X = 500 ohms. By drawing through B the line 
 B C, parallel to O X, on to O Y, and measuring Y C, 
 we find Y C is 2-5ths of O Y, and therefore represents a 
 loss of 20 volts ; O A therefore represents 2-5ths of 500 
 ohms, or 200 ohms. The resistance represented by 
 
 A X = 300 ohms, and the pressure at A = 30 volts. 
 Similar measurements might be made at any other 
 point in O X or O Y. 
 
 Suppose we desire to show the relation between 
 armature speed and electrical pressure. Let OX, 
 Fig. 84, represent the revolutions per minute, and O Y 
 the pressure. The curve shows clearly the relation 
 
 
 
 
 2o8 
 
 
 zoo. 
 
 
 
 
 
 
 
 us s 
 
 
 
 
 to\/' 
 
 
 
 
 /'• 
 
 
 
 
 
 
 
 o- 
 
 / . , . — . — Ll_ 
 
 
 __. 1 — 
 
 
 500 (OOO -A- 
 
 o 
 
 Fig. 84. 
 
 between speed and pressure. Theoretically, the speed- 
 pressure curve should be a straight line provided the 
 field through which the armature rotates is constant, 
 but the curve shows that this is not quite so, and it be- 
 comes important to interpret this difference between 
 theory and practice. In this case the difference is 
 caused because as the armature current becomes 
 stronger its interactions with the field are different, and 
 currents of different strengths act unequally on the 
 intensity of the field. Perhaps the curve which is 
 
 200 
 
 10 15 
 
 CURRENT 
 
 Fig. 85. 
 
 called the characteristic curve possesses the most 
 importance to the student of dynamos. Let the ma- 
 chine under consideration be a series-wound machine, 
 run at constant speed with a load varying from the 
 largest to the smallest that the machine will carry 
 without injury. Measure for each resistance the 
 difference of pressure or potential at the terminals, and 
 the current. Let the currents be represented along 
 O X, and the pressures along O Y. The intersection of 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 C ) 
 
 H 
 3 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 M 
 
 86 B.P. REC^ENZAUN MOTOR — ^-SIZB. 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 41 
 
 the horizontal and vertical lines which represent the 
 current and the pressure give a series of points, which 
 when joined give the curve called the characteristic of 
 the machine. In Fig. 85, let A B C represent such a 
 curve, and the interpretation of its peculiarities is the 
 object of the construction. The curve shows that when 
 the currents are small the pressure increases rapidly, 
 and as this part of the curve is almost or quite a straight 
 line, the pressure and current increase in the same 
 proportions. When the current has increased to a 
 certain value in the diagram indicated at or about the 
 point A, the pressure increases less rapidly, and soon 
 reaches its maximum limit, past which it will probably 
 slightly decrease. Now the practical value of having 
 such a curve as this depends upon the ability to inter- 
 pret. The complete interpretation can only be under- 
 stood after a course of study of the dynamo, but 
 sufficient can perhaps be said here to indicate the 
 enormous value of such curves, and to induce the 
 student to closely examine all that may be said about 
 them. The curve given is that known as the external 
 characteristic of a series dynamo. Tt has already been 
 mentioned that the electrical pressure of the machine 
 increases pretty uniformly according to the number of 
 revolutions. Here we have assumed the revolutions to 
 be constant. Part of the pressure generated is used to 
 overcome the resistance of the armature itself, tbe 
 remaining part — viz., the difference of pressure or the 
 difference of potential at the terminals of the dynamo — 
 is that available for use in the external circuit. The 
 characteristic curve can take into account the total 
 pressure, or, that available in the external circuit only. 
 From the curve given we should infer, putting other 
 considerations aside, that this machine would not be 
 run to the best advantage if run for pressure or cur- 
 rent less than those indicated by point A; neither 
 should it be run much, if any, beyond point B, as that 
 is the maximum point for pressure. Further, for value 
 of currents between A and C, the machine gives nearly 
 constant pressure for different loads. Of course, the 
 nearer the curve A C is to a straight line the more con- 
 stant the pressure. If, therefore, it is desired to con- 
 struct a machine to give a constant pressure, it is 
 necessary to have one that gives a characteristic with 
 AC as nearly a straight line as possible. 
 
 How comes there to be such a sharp bend in the 
 curve at A ? Before answering this question let us see 
 how the total characteristic is obtained. 
 
 Assume, though of course in an actual case the 
 resistance would be measured, the resistance of the 
 armature to be '5 ohm. 
 
 To drive 5 amperes through '5 ohm requires 2"5 volts. 
 „ 10 „ „ "5 „ „ 5-0 „ 
 
 „ 20 „ „ -5 „ „ 10-0 „ 
 
 From O, Fig. 86, draw the line OI at an angle with 
 O X, whose tangent = | — that is, 
 _numeri cal value of volts = in thig case 2J> QJ , 5 
 numerical value of amperes 5 10 
 
 The ordinates of this line represent the pres- 
 
 sure or electromotive force or volts lost in the 
 armature. Add the ordinates of O I to the corres- 
 ponding ordinates of OABC, and we get a series 
 of points which, when connected, give the curve O T. 
 Then the length of ordinates from O X to O T give the 
 total voltage induced, or the curve O T may be called 
 the total characteristic. In practice account must be 
 taken of the rise in temperature causing an increased 
 resistance in the armature, so that the real loss in volts 
 is greater than is depicted in this diagram. 
 
 200 
 
 C 5 TO" 15 
 
 CURRE NT 
 Fig. 86. 
 
 To return to the question, How comes the bend in 
 the curve at A? It has been stated that a current 
 through a coil around a core of soft iron evokes mag- 
 netic lines or loops of force. We have seen, p. 23, that 
 
 , T , ,., ampere-turns 
 
 Number of loops = — £ ; :-, 
 
 resistance oi magnetic circuit 
 
 that is, if the resistance of the magnetic circuit remained 
 constant, the number of loops would vary directly as the 
 ampere-turns, but the resistance is not constant. It 
 increases with the magnetisation, and after a time the 
 sending of additional current through the coil does not 
 increase the number of loops ot force through the iron 
 core. When the magnetism of the iron core does not 
 respond to the increase of current, the iron is said to be 
 saturated. In a series-wound machine the magnetic 
 field, in which the armature revolves, increases rapidly 
 in strength till the iron of the field-magnets is saturated. 
 The portion of the curve OA shows the condition 
 while the core is not saturated. An increase of the 
 current in the series coils after A is reached — that is, 
 when the iron is saturated — has little effect in increas- 
 ing the magnetism, and therefore the pressure. 
 
 Just as we have drawn the characteristic between 
 pressure and current, in the same way characteristics 
 
42 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 may be drawn between the external resistance and the 
 pressure, or between resistance and current, or between 
 the speed and current, or between the speed and pres- 
 sure. In general, however, characteristics involving 
 changes in speed are less useful, though often very in- 
 structive, than the other characteristics, for most engi- 
 neers consider the question of speed in its relation to the 
 mechanical, rather than the electrical, properties of the 
 machine. The student cannot make himself too 
 familiar with these curves, and should take every 
 possible opportunity of plotting them for himself, and 
 compare the curves taken from one machine with those 
 taken from other similar machines. The characteristics 
 must be drawn or reduced to the same scale before 
 being compared. A characteristic curve can be obtained 
 for any two varying quantities which depend upon 
 each other, and for only two. 
 
 current. Putting in a very large resistance and gradu- 
 ally decreasing it, we get more and more current with 
 the pressure falling. The first part of the curve 
 E AB C O is almost a straight line, and during this part 
 of the curve the fall of pressure is proportional to the 
 current as the resistance decreases. At A the pressure 
 commences to fall more rapidly, and at B still more 
 rapidly, till at C the maximum current O Cj is reached. 
 If we decrease the resistance past C continuously till 
 the machine is short-circuited, the pressure and the 
 current both fall till, when short-circuited, both are at 
 zero. The curve shows graphically what is known tc 
 all after a moment's thought, that maximum pressure 
 is when the machine is run on open circuit— that is, 
 when all the current goes to excite the field-magnets, 
 and no current to the external or working circuit; 
 that when no current goes into the shunt circuit, 
 there is no field, and consequently no pressure 
 and no current. This happens when the machine 
 
 50 100 ISO 
 
 CURRENT 
 
 Fig. 87, 
 
 c, 
 
 CURRENT 
 
 Fig. 88. 
 
 Of course the curves obtained from a series machine 
 are altogether different from those obtained from a 
 shunt or a compound machine. To obtain similar 
 diagrams to those above given in a series machine, the 
 armature speed must be kept constant. First, with 
 terminals open, there is no current in the external cir- 
 cuit, therefore none in the field-magnet coils, therefore 
 no magnetisation and no magnetic field. Connecting the 
 terminals by a large resistance, a small current is 
 obtained ; reducing this resistance, we get a larger and 
 a larger current, and a stronger and stronger magnetic 
 field, till the poini of saturation is reached. With a 
 shunt machine, on the contrary, when the terminals 
 are open, the whole current goes through the shunt 
 coils, the field-magnets are excited to their highest point, 
 and we get maximum pressure. 
 
 The general characteristic for current and pressure 
 of a shunt machine for different resistances is shown in 
 Fig. 87. Let X represent current, and Y pressure. 
 Starting with open resistance between the terminals, 
 the machine gives its maximum pressure E and no 
 
 is short-circuited. The curve shows there is a 
 maximum current = G lt and if the armature will 
 stand this maximum current no alteration of the 
 external circuit will injure the machine. In practice, 
 however, no one runs a shunt machine between C and 
 — it would be very uneconomical. The best machines 
 would be limited to a part of EAB. If any part of 
 EAB was perfectly straight, then the current might 
 be varied at will within the limits of that straight part 
 without' alteration of pressure. The inclination of 
 E A B is due to the resistance of the armature, some 
 part of the pressure being used to overcome this resist- 
 ance. If the armature had no resistance a portion of 
 the curve, say E A, would be horizontal, and then we 
 could have current varying, as mentioned before, with 
 constant pressure. But this part of the subject will be 
 fully treated hereafter. It will be sufficient to say here 
 that the difficulty indicated was avoided or neutralised 
 by the introduction of compound winding, or of the 
 compound dynamo. The characteristic of the com- 
 pound machine should be a combination of the charac- 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 43 
 
 teristios — the resultant characteristic, we might call it, 
 of the series and shunt machines. 
 
 In the characteristic of the series machine the pressure 
 rises in a regular manner, in the characteristic of the 
 shunt machine it falls in a regular manner — hence, by 
 a proper combination of the two the action may be to 
 keep the pressure constant. In other words, we have only 
 to design the windings of the shunt and series coils 
 so that the action due to the series coils in increasing 
 the magnetism of the field-magnets shall be just equal 
 to the action of the shunt coil in diminishing that 
 magnetisation. The resultant action will give con- 
 stant pressure, or potential, within certain limits. 
 We can show what we mean by a diagram, Fig. 88. 
 Let B represent the rise of pressure due to the series 
 coils and E A the fall of pressure due to the shunt coils. 
 
 If B rises in the exact proportion as E A falls, the 
 pressure will be represented by the straight horizontal 
 line E Aj, and up to that limit the current can vary 
 without variation of pressure. It can be shown that a 
 compound machine should only be used in such parts 
 of E Aj B as are nearly straight lines ; also that the 
 bend of the curve E Aj at A T is due principally to satura- 
 tion. Thus we learn that a compound machine should 
 be run below the saturation point. Sufficient has now 
 been said to prove our contention that the whole sub- 
 ject of electro-graphics is of the utmost importance, 
 and although the reader may start without much 
 mathematical skill, there is no reason why, with care- 
 ful attention, he should not be able to plot and interpret 
 diagrams to such an extent as to be of very great 
 service to him in practical work. 
 
 CHAPTER VI. 
 
 &s* 
 
 THE CENTEAL STATION. 
 
 HE very name central station indicates its 
 importance. Here the electrical engineer's 
 work is collected, and here most of his 
 troubles begin. The electrical energy which 
 has to be distributed over a large or a small 
 area must be generated in large quantities,' 
 and at a convenient point, in order to make its com- 
 mercial production possible. This means the collection 
 of large masses of machinery upon that spot, and the 
 provision for their due maintenance and work. 
 
 Position, Initial Cost, and Cost of Maintenance. 
 
 In most cases the cost of maintenance, which, for 
 the moment, may be taken not merely to imply the 
 keeping of the machinery in repair and in working 
 order, but to include the cost of the fuel, water, and 
 other materials required in the daily work, is of far 
 greater importance than initial cost. In choosing a 
 site, therefore, for a central station, it is necessary to 
 consider the position in regard to the cost of obtaining 
 the requisite daily supplies. Thus one site may cost a 
 thousand pounds more than another, yet by utilising 
 the dearer site a saving of a hundred pounds a year on 
 cartage may be effected. It will be better to pay the 
 higher sum once than to pay the higher cartage, which 
 affects the working expenses as long as the station 
 runs. 
 
 Another consideration of great importance in the 
 selection of a site is the room for extension. Dur- 
 ing the discussion of a recent paper on central stations, 
 a speaker remarked, " I think I speak advisedly when 
 I say that almost every central station man through- 
 out the country has, upon finding his station too 
 
 small, enlarged it to meet the requirements of all time, 
 got it quite completed, and found it was still too 
 small." Again, the position selected will depend 
 largely upon the system adopted, whether the system 
 be the generation of high pressure or the generation of 
 low pressure. It will be seen, then, that the selection 
 of a site is not so simple a proceeding as it may some- 
 times seem. It is not the seeking of a vacant plot of 
 ground in a populous part. A vacant plot in the very 
 midst of a population anxious to obtain light and power 
 may not be so economical as a plot at some distance, 
 or on one side of the area of distribution. The selec- 
 tion of a site for a large central station is, however, 
 more of a financial problem of a complex character than 
 a purely engineering problem. 
 
 The Building. 
 The style of building will depend upon the part of 
 the world where it is erected, and the available building 
 materials in the district. It is not our intention to 
 enter into any discussion as to the value of brick, stone, 
 iron, or wood, or thickness or height of walls, but, like 
 Franklin, being fully imbued with the idea that the 
 foundation is the most important part of the building, 
 a subsequent chapter is devoted to a brief survey of the 
 factors that enter into the making of a good foundation. 
 As steam will probably constitute the principal power 
 for driving dynamos, and as the best steam engines 
 will be constant sources of worry and trouble if their 
 foundations go amiss, although the engines themselves 
 may be excellent in every sense of the word, the wise 
 engineer will pay the most careful attention to all that 
 concerns his foundations. Then, again, comes into 
 consideration the system employed and the insulation 
 of the dynamos. Surely too little attention has been 
 
44 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 paid to the use of insulating material in foundations, vicinity of residential houses, has to be considered, as 
 vioration, and especially when the station is in the has noise, and the transmission of both noise and 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 45 
 
 S 
 
 
 I 
 
 -d 
 
 a 
 
 W 
 
 vibration depends largely upon 
 unsuitability of the foundations, 
 
 the suitability or 
 Of course the walls 
 
 and roof must be erected with due consideration of the 
 duties they have to fulfil, 
 
46 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 ffio. 91,— Mentmore. — Section through Engine Room. 
 
 Fi<j. 92.— Mentmore.— Section through Boiler Room, 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 47 
 
 The arrangement of the machinery in the building is 
 a test of an engineer's abilities. The plans of one engi- 
 neer will be far preferable to those of another under 
 the same conditions. There is no royal road to success. 
 The younger generation must learn from the experi- 
 ence of the older — hence too much cannot be known of 
 how others have solved the problem as presented to 
 them. The conditions of each separate installation 
 will have to be studied and arranged for. So far as we 
 can we shall aid the engineer by giving examples of 
 installation arrangements over as wide and varied a 
 field as possible. We may here refer to the arrange- 
 ments made at one or two typical installations to show 
 the great divergency of requirements. One station 
 mentioned is a large central station, the other is a 
 small isolated station for a gentleman's mansion. The 
 Edison Brooklyn station may be termed a repre- 
 
 sentative station for lighting large towns. The 
 engines in this station are upon the ground floor, 
 the dynamos on the floor above. Of the latter 
 there are 24, each with a capacity of 750 amperes and 
 140 volts. Through the centre of the dynamo-room 
 runs an electrical gallery where the electrician controls 
 the switch arrangements. The longitudinal section 
 and the plan, Plates N and 0, show clearly the whole 
 arrangement. 
 
 Figs. 89, 90, 91 and 92 show the arrangements at the 
 Earl of Bosebery's mansion at Mentmore. This is a 
 private installation, carried out by Messrs. B. Verity and 
 Sons, and is similar to many other stations erected for 
 supplying private mansions or isolated buildings. The 
 illustrations show plans and sections of boiler and 
 engine rooms. The measurements are given, so that 
 further explanation is unnecessary. 
 
 CHAPTER VII. 
 
 FOUNDATIONS CONSIDEBED PBAC- 
 TICALLY. 
 
 H H'. importance of substantial and enduring 
 foundations to heavy buildings, and es- 
 pecially such buildings as are to contain 
 heavy machinery, cannot be overstated; yet, 
 in practice, frequent failures occur because 
 due attention is not paid to this matter. 
 The precise relation of the loaded superstructure to the 
 nature and extent of the stability of the substratum 
 should be the fundamental principle in determining the 
 most suitable class of design of foundation to employ, 
 the character of the provisions and precautionary 
 measures which should be adopted, and as well should 
 command the requisite degree of attention to all im- 
 portant practical details during the execution of the 
 work which would ensure its sufficiency in all respects, 
 and hence its permanency. Foundations should not 
 be liable to be affected injuriously by frost, water, air, 
 weather, or other source of disturbing action, whether 
 mechanical or chemical in its nature, and consequently 
 should be immovable by any other conditions to which 
 they will be subject in the ordinary course of events. 
 Displacement of foundations may occur vertically or 
 horizontally — the one by compressive yielding of the 
 subsoil, in which, however, there is more or less lateral 
 displacement amongst the substances composing the 
 substrata ; the other by the sliding over one another of 
 the layers, which lie, obliquely of different bands of over- 
 lying, interlying, or parting soils. The natural soil in 
 situ is not always substantial, and is very frequently 
 deceptive and irregular in thickness, compactness, co- 
 
 hesion, etc., at different parts of a foundation site, and 
 especially so when in the vicinity of existing or old 
 waters of rivers, morasses, estuaries, lakes, etc., which 
 may have left indications of its character on the sur- 
 face, or all may be concealed in subterraneous forms 
 only, and therefore be the more likely to prove danger- 
 ously insidious. 
 
 All structures built of masonry, in stone or brick hori- 
 zontal courses, laid with yielding mortar bed-joints, must 
 settle to some extent in height depending upon the 
 frequency and thickness of the bed-joints occurring in 
 the work. And likewise all soils are more or less com- 
 pressible. Hence, in all cases subsidence of the soil to 
 some extent is unavoidable, but by judicious treatment 
 it can be greatly minimised and equalised in both 
 cases, instead of being exaggerated and distorted, as 
 frequently happens by neglect of intelligent treatment 
 and adaptation. Foundations are either (1) natural, or 
 (2) artificial, considered in a restricted sense of the 
 term, which implies in one case the natural virgin solid 
 stratum of soil, rock, etc., in situ, which underlie the 
 expanded footing base, and which may of itself be 
 of sufficient depth and compactness to support the 
 building when ultimately loaded if it be a ware or 
 store house, or in the case of mills, factories, work- 
 shops, etc., with full equipment of going machinery. 
 The methods and expedients resorted to for the pur- 
 pose of rendering permanent the stability of the natural 
 soil by preventing lateral movement or escape of any 
 part of the supporting substance when of a semi-fluid 
 character, such as sands, clays, etc., and by adequately 
 increasing the supporting area when the strata consists 
 of soft or loose materials by means of a monolithic bed 
 
48 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 of concrete, must be studied in the case of natural 
 foundations. 
 
 In the other case examination should be made of all 
 the devices which have for their object to provide, by 
 artificial means, a reinforcement of the natural bearing 
 capacity of soft soil by consolidation, by studding it with 
 piles, or by providing skeleton supports, as timber or 
 iron (screw) piles, sand or concrete piers, penetrating 
 the soft soil to reach the lower solid strata, or con- 
 structing gridirons of iron or steel beams embedded 
 in concrete. 
 
 The first consideration must be drainage. In districts 
 where there is no town drainage system available with 
 which to connect and relieve the foundation excava- 
 tions and trenches, of water accumulations from rain, 
 or from water surcharging adjacent lands, which 
 should be collected in a receptacle in a convenient 
 position for being discharged, etc., by " gripes " (small 
 channels branching in different directions), adequate 
 provision should be made in advance, either by tempo- 
 rary outlets, pumps, etc., as any flooding or submerging 
 of the foundation area or trenches should be avoided. 
 Surface waters should be prevented 'from entering the 
 excavations and trenches. The foundation strata for 
 all important structures should be thoroughly tested 
 either by deep trial pits at different parts of the foun- 
 dation site, or by wrench auger borings, by which 
 samples of the various soils passed through are brought 
 up for examination at convenient intervals. 
 
 Foundation strata are sometimes tested by applying 
 to a limited area of the footing base a sufficient unit 
 pressure per square foot, by means of loading a frame- 
 work with heavy weights of iron, stone, etc., to 
 adequately represent the ultimate unit load to be • 
 imposed by the building, etc., with exact observations 
 and notes of the time-measured subsidences which it 
 produces. It should be observed that an ordinary 
 rectangular arrangement of testing frame, with four 
 supporting areas at the four corners, is liable to give 
 confusing results if the points of application, and 
 amounts of the weights, and the leverage of their action 
 at all times upon the supports, produce not equal 
 moments. As this would be practically unavoidable in 
 case of undue subsidence of any one of the four 
 supports, which would throw an uncertain surplus pro- 
 portion of the load upon the two diagonals of the 
 remaining three supports, it is preferable to have only 
 three supports in triangular arrangement. 
 
 The site of foundations may, by these means, be dis- 
 covered to require precautionary treatment in the 
 manner already alluded to, or protection from the 
 action of a high-water level of water-bearing beds, or 
 running waters, underground springs, which may be 
 the outlets of natural reservoirs of a distant locality 
 which may retain water above the level of the outfall 
 in proportion as the capillary action and the friction 
 amongst the particles in passing through the interstices 
 of the permeable strata exceeds the hydrostatic pres- 
 sure of the accumulating rainfall, whereby several feet 
 deep of water may occur at certain seasons. In the 
 
 body of extensive hills of permeable surface strata, 
 as sand, oolitic freestone, fissured limestone, chalk, 
 etc., natural reservoirs of water are heaped up, as it were, 
 above the level of outlet, having a varying curvilinear 
 surface rising from the outlet, and accordingly there are 
 cases of wells occurring on chalk hills, many miles from 
 valleys, which have a water level 40ft. or 50ft. higher in 
 spring than in autumn. Such conditions prevailing on 
 a site might prove a matter of very serious import, as 
 may readily be imagined. The presence of quicksands, 
 more or less saturated, when underlying the surface 
 strata, are thus discoverable. Stagnant, intermittent, or 
 permanently waterlogged areas adjacent to the site 
 may by intermediate geological features well up from 
 time to time, as " land tides," by the action of which the 
 supporting or lateral materials, which may include 
 water itself, naturally present, of the foundations may 
 be disturbed or dissipated. Such cases require special 
 expedients, such as sheet-piling, puddled clay trenches, 
 etc., for effectually retaining the supporting materials 
 immovable. Sheet-piling alone is insufficient for this 
 purpose when the substratum is "waterlogged." All 
 rocks and earths being more or less porous are liable 
 to be permeated by water from rain, etc. In some 
 geological local conditions intermittent "welling up" 
 may occur at a considerable distance from the im- 
 mediate source of the supply or " catchment basin." 
 Intervening " dykes," " faults," etc., may either, accord- 
 ing to circumstances, have the effect of precipitating or 
 of temporarily preventing the access of distant ground 
 waters. Hill slopes and valleys are thus frequently 
 affected when the strata dip towards the site. A know- 
 ledge of the surface geology of the district, the strike or 
 trend and inclination of dip, etc., of the strata, their 
 superposition, the kinds and character of rocks and 
 other soils to be encountered in excavating foundation 
 trenches, cellars, etc., their outcrop, water-bearing 
 capacity, and other prevailing accidental and incidental 
 circumstances which have more or less effect upon the 
 permanence and stability of the work should be ob- 
 tained from proper sources, or by the methods known 
 as " field geology." 
 
 Ridges and escarpments consist mainly or partly of 
 water-bearing beds, of limestone, chalk, sandstone, 
 etc. The softer clays being more readily denuded, 
 gravitate to the lower grounds. Strata generally dip 
 towards and pass underneath the higher grounds ; the 
 beds are not often found quite horizontal. The 
 anticlinal strata, like A> are more readily denuded than 
 the synclinals, like letter V. hence the latter remain 
 while the higher intervening anticlinals are washed 
 away, leaving valleys occupying the sites of the pre- 
 existing hills, and escarpments of trough-like beds 
 underneath the present ridge. "Water levels will rise 
 in them to the top of the exposed outcrop or escarp- 
 ment of the impervious strata, and escape in perennial 
 or intermittent springs, according to the extent and 
 capacity of the underground reservoir formed by the 
 impervious bed. It is preferable to divert springs and 
 water channels than to dam them out from founda- 
 
N 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 WET* 
 
Practical electrical ENc-iNEERiNG. 
 
 49 
 
 tions. Damming out of springs or surcharging waters 
 from underground basements and deep cellars and areas 
 requires very carefully executed puddled walls and 
 bottom without any breaks, or walls and basement or 
 cellar floors should have an impervious coat of good 
 hydraulic cement or asphalte, etc. Hence the same 
 geological knowledge that is required to obtain water 
 supply by easiest means is also available for devising the 
 readiest expedients for discovering its presence or dis- 
 persing it. In the reverse geological conditions, an 
 impervious surface sheds rain and surface water to, and 
 received by, lower pervious surface beds . When pervious 
 beds are sandwiched between impervious beds, they are 
 in a condition to act as a syphon in conveying under- 
 ground waters to considerable distances underneath 
 valleys and undulating ground when it is required in 
 connection with the excavation of foundation trenches, 
 cellars, or deep areas, etc., to pump the water from 
 water-logged ground in the neighbourhood of existing 
 buildings. It is unsafe in most cases to use a steam 
 pump to keep the water low to permit of work of con- 
 tinued excavation, or of construction of concrete, stone 
 or brick masonry, etc., as the sudden withdrawal of a 
 large quantity of water may deprive neighbouring 
 foundations of its supporting property, and thereby 
 cause injurious settlements of these buildings. 
 
 From a failure to acquire some or all of such pre- 
 liminary information, a sound and capable substratum 
 has been frequently known to be rendered insecure by 
 blindly cutting the excavations, trenches, etc., too deep 
 into a stratum already sufficiently limited in thickness. 
 
 Characteristics of Sites as regards Soils, Boeks, ete. 
 
 Differences in the character of the soils of the sub- 
 jacent strata in foundations produce corresponding 
 differences in the amount of resulting settlements of 
 the future building, and, hence, if they vary in different 
 parts of the site instead of being of homogeneous com- 
 position, and of equal thickness and compactness 
 throughout, which is rare in extensive and occasionally 
 so even in moderate sized sites, proper expedients must 
 be adapted to improve its bearing power in those parts 
 which are the more compressible, and thus ensure equal 
 stability throughout, or by a corresponding enlarge- 
 ment of footing areas, or by deepening the trenches to 
 reach a more solidified condition in the underlying 
 layers when such is to be found, by drainage of sur- 
 charging waters, by spreading layers of sand, gravel, 
 broken stone, concrete, etc., of sufficient thickness, by 
 ramming, pinning, or grouting in the case of coarse, 
 loose gravel substrata. 
 
 The earths met with may have characteristics which 
 render them of commercial value in certain localities, or 
 they may be of present utility in connection with the 
 building, or, on the other hand, they may have qualities 
 to be guarded against in building operations, or such as 
 may render a site for certain purposes extremely un- 
 desirable. Eock seldom covers uniformly the entire 
 area of a foundation site. Sand, gravel, clay, silt, mud, 
 etc., may occupy the remainder ; the rock may ledge or 
 
 berme up at one side or end of a site, or even under 
 one side of some of the walls only. The surface of the 
 rock may have large cavities, sand or potholes, or 
 the strata may be inclined, upturned, inverted, dislo- 
 cated, or may have large deep fissures filled with 
 loose parting bands or beds of sand, gravel, and clay, 
 which the superincumbent weight of soil, or the 
 presence of water, or the action of the weather, 
 may displace in the course of time. 
 
 In the occurrence of successive beds, or layers of 
 soils, the larger stones, or particles of matter, are 
 usually found in the lowermost positions in the series ; 
 thus the successive beds of boulders, gravel, sand, silt, 
 mud usually occur in the order named when enumerated 
 upwards. The beds are thus formed by the action of 
 water, or sometimes by that of wind, bat in the latter 
 case are less regular. In the former mode there is a 
 considerable approach to parallelism amongst the suc- 
 cessive layers, but with less exactness than occurs in 
 the cleavage planes of sedimentary deposits in most 
 cases, when they are said to be " conformable." 
 Strata deposited by water contain fossil remains of 
 animals and plants embedded in them while they were 
 soft. Metamorphism by heat, chemical action, or 
 pressure, generally destroys those organic remains, and 
 produces crystalline structure, but it leaves the stratifi- 
 cation undisturbed. Stratum may thin out in places 
 where there has been a previous deposit at a higher 
 level. As a rule, the lowest strata are the oldest, 
 except when they have been tilted over, inverted, etc., 
 in limited areas by great natural convulsive agencies. 
 
 In situations where large cavities, fissures, ravines, 
 in rock sites, etc , occur, sound hard-setting concrete 
 should be freely used in a manner to ensure permanency 
 and avoid future subsidences when the full weight of 
 structure is imposed upon it. All kinds of rock, or, 
 indeed, the same kind of rock in different situations in 
 a quarry or districts, are not equally permanent. 
 
 When the laminae or beds lie obliquely to each other, 
 as when produced by the action of tidal currents, it is 
 called " false bedding "; or when the strata are " un- 
 conformable," i.e., the lower strata have been bodily 
 upturned or overturned, and succeeding strata are 
 deposited upon the upturned edges, with each of the 
 successive strata overlapping the one beneath, whereby 
 special conditions endangering instability may prevail. 
 When there occur large cavities in the surface of rocks 
 it is difficult to fill it up as solidly as the rock itself, so 
 as to proceed with the masonry work, etc., without 
 waiting until it is set perfectly hard. In favourable 
 conditions arching can be usefully resorted to ; but all 
 such preparatory work should be done as far in advance 
 as possible to ensure the least subsidence. The most 
 recent and superficial strata deposits, which consist of 
 sands, gravels, clays, boulder clay, silt, mud, etc., which 
 generally overlie the igneous and stratified rocks, are 
 composed of rock materials more or less ground up, 
 weathered, and decomposed. It maybe either alluvial, 
 as produced by the action of existing or recent waters, 
 wind, rain, and weather, or may be detritus, or northern 
 
50 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 glacial drift, to which the name diluvium was formerly 
 applied on the assumption that it was produced by the 
 historic deluge, but is now referred to subsequent 
 agencies of the glacial period, when the northern hemi- 
 sphere, down to between the 40th and 50th parallels, 
 was covered with ice, or to the action of extraordinary 
 currents of water, or to submarine earthquakes, etc. 
 The drift is usually composed of mixtures of abraided 
 rocks, boulder clay, earth, stones, boulders, etc., brought 
 from the north by ice drifts, and sands, marls, loams, 
 gravels, etc., which-have been more recently deposited 
 by overflowing water. The former occupy the more ex- 
 tended districts. These drift deposits usually cover the 
 older solid rock formations which occur in alternating 
 rock layers, various in kind, thickness, and extent, but 
 always in regular if not in constant sequence. These 
 layers are rarely horizontal, but incline or dip with 
 various slopes, and may strike in the same or in 
 different cardinal directions, according to the character 
 of the disturbing agencies by which they have been 
 affected by upheaval of old sea, lake, valley, etc., beds, 
 and which may have been subsequently partly eroded 
 by floods or the depression of pre-existing mountains. 
 By such agency the (1) primary and lower strata in the 
 natural order of formation of igneous or unstratified 
 rocks, such as granite, traps (varieties of dolerite or 
 basalt, etc.), have been brought to the surface. Bocks 
 are classified according to the nature of the prevailing 
 mineral constituent, as siliceous (flinty), argillaceous 
 (clay), calcareous (carbonate of lime). They are com- 
 posed chiefly of quartz (silica and oxygen), felspar, 
 which is of several family varieties (potassium, which 
 may be replaced by sodium, or by calcium, aluminium, 
 silicon, and oxygen), and mica, a glittering substance, 
 composed of the same elements, commonly called clay 
 and flint, with magnesia and oxide of iron, but of 
 different ratios of combination, with the addition of iron 
 and magnesium. Granite proper has the minerals dis- 
 posed in granular irregular aggregation. Granites are 
 in varieties of grey and red ; that in which the minerals 
 are aggregated in more or less parallel layers, and which 
 is stratified, as gneiss. In some varieties of granites, 
 as found in Aberdeen, etc., called syenite (from the 
 islaDd Syene in Egypt), hornblende replaces the mica. 
 Hornblende is a dark crystalline substance composed 
 of flint, alumina, magnesia, and black oxide of iron. 
 Syenite does not split well. When granite contains 
 large distinct crystals of felspar, it is the porphyritic 
 variety. When quartz is disposed in parallel lines it is 
 graphic granite. Schists are composed of finer mineral 
 grains disposed in a laminated arrangement, indicating, 
 a tranquil sedimentary deposit, parallel to the lines of 
 stratification, as in slates, flagstones, etc., or may be 
 foliated, as in the crystalline variety, whereby the 
 alternating layers of different minerals of which gneiss 
 and other metamorphic schists are composed. 
 
 The sedimentary or stratified rocks are composed 
 of the fragments of pre-existing rock formations, either 
 in the form of gravel, sand, clay, etc., or these 
 materials consolidated. The more complex the com- 
 
 position of the components of minerals or of rocks, the 
 more readily do they yield to chemical destruction ; the 
 more a compound is acid the less is its chemical 
 stability. Water and meteoric influences are the great 
 agents of destruction. Minerals containing fluorine 
 undergo decomposition. 
 
 All rocks when taken from a considerable depth in 
 the quarry are highly charged with water, called by 
 masons " quarry sap," and are soft, called " green," 
 and easily worked. As the water evaporates, a 
 hard crust is formed on the surface, which resists the 
 chisel — but preserves the surface of the stone. 
 
 Gravel consists of fragments of rocks, in many 
 instances mainly consisting oi quartz, which were first 
 in large boulder masses, then pebbles smoothly rounded 
 by the action of water. Pebbles in accumulations form 
 shingle of sea and river shores and shoals. Pebbles, 
 round or angular, which are usually larger, intermixed 
 with flints, sand, clay, loam, etc., form the gravel of 
 old sea, lake, or river-raised beds, is a very desirable 
 substratum when sufficiently compact and of adequate 
 thickness. When it occurs overlying softer strata, it 
 occurs in several varieties and situations, in high or 
 low levels, valley, terrace, or plateau, and belongs 
 principally to recent and glacial formations. The high- 
 level gravels are generally coarser, and contain fewer 
 subordinate beds of sand or clay than those on lower 
 levels, which also consist of beds alternating more 
 commonly with beds of sand and clay, and sometimes 
 containing veins or lenticular beds of clay. Loose 
 coarse gravel may sometimes be solidified in a pretty 
 thick top layer by grouting. 
 
 Sand is loose aggregates or grains of quartz (silica), 
 often with minute plates of mica and green grains of 
 silicate of iron, frequently slightly coloured with iron 
 and manganese peroxides. Sand is of several varieties, 
 according to the predominance of clay (" argillaceous "), 
 quartz ("siliceous"), lime ("calcareous"), magnetic 
 iron sand, coral sand, and other varieties. The grains 
 may be fine, gritty, coarse, angular, smooth, and 
 rounded, as in quicksand lying in water, which is so 
 mobile as to afford no stability, and hence its presence 
 must be carefully sought out. In some districts a 
 leaden-coloured silt or soapstone slime, a kind of marly 
 substance, which absorbs a large amount of water, is 
 found in quicksands. It is mostly found in the vicinity 
 of lakes, rivers, and seas, and in valleys and marshes. 
 Previous to attempting to excavate quicksand, the 
 water must be pumped away from it. When sand is 
 in compact beds and not overlying a softer stratum, and 
 not liable to the action of moving water, and prevented 
 from spreading, it generally makes a good foundation. 
 Its presence may often be unsuspected in narrow 
 valleys, which may in some places be filled up with a 
 rain or weather wash from the adjacent hills, composed 
 of their sand, clay, etc., so that both the valley and 
 these hills appear to be of the same order and arrange- 
 ment of strata, with a continuous substratum of the same 
 rock, the resemblance being so close as to be mis- 
 leading, even though trial pits have been sunk, but 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 51 
 
 notwithstanding these points of resemblance the wash 
 may overlie a treacherous bed of quicksand, or, perhaps, 
 peat, silt, etc. In extensive alluvial flat areas ancient 
 river sites may exist, parts of which, in some instances, 
 have been covered over to any .depth up to even 100ft., 
 with probably only a comparatively shallow compact 
 crust overlying a deep bed of soft silt. 
 
 Clays and shale, mud, ooze, etc., are of so numerous 
 a description, each varying in composition and charac- 
 teristics — very widely varies from slate or shale to soft 
 oozing mud, but none possessing perfect compactness, 
 and all more or less alterable in elasticity and fluidity 
 by air, moisture, of which it is very retentive, etc. — that 
 it is regarded as the most treacherous and troublesome 
 of all strata for a foundation. The London blue clay 
 is no exception. It and other clays expand with such 
 enormous force as to crack and crush large scantling 
 timbers and the struts used in shoring up the sides 
 of excavations. If a compact dry bed of clay be covered 
 with concrete, and be sufficiently below the surface, it 
 may make a safe foundation. Silicate of alumina is the 
 basis of all clays. They usually contain oxide or 
 sulphide of iron, or some carbonaceous matter which 
 imparts a dark bluish-grey colour. Some clays are 
 calcareous, and contain septaria, masses of impure car- 
 bonate of lime, and sulphate of lime (selenite), as in the 
 London clay, Reading or mottled clay, and Oxford clay. 
 The pure clay is the China clay, and is derived from 
 decomposed granites and other felspar bearing rocks, 
 but generally clays are largely admixtures with various 
 impurities. It is fuller's earth with an excess of silica. 
 Loam, with a certain proportion of fine sand, as brick 
 earths. Shale is clay and marl hardened and 
 laminated ; if mica is present it is micaceous, carbon- 
 aceous if carbonaceous or bituminous matter present. 
 Marl is clay that contains a considerable proportion of 
 lime ; if it is hardened, indurated, it is marl rock, which 
 decomposes on exposure to the weather. The varied 
 colouring of clay is due to iron in various states of 
 oxidisation, and to organic matter — the latter impart- 
 ing colours varying from light grey to black. The iron 
 imparts red, yellow, brown, and purple. Fireclay con- 
 tains an excess of silica. Beds of blue marl, though 
 tough and hard, do not afford a good foundation 
 stratum. 
 
 When the excavation, either for cellars or footing 
 trenches, is made in compact soil, short rough boards, 
 called " poling boards," are laid vertically against the 
 excavated sides at intervals on the opposite sides, and 
 kept in place by cross struts. When the ground is 
 loose, the poling boards, 2ft. to 3ft. long, are placed 
 close together, and kept in place by stout horizontal 
 planks called " wales," which are kept apart by square 
 timber struts of scantling suited to the width of the 
 cutting. This operation is repeated lower down as the 
 excavation proceeds. If it be a cellar, the sides must 
 be sustained by raking shores footed upon the excavated 
 bottom. In very loose unstable soils, as sloppy clays, 
 dry sand, etc., long poling boards are placed hori- 
 zontally, and strutted. As the excavation gains depth, 
 
 short vertical wales can be laid across several of the 
 horizontal poling boards, and thus release some of the 
 struts to be used lower down. 
 
 All excavations for footing trenches should be 6in. 
 wider on each side than the figured width of the footing 
 base, and should be made perfectly level in one plane, 
 or, if on rising ground, in as long benches or steps as 
 the gradient will admit of. It is advisable to make the 
 step, or rising face of the bench, in the lower one-third 
 point of a bay under voids, to avoid any sliding of the 
 face of the soil terrace by the pressure of the solid pier 
 footing if placed too near the edge of the bench. 
 Trenches for drains to be 1ft. wider than the diameter 
 of the pipes. Deep trenches should be 2|-ft. to 3ft. wide, 
 and 4ft. wide if over 9ft. deep. 
 
 Underfooting. — The trenches are sometimes filled to 
 a depth of 3ft. or 4ft. or more, with layers of broken 
 stone, gravel, sand, and concrete, well rammed in con- 
 venient layers. The soil may thereby be greatly com- 
 pacted, and the ultimate subsidence of the building 
 greatly reduced. Where concrete is used it should be 
 made of good hard-setting lime or hydraulic cement, so 
 that it may become and act as a monolithic structure, 
 as otherwise it will not spread the pressure of the foot- 
 ing base over a larger area as intended. 
 
 Increase in Bulk of Excavated Material. 
 
 Earth and clay (before subsequent subsidence) 
 
 about i more. 
 
 Sand and gravel, or earth and clay, after 
 subsidence 
 
 i 
 
 Chalk 
 Eock 
 
 'depends on the -size and ruling 
 shape of the~pieces, and the conse- 
 quent proportion of voids to solids 
 which they make in the heap . . 
 
 i 
 
 Loose soils compress into less space than occupied in 
 their natural deposition. Thus clayey earth occupies 
 about one-tenth part less volume in embankment than 
 before digging, gravelly earth one-twelfth less. Eock 
 in large pieces about five-twelfths more. Eock in 
 small pieces about three-fifths more. Light sound 
 earth occupies about the same space. 
 
 Estimating. — Cost depends upon the nature and stiff- 
 ness of soil and depth of the excavation, for each throw 
 up or staging of 6ft., when the stuff is not carted or 
 derricked out. In loose ground a man can throw up 
 about 10 cubic yards per day of 10 working hours. In 
 stiff clay or firm gravel about 6 yards. In hard ground, 
 where picking is required, 4 yards. Three men will 
 remove 30 yards of earth to a distance of 20 yards in a 
 day. 
 
 Excavating is measured and priced by the cubic yard. 
 Carting is charged by the single or double load — the 
 single load of sand, earth, rubbish = 1 cubic yard. 
 
 When the footing courses are laid and the foundation 
 wall well forward, the mortar joints properly set and 
 the work dry, the filling in should be done by careful 
 ramming of the earth in the trench round the footings. 
 
 Foundation Drainage. — Should the conditions require 
 water to be carried away from the footings, either drain 
 
52 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 tiles or broken stone, covered with slate or with small 
 slabs, should be laid outside the footings for this purpose 
 before the refilling is proceeded with. 
 
 Protection from Surface Water. — When the ground 
 is levelled up against the walls it should slope down 
 from the wall to shed off the surface water, and prevent 
 its penetrating to the foundation walls, or other needful 
 precautions should be adopted for the purpose. Stone 
 has about three times the capacity of conducting 
 moisture that brick has. Wall plaster is about one-fifth 
 less than brick, wood is one-fourth that of plaster. 
 
 Design of Footing Courses. 
 
 As the enduring safety of a building depends upon 
 the adequate apportionment of the footing base, it 
 necessarily demands very careful attention in design 
 and execution. The area of the base of the foundation 
 footing should be proportional to the bearing capacity 
 of the strata upon which the ultimate loads are to be 
 imposed. In this connection there are three important 
 considerations which are essential in all foundations : 
 
 (1) The area of the base of the foundations should be 
 so extended as to impose only a safe unit of pressure 
 of the loaded structure upon the subsoil suitable to its 
 bearing capacity. 
 
 (2) That the centre of the footing area should coin- 
 cide with the centre axis of the loaded pier wall, i.e., 
 the foundation area should support its load centrally, 
 whether on the continuous or on the isolated 
 principle. 
 
 (3) That the upper surface should be made truly 
 horizontal, in one plane where possible, and where 
 rising or hilly ground occurs, the number and extent 
 of the planes, or benches, should be regulated by the 
 amount of inclination. It is advisable to make the face 
 of the step, or rising of the bench, in the lower one- 
 third of a bay, i.e., underneath the voids, so as not to 
 endanger any sliding of the face of the soil at the 
 change of level, by the pressure of the solid pier footing 
 if placed too near to the edge of the raised bench If 
 the difference in level of these benches be considerable 
 it may be necessary, in compressible soils, to allow 
 additional extension of the footing base to correspond 
 with the added weight of the extra depth of founda- 
 tion walls, etc. In all benched, or stepped, founda- 
 tions, the walling and footing courses should be of 
 thick stones, and the joints should be very closely laid 
 in hard-setting mortar or cement, to ensure as far as 
 possible that the entire extent of the structure shall 
 subside equally. Similar precautions should bo 
 observed when one portion of a building is higher than 
 other adjoining parts. 
 
 Also when only a portion of the site is covered by 
 so id rock, for it seldom occurs that rock, or equally 
 solid portions extends all over the entire building arel 
 of a site and hence some portions of the structure 
 require to be supported upon, it may be, loose gravel 
 clay or other material of a very different character Ind 
 bearing capacity. The surface of the rock if £ Tay 
 be weather-worn, exposing more or less sloping smooth 
 
 surfaces on which it would be unsafe to build without 
 resort to some effectual method to prevent sliding or 
 canting of any portion of the structure. In such cases 
 it is usual to chisel out steps, checks or channels, etc., 
 to hold dowels, keys, etc., as there are many circum- 
 stances in which blasting operations would hot be 
 permissible. In some rocks there occur sand, or pot 
 holes, or pockets, which may so occur as to be utilised 
 for this purpose. In the softer rocks the surfaces may 
 be loose and disintegrated, all of which should be care- 
 fully removed. Large mass boulders may occur in the 
 way of the footings, and if not judiciously treated may 
 incur possibilities of unequal settlement, it may require 
 to be considered as a part rock foundation. Irregu- 
 larities occurring in firm ground should be levelled up 
 with concrete. 
 
 The Principles of Footing Courses, Continuous v. Isolated, 
 
 The fenestrated bays of buildings, by which void is 
 placed over void and solid over solid, naturally resolves 
 the wall into imposts and spandrels. The portions of 
 the solid wall intervening between the bays act as piers. 
 Yet, in violation of this very evident principle of design 
 of foundation, many architects and builders give to this 
 spandrel or bay portion of the wall the same width of 
 footing base as to the pier portion, though, in the ma- 
 jority of instances, it does not support the one-twentieth 
 part of the weight of the loaded wall per foot run in 
 comparison with that sustained by the intervening 
 piers. The part of foundation underneath the bay or 
 spandrel supports merely that lowest part of wall which 
 lies immediately underneath the lowest void or window 
 opening ; all the remaining part of the wall above and 
 vertically between the voids which forms so many 
 spandrels is imposed upon the piers, one-half upon the 
 pier on each side of the void. The frequent result of 
 this violation is to be seen all over London as well as in 
 the provinces in disfiguring cracks, broken walls, sills, 
 spandrels, window arches, lintels, stone storey courses, 
 or other horizontal stone bands, mouldings, corners, 
 widely opened stone joints, etc., and in many cases 
 where it does not appear its salvation is due rather to 
 the extra strength given to the masonry, or to the in- 
 compressibility of the substrata, concrete, hoop iron 
 bond, etc., rather than to design upon the proper 
 principle. 
 
 The footing base of ordinary foundations are of the 
 same continuous width throughout for the main walls 
 of the building, but the footings of fenestrated walls 
 designed on the isolated principle would have, for voids 
 in vertical tiers and equidistant horizontally, indents in 
 the plan of the footing base nearly equal to the width 
 of and immediately underneath the spandrel bays, upon 
 both the inside and the outside of the wall. The fact 
 of the weight of the spandrels being transferred to the 
 imposts of the intervening pier walls requires that the 
 footing base underneath the impost should extend 
 sufficiently in the shape of a " return " under the sides 
 of the bays to equalise the insistent pressure upon the 
 subsoil. 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 53 
 
 Off-Sets of Masonry Footings. 
 
 The area of the foundation having been determined 
 and its centre having been located with reference to 
 the axis of the load, the next step is to determine how 
 much narrower each footing course may be than the 
 one next below it. The projecting part of the footing 
 resists as a beam fixed at one end and loaded uniformly. 
 The load is the pressure on the earth or on the course 
 next below'. The off-set of such a course depends upon 
 the amount of the pressure, the transverse strength of 
 the material, and the thickness of the course. 
 
 To deduce a formula for the relation between these 
 quantities, let 
 
 P = the pressure, in tons per square foot, at the 
 bottom of the footing course under considera- 
 tion ; 
 
 K = the modulus of rupture of the material, in 
 pounds per square inch ; 
 
 p = the greatest possible projection of the footing 
 course, in inches ; 
 
 t = the thickness of the footing course, in inches. 
 
 The part of the footing course that projects beyond 
 the one above it, is a cantilever beam uniformly loaded. 
 From the principles of the resistance of materials, we 
 know that the upward pressure of the earth against the 
 part that projects multiplied by one-half of the length 
 of the projection is equal to the continued product of 
 one-sixth of the modulus of rupture of the material, the 
 breadth of the footing course, and the square of the 
 thickness. Expressing this relation in the above 
 nomenclature and reducing, we get the formula 
 
 p = t\/ 
 
 E 
 41-6 P' 
 
 or, with sufficient accuracy, 
 
 p = it 
 
 4 
 
 E 
 
 Hence the projection available with any given thick- 
 ness, or the thickness required for any given projection, 
 may easily be computed by the latter equation. Notice 
 that, with the off-set given by the above formula, the 
 stone would be on the point of breaking. 
 
 The margin to be allowed for safety will depend upon 
 the care used in computing the loads, in selecting the 
 materials for the footing courses, and in bedding and 
 placing them. If all the loads have been allowed for 
 at their probable maximum value, and if the material 
 is to be reasonably uniform in quality and laid with 
 care, then a comparatively small margin for safety is 
 sufficient ; but if all the loads have not been carefully 
 computed, and if the job is to be done by an unknown 
 contractor, and neither the material nor the work is to 
 be carefully inspected, then a large margin is necessary. 
 As a general rule, it is better to assume, for each par- 
 ticular case, a factor of safety in accordance with the 
 attendant conditions of the problem than blindly to 
 use the result deduced by the application of some 
 arbitrarily assumed factor. The following table is 
 
 given for the convenience of those who may wish to 
 use 10 as a factor of safety : 
 
 Safe Off-Set for Masonry Footing Courses, in Terms of the 
 Thickness of the Course, using 10 as a Factor of Safety. 
 
 Kind of Stone. 
 
 P., in lbs. 
 
 per square 
 
 inch. 
 
 Off -set for a Pressure, in 
 
 tons per square foot, 
 
 on the 
 
 Bottom of ihe Course, of 
 
 
 0-6 
 
 1-0 
 
 2-0 
 
 
 1,800 
 1,500 
 1,200 
 5,400 
 1,500 
 800 
 
 150 
 
 2-9 
 
 2-7 
 26 
 5-0 
 2-7 
 1-9 
 
 0-8 
 
 21 
 1-9 
 1-8 
 3 6 
 1-9 
 1-4 
 
 06 
 
 1-5 
 1-3 
 
 Slate 
 
 1-3 
 
 2-5 
 
 Best Hard Brick 
 
 1-3 
 
 
 0-8 
 
 [ 1 Portland") 1A ■■ 
 Concrete^ 2 sand \ W ?1 7 
 (3 pebbles j old 
 
 0-4 
 
 To illustrate the method of using the preceding 
 table, assume that it is desired to determine the off-set 
 for a limestone footing course when the pressure on 
 the bed of the foundation is 1 ton per square foot, 
 using 10 as a factor of safety. In the table, opposite 
 limestone, in next to the last column, we find the 
 quantity 1*9. This shows that, under the conditions 
 stated, the set-off may be l - 9 times the thickness of 
 the course. 
 
 If it is desired to use any other factor of safety, it is 
 only necessary to substitute for E, in the preceding 
 formula, the desired fractional part of that quantity as 
 given in the second column of the above table. For 
 example, assume that it is necessary to use limestone 
 in the foundation, and that it is required to draw in 
 the footing courses as rapidly as possible. Assume also 
 that the pressure P on the base of the foundation is 
 2 tons per square foot. If the limestone is of the 
 best, and if it is laid with great care, it will be sufficient 
 to use 4 as a factor of safety. Under these conditions 
 the equation gives 
 
 P = *«. 
 
 >A?-h>A 
 
 1500 
 
 2 
 
 2-3 t. 
 
 That is, the projection may be 2 - 3 times the thickness 
 of the course. 
 
 Strictly, the above method is correct only when the 
 footing is composed wholly of stones whose thickness 
 is equal to the thickness of the course, and which pro- 
 ject less than half their length, and are also well 
 bedded. The values in the table agree very well with 
 the practice of the principal architects and engineers 
 for hammer-dressed stones laid in good cement mortar. 
 
 The preceding results will be applicable to built 
 footing courses only when the pressure above the 
 course is less than the safe strength of the mortar. 
 The proper projection for rubble masonry lies some- 
 where between the values given for stone and those 
 given for concrete. If the rubble consists of large 
 stones well bedded in good, strong mortar, then the 
 values for this class of masonry will be but little less 
 than those given in the table. 
 
 If the rubble consists of small, irregular stones with 
 Portland cement, the projection should not much exceed 
 
54 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 that given for concrete. If the rubble is laid in lime 
 mortar, the projection of the footing course should not 
 be more than half that allowed when cement mortar is 
 
 used. 
 
 Notice that drawing in the footing courses decreases 
 the area under pressure, and consequently increases 
 the pressure per unit of area ; hence the successive pro- 
 jections should decrease from the bottom towards the 
 
 top. 
 
 Foundations for Steam Engine Beds, ete. 
 
 Foundations for beds of steam engines and other 
 stationary motors, and for alternating and revolving or 
 percussing heavy machinery, tall chimney shafts, 
 towers, etc., should in all cases be laid quite indepen- 
 dently of all walls of buildings, and be in no wise 
 attached thereto. 
 
 Elastie Foundations for Engines or Machines. 
 
 Such foundations require complete isolation of con- 
 structions, with the object in view of deadening shocks, 
 avoiding transfer of vibrations, and of lessening the 
 incessant noise which often constitute a legal nuisance, 
 involving litigation for its abatement. This problem 
 has had many solutions offered, but none entirely 
 satisfactory and generally applicable. Neither rigid 
 foundations nor the interposition of elastic bodies has 
 succeeded, though this latter direction promises well. 
 The holding-down bolts of the engine or machinery bed 
 transmit vibrations to the parts bolted to them. A 
 recent experiment at an electrical station by M. 
 Juppont has proved the value of judicious methods of 
 introducing elastic bodies. It consisted of placing in a 
 rectangular excavation of suitable depth two sheets of 
 plate iron having a series of discs of caoutchouc between 
 them, causing isolation, placed upon a platform laid 
 upon the bottom, and having a platform rivetted to the 
 top plate to prevent deformation. Upon this upper 
 platform the foundation proper was erected, having a 
 clear trench all round it, with provision for bolts and 
 space for cleaning, etc. The mass of the foundation 
 need not be of masonry, and may with advantage in 
 certain cases be replaced by a caisson filled with sand. 
 The steam and exhaust pipes were wound in a spiral at 
 their top part to give elastic accommodation without 
 forcing the joints. The amplitude of oscillation was 
 eight millimetres, but no vibration was communicated 
 outside of the trench. 
 
 Suspended Foundations for Machines. 
 
 A sugar refinery in Philadelphia, U.S.A., has several 
 steam engines distributed all over the building, and 
 upon different storeys, some being upon the second 
 and third floors. Some of these engines were bolted 
 to iron girders, or heavy beams, and some of them ran 
 smoothly and silently, while others of them produced 
 vibration and disturbing "rattle." To correct this 
 " rattle," as an experiment, the mass of the engine 
 foundations was suspended from the bottom of the 
 engine, whereby, in consequence of the inertia of the 
 weight of the mass suspended, the vibrations and 
 " rattle " were absorbed. 
 
 Strength and Stability. 
 
 Strength and stability of masonry begin with the 
 foundations, and hence it is essential to give early 
 attention to the actual conditions of the mode upon 
 which the stability of masonry structures depend. It 
 is often erroneously treated as a mere question of the 
 simple moment of resistance against overturning, 
 which is thus made to depend upon the assumption 
 that the leverage resistance of the entire actual width 
 of the footing base, out to the extreme edge of the toe, 
 is supposed to be effective, thus treating the footings 
 and wall as an entire, perfectly rigid monolith, like a 
 piece of solid iron. But no masonry wall of ordinary 
 building materials and structure possesses this un- 
 attainable degree of incompressible and tenacious 
 strength, sufficient to support its entire weight and 
 imposed loads, upon only one mere edge of its footing 
 base. Indeed, the generality of walling does not 
 possess any appreciable extent of reliable cohesion, and 
 the measure of its strength is mainly derived from the 
 weight and frictional resistance to displacement, which 
 the blocks of stone, etc., of which it consists, exert upon 
 each other. But, in practice, the mode of overturning 
 involves three general cases. Thus : (1) If we con- 
 sider the case of the overturning of a wall of ordinary 
 construction of brick footing consisting of two bottom 
 courses, and to be tilted, or canted over, experi- 
 mentally by a horizontal force, as wind pressure, 
 whereby the centre of gravity of all its weight is 
 vertically over the extreme outside edge of the footing 
 base, thereby depressing it into a yielding soft subsoil — 
 in order to preserve stability in such a case, the resul- 
 tant of the weight acting vertically, and the wind 
 pressure acting horizontally, multiplied by their respec- 
 tive leverages, are adjusted so as to pass within the 
 middle half width of the footing base, i.e., a full quarter 
 of the width inside of the toe. (2) If the subjacent 
 soil or bed be of solid rock, and the walls were similarly 
 canted over (a) , the edge of the toe would be crushed 
 backwards until the area of support would be enlarged 
 equivalent to the crushing resistance of the brick, 
 assuming them to act as a monolith, or (6) a vertical 
 section of the toe would be ruptured by shearing at 
 some critical inside point, or (c) by the cross rupture 
 of the bricks and by simultaneous opening of the 
 corresponding back joints of the underside of the base. 
 (3) If the footings remain rigid as well as the subjacent 
 foundation bed, then overturning takes place by the 
 opening at a critical point in the height of vertical 
 section of a horizontal joint on the windward side of 
 the wall, which is thus said to be in tension. 
 
 In view of these actual modes of operation of the 
 active forces and of the building materials in the act of 
 overturning, and the application of the corresponding 
 principles to determine the stability of walls, there are 
 two obvious provisions which the design of walling 
 should contemplate. 
 
 1st. It is evident that for a high masonry structure 
 it should be designed ior a convenient safe limit of 
 crushing pressure upon the materials, instead of for a 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 55 
 
 mere overturning leverage moment — overturning tak- 
 ing place if the deviation of the resultant of the hori- 
 zontal and vertical moments be half the thickness of the 
 wall from its centre on plan. 
 
 2nd. When the adhesion of the mortar is not suffi- 
 cient to be considered, the windward joints would be 
 in tension, and have a tendency to open if the resul- 
 tant of the moments of the vertical and horizontal 
 forces acting at any joint in the height deviates from 
 the centre of that joint by only one-sixth of the thick- 
 ness of the wall. This would occur before the leeward 
 edge of the joint would have any tendency to be 
 
 crushed. 
 
 Building Materials.— Stone. 
 
 The position of the occurrence of rock beds in 
 quarries or quarry sites to be opened, suitable for build- 
 ing stone, requires practical skill to discover its precise 
 value. The appearance and qualities of the same kind 
 of stone, and even that from the same quarry, vary 
 as successive beds or layers are reached, the deteriorated 
 and less compact forms lying at the top ; the more 
 dense but " green," i.e., soft and containing moisture 
 called quarry sap, lie towards the bottom, or further 
 from the surface. These latter, however, harden upon 
 exposure to the atmosphere, the moisture coming to the 
 surface in the form of a hard crust, which makes it 
 more expensive to work. 
 
 The position in a quarry of the best beds or stratum 
 for furnishing stone of a quality and characteristics 
 suitable for exterior building purposes is of importancs 
 to be specified for use in an intended building, other- 
 wise a stone of identical appearance, but possessing 
 inferior weathering characteristics, may be substituted 
 without the possibility of present detection. In quarries 
 of different kinds of stone, and in different localities, 
 the successive beds of the same stone are known by 
 different names. 
 
 Thus, in a quarry of Portland stone, the best bed of 
 which is one of the most durable of stones, the succes- 
 sion and names of the beds will be somewhat thus in 
 the downward order, until the useful bed, No. 11, 
 forming the upper beds of the oolitic limestone series, 
 is reached about 30ft. below the surface : 
 
 1. The vegetable mould, with various debris of rock 
 admixture on the ground surface. 
 
 2. Clay and shingle from debris of purbec (fresh- 
 water) limestone, which is only one of a series of lime- 
 stones, clays, shales, and sandstones, of which the 
 purbec beds consist, which is divided into an upper, a 
 middle, with cinder beds, a lower series, with dirt 
 beds. The dirt beds being dark-coloured loam like, 
 which are interstratified with oolitic limestones and 
 sandstones of Portland. 
 
 3. Slaty beds of stone. 
 
 4. Bacontier, with layers of stone. 
 
 5. Aish stone. 6. Soft burr. 
 
 7. Dirt bed, with fossil trees. 
 
 8. Cap rising. 9. Top cap. 10. Skull cap. 
 
 11. Eoach (true), 2ft. or 3ft. thick. A useful 
 weather stone, a conglomerate of fossils cemented 
 
 together by carbonate of lime ; cavities, large and 
 small, are very numerous in it. It is distinguished by 
 the shell-cast, known as the "Portland screw," which 
 is peculiar to this bed only. 
 
 12. Whitbed, 8ft. to 10ft. thick, fine even grained, 
 one of the most durable building stones in England, 
 and is the most valuable of the Portland quarries. It 
 consists of fine oolitic (roe-like) grains durably cemented 
 together with a hard, crystalline material, and inter- 
 spersed occasionally with shelly matter. 
 
 13. Curf or kerf, flinty. A bastard roach surmounted 
 by a layer of rubbish. 
 
 14. Curf and basebed or bastard roach. 
 
 15. Basebed, a substantial stone 5ft. or 6ft. thick. 
 This, with the bastard roach or basebed roach, are very 
 similar in appearance to the true roach and whitbed, 
 but they do not weather well, and therefore are only 
 fitted for inside work, or where only exposed to un- 
 varying conditions in foundations. 
 
 16. Flat beds or flinty tiers. 
 
 The true roach is remarkably tough and strong, 
 weathers well and resists the solvent action of water ; 
 is a very light brown colour. 
 
 The first eight numbers are excavated with shovel 
 and pick, 9, 10, and 11 require to be blasted, 12 to 16 
 are quarried by wedges and lever bars. Though Nos. 
 11, 12, 13, 14, and 15 can hardly be distinguished from 
 each other, even by the most practised eye, and are, as 
 stated above, very different in characteristics, while 
 their chemical composition is almost the same, and 
 consists of (in 100 parts) silica, 1'20 ; carbonate of lime, 
 95'16 ; carbonate of magnesia, 1'20 ; iron and alumina, 
 0'50 ; water and loss, 1"94; and a trace of bitumen. 
 
 The quality of resisting the deteriorating influences 
 of the atmosphere, rain, dampness, alternate wet, dry, 
 sun, frost, etc., is of the utmost importance. The 
 durability depends upon its physical properties and 
 structure as well as upon the chemical composition of 
 its mineral constituents. The nature of the substance 
 which cements the minerals together in the rock mass, 
 to be durable, should be solid and in a half crystalline 
 state. Thus some durable sandstones are rendered so 
 by having a cementing matrix chiefly of silica, but when 
 the matrix contains alumina, the principal ingredient 
 of clay or lime, the sand stone is less durable, and may 
 even be very perishable — the least durable being in an 
 earthy powdery state. The durability is also affected 
 by the nature of the exposure which the position of the 
 stone in the building subjects it to. A "weathering" 
 exposure subjects it to the deteriorating effects of the 
 weather, consisting of alternations of rain, dew, wind, 
 sunshine, frost, etc., with all their attending concomi- 
 tants, as well as the destructive chemical action of 
 the atmospheric gases, acids, etc., especially of those 
 peculiar to large and manufacturing cities, sea 
 coasts, etc. The action of the freezing of the 
 water absorbed into the pores of the stone, car- 
 bonic acid, both the natural and that of artificial 
 production in the air, sulphuric and hydrochloric acids 
 in rain, nitric acid, all act rapidly in effecting de- 
 
56 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 composition either of the mineral constituents, or of 
 its cementing materials, if of carbonate of lime or 
 of magnesia, alumina, etc. The oxygen of the air 
 acts upon the iron salts in the stone. The sulphurous 
 and other waste acid products of factories of different 
 kinds as of bleach works, chemical works, etc., 
 send forth clouds of acid-laden fumes that are quickly 
 destructive of the structure of stone. The crystallisa- 
 tion in the pores of the stone of the sulphates thereby 
 formed produces fracture by expansion, whereby large 
 flakes and surface fragments are loosened. Earn that 
 is absorbed in excess by capillary action on the lower 
 horizontal surfaces or undersides, as in soffits of 
 lintels, arches, cornices, etc., by solvent action and 
 by expansion is destructive. These are the trying 
 positions to give particular attention to when investi- 
 gating the durability of different kinds of stone, whether 
 they be in a building, or strewn about a quarry, stone- 
 yard, etc. Building stone is less destroyed in a dry than 
 in a rainy climate, or on the side of a building exposed 
 to the prevailing rains and sheltered from the sun and 
 from drying winds. Light winds dry out dampness 
 and distribute dust, which by accumulation retains 
 moisture, while high winds force further into the pores 
 any surface dampness, dust, etc., besides, by blowing 
 forcibly about heavy sharp dust particles, producing a 
 wearing action. Sudden variations of cold and heat, 
 by producing alternating contraction and expansion of 
 the different substances of the composition, which may 
 have differing expansion units, are also destructive. 
 Quoins, corner, or arris stones, when exposed to diffe- 
 rent degrees of heat on their different faces, are liable 
 to crack, from the unequal expansion and contraction 
 thereby caused. Exposure to continuous dampness, as 
 in foundations, may subject stone to the action of 
 mineral salts in the local underground waters, where 
 it is employed, and if frost penetrates toit, deterioration 
 is hastened. Frost penetrates damp solid earth deeper 
 than dry porous soils ; it will also follow iron, which 
 acts as a conductor. 
 
 Physical structure contributes greatly to the dura- 
 bility or otherwise of stone. Thus, chalk and marble 
 have the same chemical composition (pure carbonate of 
 lime), but are of different structures. Marble, especially 
 when polished, will endure much longer than chalk. 
 Hence stones which are crystalline in structure, as 
 marbles, granites, etc., " weather " better than the non- 
 crystalline varieties, as slaty stones, or the granular 
 class, as chalk, limestone, sandstone, etc. Porous stones 
 absorb moisture more largely than close-grained dense 
 stones ; and hence acid-charged dampness, as rain, etc., 
 and the freezing of the moisture in the pores tends to 
 disintegrating and rupturing the structure of the stone. 
 
 "When both the mineral grains and the uniting sub- 
 stance are alike durable, the rock partakes of their 
 durable characteristics ; but when the mineral grains 
 easily decompose, while the cementing substance 
 remains, the structure becomes porous. If, on the 
 other hand, the cementing material is dissolved, the 
 mineral grains separate. Stones often contain soft 
 
 patches and inequalities of structure or of chemical 
 composition, which are frequently indicated by blotched 
 or mottled colour, when the one part will wear or 
 weather away faster than the other, leaving the pro- 
 jecting portions exposed to catch accumulations of 
 dust, rain, etc., and hasten decay. Lichens and other 
 like vegetable growths upon the faces of stone tends to 
 deteriorate it, notwithstanding that for sandstones they 
 are sometimes accounted a protection from the 
 mechanical action of the weather. 
 
 The best stones absorb the least water, and are 
 therefore the less liable to the expansion of being 
 frozen in the pores of the stone, by which it is disin- 
 tegrated. Thus, in 24 hours, trap and basalt will 
 only absorb up to about one-fifth per cent, of their 
 volume. Good granites, half per cent. ; indifferent 
 granites, 1 to 3 per cent. ; the harder sandstones, 
 less than 7 per cent. ; those of a very durable kind, 
 8 per cent. ; moderately durable sandstones, 10 
 per cent. ; very bad standstone, 20 per cent, of its 
 volume. Limestones vary from about 8 per cent, 
 upwards ; Portland, very durable, 13 per cent. ; 
 Ancaster and Eoche Abbey, durable, 16 to 17 per cent. 
 
 A few fresh thin chips of the stone to be tested, 
 immersed for several hours in rain-water, will, when 
 afterwards shaken up, show by its milky or murky 
 appearance to what extent the constituents of the stone 
 are not stable. When the chippings are broken off the 
 stone should be thoroughly damp. 
 
 A solution of about 1 per cent, of sulphuric and hydro- 
 chloric acids will suffice to indicate the behaviour of a 
 stone subject to a city atmosphere ; or a few drops of 
 acid upon a stone will produce effervescence if carbonate 
 of lime or of magnesia be present in large proportions. 
 But the observation of how similar stone has worn is a 
 test of the greatest value. 
 
 Hard stone should always be put where exposed to 
 rubbing, as in jambs and sills of doors, pavements, 
 exposed quoins and parts of arrises, plinths, etc., also 
 where they receive dripping water, the running water 
 of rivers, shore waves, etc. Hard stone, however, may 
 be chemically inferior, and may not weather as well as 
 a soft stone. The granites have about three times the 
 hardness, or resistance to abrasion, that the hardest 
 sandstones possess, about 10 times that of marble, and 
 about 80 times that of soft limestone, as the Bath stone 
 used for insides of walling, etc. 
 
 Tough stone is not liable to splinter and crack, and- 
 is of importance in places where bolted to machinery, 
 especially that which has a percussive action. 
 
 Strength, stress, is usually denoted by the ability to 
 resist great compressive stress or weight. Although 
 stone in its ordinary position in overlapping courses 
 and crosslapping bond is subject principally and in the 
 normal conditions to equal and uniformly distributed 
 pressure and likewise to equal subsidence, but it is 
 very liable, not only during construction, but after- 
 wards, to be severely affected by unequally distributed 
 or excentric pressures, producing oblique acting or 
 cross stresses, in which its cohesive properties of resis- 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 SECTIONAL ELEVATION, 
 
 PAVEY-PAJCMAN'S ESSEX BOILER, 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 SECTION ON LINE ABCD. 
 
 DAVEY-PAXMAN S ESSEX BOILEK. 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 57 
 
 tance, more than its shearing strength, are subject to 
 critical stress, and which is equivalent to tensile action 
 upon the material of the stone, as in unequal settle- 
 ments of different portions of the building or of the 
 wall, or of the facing and backing portions, vibratory 
 or oscillatory motion 'of machinery. In some such 
 cases the outside edges are pinched off in large flakes. 
 The ample dimensions usually given to walls, aud 
 even to piers in ordinary sizes of buildings, will never 
 likely subject the stone to an unsafe pressure simply, 
 as the weakest sandstones which would be admitted to 
 a building will safely bear even 10 tons per square foot 
 of area of horizontal section. It is only in some of the 
 piers or columns of a few mediaeval churches or large 
 Gothic buildings that 10 tons per square foot is ex- 
 ceeded. "When a heavy weight is concentrated upon 
 only a small part of the surface of a block of stone, its 
 full nominal strength is reduced, as when an iron 
 column stands on a stone base, or the end of an iron 
 girder rests upon a bonding stone, and only covers 
 half the block, its effective resistance is only about two- 
 thirds of its nominal strength, and if the column base 
 only occupied one-quarter of the surface of the stone 
 block, its effective resistance would be reduced by one- 
 half. This is very important to bear in mind. For 
 engine bedding a heavy stone is essential, also for 
 retaining walls, or where subject to the action of tide 
 waters. 
 
 It is important to the durability in the wall of all 
 stone (except the unstratifled and unlaminated varieties) 
 that it should be laid in the wall in the position corres- 
 ponding to that of its original deposit in the rock for- 
 mation in the quarry, whether the layers have been 
 subsequently tilted or inverted or not — i.e., with the 
 direction of its structural natural layers or laminae in 
 a horizontal position — otherwise, if laid perpendicularly, 
 the edges of the layers will be turned upwards, and will 
 the more readily receive and retain dampness between 
 them, whereby the disintegrating acids, frost, etc., will 
 scale off the face laminae, and the stone be thus dis- 
 figured and further exposed to rapid deterioration. 
 
 Bricks. 
 
 London bricks are nominally 9in. by 4Jin. by 2£in., 
 but really only about 8Jin. by 4£in. by 2£in. to 2fin. 
 
 " Eubbers," or " cutters," are made of washed clay, 
 freed from lumps and pebbles, and of a uniform com- 
 position, and only burned sufficiently to admit of cut- 
 ting, carving, rubbing, moulding, to any required 
 design of architectural features. The best kind is 
 burnt to a point just short of vitrification. Inferior 
 kinds are less burnt, or under burnt. Eubbers, how- 
 ever, are not so durable as "purpose-made" bricks, 
 which are moulded to any required shape, and 
 sufficiently burnt. 
 
 " Facings," for fronts, are the best selected of the 
 clamp or kiln production, combining uniform colour, 
 sufficient burning, perfect in shape, sound and free from 
 pebbles or lumps and flaws of any kind. Bach 
 district and each brickfield has its own quality of 
 production. 
 
 Common bricks are " unwashed," and frequently of 
 very unfit clay, carelessly treated, unequally burnt, 
 brittle, cracked, and generally unsound, soft, and unfit 
 for house building of any structural importance. They 
 are, however, classed into shippers, the best, next 
 stocks, grizzles, rough stocks, and place bricks. 
 
 Malms, originally made from the malm clay which 
 in early times was found in the neighbourhood of 
 London, and corresponds to that which is suitable for 
 hop-growing districts. The present so-called malms 
 consist of tempered clay with cream of lime, or about 
 one-sixteenth part of ground chalk in pulp, and breeze 
 (or refuse cinders) incorporated before moulding. 
 The methods and care in the manufacture differ in 
 different brickfields and districts, with corresponding 
 varieties in the qualities and grades of production. 
 
 Good bricks should be free of cracks, and pebbles of 
 lime, however small, which slake on being wet or in 
 wall and burst. They should also be free of lumps and 
 pebbles of other rock, which cause unequal expansion 
 and contraction, and have a good even shape edge 
 unbroken, with ordinary roughness of bed surface 
 (both upper and lower sides) for adhesion of mortar, 
 show a glassy fractured surface, give a clear sharp ring 
 when struck together. The length should be = 2 
 breadths + a mortar joint of x^in. to fin., according 
 to the kind of the brick and brickwork, x 2 tin. thick, 
 weigh about 71b. It is well to remember that a £in. 
 thinner brick than 2^in. will require an extra course in 
 every 20 courses, say, in every 4ft. 6in., if laid with a 
 £in. bed-joint. 
 
 Masonry Walling 1 . 
 
 Uniformity of construction of walls is necessary to 
 impart the utmost strength, by ensuring uniformity of 
 subsidence during construction, and stability when 
 exposed to the action of excessive heat when the build- 
 ing is on fire. 
 
 When there is an exterior facing of ashlar, backed 
 internally by rubble or brickwork, there is a diversity 
 of subsidence in proportion to the difference in the 
 number and thickness of the mortarbed joints in the 
 ashlar and in the backing, and when a fire occurs in 
 such a building the walls crack and rend by reason of 
 the difference of the expansion units of the two classes 
 of materials, and endanger the safety of the entire 
 building. Ashlar as usually constructed has never a 
 thorough bonding, but may have a three-quarter bond 
 stone nominally, but which in practice often falls far 
 short of three-quarters of the thickness of the wall, and 
 be merely an angular point of overlap. These bond 
 stones ought to be sufficiently distributed in every few 
 courses, depending on their height, and on the duty of 
 the wall. 
 
 Facing brick, backed with common " stocks," will 
 not be conformable in height of courses, as it may re- 
 quire, say, eight of the facing bricks to correspond with 
 seven courses of the stocks. Thus, if stocks laid up 
 four courses to the foot high, and that facing brick laid 
 up fin. less in each course — i.e., each course of facings 
 would lay up 2fin., and so on for other differences — it 
 
'58 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 would, therefore, only be in such a case at every eighth 
 course that a heading course could be laid with which 
 to cross-bond the facing with the backing. Such a 
 heading course, however, is not laid as in common 
 brickwork, but merely an inside corner is clipped off 
 the face brick, which is usually laid in Flemish bond — 
 i.e., alternate header and stretcher in each course, 
 breaking joint — but the header is false and only a half 
 brick, a corner of the backing brick projects into the 
 clipped space of the facing, like a king closer, which 
 makes a very imperfect bond and a necessarily weak 
 wall. It especially becomes weak in narrow piers, when 
 closers are required in each header course, so as to 
 bring the header to break joint in the centre of each 
 stretcher, a three-quarter bat being required in each 
 stretcher course, whereby the opportunity for bond is 
 proportionately lessened. In Flemish bond the closer 
 and three-quarter bat would both occur in the same 
 course, but break joint in the alternate courses. When 
 the width of the pier suits the working in of a certain 
 number of whole bricks better bonding is secured than 
 if otherwise, when there is more clipping of bricks 
 necessarily resorted to. English bond is composed of 
 all header and all stretchers alternating in the suc- 
 cessive courses, but there are varieties in the methods 
 of laying both the Flemish and English bonds. 
 
 When the number and thicknesses of the external 
 and internal bed joints is unequal, the backing should 
 be laid in Portland hydraulic cement, or quick-setting 
 blue lias lime mortar, properly gauged with sharp sand, 
 free of loam, dust, or clay. 
 
 For mills, factories, or where machinery is employed, 
 in order to localise and absorb the vibratory action of 
 driving shafting and of machines distributed through a 
 building on several of its upper floors, deep pilasters or 
 piers should be built with light bonded walls of whole 
 brick between them in the nature of panels. This 
 method is effectual in imparting a longer life to 
 machinery as well as to the building, and minimises 
 the expense of repairs in both. Besides this arrange- 
 ment, it is likewise essential that equally balanced 
 foundations for walls or piers should be constructed 
 in order to preserve the true level of the building in 
 all its parts throughout its extent, so that shafting and 
 connected machinery and power engines may be laid 
 with all probability of retaining an accurate level, 
 without which it is impossible to work smoothly. 
 
 No outside wall of ordinary rubble should be less 
 than 16 or 18 inches thick. Some slaty districts pro- 
 duce a flag or rag stone, which will quarry in regular 
 layers of only a few inches thick. Walls built of these 
 could be laid up of less width, if not too high— i.e., 
 14 to 16 times its width without intermediate tie 
 beams, etc. 
 
 Hollow walls, with a vertical air cavity of 2in., are 
 sometimes built to keep basements and cellars dry ; 
 but, in order to be effective, they should be very care- 
 fully executed, and no mortar allowed to drop between, 
 which in ordinary work, and without a special pre- 
 cautionary device, is practically impossible. As a means 
 
 of preventing it, iron tubes, wrapped round with hay 
 bands, are sometimes resorted to. These rest on the 
 last bond ties in the cavity, and are moved out before 
 the next row of bond ties is laid. There are several 
 forms of cast-iron wall ties, with claws and turned-up 
 ends and a rising bend to cross the cavity, to prevent damp 
 passing along it by obliging it to ascend, Which gravita- 
 tion will prevent. The cavity must be commenced at 
 the horizontal damp course, and be made continuous 
 round the corners and angles, and brought close up to 
 the reveals of all openings. There should be weep 
 holes provided at bottom of the cavity, to carry off any 
 injected moisture and to afford ventilation. 
 
 In surcharged sites, underground basements and 
 cellars may require damp-proofing all underneath the 
 floor and all round the walls up to and above the 
 highest water level of the district waterlogged. Hygeian 
 rock asphalte is considered one of the best waterproof- 
 ing materials for underground walls. It is poured in 
 a molten state into a fin. continuous open vertical 
 joint, left in the thickness of the work at every third 
 or fourth course as the wall rises. It sets hard, and 
 strongly combines the inside and outside portions into 
 one compact wall, greatly increasing its strength, and 
 rendering it perfectly waterproof. 
 
 Masonry Footing Courses. 
 
 The metropolitan by-laws require at least 6in. thick 
 layer of concrete spread over all building sites, unless 
 of gravel, sand, or natural virgin soil. 
 
 Foundations of walls of houses to have not less than 
 9in. thick layer of concrete extending 4in. beyond each 
 side of the footings, unless the site be on a natural bed of 
 gravel, in situ, does not apply to a building to be used 
 as a stable or shed, provided such be not used for 
 public assemblies or entertainments, or as dwelling or 
 sleeping places. Therefore workshops, etc., which may 
 have in confined employment large numbers of persons 
 for 12 hours in the 24, are curiously exempted, what- 
 ever may be the insanitary conditions of the site. 
 
 When the substratum consists of solid rock, or a 
 thick bed of gravel, dry sand, or gravel and sand, the 
 footings may consist of two courses in each inset. The 
 header course should be above, and the stretcher course 
 laid below, all overlying courses should carefully break 
 joint with the course immediately beneath it. It is 
 usual to make the insets (in the act of laying the bricks) 
 I brick = 2 Jin. When the insets consist of one course 
 in each inset, the lower tier usually consists of a header 
 and a stretcher, but it would be better to have two 
 header courses with a stretcher between them. When 
 the substratum is of compressible soil, large thick 
 sound stone flags, as used in Yorkshire, etc., some- 
 times called " landings," or a thick bed of sound concrete 
 of Portland cement (see under Concrete), should be used 
 to impart a monolithic character under the imposed 
 pressure. But in the absence of such provision the thick- 
 ness of brickwork under the toe of the footing should 
 be increased very much more than is commonly done on 
 compressible subsoils. According to experiments by 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 m 
 
 J. H. Apjohn, B.E., at Akra, Indian Government brick 
 works, Calcutta, four courses of brickwork under the 
 toe of splayed pier footings cracked badly with If ton 
 per square foot of area on a soft alluvial soil, and he 
 recommends that for similar soils to those experi- 
 mented on, in order to sustain a pressure safely of 1 
 ton per square foot, the thickness at the toe of the 
 footing should be at least 1ft. 6in., or, say, six courses 
 of brick. Any rupture under the toe by shearing, 
 or opening of the back joints, means a lessening of 
 the area of the bearing surface, and a consequent 
 liability to an excessive subsidence of the wall or 
 pier. 
 
 Bricks should be hard burned, well shaped, whole 
 and sound, laid in strong hydraulic lime, mortar, or 
 Portland cement, with all the joints and vacuities solidly 
 filled. 
 
 The Staffordshire blue bricks (with rough surface), 
 which are proof against the chemical decomposition of 
 damp soils or ground waters, and are of the utmost 
 strength, should be more used in footings than they 
 are. 
 
 Failures in footings are always expensive to remedy, 
 and destroy the value of a building ; and therefore 
 strong, sound materials of sufficient section should be 
 employed, and only the best hydraulic lime or cement 
 used. 
 
 Conerete. 
 
 What is called the " aggregate " is of broken brick, 
 slag, or stone (to pass through a ljin. to 2Jin. ring); 
 burnt clay, gravel, ballast, breeze (refuse coal cinders), 
 are also used. There should be no adhering dust 
 or mud coating, which prevent the adhesion of the 
 cement. Shingle, or round smooth pebbles, are not 
 so desirable. 
 
 Sand is requisite to help fill the voids, and is 
 desirable for improving the strength, and is necessary 
 to make it waterproof. 
 
 It is more economical to have stones in relative 
 quantities broken to different sizes, so that the smaller 
 ones will nearly fill the voids of the larger. The nett 
 voids for different sizes are as follows : 
 
 Voids in a Cubic Yard. 
 
 Stone broken to 2£in. gauge = 10 cubic feet. 
 
 Ditto 2 „ „ = 10| » 
 
 Ditto 1J „ „ = HJ „ 
 
 Shingle... ... ... ... = 9 „ 
 
 Sand (coarse) ... ... ••• = 6 „ 
 
 Thames ballast (obtained above 
 the bridges with the neces- 
 sary proportion of sand and 
 
 shingle) ... ■•• ••• = 4J „ 
 
 When the concrete is required in large thick masses, 
 the centre portion, or core, is often made with large 
 pieces or chunks of stone or old bricks. 
 
 The proportions of the ingredients in any particular 
 case depend on the composition of the aggregates, 
 without estimating the sand. Thus, for shingle 1 cubic 
 yard would require 4 cubic feet of Portland cement, 8 
 
 cubic feet of sand, and 7 or 8 gallons of water, which 
 should always be clean, and not muddy. For ballast, 
 1 cubic yard, having already 4| cubic feet of sand 
 mixed in it, would require 4 cubic feet of cement, 3 
 cubic feet of sand, and 8 gallons of water, and so on. 
 
 Ordinary proportions are 1 part of Portland cement 
 to 6 or 8 parts of gravel, according to the coarseness oi 
 the gravel and the compressibility of the soil under- 
 neath, which, by irregularity, may need a larger 
 factor of safe strength. 
 
 The materials are mixed in a dry state upon a 
 boarded platform, then tempered with the water, 
 sprinkled over gradually, then the whole turned over 
 three times. It is then laid in the footing trench, lOin. 
 to 12in. deep, in horizontal layers in convenient 
 lengths, boarded or stopped off so that it can be 
 rammed before any setting takes place. The space to 
 be thus filled at a time should correspond with the 
 quantity in the batch made each time — none should be 
 left mixed over night, or even over meal hour. The 
 top and end of each layer should be swept clean, and 
 the exuded skin removed by a pick, etc., as it would 
 prevent adhesion of succeeding layer. To ensure proper 
 adhesion between layers, it may be necessary to give 
 the lower one a thin coating of fresh cement just before 
 applying the over layer. When hydraulic lime is used 
 instead of Portland cement, it should be ground, and 
 not be used as sent from the limekiln. 
 
 Mortar. 
 
 Pat limes should not be used, but being cheap and 
 expeditious in slaking tempering, they are preferred by 
 workmen and contractors, notwithstanding their known 
 unfitness for mortar to be used in damp situation as for 
 foundations. In dry positions only a thin outer crust 
 or edge of the joint sets and becomes dry, the interior 
 remaining soft, and meanwhile all the weight of super- 
 structure rests upon these dry outer edges of the joints. 
 In brickwork this same cause produces fracture of head- 
 ing courses, leaving the wall liable to split up and the 
 outer to be detached from the inner face. Moderately 
 hydraulic lime or Portland cement should be used in 
 all ordinary foundations below the damp-proof course, 
 not only for the masonry, but also for the concrete 
 bedding. 
 
 Por foundation masonry, etc., exposed to water or to 
 water-logged soil, only a strongly hydraulic lime or 
 heavy Portland cement should ever be used, as alone 
 possessing enduring strength. 
 
 General Safe Bearing Power of Soils, etc. 
 
 (The factor of safety being 3 to 4) 
 
 Pock, the hardest in thick layers or strata in native 
 bed, has been loaded with 200 tons per square foot. 
 
 Gravel and coarse sand, the particles well 
 cemented together with clay, protected 
 from water... ... ... ... 4 to 6tons 
 
 Sand, well compacted and not liable to 
 lateral disturbance ... ... ...8 „10 „ 
 
 Sand, clean, dry do. ... ... 2 ,, 4 „ 
 
60 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Clay in thick beds and always dry ... 4 to G tons 
 
 Clay moderately dry ... •■• ••■ 2 ,, 4 „ 
 
 Clay, Central Illinois, U.S.A. without appre- 
 ciable settlement ... ■•• ■•• l^j, 2 „ 
 Clay, soft ... ... •■• ••• 1 » 1J » 
 
 The stability of clay is increased by ad- 
 mixture of sand and gravel. 
 Alluvial soil, quicksand, according to damp- 
 ness, etc. ... ... ••• ■•• J » 1 » 
 
 Tall chimney shafts, towers, steeples, etc., should 
 impose a less unit of pressure on substratum than 
 ordinary buildings because of wind leverage range. 
 
 Examples of Loads on Foundation Soils. 
 
 Blue clay mixed with sand and water ... 1 to If ton. 
 Yellow sandy clay ... ... ... 2 ,, 2f ,, 
 
 Compact clay ... ... ... 1£„ ,, 
 
 Compact clay, stony ... ... ... 5J,, „ 
 
 Clay and sand ... ... ... 3 „ 9 ,, 
 
 Alluvial soil and quicksand ... ... 0|„ Of ,, 
 
 Unstable sand ... ... ... If,, ,, 
 
 Compact sand with slight mixture of clay 2 J ,, 2f ,, 
 Wet sand, not very compact, cracked with 4 ,, ,, 
 Compact gravel and sand ... ... 7 -J- „ 8 ,, 
 
 Coarse gravel ... ... ... 4-i ,, ,, 
 
 A thin hard layer of clay resting on soft mud 
 which overlies a stratum of quicksand 1£ to 2 tons, 
 which produces subsidence of lin. per ton in one year. 
 London clay, blue, l£ to 2 tons ; red ... 4 to 6 tons 
 Blue clay \ to xs nne sand and £ water, 
 
 weighing 801b. to 1001b. per cubic foot „ 2 „ 
 Yellow clay mixed with sand ... ... ,, 2i „ 
 
 Stiff clay when kept dry, 4 to 6 tons, but 
 
 same clay saturated... ... ... 1£„ 2 ,, 
 
 Light structures with heavy running machinery or 
 rolling loads should have a minimum unit of pressure. 
 
 Heavy structures with quiescent loads may have 
 maximum unit pressure. 
 
 Laws of Subsidence in Alluvial Soft Soils. 
 
 1. "When the pressure upon the soil increases in an 
 arithmetical ratio the subsidence increases in a 
 geometrical ratio. 
 
 2. When the depth of the foundation from the 
 surface of the soil increases in arithmetical pro- 
 gression the subsidence decreases according to a 
 geometrical series. 
 
 Small and Large Pier Subsidences. 
 There is a common, but erroneous, idea that equal 
 areas of foundations should sustain equal pressures in 
 compressible soils, but walls subside more than piers, 
 and large piers more than small piers transmitting 
 equal loads per square foot of subjacent soil, because of 
 the difference ol support derived in each case from the 
 friction of the soil upon the sides of the footings. 
 Those, therefore, having the most perimeter in contact 
 with the soil per square foot of horizontal area derive 
 the most frictional support. For the same reason the 
 
 greater the depth to which foundations are sunk in soil 
 in contact with the foundation masonry the greater is 
 the sustaining power thus derived. Thus a long wall 
 footing has only its external and internal sides or edges 
 exposed to contact with the soil, but a pier has all its 
 sides. Then comparing piers, a square pier has the 
 maximum periphery — two squares, close side by side, 
 present a perimeter only of one-fourth less, and so on 
 for other rectangular proportions. 
 
 Piles are of whole timber or timber poles, or spars of 
 oak, teak, beech, elm, larch, fir, straight grained, free of 
 bark and projections of knots, etc. Large and diagonal 
 knots must be avoided, as they are apt to break off at 
 these in the driving. The butt end is driven lowermost ; 
 the head has a wrought-iron strap or hoop to prevent 
 splitting while being driven by the monkey. The 
 lower end is pointed and shod with a wrought-iron 
 pointed shoe with flaring straps by which it is spiked 
 to the pile. Some shoes are made with cast-iron 
 points. 
 
 Bearing piles 9in. to 18in. diameter of whole timber 
 are employed to support foundations upon soft soils, 
 either by reaching to a hard stratum underlying the 
 upper soil or by the frictional support of their sides 
 upon the contact soil ; in order that piles may not bend 
 by driving their length they should not be more 
 than 20 times their diameter. 
 
 Semi-Liquid Soils. 
 
 With a soil of this class, as mud, silt, or quicksand, 
 it is customary (1) to remove it entirely, or (2) to sink 
 piles, tubes, or caissons through it to a solid substra- 
 tum, or (3) to consolidate the soil by adding sand, 
 earth, or stone. Soils of a soft or semi-liquid character 
 should never be relied upon for a foundation, when 
 anything better can be obtained ; but a heavy super- 
 structure may be supported by the upward pressure 01 
 a semi-liquid soil, in the same way that water bears up 
 a floating body. 
 
 According to Eankine, a building will be supported 
 
 when the pressure at its base is w h ( , + sm a ) per 
 
 \1 - sin a/ 
 
 unit of area, in which expression w is the weight of a 
 unit volume of the soil, h is the depth of immersion, 
 and a is the angle of repose of the soil. If a = 5 deg., 
 then, according to the preceding relation, the support- 
 ing power of the soil is l - 4 wh per unit of area; if 
 a = 10 deg., it is 2"0 w h ; and if a = 15 deg., it is 2'9 
 to h. The weight of soils of this class, i.e., mud, silt, 
 and quicksand, varies from 100 to 1301b. per cubic foot. 
 Eankine gives this formula as being applicable to any 
 soil ; but since it takes no account of cohesion, for 
 most soils it is only roughly approximate, and gives 
 results too small. The following experiment seems to 
 show that the error is considerable : "A 10ft. square 
 base of concrete resting on mud, whose angle of repose 
 was 5 to 1 (a = 11-Jdeg.), bore 7001b. per square foot." 
 This is 2} times the result by the above formula, using 
 the maximum value of w. 
 
STEAM BOILERS. 
 
 61 
 
 CHAPTER VIII. 
 
 STEAM BOILERS. 
 
 BOILER is a closed vessel, which contains 
 partly steam and partly water. Steam is pro- 
 duced by applying heat to the water, whereby 
 the temperature of the water rises, becoming 
 greater and greater the more heat is added. 
 1> It is only one part of the heat, the sensible heat, 
 which is spent in raising the temperature of the water ; 
 the rest of the heat, the latent heat, is spent in chang- 
 ing the molecular state of the water into that of steam. 
 Therefore, starting with cold water in the boiler, the 
 heat generated by the fire will first raise the temperature 
 of the water, and then after some time, we shall find 
 that the hand of our pressure-gauge will begin to move, 
 indicating that the pressure of the steam within the 
 boiler has become greater than the pressure of the 
 atmosphere outside the boiler. If we continue apply- 
 ing heat, the pressure of the steam will become so 
 great that it lifts the safety-valve, thus telling us that 
 we have reached the maximum pressure at which our 
 boiler ought to be worked. 
 
 The temperature of the water in the boiler will rise 
 with the pressure, but if we keep the pressure constant 
 the temperature will also keep constant. The relation 
 between temperature and pressure of steam, generated 
 in a boiler, has been determined by very accurate ex- 
 periments by the French philosopher Regnault, and 
 the result, of these experiments are shown in the table 
 given below. 
 
 Suppose we have a cylinder with a tight -fitting 
 piston, and that on the one side of the piston we have 
 steam, in direct connection with a boiler, at a certain 
 temperature, and on the other side of the piston we 
 have a perfect vacuum ; then the pressure, in pounds 
 per square inch, by which the piston is pushed along 
 in the cylinder, is called the absolute pressure of the 
 steam, and will in this book be denoted by p a . If, 
 instead of a perfect vacuum, we had the atmospheric 
 pressure — i.e., 14-71b. per square inch — on the piston, 
 then the pressure, by which the piston would be 
 pushed along, is called the effective pressure of the 
 steam, and will be denoted by p e . In both cases we 
 must imagine that there is no friction between the 
 piston and the cylinder. 
 
 Steam, as generated in an ordinary steam-boiler, is 
 probably never perfectly dry, but contains suspended 
 water which has been carried away by the steam, just 
 at the moment it left the water from which it is 
 generated. We can, however, dry the steam by carry- 
 ing it into a closed vessel, which is connected by a pipe 
 to the steam-room of the boiler, and there heat it. We 
 ghall then find that the temperature of the steam will 
 
 remain the same as that in the boiler, as long as the 
 steam contains suspended water. Just when the last 
 drop of suspended water is evaporated, the steam being 
 perfectly dry, it is said to be dry saturated steam, 
 whereas the steam in the boiler is called saturated 
 steam simply. 
 
 If we continue to apply heat to the steam after it is 
 dried, we shall find that its temperature will rise, and 
 the steam will expand ; steam in this condition is called 
 superheated steam. The pressure of the superheated 
 steam will, of course, remain the same as that of the 
 steam in the boiler, on account of the connection with 
 the boiler. It is evident that the mass of the super- 
 heated steam contained in the small closed vessel, will 
 be smaller than that of the dry saturated steam which 
 the vessel previously contained. We may therefore 
 say that dry saturated steam is steam of maximum 
 density, and that its temperature is the lowest possible 
 of steam at a particular pressure. 
 
 It is of importance to engineers to know the 
 volume, in cubic feet, of one pound of steam — i.e., the 
 specific volume of steam ; and also the mass in pounds 
 of one cubic foot of steam — i.e., the density of steam. 
 The former will be denoted by v, and the latter by w. 
 They both vary with the pressure, and some of their 
 values are given in the annexed table, together with the 
 corresponding pressures and temperatures. 
 
 Table showing Relation between 
 
 Pressure 
 
 Temperature, 
 
 Specific Volume, asd Density or Dry Saturated Steam. 
 
 'a 
 
 t 
 
 V 
 
 w 
 
 P 
 a 
 
 t 
 
 V 
 
 w 
 
 15 
 
 213 
 
 25-87 
 
 •0387 
 
 100 
 
 327-6 
 
 
 
 
 
 20 
 
 228 
 
 19-74 
 
 ■0506 
 
 110 
 
 334-5 
 
 3-986 
 
 •2509 
 
 25 
 
 240 
 
 1601 
 
 •0625 
 
 120 
 
 341 
 
 
 
 — 
 
 30 
 
 2502 
 
 13-49 
 
 •0741 
 
 130 
 
 347 
 
 — 
 
 — 
 
 35 
 
 259 
 
 — 
 
 — 
 
 140 
 
 352-8 
 
 3-177 
 
 •3148 
 
 40 
 
 267 j 10-30 
 
 •0971 
 
 150 
 
 35S-1 
 
 — 
 
 — 
 
 45 
 
 274-3 j — 
 
 — 
 
 160 
 
 363-3 
 
 — 
 
 — 
 
 50 
 
 280 9 i 8-347 
 
 •1198 
 
 170 
 
 368-1 
 
 2-645 
 
 •3780 
 
 55 
 
 287 
 
 — 
 
 180 
 
 372-8 
 
 — 
 
 — 
 
 60 
 
 29-2-5 7-037 
 
 ■1421 
 
 190 
 
 377-3 
 
 — 
 
 — 
 
 65 
 
 297 8 
 
 — 
 
 — 
 
 200 
 
 381-6 
 
 2-270 
 
 ■4404 
 
 70 
 
 302-7 
 
 6-090 
 
 •1642 
 
 210 
 
 385-5 
 
 — 
 
 — 
 
 75 
 
 307-4 
 
 — 
 
 — 
 
 220 
 
 389-7 
 
 — 
 
 — 
 
 80 
 
 311-8 
 
 — 
 
 — 
 
 230 
 
 393 5 
 
 — 
 
 — 
 
 85 
 
 316 
 
 — 
 
 — 
 
 240 
 
 397 2 
 
 — 
 
 — 
 
 90 
 
 320 
 
 4-810 
 
 •2079 
 
 250 
 
 400-8 
 
 1-841 
 
 ■5432 
 
 95 
 
 323-9 
 
 — 
 
 — 
 
 260 
 
 404-3 
 
 — — 
 
 We have now to consider the quantity of heat 
 required to raise the temperature and to evaporate the 
 water in our boiler. This has also been most accu- 
 rately determined by Regnault, and the result of his 
 experiments is, that the amount of heat, H, required to 
 raise the temperature of one pound of water from 
 32deg. F. to t deg. F., and then evaporate it at that 
 temperature, is 
 
 H =1082 + 0-305 t (1) 
 
62 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 H is called the total heat of evaporation of water, and 
 is expressed in British heat units. We must remember 
 that a British heat unit is the amount of heat requisite 
 for raising the temperature of one pound of water from 
 82deg. F, to 83deg. F., at a pressure of one atmosphere. 
 The temperature of the feed-water which we pump 
 into our boiler is never as low as 32deg. F., but is, say, 
 £ideg. F. ; we have, therefore, only to raise the tempera- 
 ture of the water from ^deg. to t deg., and the heat we 
 have to apply, in order to turn one pound of feed-water 
 into steam, will be 
 
 H - (^ - 32) = 1114 + 0'305 xf-i, 
 
 (2) 
 
 The temperature of the water in the different classes of 
 boilers varies from about 300deg. F. to 390deg. F. ; 
 the total heat of evaporation will therefore vary 
 between 1,173 and 1,200 heat units, or less than 3 per 
 cent. The heat required for producing a certain 
 amount of steam, will therefore practically be inde- 
 pendent of the pressure to which the steam has to be 
 raised. 
 
 Combustion and Fuel. — We have now to consider 
 the best way of producing the heat which is required 
 for our boiler. For this purpose we burn fuel. Gene- 
 rally speaking, burning or combustion is a rapid chemi- 
 cal combination, but the only kind of combustion 
 which is used to produce heat for driving steam-engines, 
 is the combination of the constituents of various kinds 
 of fuel with the oxygen which is contained in the 
 atmosphere. 
 
 The chief combustible constituents of fuel are carbon 
 and hydrogen. Of these, hydrogen is the best; for one 
 pound of hydrogen will, by burning with oxygen, de- 
 velop 62,030 heat units, whereas one pound of carbon, 
 when completely burnt, will only develop 14,500 heat 
 units. 
 
 Sulphur is also often met with in ordinary fuel, but 
 the heat developed by burning it is comparatively 
 small, and the product of combustion is bad for the 
 boiler. 
 
 The heat developed by the complete burning of a 
 fuel is called its total heat of combustion. 
 
 As one pound of water at 212deg. F. requires about 
 966 heat units for its complete conversion into dry 
 saturated steam, it is evident that one pound of hydro- 
 gen by its combustion with oxygen can completely 
 
 62 030 
 evaporate -^__ = 641bs. of water at 212deg. F. The 
 
 number 64 is called the evaporative power of hydrogen. 
 If we therefore say that the evaporative power of 
 ordinary coal is 15, we mean, that if all the heat pro- 
 duced by the complete combustion of one pound of this 
 coal could be made useful, it would be able to evapo- 
 rate 151b. of water at 212deg. F. In practice, however 
 this result is impossible, because the combustion can- 
 not be perfectly completed, nor can all the heat be 
 made useful. It is, however, of great importance to 
 know the total heat of combustion of a fuel, in order to 
 judge whether a furnace is well constructed or not. 
 The available heat of combustion of a fuel, when 
 
 burnt in a boiler-furnace, is that part of the total heat 
 of combustion which is actually absorbed by the water, 
 and the efficiency of a boiler for a certain kind of fuel, 
 is the ratio of the available heat to the total heat, when 
 the given kind of fuel is burnt in the boiler-furnace. 
 We shall see later on that one boiler may be more 
 efficient for one kind of fuel than for another ; and it 
 may also be more or less efficient for the same kind of 
 fuel, according to the way the firing and the boiler on 
 the whole are managed. 
 
 The following are the principal causes why the 
 available heat is smaller than the total heat of com- 
 bustion : 
 
 (1) By careless stoking and handling of the fuel with 
 the shovel, part of the coal is made to break into small 
 pieces, which will fall through the openings between 
 the firebars. 
 
 To prevent this waste of unburnt fuel, (a) the coals 
 should be thrown evenly over the fire with the shovel, 
 and (b) the layer of fuel should not be so thick as to 
 make it necessary to stir the fire from above, (c) The 
 firebars should be cleared from below, (d) the ashes 
 should be riddled, and the small coals should be thrown 
 on the fire again. 
 
 (2) If the supply of air to the fire is insufficient, a waste 
 of fuel will take place in the form of smoke, which is fuel 
 in waste state. It is therefore of great importance to 
 produce a necessary draught in the furnace. In the 
 annexed table, W denotes the mass of one cubic foot of 
 fuel, H t the total heat of combustion per pound of fuel, 
 and A the mass of air, in pounds, which is theoreti- 
 cally required for the complete combustion of one 
 pound of fuel. It is, however, necessary to furnish an 
 additional supply of air for carrying away the gaseous 
 products of combustion at the moment they are formed, 
 in order to give free access of air to the fuel. 
 
 The greater the velocity of the air and the smaller the 
 fuel, the smaller is the additional air required for dilu- 
 tion. It has been found in practice, that the amount of 
 additional air to be supplied to a boiler-furnace with 
 ordinary chimney draught, must be equal to the amount 
 of air required for the complete combustion of the fuel. 
 The total amount of air required per pound of bitu- 
 minous coal will therefore be 221b. 
 
 Fuel. 
 
 Wood, dry 
 Peat, dry .. 
 Bitum. coal 
 Coal, dry .. 
 Coke 
 
 AY. 
 
 25 
 
 20 
 47 
 55 
 25 
 
 H t 
 
 5,950 
 
 7,750 
 
 15,000 
 
 11,500 
 
 12,800 
 
 5-5 
 60 
 
 11 
 
 101 
 
 11 
 
 (3) In order to produce a draught in a boiler-furnace 
 by means of a chimney, the gases which escape by the 
 chimney must have a high temperature — hence a con- 
 siderable loss of heat. 
 
 (4) Waste by Badiation of the Incandescent Fuel— 
 According to Peclet, the heat radiated from incandes- 
 cent fuel is, from wood 0'25, and from coal and coke 
 0"5 of the total heat of combustiou. 
 
STEAM BOILERS. 
 
 63 
 
 For this reason it is of the greatest importance that 
 the firebox of the boiler should be constructed in such 
 a manner, that the radiated heat is carried into the 
 furnace and made useful. This is generally done in 
 one of two ways — viz. : (a) The firebox and ashpit are 
 contained within the boiler, whereby all the radiated 
 heat will be intercepted by the boiler, and is thus made 
 useful ; (6) the firebox and ashpit are outside the 
 boiler. In this case the firebox is built of brickwork 
 lined with firebrick. As both are bad conductors to heat, 
 the temperature of the firebrick lining will be very 
 high. The heat will therefore be given to the gases 
 which strike the lining, and thus carried into the 
 furnace ; the greater part of the heat radiated down to 
 the ashpit will be made useful in heating the fresh air 
 while it is passing through the ashpit to the grate. 
 
 To prevent heat from radiating through the holes in 
 the fire-door, it is usual to fix a plate to the inner side 
 of the door ; the heat absorbed by this plate will partly 
 be carried into the furnace by the air entering through 
 the holes. 
 
 (5) Some heat will be carried away through the brick- 
 work, although this is a bad conductor. As the boiler 
 cannot altogether be covered with good insulators, 
 some heat will be lost through the naked parts of the 
 boiler. 
 
 Practical Determination of the Available Heat of 
 Combustion. — The fuel should be burnt in a well-con- 
 structed boiler, and the water should either be evapo- 
 rated under the pressure of the atmosphere — i.e., the 
 generated steam should be allowed to escape into the 
 air — or else the steam should work a steady-running 
 engine. Variations in steam pressure in the boiler 
 should be avoided. 
 
 Before starting the experiment, the fire must be in 
 the normal condition, and the height of the water 
 in the boiler, as well as the temperature of the feed- 
 water, must be noted. The firing should then be kept 
 up very regularly, and the masses of the coal and feed- 
 water are measured. The experiment is then carried 
 on in this manner for several hours. In finishing the 
 experiment, the fire must be in exactly the same condi- 
 tion as when the experiment began, and the height of 
 the water in the boiler must also be the same. 
 
 Let the mass of feed- water evaporated be F, and 
 
 its temperature be t v the temperature of the steam be 
 
 t, and the mass of the coal consumed be C, then the 
 
 available heat of combustion, H a , of lib. of fuel, will be 
 
 F(1114 + -305 t - y 
 
 In order to find the relative available heat of combus- 
 tion of various kinds of fuel, these must be burned in 
 the same furnace and under the same conditions. In 
 the same way we can find the relative efficiency of 
 various kinds of furnaces, by burning the same kind of 
 fuel in the furnaces to be examined. 
 
 These experiments, however, will not give a useful 
 result, unless the following precautions are taken : 
 
 (1) That the engine must be worked at constant speed 
 and power, 
 
 (2) That the firing is kept up very regularly, and the 
 same amount of coal is burnt in the same interval of 
 time. 
 
 (3) That the temperature of the feed-water is con- 
 stant. 
 
 (4) That the steam-room of the boiler be large enough 
 to avoid variations in steam pressure, as it is otherwise 
 difficult to observe the exact height of the water in the 
 boiler. A small steam-room also causes water to be 
 carried over to the cylinder by the steam, whereby the 
 exact quantity of water evaporated cannot be measured. 
 
 By the temperature of the fire is understood the 
 temperature of the gaseous products of combustion and 
 the additional quantity of air with which they are 
 mixed. The temperature of the fire can be measured 
 by means of a pyrometer, but it can also be calculated ap- 
 proximately. The mean specific heat o/ the products 
 of combustion, including air and ashes, may be taken 
 as 0"24 ; the total amount of air required to burn one 
 pound of fuel, when the draught is produced by means 
 of a chimney, is 2 A, the temperature of the fire will 
 therefore be 
 
 H t 
 
 ... (4) 
 
 t = 
 
 0-24 {-J, A + 1) 
 
 The values of Ht and A to be taken from table on 
 preceding page. 
 
 If the air be supplied at a temperature of ^deg. F., 
 then the temperature of the fire will be (t + tj deg. F. 
 
 General Description of Steam Boilers. 
 
 As fuel, coal, coke, wood, and peat have been chiefly 
 used, but in later years liquid fuel has been introduced, 
 especially in Bussia and in America. The right fuel to 
 be used at- any locality is that of which the ratio be- 
 tween the total heat of combustion and the price is a 
 maximum, with due consideration to the fact that the 
 boiler must be constructed to suit the fuel. Thus in 
 certain parts of Central America rosewood is burnt 
 under the boilers. In Cuba the refuse of sugar-cane is 
 used. According to Messrs. Babcock and Wilcox, their 
 boilers at the Chicago cable railways are worked regu- 
 larly on the offal from the stables of the horse-roads, a 
 very small proportion of coal being used to keep it 
 alight. " Slack " is also often burnt under boilers with 
 economy. 
 
 The heat which is contained in the gaseous products 
 of combustion, which latter hereafter will be called the 
 gases, is conveyed from the firebox to the boiler. 
 This is done by letting the gases pass through the flues 
 on their way to the chimney. These may be external, 
 and are then partly bounded by the outside of the 
 boiler-shell, and partly by the brickwork surrounding 
 the boiler ; or else they may be internal, and are then 
 wholly contained within the boiler. 
 
 That part of the boiler-surface which is in contact 
 with the gases is called the heating-surface, and it is 
 through this that the heat of the gases is communicated 
 to the water. The temperature of the gases will thus 
 diminish in proportion as the heating-surface is increased, 
 a.nd will reach the chimney at a temperature which is 
 
64 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 higher than that of the boiler, but will approach it the 
 closer, the greater the heating-surface. The chimney 
 not only carries away the gases, bat also produces a 
 draught by means of the high temperature of the gases. 
 We have seen that the total heat of combustion of 
 one pound of good coal is about 15,000 heat units. The 
 temperature of the fire will therefore be 
 15,000 
 
 t 
 
 •24 x 23 
 
 = 2,700° F. 
 
 and the heat-waste will be, taking the temperature of 
 the chimney as 500 deg. F., 
 
 •24 x 23 x 500 = 3,000 heat units, 
 
 or 20 per cent, of the total heat produced by the com- 
 plete combustion of the fuel. In order to diminish 
 this heat-waste, the combustion must be completed 
 with as small a quantity of air as possible. 
 
 The Fireplace of a boiler consists of three parts — viz.: 
 (1) the grate on which the fuel is placed ; (2) the fire- 
 box, above the grate, in which the constituents of the 
 fuel are burnt ; and (3) the ashpit, below the grate, 
 into which the ashes fall and the supply of air is 
 admitted. 
 
 The grate is composed of firebars, which support 
 the fuel, and open spaces between the bars for the 
 admission of the necessary air to the fire. The firebars 
 are made of cast iron, in length from 2ft. to 3£ft. 
 Their thickness diminishes towards their lower edge, in 
 order to admit of the free entrance of air and the better 
 escape of the ashes. At each end, as well as at the 
 middle of the bar, are projections whereby the bars 
 are kept apart, leaving the proper spaces between them 
 for the admission of the air. The firebars are laid on 
 the cross bearers. It is necessary to allow an end play 
 for the expansion of the bars : this play is generally 0"02 
 of the length of the bar. 
 
 For the purpose of saving time when removing the 
 bars, they are often cast in pairs ; this is hardly a good 
 plan, as the two bars seldom expand equally, and will 
 therefore bend, leaving uneven spaces between them. 
 
 The surface of the grate is either horizontal or, more 
 usually, a little inclined towards the back. The length of 
 the grate must never be more than 7ft. and the width 
 not more than 5ft., in order to make the firing easy. 
 "When the grate is placed in an internal flue, then the 
 width is determined by the diameter of the flue. The 
 total open space between the firebars should be 
 
 For coal and cokes from \ to \ of the total grate area. 
 For wood and peat from \ to | ,, ,, ,, 
 
 The size of the grate area depends upon the amount 
 of fuel which is to be burnt during a certain interval of 
 time. If, now, the velocity of the air passing through 
 the grate is 4ft. per second, and the amount of air 
 necessary for the complete combustion of the fuel is 
 twice the theoretical, then the weight of fuel which can 
 be burnt per square foot of grate-surface per hour will 
 be 
 
 For coal and coke 141b. to 181b. 
 
 For wood and peat 181b, to 221b, 
 
 By using forced draught, the amount of fuel which 
 can be burnt per hour per square foot of grate-surface 
 may be more than doubled. 
 
 The firebox is bounded at the top and the three 
 sides by firebrick, or by part of the boiler-surface, or 
 by a combination of both. In the front part of the 
 firebox is the mouthpiece, being a cast-iron frame, 
 through the opening of which the fuel is introduced. 
 The mouthpiece is closed by the fire-door. When 
 single doors are used, the doorway is taken from llin. 
 to 14in. wide, and 9|in. to llin. high. With double 
 doors the total doorway is taken from 17|in. to 21in. 
 wide, and llin. to 14in. high. 
 
 Fig. l. 
 
 Fig. 1 shows cross-section of a fire-door, which is pro- 
 vided with air-holes at the top, so that the air, entering 
 through the holes, will carry some of the heat contained 
 in the baffle-plate into the furnace, and strike the fire 
 on the surface just where the hydrocarbons are liberated. 
 
 Fire-doors are generally provided with a circular 
 slide, which, by being turned, will more or less close 
 the air-holes, thus regulating the admission of air into 
 the furnace. When bituminous coal is used, each fresh 
 charge of coal should be laid in the front of the fire 
 until coked, then pushed further back ; the hydro- 
 carbons contained in the coal will thereby be more 
 completely burnt, as more air can get through the 
 grate. For this purpose boiler furnaces are provided 
 with a cast-iron plate, the dead plate, which is not 
 perforated, and which is placed between the fire door- 
 way and the firebars. The fresh coal is placed on the 
 dead plate, and coked by the radiated heat from the 
 fire. 
 
 A.t the back of the firebox is the bridge, over which 
 the gases pass to the flues. The object of the bridge 
 is partly to cause the air, which enters at the back of 
 the grate, to rise in a perpendicular direction through 
 the fuel, and partly to mix the gases over the bridge, 
 and thereby effect a more complete combustion. The 
 area of the opening above the bridge should be three- 
 fifths to four-fifths of the total air passage through the 
 grate. 
 
 The ashpit is often provided with one or two doors 
 in the front, which can be opened more or less, and thus 
 regulate the supply of air to the fire. 
 
 Flues. — The shape and sectional dimensions of the 
 flues mugt be such that the gases, while passing through 
 
STEAM BOILERS. 
 
 65 
 
 the flues, can come into contact with the heating- 
 surface. 
 
 From this follows that the current of the gases 
 should be as much as possible at right angles to the 
 heating-surface, and that the passages through the 
 flues must not be too wide ; on the other hand, narrow 
 passages increase the resistance to the draught. The 
 proper sectional area of the flues depends upon the 
 amount of the fuel to be burnt per hour, and upon the 
 volume of the gases. 
 
 The latter again depends upon the force of the 
 draught, whereby more or less air is required for the 
 complete combustion of the fuel. In practice the sec- 
 tional area of the flues is made about equal to the total 
 area of the spaces left between the firebars. The flues 
 may be arranged as shown in Fig. 2, where the gases 
 pass first through the internal flue to the back of the 
 
 M H 
 
 the cost of the boiler. This resistance is further in- 
 creased when the flues are dirty ; doors for cleaning 
 should therefore be placed in such a manner that all 
 parts of the flues can easily be swept. The flues should 
 have no sharp bends, and their surface should be 
 smooth. 
 
 As already mentioned, the gases will give off their 
 heat to the boiler while passing through the flues on 
 their way to the chimney. When the flues are partly 
 bounded by brickwork some of the heat will be 
 absorbed by the latter, which is a bad conductor for 
 heat, and the heat-loss through the brickwork is there- 
 fore small. 
 
 The boiler-surface, on the contrary, is a good con- 
 ductor for heat, and, being in contact with the water- 
 on the other side, will give off to the water during any 
 interval of time just as much heat as it receives during 
 the same time. 
 
 The amount of heat given to the water in unit of 
 
 boiler, where they divide into two currents, one flowing 
 through each lateral flue towards the front of the boiler, 
 then uniting into one current, they pass through the flue 
 underneath the boiler, and then into the chimney. 
 Sometimes, however, the gases do not divide after hav- 
 ing left the internal flue, but pass first through the 
 underneath flue, then divide and flow through the 
 lateral flues into the chimney. In both cases the flues 
 are said to be arranged as a split draught. 
 
 The arrangement of the flues as shown in Fig. 3 is 
 called a wheel draught. The gases flow first through 
 the internal flue to the back of boiler, then through one 
 of the lateral flues towards the front, and then through 
 the other lateral flue into the chimney . 
 
 By increasing the length of the flues we increase the 
 heating-surface, but at the same time we also increase 
 the resistance to the free passage of the gases, and also 
 
 time depends upon the amount of fuel burnt in unit of 
 time, upon the size of heating- surface and upon the 
 temperature of the fire, which will be the greater the 
 stronger the draught is, as then less air is required for 
 the complete combustion. Call, therefore, the tempera- 
 ture of the gases t, and that of the water t v then the 
 heat received by the water in unit of time will be 
 proportional to (t - tj, and to the conductivity of the 
 material of which the heating-surface is composed. 
 In new boilers this material is simply the steel or 
 wrought iron of which the boiler is made, and which 
 are both good conductors, and therefore the thinner 
 the shell is the greater will be the amount of heat 
 transferred from the gases to the water. A thin shell 
 will for the same reason deteriorate at a slower rate, as 
 its temperature will approach more to that of the 
 water. 
 
 In boilers which have been used for some time the 
 flue-shells are covered with soot on the gas side, and 
 
66 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 with a hard incrustation on the water side, both of 
 which are bad conductors. The layer of soot and 
 ashes may become so thick, when the flues are not 
 properly cleaned, that hardly any heat will be trans- 
 ferred to the shell. The incrustation resists the heat 
 being transferred from the shell to the water, 
 whereby the former may become red hot, and will 
 deteriorate quickly. But even if the flue-shells are 
 kept clean in every respect, the heating-surface of an 
 old boiler can never be so efficient as that of a quite 
 new one, as the incrustation and the soot cannot be 
 entirely removed from the flues. 
 
 It is well known that water is a bad conductor for 
 heat, and must therefore be heated by convection. 
 The flues must therefore be placed in the boiler in 
 such a manner that a rapid circulation will be produced. 
 It is also necessary that the heated water and the 
 generated steam should be able to escape easily from 
 the heating'Burface. For this reason the top part of 
 an internal flue is more effective than the lower part, 
 and the lower part of an external flue is more effective 
 than the lateral part. We may therefore distinguish 
 between effective-heating surface and total heating- 
 surface. By heating-surface in this book is always 
 understood total heating-surface. 
 
 The following table gives the relation between the 
 heating-surface, H s , in square feet, of a boiler ; the 
 number of pounds, S, of steam generated per hour, and 
 the number of pounds, B, of coal burnt per hour. 
 
 
 H 3 
 B 
 
 S 
 B 
 
 Ordinary boilers 
 
 2 
 
 6 7 
 
 
 
 Economical boilers 
 
 3J 
 
 9-10 
 
 
 
 Locomotive boilers with 
 forced draught 
 
 1 
 
 a 
 
 
 
 The Chimney.— The object of the chimney is partly 
 to get rid of the gases, and partly to produce a draught 
 by means of the hot gases within the chimney, 
 which have, on account of their high temperature, a 
 smaller density than the air outside the chimney. 
 
 The pressure of the draught, expressed in pounds per 
 square foot, must therefore be equal to the weight in 
 pounds of a column of cool air outside the chimney, 
 with a base of one square foot and of the height of the 
 chimney, minus the weight in pounds of a column of 
 the hot gases inside the chimney, with a base of one 
 square foot and of the height of the chimney. 
 
 Let h denote the height of the chimney and y the 
 weight of one cubic foot of air at an absolute tempera- 
 ture T 32 , corresponding to 32 deg. F., and at a pressure 
 of one atmosphere ; then the weight of the column of 
 air outside the chimney will be 
 
 T 
 v x h x =A 2 
 T 
 
 (1) 
 
 where T a is the absolute temperature of the air, 
 Assuming that the density of the hot gases within the 
 chimney is the same as that of the air outside the 
 chimney at the same temperature and pressure, then 
 the weight of the column of hot gases within the 
 chimney will be 
 
 yxhxp! (2) 
 
 where T c is the absolute temperature of the hot gases. 
 
 The pressure of the draught in pounds per square 
 
 foot will then be 
 
 T - T 
 y x h x T 32 x rj/ x T " . ... (3) 
 
 The draught could also be expressed by the height, hi, 
 of a column of air at the absolute temperature T c and at 
 a pressure of one atmosphere. The weight of such a 
 column of air must be equal to (3) , or we must have 
 
 y x h x T 32 x T? ^L = y x h' x 2-ffi 
 
 y T«xT, Y T c 
 
 which gives us 
 
 K 
 
 T„ 
 -hx. — - 
 
 T a 
 
 T„ 
 
 (4) 
 
 (5) 
 
 The resistance to the draught can be expressed by the 
 height, h", of a column of air at temperature T c , and 
 Peclet has found that 
 
 *-£ [*♦"■!] 
 
 (6) 
 
 where v is the velocity of the draught, in feet, in the 
 chimney, d the inner diameter of the chimney at the 
 top, I the whole length of the chimney and the flues, G 
 a factor of resistance for the passage of the air tbrough 
 the grate and the layer of fuel, the value of which ac- 
 cording to Peclet is 13 for ordinary firing with coal, / a 
 factor of resistance for the passage of the gases over 
 sooty surfaces of the flues, the value of which is, accord- 
 ing to Peclet, 0'05. We must now have h'=\", which 
 will give us 
 
 hx 
 
 i-C 1(1 
 
 2.7 L 
 
 13+ 0-05 x 
 
 I 
 
 ] 
 
 (7) 
 
 T„ 2gL~ d. 
 
 From this equation we can find either h when v is 
 given, or v when h is given. 
 
 The quantity of gases which can escape through a 
 given chimney must be proportional to v, the sectional 
 area, A, of the chimney, and the density of the gases. 
 The volume of the gases which escapes through the 
 chimney per second is, when reduced to T 32 , 
 
 Q-^^" (8) 
 
 T„ 
 and the weight in pounds of same 
 
 yxQ = yx Ax 
 
 T 
 
 82 
 
 T. 
 
 Jg_ha\_-Ta) 
 
 T a [lc 
 
 13 + 0-05 
 
 dJ 
 
 (9) 
 
 This equation shows that there exists a certain value 
 of T„ which would make y x Q a maximum. 
 
STEAM BOILERS. 
 
 67 
 
 The diagram, Pig. 4, has been given to the author by 
 the Babcock and Wilcox Company, and shows the 
 draught, in inches, of water for a chimney 100ft. high, 
 under different temperatures, from 50 deg. F. to 
 800 deg. F. above the external atmosphere, which 
 is assumed at 60 deg. F. Bach vertical division 
 represents 0'05 of an inch. It also shows the 
 relative quantity, in pounds of air, which would be 
 delivered, in the same time, by a chimney under the 
 same temperature. It will be seen that practically 
 nothing can be gained by carrying the temperature of 
 the chimney more than 350 deg. F. above that of the 
 external air. It also shows the temperature at which 
 the quantiby of air delivered would be a maximum; 
 in this case about 575 deg. F. 
 
 DIAGRAM OF DRAFT AND CAPACITY OF CHIM 
 
 Fi|.4. 
 
 The draught produced by a chimney varies, however, 
 with the force and direction of the wind, and with the 
 temperature of the air outside ; and the mass of a certain 
 volume of air varies with the barometric pressure and 
 the temperature. For these reasons it will be neces- 
 sary to have means by which the draught can be regu- 
 lated. This is done by raising or lowering the damper, 
 an iron plate which can slide in a frame in the 
 brickwork, and which is placed at the back of the flues, 
 where the gases enter the chimney. The ashpit door 
 can also be used for the regulation of the draught. 
 
 The height of the chimney for stationary boilers 
 should not be less than 60ft., but is generally much 
 more — exceeding 100ft. The internal sectional area 
 of the chimney at the top is usually made equal to the 
 sum of the openings between the firebars ; lower down 
 in the chimney the area is larger. 
 
 Chimneys are built of brickwork, which lasts long, 
 and is a bad conductor for heat. For stability's sake 
 the wall-thickness is made to increase from the top 
 downwards. Thus the chimney may be half a brick 
 thick for the first 20ft. from the top, then one brick 
 thick for the next 20ft., etc. In Fig. 5 is shown a 
 chimney made of brickwork. 
 
 Iron chimneys are used with marine boilers, loco- 
 motives, and for temporary purposes. A chimney 
 must be so proportioned that it can withstand the 
 turning force of the wind. Let W denote the weight, 
 in pounds, of the chimney, b the width at base, d the 
 average diameter, and h the height of the chimney, all 
 in feet, then the relation between these quantities 
 must be 
 
 where C is a coefficient of wind pressure per square 
 
 
 
 
 Plaru at Cap 
 
 ilOfi 
 
 Z3 
 
 Plaa/a^fLearlim 
 i 
 i 
 
 Fig. 5. 
 
 Seetivn/ 
 
68 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 foot, and is, according to Messrs. Babcock and Wilcox, 
 56 for a square, 35 for an octagon, and 28 for a round 
 chimney. Brickwork weighs from 1001b. to 1301b. 
 per cubic foot. 
 
 Forced draught is always used with locomotives, 
 torpedo boats, and generally with boilers which have 
 a short chimney. For this purpose may either be used 
 a blast-pipe in the chimney, or a centrifugal blower, by 
 which air is forced up through the grate. In the first 
 case, the draught is produced by the kinetic energy of 
 the steam, which passes through the blast, being com- 
 municated to the gases. 
 
 When the draught is produced by a blower, the ash- 
 pit door as well as the fire-door must close air-tight ; or 
 the boiler-room must be air-tight, the blower is in the 
 latter case placed outside the boiler-room. The pressure 
 of the draught is usually 2|in. — 5in. of water. 
 
 The Figure of Boilers. — The conditions under 
 which boilers are required to work vary largely, and 
 have therefore given rise to a great many different 
 boilerforms. The shape of the boiler has to be con- 
 sidered, partly with regard to the strength, and partly 
 with regard to the ratio between the heating-surface 
 and the volume of the boiler ; the latter is divided into 
 two parts, the water-room and the steam-room. 
 
 The best figures for boilers are the sphere form and 
 the cylinder. The sphere form would be perfect, as 
 the pressure of the steam would not tend to alter the 
 shape, but it would be expensive to make and incon- 
 venient to use. The cylinder form will also be un- 
 altered, but the ends have to be closed. This could 
 best be done by half-spheres, but as they are expensive, 
 flat ends are used, which, on account of their small 
 dimensions, can be made strong enough. 
 
 The water-room and the steam-room act as regulators 
 for the steam pressure, and the better, the larger they 
 are. The steam is never taken regularly out of the 
 boiler, and it is therefore evident that the larger the 
 steam-room is, the less will the pressure vary. The 
 water-room, however, has a still greater regulating 
 effect. This can best : be understood by an example. 
 
 Assume we have a boiler with a steam-room equal 
 to the water-room, and let the absolute pressure of the 
 steam be p a = 1101b., the temperature will then be 
 334-5 deg. F. The question is now, How much steam 
 can we take out of the boiler at once without diminish- 
 ing the pressure more than 2 per cent. ? 
 
 The diminished pressure will then be 107'81b., and 
 the corresponding temperature will be about 333 deg. F. 
 On account of the lower pressure, the water will give off 
 so much steam that the temperature sinks to 333 deg. F 
 Now the heat required for evaporating one pound of 
 water at 333 deg. F. will, according to art. 5, formula 
 (2), be 
 
 1114 + 0-305 x 333 - 333 = 883 heat units. 
 As the temperature of the water has fallen 1-5 deg F 
 each pound of water will give off 1-fi heat units to be 
 spent m producing steam. Let V denote the volume 
 m cubic feet of the water-room, which is equal to the 
 steam-room, then the mass of the water in the boiler 
 
 will be 62 - 4 x V pounds, and the amount of steam pro- 
 duced will be 
 
 1/5 
 
 883 
 
 x 62-4 x V pounds = 0-106 x V pounds. 
 
 As the [specific volume of steam at 333 deg. F. is 
 4-078, the total volume of steam produced by reducing 
 the pressure 2 per cent, will be 
 
 0-106x4-078 V= 0-432 xV cubic feet. 
 
 The steam-room could only give off 0'02 x V cubic 
 feet of steam, the water-room can therefore give off 
 about 22 times as much steam. 
 
 It will thus be seen that the property of the boiler 
 to regulate the steam pressure depends chiefly upon a 
 large water-room. It must, however, be well under- 
 stood that as the generated steam must leave the water 
 at the surface of the water, the regulating effect of the 
 water-room depends upon the size of the water-sur- 
 face, and also upon the depth of the water. The 
 greater the former is, and the smaller the latter, the 
 better will the boiler regulate. 
 
 By having a large water-room the cooling of the 
 water when fresh water is pumped into the boiler will 
 be greatly diminished, and thus the variation in steam 
 pressure will also be diminished. The water-room of 
 portable boilers is made smaller than that of stationary 
 boilers, for the purpose of diminishing the bulk of the 
 boiler. A small water-room is necessary when it is 
 of importance to get steam up quickly. 
 
 The steam-room should be high, in order to prevent 
 water from being carried into the steam-pipe by the 
 steam. For this purpose boilers are often provided 
 with a dome, being a closed vessel on the top of the 
 boiler, into which the steam must pass before entering 
 the steam-pipe. 
 
 Classification of Boilers. 
 
 It is usual to class boilers as stationary boilers and 
 portable boilers. But the portable form is often used 
 for stationary purposes, as it combines a large heating 
 surface with a small bulk. As, however, all boilers at 
 the present day are cylindrical, the author proposes to 
 class them as horizontal boilers and vertical boilers, 
 according to the direction of the axis of the shell. 
 
 For the purpose of assisting the reader in under- 
 standing the drawings shown in this book, the following 
 lettering has been adopted to indicate the various parts 
 of the boilers : 
 
 Butt-joint B J 
 
 Steam-dome SD 
 
 Main shell . . . . MS 
 
 Dome neck DN 
 
 Standpipe SP 
 
 Steam outlet ... SO 
 
 Fusible plug FP 
 
 Gusset stay G S 
 
 Longitudinal stay L S 
 
 Suspension stay ... SS 
 
 Stay-tube ST 
 
 Crown stay C S 
 
 Stay-bolt sb 
 
 Feedwater-pipo . . F W P 
 
 Fire-door 
 
 ... FD 
 
 Fire-doorway 
 
 FDW 
 
 Grate 
 
 G 
 
 Ashpit 
 
 ... AP 
 
 Ashpit door ... 
 
 ...APD 
 
 Firebox 
 
 ... FB 
 
 Bridge 
 
 ... Br 
 
 Corrugated firebox 
 
 crown 
 
 ...CFB 
 
 Corrugated flue 
 
 ... CF 
 
 Galloway tube 
 
 ... GT 
 
 Field-tube ... 
 
 ... Ft 
 
 Flue-tube 
 
 ... F T 
 
 Tube-plate 
 
 ... TP 
 
STEAM BOILERS. 
 
 69 
 
 Water-tube ... 
 Dead-plate . . . 
 Smoke-box . . . 
 
 Uptake 
 
 Damper 
 
 Cbimney 
 Cleaning-door 
 Combustion chamber 
 Smoke-box door ... 
 Flue-tube hole 
 
 Steam-room 
 
 Firebrick 
 
 Manhole 
 
 Mudhole 
 
 Mud-drum 
 
 Non - conducting 
 
 covering 
 
 Lap-joint 
 
 WT 
 
 DP 
 
 SB 
 
 UT 
 
 D 
 
 Cy 
 
 CD 
 
 r C C 
 
 SBD 
 
 TH 
 
 SB 
 
 Fb 
 
 MH 
 
 m h 
 
 MD 
 
 NC 
 
 L J 
 
 Steam-pipe 
 
 Pressure-gauge ... 
 
 Water-gauge 
 
 Scum-cock 
 
 Blow-off-cock 
 
 Steam- valve 
 
 Check- valve 
 
 Injector 
 
 Bamsbottom's safety- 
 valve _ BSV 
 
 Dead-weight safety 
 valve 
 
 Lever safety-valve 
 
 Hopkinson's com- 
 pound safety- 
 valve 
 
 Steam drier 
 
 Steam-whistle . . . 
 
 sp 
 
 PG 
 
 WG 
 
 SO 
 
 BC 
 
 sv 
 
 cv 
 
 I 
 
 DWS 
 LSV 
 
 CSV 
 
 Sd 
 
 S W 
 
 Horizontal Boilers. 
 
 The Egg-Ended Boiler. — This form of boiler is 
 illustrated in almost all text-books, but is hardly ever 
 
 boiler will only be economical when worked continu- 
 ously. It is a serious defect with this class of boilers 
 that the bottom of the shell, where sediments always 
 collect, is the part exposed to the most intense heat, 
 and is therefore liable to get burnt. On account of 
 the great length, the brickwork is costly, and the ex- 
 pansion lengthways of the boiler becomes serious. 
 
 The Cornish Boiler (Fig. 6) is a horizontal boiler, 
 with one internal cylindrical flue. Its form is a cylinder 
 with flat ends. The external flues are usually arranged 
 as a split draught, the gases flowing through the lateral 
 flues, after having left the internal flue, and then dip 
 underneath the boiler and into the chimney. By this 
 arrangement the bottom of the boiler, where sediments 
 tend to deposit themselves, is the coolest, whereas the 
 hottest part of the flues is near where the steam is 
 generated. Messrs. Galloways, however, consider that 
 the circulation of the water obtained in this way is not 
 sufficient, and therefore advise that the flame should turn 
 under the bottom first, so as to ensure a more equable 
 
 used now, and can therefore be considered as belong- 
 ing to the past. The form is a cylinder, whose ends 
 are closed by two hemispheres. The grate is below the 
 boiler, and the flues form a wheel draught, the current 
 of gases flowing first underneath the shell. The boiler 
 is very strong; it has a large water-surface, and will 
 therefore give dry steam. 
 
 The heating-surface, however, is small compared with 
 the water-room and the bulk of the boiler. The water- 
 room varies according to the square of the diameter, 
 and directly as the length of the boiler, whereas the 
 heating-surface is in direct proportion to both diameter 
 and length. It therefore follows that the boiler must 
 be made long in proportion to its diameter. On 
 account of the large quantity of water it contains, this 
 
 temperature of the water in the boiler, the gases after- 
 wards rising and returning to the back by means of the 
 lateral flues. This, however, necessitates two dampers 
 to each boiler, which may be considered objectionable. 
 This boiler combines a smaller water-room with a 
 larger heating-surface, and therefore takes up less space 
 than the old egg-ended boiler for the same evaporative 
 capacity. The heating surface is usually increased, 
 without increasing the size of the boiler, by introducing 
 into the internal flue a number of the so-called Galloway 
 tubes. Fig. 7 is an iilustration of the Galloway tube. 
 The largest diameter of the tube is about llfin., and 
 the smallest about 6fin. The diameter of the smaller 
 flange must of course be somewhat smaller than llfin., 
 so as to allow the tube to be inserted in the flue. The 
 
70 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 larger flange is rivetted on the outside, and the smaller 
 on the inside of the flue. The Galloway tubes are 
 made of either iron or steel, and are lap- welded, the 
 flanges being forced up at one heat. There is not only 
 the advantage of the additional heating-surface, but 
 the increased circulation of water and the stiffening of 
 the flue. As the tubes are placed at right angles to 
 the current of gases, the latter will be more completely 
 broken up, and thus the efficiency of the boiler will be 
 increased. 
 
 For the purpose of cleaning the boiler, its dimensions 
 must be such that a man can move inside the boiler. 
 For this reason, the distance between the crown of the 
 internal flue and that of the boiler should not be less 
 than 22in., and the space below the flue must be at 
 least 5in. The diameter of the internal flue is about 
 
 J?iu. 7. 
 
 Let D b denote the latter, then 
 + 22in., or D b must not be less 
 
 half that of the boiler 
 ft "^ +f 
 
 than 54in. The diameter of the internal flue will 
 therefore be at least 27in. 
 
 The crown of the firebox and part of the internal 
 flue are usually lined with an arch of firebrick, to prevent 
 the cooling of the flame before the combustion is com- 
 pleted. 
 
 Figs. 8 and 9 show a Cornish boiler with double 
 mud-drum or heaters, as made by Messrs. Musgrave 
 and Sons, of Bolton. It will be seen that the grate is 
 underneath the boiler, and that the gases pass first 
 below the shell, then through the internal flue 
 descending by the lateral flues to the mud-drums, and 
 then into the chimney. This construction of boiler is 
 chiefly used for burning wood. The feed-water is 
 pumped into the mud-drums, where it receives the first 
 
 heat from the gases ; hence they are also called 
 heaters. Much of the impurity contained in the water 
 is deposited in the mud-drums, which is of special im- 
 portance in boilers fired from below. 
 
 o.w.s. 
 
 s.v. 
 
 The dimensions of a Cornish boiler, which is re- 
 quired to evaporate Q pounds of water per hour, can be 
 calculated approximately in the following way. Let 
 the feed-water enter the boiler at a temperature of 
 about 100 deg. F., and let there be 3£ square feet of 
 h eating-surface for each pound of coal burnt per hour, 
 then we may expect that the boiler will evaporate 
 about 91b. of water per pound of coal. We must there- 
 fore burn ^ pounds of coal per hour. We can now 
 
 calculate the grate-surface according to page 64. The 
 diameter of the internal flue can be found by remem- 
 bering that it must not be less than 27in., and that the 
 length of the grate ought not to be more than 7ft. 
 The diameter of the boiler is to be about twice that of 
 the internal flue. Distance a, Fig. 10, should be 
 5in. to 6in., b about 4in. The line between points c 
 should be about 2in. above the crown of the internal 
 flue, the water-line being 5in. to 6in. above the latter. 
 
 The total heating-surface must be 3i x ^=J_Q 
 
 9 18^ 
 square feet, and this, again, must be equal to the 
 length of the boiler multiplied by the circumference of 
 the cross-section of the heating-surface, as shown in 
 
 D.W.S. 
 
 Fig 9. 
 
 Fig. 10. In this way we obtain the length of the 
 boiler. We must, however, remember that the length 
 of the external flues will be about 1ft. shorter than the 
 boiler. The boiler will also be shorter when using 
 
STEAM BOILERS. 
 
 71 
 
 s.v 
 
 c.s.v. 
 
 FiglO. 
 
 Rg.!2. 
 
STEAM BOILERS, 
 
 73 
 
 Galloway tubes, as the internal heating-surface is 
 thereby increased. 
 
 The Lancashire Boiler is a horizontal boiler, with 
 two internal parallel flues. By this construction, the 
 ratio of the heating-surface to the bulk of the boiler is 
 still greater than in the Cornish boiler. A boiler of this 
 class, as made by Messrs. G-alloways, is shown in Fig. 
 11. It will be seen that the heating-surface is further 
 increased by the introduction of a number of Galloway 
 tubes. For the same reasons as explained with the 
 Cornish boiler, Messrs. Galloways arrange the flues in 
 such a way that the gases from the internal flues join 
 at the back of the boiler, whence they dip underneath 
 it, and return through the lateral flues to the back of 
 the boiler, from where they escape into the chimney, 
 each lateral flue being provided with a damper. 
 
 meter of the flue. We can now construct the centre 
 and the diameter of the boiler. The length of the boiler 
 is determined in the same way as in the preceding 
 one, the total heating-surface being ye Q square feet. 
 
 Breeches-Flued Boiler. — The external appearance of 
 this class of boiler is like that of a Lancashire, but the 
 two cylindrical flues, instead of running through the 
 whole length of the boiler, join into one large flue just 
 behind the bridges. The gases produced in the two fire- 
 boxes will thus meet and flow together through the 
 common flue. 
 
 The best-known boiler of this class is the Gallo- 
 way boiler, which is illustrated in Fig. 13, and a back 
 view of same is- shown in Fig. 3. It will be seen that 
 the wide flue is slightly arched at top and bottom. 
 The boiler contains for the same length from two to 
 
 D.w.s. 
 
 c.s.v. 
 
 Fig. 13. 
 
 The approximate linear dimensions of a Lancashire 
 boiler required to evaporate a quantity of Q pounds of 
 water per hour, is obtained in the same manner as with 
 the Cornish boiler. We take also here the heating- 
 surface to be 3£ square feet per pound of coal to be 
 burnt per hour, for which we expect the boiler to evapo- 
 rate about 91b. of water. The grate-surface is deter- 
 mined according to page 64, and the diameter of the 
 flues, which must not be less than about 26m., is then 
 obtained. The smallest distance between the two 
 flues should be about 5in. We can now draw the two 
 internal flues in their position. The distances, a, Fig. 
 12 to be kept between the flues and the mam shell 
 should be 5in. to 6in., and the vertical distance between 
 the centre line of the flues and the centre of the mam 
 shell should be about one-fourth to one-fifth of the dia- 
 
 three times as many Galloway tubes as the Lancashire 
 boiler, the advantages of which are : 
 
 1. That the proportion of heating-surface to volume 
 of boiler is greatly increased ; 
 
 2. That the gases will be more perfectly broken up 
 in their passage through the wide flue, and thus cause 
 the heat whicb they contain to be better absorbed ; 
 
 3. That the circulation of the water will be increased 
 with the number of tubes, and this is of great import- 
 ance, as the temperature throughout the boiler will be 
 more equalised, and thereby prevent an unequal expan- 
 sion of the various parts of the boiler ; 
 
 4. That the resistance of the flue to collapse is greatly 
 increased. 
 
 The number of tubes must not be increased to such 
 
74 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 an extent that the cleaning of the boiler thereby is 
 made difficult. 
 
 The object of the side-pockets, shown in the drawing, 
 is to deflect the gases, which otherwise would pass 
 between the tubes and the sides of the flue without 
 coming into contact with the heating- surface. The 
 circulation of the water being more complete than in 
 the boilers previously described, the flues are always 
 arranged as a split draught, the gases passing last 
 through the flue underneath the boiler. 
 
 The firing of a breeches-flued ooiler may be done 
 alternately. The gases from the fire which has just 
 received a fresh charge of fuel will contain more or less 
 smoke, while the other fire will give off gases which are 
 completely burnt, and which contain an excess of air 
 at a high temperature. Then, when the gases from the 
 two fireboxes mix behind the bridges, a complete com- 
 bustion will take place. 
 
 Lancashire boilers are also sometimes fired in this 
 way, but as the gases do not mix until they get behind 
 the boikr, the temperature of the air may have fallen 
 so much that only a partial ignition takes place. 
 
 The dimensions of a Galloway boiler may be calcu- 
 lated by assuming that 101b. of water will be 
 evaporated by burning lib. of good coal, when the 
 temperature of the feed- water is about 100 deg. F. 
 The diameter of the cylindrical flues and that of the 
 main shell are determined in the same way as with the 
 Lancashire.' The radius of curvature of the top 
 " breeches " plate, Fig. 14, is about two and a half 
 times the diameter of the cylindrical flues. 
 
 As the grate-surface, as well as the heating-surface 
 will be smaller in the Galloway than in the Lancashire! 
 it will be found that the volume of the former will be' 
 
 20 to 25 per cent, less than that of the latter boiler, 
 when evaporating the same quantity of water per hour. 
 
 In Fig. 15 is shown a pair of Galloway boilers with 
 external fireplaces. By this arrangement the firebox 
 may be made of any required size, and thus is avail- 
 able for burning wood, shavings, or other fuel which 
 requires greater space than is possible in the ordinary 
 internal fireplaces. 
 
 Another modification of the Galloway is shown in 
 Fig. 16, which removes an objection sometimes 
 made to the preceding design — namely, that a portion 
 of the heat is communicated directly to the brickwork 
 instead of all being passed into the boiler. In this case 
 the firebox is of the locomotive construction, the top 
 being firmly stayed to the shell, and the sides being 
 secured by stay-bolts. 
 
 The setting of the boiler in its proper brickwork is a 
 matter of considerable importance, as it is necessary 
 that the flues should be of proper proportions through- 
 out, and that ample room should be given for the free 
 passage of the gases. In preparing the foundations, a 
 matter of great importance is that good drainage 
 should be ensured, as it is found that serious damage 
 often occurs to boilers owing to dampness which arises 
 from the foundations, such outside corrosion not being 
 caused by leakages from the boiler itself, but simply 
 from being built either on wet ground or where surface 
 water is allowed to accumulate. Foundations should 
 
 also be built in hard, well-burned bricks, and the 
 internal parts of the flue, where exposed to the. fire, 
 
STEAM BOILERS. 
 
 75 
 
 lined with firebrick, the boiler itself being carried 
 upon blocks of the same material, and it should be set 
 having a fall towards the front of £in. when the boiler 
 is not longer than 20ft., and of lin. if the boiler is more 
 
 surface to the volume of the boiler step by step, from 
 the egg-ended boiler to the Galloway, but taking into 
 consideration the necessity of a fairly large water- 
 room for regulating the pressure, without having to 
 
 Fig. 17. 
 
 Fig. 18. 
 
 than 20ft. long, so as to ensure the boiler being depend too much on the stoker, and the water-sur- 
 thoroughly drained of its contents when it is emptied face, where the steam is liberated from the water, has 
 for cleaning. The top of the lateral flues should be 
 
 "S§§§& 
 
 made up by flue-covers (see Fig. 3) of fireclay in such a 
 
 form as generally to allow a flue of about 6in. at each 
 
 side of the boiler, the flues themselves being of such 41 > -V <?' <i> 6 fyr f ; i 
 
 dimensions as will allow of a man getting round for the 
 
 purpose of cleaning and examining the boiler. 
 
 The Locomotive or Multitubular Boiler— -This boiler also been taken large enough for producing dry steam, 
 originally designed as a portable boiler, and for But with these boilers considerable time is required for 
 
 was 
 
 Fig. 20. 
 
 + -™u, err, Q ll hulk and lightness getting steam up, on account of the comparatively large 
 
76 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 necessary to be able to get steam up at short notice, 
 and consequently the heating-surface must be large and 
 divided into small parts, so to speak, scattered about in 
 all parts of the comparatively small water- room. This 
 object is obtained by employing a great number of nar- 
 row flue-tubes, through which the gases must pass on 
 their way from the firebox to the chimney. 
 
 On account of its quick steaming qualities, the loco- 
 motive boiler has been used, and is still used, a great 
 deal at temporary installations, and also where the light 
 is only required for a short period every day — for 
 instance, at exhibitions. Figs. 17, 18, 19, and 20 show 
 a locomotive boiler as made by Messrs. Eobey and 
 Co., of Lincoln. Figs. 21 and 22 show a similar 
 boiler, as made by Messrs. Garrett and Son, of 
 Leiston. 
 
 A locomotive boiler consists of : (1) The "fireplace," 
 which contains the firebox (F B) in which the combus- 
 tion takes place, the fire-grate (G), and the ashpit 
 (A P) . (2) The ' ' flue-tubes "(FT), which begin at the 
 back of the firebox, where they are fixed to the firebox 
 tube-plate (FTP), and through which the gases pass 
 to the chimney. (3j The " smoke-box " (SB), at the 
 back ; this is bounded by the smoke-box tube- 
 plate (S T P), which forms the back of the boiler, and 
 by an outer shell which has a door, smoke-box door 
 (S B D), on the outside for permitting access to the flue- 
 tubes. The smoke-box is connected at the top by the 
 uptake (U T) with the chimney. (4) The outer boiler 
 shell, which consists partly of a cylindrical barrel, 
 through which the tubes pass, and partly of a shell sur- 
 rounding the firebox, which has the same shape as the 
 latter. 
 
 As a portable boiler, the locomotive boiler cannot be 
 set in brickwork, as this could not be carried about ; there 
 can, therefore, be no external flues, and the heat which 
 the gases ought to give off to the boiler must, there- 
 fore, be transferred to the water while they pass 
 through the comparatively short tubes. The tempera- 
 ture of the gases must therefore be high, which necessi- 
 tates the smallest quantity of air possible for completing 
 the combustion of the fuel, but this again requires an 
 artificial draught, as explained on page 67. 
 
 The layer of fuel in railway boilers is taken from 
 18in. to 36in., the draught being so strong that a current 
 of air at high velocity is produced ; 401b. to 501b. of coal 
 can then be burnt per square foot of grate-surface per 
 hour; the heating-surface being one square foot per pound 
 of coal burnt per hour, whereby 81b. of steam are pro- 
 duced. The effective steam pressure in such boilers 
 may be as high as 1601b. per square inch. 
 
 The effective pressure of the steam produced in loco, 
 boilers used for stationary or portable purposes averages 
 801b. to 1001b. per square inch, sometimes, however, 
 more. Consequently, the draught will not be so strong, 
 a greater heating-surface is required, and the layer of 
 coal on the grate must be thinner. About 201b. 
 of coal can be burnt per hour per square foot. The 
 heating-surface is taken between T5 and 2'6 square feet 
 per pound of coal burnt per hour, whereby from 71b, 
 
 "flff 
 
 jS* 
 
 LtifrUl 
 
 -f-M>-4t-f-"-^ 
 
 #+l*-!-t. 
 WW 
 
 ?: p--?:Lt.-?- 
 
 •# 
 
 # 
 
 m 
 
 - *f ^ -^ -y '*£-- -f -^ ^p 
 f -f -Hf-f- l-fi || 
 
 ' ' I < I i 
 
STEAM BOILERS. 
 
 77 
 
 to 81b. of water will be evaporated, the feed-water 
 being about 100 deg. F. 
 
 If the locomotive boiler is worked witb ordinary 
 chimney draught, the heating-surface ought to be, as in 
 an economical boiler, 3J square feet per pound of coal 
 burnt per hour. This, of course, would necessitate a 
 great length of boiler. It would, therefore, be better 
 to set it in brickwork, as shown in Fig. 16, where we 
 must imagine the Galloway flue taken away and sub- 
 stituted by flue-tubes as in an ordinary loco, boiler. 
 
 The flue-tubes should be made of the best iron, and 
 should be lap-welded ; their diameter runs from 2in. to 
 4in. A certain portion of the tubes should be screwed 
 at the ends, and provided with nuts for the pur- 
 pose of acting as stays, stay-tubes (S T), in order to 
 assist the tube-plates in resisting the pressure to which 
 they are subjected. The tubes are more or less liable 
 to leak at the junction with the tube-plates, especially 
 where the water contains any impurities. Their 
 diameter being small, the tubes can be made thin, and 
 still be strong enough to withstand the steam pressure. 
 If a tube breaks, the result will be the extinction of the 
 fire. 
 
 For the purpose of preventing the gases from rush- 
 ing through the tubes before the combustion is com- 
 pleted in the firebox, an arch of firebrick is placed 
 over the fire, between the grate and the tubes, at about 
 the same height as the fire-door. The arch will deflect 
 the gases as they rise from the fire, and thus cause the 
 smoke and air to mix, and thereby secure a more com- 
 plete combustion. In order to reduce the radiation of 
 heat, loco, boilers are always lagged — i.e., covered with 
 some non-conducting material, as hair, felt, silicate 
 cotton, etc., lined with a layer of wood, and held 
 together by sheet iron. 
 
 The Bobey Loco. Boiler (Figs. 17, 18, 19, 20).— The 
 boiler barrel and external firebox are made of best mild 
 steel plates, Jin. thick; the smoke-box tube-plate is 
 flanged and made of the same material, but is fin. thick. 
 There are three longitudinal stays for taking up the 
 end pressure on the boiler ; these, as well as the stay- 
 bolts, are made of best mild steel. All longitudinal 
 seams are double rivetted, and butt-jointed, with inside 
 and outside covers. The manhole is strengthened by 
 a compensating ring and a raised mouthpiece of steel, 
 which are both rivetted round the opening, and the 
 latter is fitted with a strong cast-iron cover with bolts. 
 The cover is fitted with a Eamsbottom's safety-valve. 
 The firebox is also made of best mild steel plates, Jin. 
 thick, the roof is well strengthened with deep crown 
 stays, and the sides and ends are stayed by steel screw- 
 bolts. The top and sides are in one plate. The corners 
 of the firebox are made of large radius, so as to admit 
 of the mudhole plugs being placed in the corners. By 
 this arrangement a scraper can be passed between each 
 row of stays in the firebox shell, thus making it conve- 
 nient for cleaning. A fusible plug is inserted in the 
 crown. All plates are planed and turned on the edges, 
 before being put together, and the rivetting is done by 
 hydraulic machinery. 
 
 The flue-tubes are made of best wrought iron and 
 lap-welded. They are arranged in vertical lines, with 
 wide spaces between each row, so as to diminish the 
 formation of scale upon them. Six of the tubes are 
 stay-tubes. They are expanded by tube-expander and 
 beaded over at the firebox end, but they are simply 
 expanded at the smoke-box end. The removal of a 
 tube is done by cutting off its firebox end, and then pull- 
 ing out the tube through the smoke-box, the holes in the 
 smoke-box plate being about Jin. greater in diameter 
 than that of the tubes. The tubes project at least Jin. 
 through the smoke-box tube-plate. The plates for the 
 smoke-box and door are of steel, and have a smooth 
 surface. There is a hole at the bottom of the smoke- 
 box for the soot to fall through into the pillar, whence 
 it is removed through the cleaning-door at the bottom. 
 The chimney is made of steel plates, rivetted to an angle- 
 iron ring at the bottom, which is bolted to a cast-iron 
 uptake, attached to the top of the smoke-box. The 
 inner and outer firebox shells are jointed at the bottom, 
 and also at the fire-door, by wrought-iron caulking 
 rings. 
 
 The boiler is fitted with steam-valve, pressure-gauge, 
 two water-gauges, two blow-off-cocks, steam-whistle, 
 injector, and check-valve. The object of the cock below 
 the check-valve is for the release of air. The boiler is 
 lagged with well-seasoned pine, and covered with 
 smooth sheet iron, secured by hoops. The boiler can 
 be worked up to 1501b. per square inch. The grate- 
 surface is 16 square feet, and 3001b. of coal can be 
 burned per hour. The heating-surface of the firebox is 
 69 square feet, and that of the tubes 424 square feet, 
 thus making a total of 493 square feet. The water 
 evaporated per pound of coal burnt on the grate depends 
 upon the force of the draught, which is generally pro- 
 duced by the exhaust steam. 
 
 The Garrett Loco. Boiler (Figs. 21 and 22).— The 
 outer boiler shell, inner firebox, stay-bolts, longitudinal 
 stays, standpipe, manhole-cover, manhole saddle, and 
 smoke-box are made of Siemens-Martin steel. The 
 tensile strength of the steel used for the firebox is 
 
 24 to 28 tons per square inch, elongation in lOin. equal to 
 
 25 per cent., the steel contains from 12 to 14 per cent, 
 carbon. The total length of the boiler, including smoke- 
 box, is 16ft. 23in., and its total height is 6ft. 10in., the 
 uptake being 2ft. high. The distance between the tube- 
 plates is 10ft., and the firebox measures inside 5ft. long 
 Dy 3ft. 7|m. wide by 4ft. greatest height. It will be 
 seen that the firebox roof is corrugated, whereby crown 
 stays are made unnecessary, and the heating surface is 
 increased. The firebox consists of three plates — viz., 
 the corrugated roof and the two sides forming one, and 
 being Jin. thick, a flanged tube-plate fin. thick, and a 
 flanged front-plate fin. thick. The firebox is jointed to 
 the outer shell at the fire-door by a wrought-iron caulk- 
 ing ring. The flue-tubes are made of iron, lap-welded, 
 and are 2in. in diameter. They are fixed in the tube- 
 plates in a similar way as in the Eobey boiler. The 
 thickness of the outer boiler shell is ^in., except the 
 front-plate and the smoke-box tube-plate, which are 
 
78 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Jin. thick. Mudholes are plaoed at each corner of the 
 firebox, as well as at the bottom of the boiler barrel 
 near the firebox, but they are not shown on the draw- 
 ing. There are five longitudinal stays, which are fixed 
 by bolts to the end-plates of the boiler as shown. The 
 grate surface is 17'9 square feet, and the total heating 
 surface is 510 square feet. 
 
 The following table, which the author has received 
 from Messrs. R. Garrett and Sons, shows the relation 
 between force of draught, coal burnt, and water evapo- 
 rated. 
 
 Draught 
 
 in inches of 
 
 water. 
 
 Coal in lbs. burnt 
 per hour per 
 square foot of 
 grate-surface. 
 
 Heating-surface 
 
 in square feet per 
 
 lb. of coal burnt 
 
 per hour. 
 
 Water in lbs. 
 
 evaporated per lb. 
 
 of coal burnt on 
 
 grate from and 
 
 at 212 deg. F. 
 
 0-3 
 0-4 
 0-5 
 0-6 
 0-8 
 
 10-48 
 14-00 
 18-18 
 23-51 
 26-60 
 
 2-71 
 
 2-04 
 
 1-57 
 
 1-2 
 
 1-07 
 
 10-25 
 
 10-50 
 
 10-19 
 
 9-65 
 
 9-68 
 
 This table shows, what has already been explained, 
 that the heating- surface of a boiler can be diminished 
 without materially altering its evaporating capacity, if 
 the force of the draught is increased in proportion. The 
 calorific value of the coal which was used in obtaining 
 the above figures is unknown to the author, nor does 
 he know whether the boiler was quite new or had been 
 used for some time. 
 
 Compound Boilers. — When a large evaporative capa- 
 city is required within a comparatively small space, and 
 the boiler is to be worked with ordinary chimney 
 draught, a loco, boiler set in brickwork might be used, 
 as explained above. But on account of the shape 
 of the firebox, the external flues cannot be 
 carried further than along the barrel of the boiler ; 
 thus the external surface of the part of the boiler sur- 
 rounding the firebox is wasted as heating surface. 
 Besides, the diameter of a loco, boiler must always be 
 small compared with the length, as otherwise the 
 firebox would be too large. The loco, boiler would 
 therefore be long, and take up too much ground space. 
 For this reason a compound boiler will do better. 
 The author understands by a compound boiler a 
 combination of cylindrical flue boilers and multi- 
 tubular boilers. A boiler of this kind consists of : 
 
 (1) A cylindrical boiler shell, with one or more 
 internal cylindrical flues, each containing a fire- 
 place, as in the Cornish and Lancashire boilers. 
 
 (2) One or more combustion chambers into which the 
 gases flow from the cylindrical flues, and where the 
 combustion is completed. The combustion chambers 
 may be contained either within or without the boiler. 
 
 (3) A number of narrow flue-tubes, as in loco, boilers, 
 running parallel with the cylindrical flues, from the 
 
 combustion chamber to the front of the boiler, where 
 there is a smoke-box. 
 
 The gases pass from the firebox through the cylin- 
 drical flues to the combustion chamber, and then 
 through the flue-tubes to the smoke-box at the front of 
 the boiler, from where they either go straight into the 
 chimney, or, if the boiler is set in brickwork, through 
 lateral flues into the chimney at the back of the boiler. 
 The following compound boilers may be noticed : 
 
 (1) Marine Boilers. — The figure of this class of boilers 
 is somewhat different to those hitherto mentioned, 
 partly because of the shape of the space in which they 
 are to be placed, and partly because the water-surface 
 should be small, in order to prevent any part of the 
 heating-surface coming out of contact with the water 
 by the oscillations of the vessel in which they are 
 placed. In Figs. 23 and 24 are shown scale drawings 
 of one of the boilers of ss. " Brighton," belonging to the 
 London, Brighton, and South-Coast Railway, and made 
 by the Leeds Forge Company, Limited, of Leeds. 
 There are two Fox corrugated furnaces, each containing 
 a fireplace, not shown, but similar to those shown in 
 the Cornish and other boilers. Each furnace has its 
 own combustion chamber, from which 118 flue-tubes 
 carry the gases into a common smoke-box at the front 
 of the boiler. The smoke-box is connected by an 
 uptake to the chimney, but they are not shown in the 
 drawings. 
 
 The corrugated furnaces are made of mild steel ; 
 their mean diameter is 3ft. 9in., their front open- 
 ing is 3ft. 7in. in diameter, and they are 6ft. long. 
 The total grate-surface which is contained within the 
 corrugated flues is 49 square feet. 
 
 The shells of the combustion chambers are made of 
 mild steel plates ^in. thick, except the tube-plates, 
 which are fin. The back plates are stayed to the back 
 plate of the boiler shell by steel bolts l£in. in diameter. 
 The stay-bolts between the two combustion chambers, 
 as well as those between the same and the boiler 
 cylinder, are l^in. in diameter, and are also made of 
 steel. Crown stays are employed for strengthening the 
 flat roofs of the chambers. 
 
 The flue-tubes are 5ft. 6in. long, and their external 
 diameter is 2|in. They number altogether 236, of 
 which 88 are stay-tubes ; the latter are made of steel 
 and the others are of best iron and lap-welded. The 
 stay-tubes are shown in the drawing by a double circle, 
 and they are fixed by nuts inside and outside the tube- 
 plates. The heating-surface due to the tubes is 917 
 square feet, and that of the furnaces, including the 
 combustion chambers, is 227 square feet. The total 
 heating-surface of the boiler is therefore 1,144 square 
 feet. 
 
 The boiler shell, being of steel, is lift. lOin. in 
 diameter and 7ft. llin. long. The boiler cylinder is 
 Jin. thick, and the end and front plates are |in., except 
 the tube-plate, the thickness of which is fin. 
 
 There are 18 longitudinal stays placed above the 
 flue-tubes, all of steel, and 2in. in diameter ; they are 
 threaded at both ends, where the diameter is increased 
 
STEAM BOILERS. 
 
 79 
 
 FujZS 
 
STEAM BOILERS. 
 
 81 
 
 to 2§in. Two nuts at each end keep the stays in their 
 position. These nuts are not tightened directly against 
 the boiler-plates, but copper washers |in. thick are put 
 between the nuts and the plates in order to prevent 
 corrosion. Three similar steel longitudinal stays lfin. 
 in diameter are placed round the mudhole at the bottom 
 of the boiler, in order to strengthen this part of the end- 
 
 The material used for making manhole rings, mud- 
 hole rings, crown stays, stay-tubes, corrugated flues, 
 and rivets is Siemens mild steel, having a tensile 
 strength of about 26 tons per square inch, with an 
 elongation of 25 to 30 per cent, in lOin. 
 
 Compound Lancashire and Multitubular Boiler. — 
 Fig. 25 gives a perspective view of Messrs. Davey, Pax- 
 
 ■w BJ 
 
 Ky24 
 
 plates. Three mudholes are placed at the sides of the 
 corrugated flues, in order to facilitate the cleaning of 
 the latter. The manhole is put on the back end-plate. 
 The effective pressure at which the boiler is worked 
 is 1601b. per square inch, and the steam is taken off 
 direct from the boiler shell, there being no steam- 
 dome. 
 
 man, and Co's. "Economic " boiler. The grates, fire- 
 boxes, and cylindrical flues are arranged exactly in the 
 same manner as in a Lancashire. The gases from 
 the two flues meet behind the boiler in a combustion 
 chamber, made in brickwork, from whence they return 
 through a series of flue-tubes to the front of the boiler, 
 where there is a smoke-box. From the smoke-box the 
 
82 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 gases may pass into the chimney, but it is preferable to 
 let them return to the back of the boiler by two lateral 
 flues, set in masonry, and which are connected to the 
 chimney. 
 
 D.W.S. 
 
 Fig 25 
 
 B.C. 
 
 Sections of one of these boilers are shown in Eigs. 26, 
 27, and 28. The shell is 14ft. 6in. long by 7ft. 6in. 
 diameter, and it is made of steel plates fin. thick. The 
 
 pansion-joint, and are provided with four Galloway 
 tubes made of steel plate fin. thick. The tensile 
 strength used for the internal flues is 25 tons, with 
 about 28 per cent, elongation. It is, therefore, more 
 ductile than the material of which the shell is made. 
 The flue-tubes number 74, of which 12 are stay-tubes ; 
 their external diameter is 3in., they are made of wrought 
 iron, and lap-welded, and the thickness of material is 
 No. 9 B.W. G. The stay-tubes are threaded, and 
 fixed to the end-plates by nuts; the other tubes are 
 expanded in the holes. There are six longitudinal stays 
 made of steel, of which the diameter at bottom of thread 
 is If in. The gusset stays are also of steel, the thick- 
 ness of the plate is £in. The smoke-box, manhole 
 ring, mudhole ring, and manhole saddle are all of 
 wrought iron. The standpipes and manhole cover are 
 of steel. The heating-surface of flue-tubes is 814 square 
 feet, and that of the cylindrical flues is 171 square feet. 
 The total heating-surface of the boiler, when not set in 
 brickwork, is therefore 985 square feet ; but when set in 
 brickwork, and having two lateral external flues, the 
 total heating-surface will be 1,219 square feet. The 
 working pressure is 1301b. per square inch. 
 
 The external view of a similar boiler is shown in 
 Figs. 29 and 30. This boiler is supplied with a Steam- 
 
 er. G 
 
 Fig. 2 6 
 
 boiler ends are made of the same material and are of 
 the same thickness. The tensile strength of the steel is 
 28 tons, with about 24 per cent, elongation. The internal 
 flues are 2ft. Gin. in diameter, and made of steel plate 
 j^in. thick ; they are strengthened with Paxman's ex- 
 
 valve (S V), one dead- weight safety-valve (D W S), 
 two water-gauges (WG), one pressure-gauge (PG), a 
 manhole (M H), a mudhole (m h) in front underneath 
 the cylindrical flues, etc. 
 
 Messrs. Galloways make a boiler of the construction 
 
STEAM BOILERS. 
 
 83 
 
 shown in Pig. 31. It consists of two independent in order to maintain an equal water level, and overhead 
 boilers ; the first is a short Lancashire, and the other by a longitudinal dome. Between the boilers is an open 
 
 M.H. 
 
 a multitubular boiler of the same dimensions. These space which serves as a combustion chamber. The gases 
 two sections are connected underneath by a large pipe pass from the Lancashire into the combustion chamber, 
 
84 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 and from there through the flue-tubes of the multi- The boiler shown in Figs. 32 and 32* is made by the 
 tubular boiler into the chimney. No external lateral Leeds Forge Company, Limited. Fig. 32 represents a 
 
 -PWJ". 
 
 flues are necessary, as the heat is thoroughly absorbed longitudinal section through the boiler shell whereas 
 when reaching the end of the second boiler. The boiler Fig. 32* gives the front view. The boiler Tenls 
 
 Fig. 32b. 
 
 being in two pieces of moderate dimensions is easily 7ft. Gin. in mean diameter, and is 15ft lon« The 
 tranSP ° rted - thi <*ness of the boiler cylinder is jin., and the end' 
 
STEAM BOILERS. 
 
 85 
 
 plates, which also act as tube-plates, are fin. thick, the flues is 3ft. ljin. The ratio between the diameter 
 There are two of Fox's corrugated flues, each having an of the flue and that of the boiler barrel is made smaller 
 external diameter of 2ft. 9in., and they are rivetted to than in a Lancashire, in order to make room for the 
 
 the flanged end-plates of the boiler shell. The grates, flue-tubes. A fasible plug (P P) is placed on the crown 
 J^W* fireboxes and bridges are contained within of each flue for the protection of the flue in case of low 
 telt^teTZr%^tt^e between centres of water. The external diameter of the flue-tubes is 3m., 
 
86 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 and their number is 76, of which 14 are stay-tubes. The 
 latter are marked in the drawings by a double circle, and 
 are fixed to the tube-plates by nuts. 
 
 A similar boiler made by the same firm is shown in 
 Figs. 32a and 32b. 
 
 The five longitudinal stays above the flue-tubes are 
 2in. in diameter, and they pass at both ends through 
 channel section stays (C S) ; the latter are rivetted to 
 the end-plates for the purpose of stiffening the plates. 
 The longitudinal stays are provided with nuts in the 
 
 by a longitudinal stay. The stop-valve is mounted on 
 the top of the dome. 
 
 The drawings further show the manhole (M H), 
 with saddle and cover, two mudholes, one above and 
 one below the corrugated furnaces, a blow-off cock 
 (BC), and a standpipe (SP), between the dome and 
 the front of the boiler, to which the safety-valve is fixed. 
 The boiler is set in brickwork, and has a combustion 
 chamber at the back of it ; there is a smoke-box at the 
 front of the boiler, and also two lateral external flues. 
 The passage of the gases is the same as in the 
 " Economic " described above. 
 
 The heating-surface of the tubes is 570 square feet, 
 and that of the two corrugated flues is 240 square feet. 
 The external heating-surface is 150 square feet, which 
 gives a total heating-surface of 960 square feet. The 
 total grate-surface is 32 square feet, and the working 
 pressure is 1601b. per square inch. 
 
 The boiler shell, corrugated flues, stays, tube stays, 
 manhole ring, manhole saddle, mudhole rings, stand- 
 pipe, dome, and dome necks, are all of Siemens 
 mild steel, made by the Leeds Forge Company, 
 Limited; the tensile strength of the steel being 26 
 tons per square inch, with an elongation of 25 to 30 per 
 cent, in lOin. 
 
 usual manner. The end-plates are further stiffened at 
 the back by a t stay (T S), and at the front by the 
 flanged triangular mudhole, which, m combination 
 with the longitudinal and channel section stays, keeps 
 the end-plates rigid, and so prevents any expansion 
 and contraction, which would be liable to cause the 
 tubes to leak. The steam-room is increased by the 
 large dome on the top of the boiler, the length of which 
 is 10ft. 8in., and its diameter 2ft. The dome is con- 
 nected with the boiler by means of two necks, one of 
 which is wide enough for a man to pass through. The 
 ends of the dome are dished, and are further stiffened 
 
 B.O. 
 Wig. 34 
 
 Compound Cornish and Multitubular Boiler.— Davey, 
 Paxman, and Co.'s "Economic" boiler with one 
 cylindrical flue is shown in Figs. 33 and 34. The 
 principle of this boiler is the same as that of the com- 
 pound Lancashire and multitubular boiler. The gases 
 may either pass direct from the smoke-box into the 
 chimney, or there may be two lateral external flues. 
 The main shell is 9ft. 6in. long by 5ft. 6in. diameter, 
 and made of steel ; the thickness of cylinder plates is 
 tin., and that of the end-plates is -^in. The tensile 
 strength of the material is 28 tons, with about 24 per 
 cent, elongation. The internal flue is made of steel 
 
'STEAM BOt'LEkS, 
 
 87 
 
 plate fin. thick, the diameter of the flue is 2ft. 9in. 
 The tensile strength of the steel used for the internal 
 flue is 25 tons, with about 28 per cent, elongation. 
 There are 40 flue-tubes made of wrought iron and lap- 
 welded ; their external diameter is 2fin., and thickness 
 of material is No. 10 B.W.G. They are fixed in the 
 
 cisely the same as in the boiler shown in Figs. 31 and 
 32, and made by the same firm. This boiler, however, 
 has only one corrugated flue, of which the internal 
 diameter is 3ft. 9in. The object of placing the flue at 
 the one side of the shell is to facilitate the cleaning of 
 the bottom of the boiler. There are 70 wrought-iron 
 
 end-plates by expansion. There are no stay-tubes. 
 The longitudinal stay is made of wrought iron, with a 
 diameter of l£in. The four gusset stays are made of 
 wrought iron, the plate being T Vn. thick. The smoke- 
 box, manhole ring and mn dhole ring are made of 
 wrought iron, and the dome is made of steel. The stop- 
 valve is mounted on the top of the dome, whereas the 
 safety-valve is mounted on the manhole cover. The 
 grate-surface is 12 square feet ; the heating-surface of 
 the flue-tubes is 240 square feet, and that of the 
 cylindrical flue is 84 square feet. The total heating- 
 surface of the boiler, when not set in brickwork, is 
 therefore 324 square feet, but when the boiler is set in 
 brickwork, and has two lateral external flues, then the 
 total heating-surface will be 423 square feet. The 
 working pressure is 701b. per square inch. 
 
 A compound Cornish and multitubular boiler made 
 by the Leeds Forge Company, Limited, is illustrated 
 in Figs. 35 and 36. The dimensions of boiler shell, 
 dome, dome necks, manhole, and standpipe are pre- 
 
 flue-tubes as well as 14 stay-tubes of steel. The heat- 
 ing-surface of the tubes is 630 square feet, that of the 
 
 L.S.V 
 
 corrugated flue 175 square feet, and the external heat- 
 ing-surface is 150 square feet, thus making a total of 
 
88 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 955 square feet. The grate-surface is 22 square feet, 
 and the working pressure is 160 pounds per square inch. 
 The end-plates of the boiler are stiffened by channel 
 section stays (C S) . The material used in making the 
 various parts of the boiler is the same as described on 
 page 86. 
 
 from where they return by the sides and bottom to the 
 chimney. A boiler of this construction would have a 
 greater evaporative capacity than a breeches-Sued boiler 
 of the same dimensions. 
 
 Water-Tube Boilers. — These boilers contain a number 
 of long water-tubes of small diameter, which are con- 
 
 Gonvpound Breeches-Flued and Multitubular Boiler.— 
 It is evident that the heating-surface of a breeches- 
 flued boiler as the Galloway, might be increased by the 
 use of flue-tubes. In the boiler illustrated in Fig. 37 
 the combustion chamber is a short Galloway flue, the 
 flue-tubes carrying the gases to the back of the boiler, 
 
 nected at two or more places to a horizontal main shell, 
 which is placed above the tubes and contains the steam- 
 room. 
 
 The best-known boiler of this class is that manufac- 
 tured by the Baboock and Wilcox Company, of 
 Glasgow. As shown in Pigs. 38 and 39, this boiler 
 
90 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Fig. 38. 
 
STEAM BOILERS. 
 
 91 
 
 Tig, 39, 
 
 Fig. 41. 
 
 Fig. 40 
 
 Fig. 42, 
 
STEAM TOILERS. 
 
 93 
 
 consists of a number of lap-welded wrought-iron tubes 
 placed in an inclined position, the ends being connected 
 at the front as well as at the back to the main shell by 
 two sets of tubes. The end connections (E C) are in 
 one piece for each vertical row of tubes (see Fig. 41), 
 and are so placed that the tubes in the same horizontal 
 row come over the spaces left between the tubes in the 
 row below. The holes in the end connections are made 
 tapering (see Fig. 40), and the tubes are fixed therein 
 by an expander. The end connections are connected 
 with the main shell by tubes, and at the back also with 
 a mud-drum (M D) below. For the purpose of cleaning 
 the tubes, a mudhole is placed opposite the opening of 
 each tube. The mudhole joints are faced, so as to do 
 away with packing. The main shell is made of wrought 
 iron or steel, and is closed at the ends by dished plates. 
 
 tinuous circulation. The steam is taken off at the back 
 of the main shell in order to separate it as much as 
 possible from the suspended water. The mud-drum is 
 made of cast iron, and is provided with mudhole and 
 blow-off-cock. The greater part of the sediment will 
 settle down in the mud-drum, from which it can partly 
 be removed by blowing-off at intervals, the frequency 
 of which depends upon the amount of impurities con- 
 tained in the feed-water. The feed-water is pumped 
 into the main shell, as shown in Fig. 40. Fig. 42 
 shows a section of a Babcock-Wilcox boiler in situ. 
 
 Figs. 43, 44, and 45 illustrate scale drawings, with 
 setting in brickwork, of two Babcock and Wilcox 
 boilers at the station belonging to the Chelsea Electri- 
 city Supply Company, Limited. The two boilers are set 
 together in masonry. Each boiler contains nine rows 
 
 Fig. 39 shows the boiler in section, and also the 
 passage of the gases. It will be seen that there are two 
 partitions, lined with firebricks, at right angles to the 
 tubes, the object of which is to compel the gases to 
 follow in a zigzag up and down the tubes. The gases thus 
 strike the tubes three times on their way to the chimney, 
 and almost at right angles every time. The boiler is 
 suspended, independent of the brickwork, from wrought- 
 iron girders resting on iron pillars, thus allowing the 
 boiler to expand freely and the masonry to be repaired 
 or removed without disturbing the b :iler. . The water 
 inside the tubes, as it is heated and evaporated, tends 
 to rise towards the higher end of the tubes and then 
 into the main shell, where the steam is given off to the 
 steam-room ; the water then descends from the main 
 shell through the tubes at the back, thus forming a con- 
 
 of five tubes each. The tubes are 4£in. in diameter, and 
 their length is 17ft. 6in. The main shell is 3ft. 
 in diameter, 23ft. long, and its axis is lift. lOfin. from 
 the ground. It will be seen that it is suspended from 
 the two girders, G t and G 2 , Fig. 43. For the purpose of 
 cleaning the boiler there are three cleaning-doors (C D) 
 built in the masonry, and large enough for a man to 
 enter through. When the inside of the tubes is to be 
 cleaned, the door in front of the boiler is opened, and 
 the mudholes at each end of the tubes are removed, 
 thus enabling the man to see right through when a lamp 
 is placed at the opening of the other end of the tube. 
 The thickness of the masonry is 1ft. bin., and the dis- 
 tance between opposite walls, which are lined with fire- 
 brick, is 3ft. 3in. 
 
 It will be seen that the gases only circulate twice 
 
94 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 round the tubes. Eising from the grate, they pass 
 through the spaces left between the front part of the 
 tubes, and strike the main shell, then descending, strike 
 that part of the tubes which lies between the partition 
 on the top of the bridge and the back end connections. 
 An underground channel carries the gases from all the 
 
 fails, by a donkey-pump. The boiler is provided with 
 one pressure-gauge, two water-gauges, one dead-weight 
 safety-valve, and a stop-valve. The total heating-sur- 
 face of each boiler is about 1,000 square feet ; the grate 
 being 6ft. long by 3ft. 2in. wide, the grate-surface is 
 19 square feet. 
 
 boilers into the chimney. A damper (D) , which can turn 
 round a horizontal axle, is placed on the top of the flue 
 which carries the gases to the underground channel 
 The opening left by the damper for the passage of the 
 gases can be regulated from the front of the boiler bv 
 moving a lever not shown in the drawings. The water 
 is pumped into the main shell by an injector, and if this 
 
 Another kind of water-tube boiler is illustrated in 
 Figs. 47, 48, and 49, which show longitudinal section, 
 cross-section, and front view respectively. The tubes 
 are here horizontal, and their diameter may be 2ft. 
 or more. The tubes are placed in three horizontal 
 rows, and the water circulates from the one row to the 
 next through a number of vertical connecting tubes. 
 
STEAM BOILERS. 
 
 95 
 
 Below is a mud-drum, the object of which is the same 
 as that in the preceding boiler. The flues are arranged 
 in a wheel draught, the gases circulating first round the 
 lowest set of tubes, then round the second set, and 
 
 Fiff/,3 
 
 Cross Section- 
 
 last round the top set, where they also help to dry the 
 steam, the temperature of the gases being then so low 
 that the part of the shell bounding the steam-room 
 cannot become red-hot. A steam-dome above connects 
 the steam-rooms of the three top shells. The steam- 
 valve, as well as two safety-vales, are mounted on the 
 
 an internal firebox, the sides and crown of which act as 
 heating-surface. As vertical boilers have a compara- 
 tively small water-surface and a high water-room, they 
 are apt to prime. 
 
 Vertical Boilers with Water Tubes. — The heat- 
 ing-surface of these boilers is increased by a number of 
 water-tubes passing through the firebox. 
 
 A vertical boiler with water-tubes is shown in Figs. 
 50 and 51, and is constructed by Messrs. E. Garrett 
 and Sons. The water-tubes are not quite horizontal, 
 which has the advantage of allowing the better escape 
 of the steam and heated water from the tubes, and 
 thus promoting the circulation of the water. There 
 are only three tubes, the internal diameter of which is 
 8in. ; a mudhole is placed in front of each tube, thus 
 permitting thorough cleaning without much trouble. 
 The crown of the firebox is stayed to the crown of the 
 boiler by the vertical uptake. 
 
 The height of the boiler, including uptake, is 9ft. 9in., 
 and its external diameter is 3ft. 6£in. The thick- 
 ness of the boiler shell is fin. There is one manhole 
 lOin. by 14Jin., and seven mudholes 3in. by 4£in. The 
 firebox shell is Jin. thick. The draught can be regu- 
 lated from outside by lifting or lowering the damper, 
 and thus closing, more or less, the opening between 
 the firebox and the uptake. The grate-surface is 7 
 square feet, and the total heating-surface is 78 square 
 feet. The average rate of consumption of coal is about 
 91b. to 101b. per hour per square foot of grate-surface. 
 The material used for all parts of this boiler is Siemens- 
 Martin steel. The gases rising from the grate will strike 
 the water-tubes, and thus be broken up and deflected 
 towards the sides and crown of the firebox. The 
 
 Fig. 49. 
 
 Fig. 48. 
 
 dome, which is also provided with a manhole. Man- 
 holes or mudholes are also provided at the front of each 
 tube and in the mud-drum. 
 
 Vertical Boilers. 
 
 Vertical boilers are only used for producing small 
 quantities of steam, and are handy where economy of 
 space has to be considered. They consist of a cylin- 
 drical barrel, the axis of which is vertical. They have 
 
 Fig. 47. 
 
 damper will also act as a deflector. The temperature 
 of the gases when flowing through the uptake must not 
 be so high as to make red-hot that part which passes 
 through the steam-room ; this part of the uptake will 
 act as a steam drier. 
 
 A vertical water-tube boiler made by Messrs. John 
 Musgrave and Sons, of Bolton, is shown in Figs. 52 
 and 53. 
 
 It has 120 tubes of 2Jin. diameter, and 3ft. 5in. long ; 
 
96 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 they are made of wrought iron and are lap-welded. 
 The tubes are placed in 12 horizontal rows, with 
 10 tubes in each row. It will be seen that the tubes 
 in any row are placed at right angles to the tubes in 
 the rows below and above, thus forming a complete 
 network for the purpose of breaking up the gases and 
 
 deflected into the tubes : a circulation is thus produced. 
 The tubes are fixed by expanding the ends in the tube- 
 plate holes. The boiler is 9ft. high and 5ft. inside 
 diameter. The main shell is welded and is made of 
 steel plates T Vin. thick, the crown being |in. thick. 
 For the purpose of cleaning the inside of the tubes the 
 
 Funnel 6 long. 
 
 M. H v 
 
 ■(rrTTOD HOLE 
 * TO EACH TUBE 
 
 F. D . W. 
 
 M. H . 
 V" MUD HOLES 
 ALL ROUND. 
 
 Fig. 50. 
 
 deflecting them to the sides. As the tubes are hori- 
 zontal, the heated water and the steam produced inside 
 the tubes would have difficulty in escaping from them ; 
 it is therefore necessary to promote circulation in 
 another way. This is done by placing a deflector, N, 
 at one end of each row of tubes, whereby some of the 
 water, while rising in the boiler, will be caught and 
 
 shell is made in two parts, having a joint 3ft. 5in. from 
 the base of the boiler. This joint is faced, and the 
 two parts of the shell are connected by 68 bolts 4]in. 
 in diameter. The crown of the shell is further fixed to 
 the uptake, as shown in the drawing, by 26 bolts ljin. 
 in diameter. 
 Fig. 54 shows the upper shell removed for cleaning, 
 
STEAM BOILERS. 
 
 97 
 
 and Fig. 55 shows the external appearance of the so close together. The soot and ashes which are col- 
 boiler, lected on the tubes are blown off by steam jets, pro- 
 
 6H 3 VERTICALCROSSTUBE BOILER 
 
 Fig. 51, 
 
 Fie. 53. 
 
 
 5* 
 
 f 1 J 111 11 i? u j,* 
 
 *[ 
 
 «t3*si> 
 
 * JKRIXKHMHK), 
 
 
 
 X 
 
 L/v 
 
 !X 
 
 
 
 Fig. 52. 
 
 The cleaning of the outside of the water-tubes with duced by letting steam into the hollow ball, B, which 
 a brush is made difficult on account of the tubes being has a number of nozzles screwed in like a porcupine. 
 
98 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 The pipe, P, which is connected at ihe one end to the 
 ball, and at the other end to the comhira^vn casting, 
 C, carries tbe steam from the steam-room into the ball. 
 There are four muc.ha.ss, 3jin. by -5in.. a: tire bottom 
 of the shell, for the purpose of removing the sediment. 
 
 The grate-surface is 15'5 square feet, and the total 
 heating-surface is 344 square feet ; 151b. to 181b. of 
 
 Fig. 54. 
 
 Fig 55. 
 
 The pressure-gauge, as well as the blast for forcing the 
 draught, are connected to the combination casting, C, 
 so as to avoid making a number of holes in the boiler. 
 
 The lower part of the firebox, surrounding the fire, 
 is a welded ring, 4ft. 5Jin. inside diameter, and is made 
 of steel plate fin. thick ; it is stayed to the outer shell by 
 stay-bolts. The upper firebox, containing the water- 
 
 coal may be burnt per square foot of grate-surface per 
 hour. 
 
 The Field Boiler, Figs. 56 and 57, is also a vertical 
 water-tube boiler ; the tubes are, however, of a peculiar 
 construction, known as the Field Tube, and named after 
 the inventor, Mr. Edward Field. Fig. 56 shows the 
 Field tube in section. The tube-plate of the boiler 
 
 tubes, is square, with corners of 3in. radius ; its inside 
 dimensions are 3ft. 4in.by3ft. 3in. high, and it is made 
 of steel plates T Vn. thick. The firebox crown is stayed 
 to the boiler crown by the cylindrical uptake, which is 
 20m. inside diameter, and made of steel plate fin. thick. 
 The tensile strength of the steel used in this boiler 
 is 28 tons, with an elongation of 20 per cent, in 8in, 
 
 is drilled with a hole of required taper to suit the 
 expanded end of the outer tube ; the latter is a lap- 
 welded tube, which hangs down into the fire from the 
 tube-plate, the pressure of the steam tending to keep it 
 always tight ; it supports on its top the edge of the 
 inner tube, which is open at both ends, and provided 
 with a. cone-shaped deflecting top. The effect of 
 
STEAM BOILERS. 
 
 99 
 
 this combination is to produce a natural and powerful 
 circulation of the water, due to the fact that the hottest 
 water in a boiler is always that near the heating- 
 surface. 
 
 The boiler consists of an outer cylindrical shell, and 
 an inner cylindrical firebox, the crown of which serves 
 as tube-plate for a number of Field tubes. The two 
 
 in masonry, and the draught can be regulated by the 
 damper. 
 
 The advantage in this boiler, besides the powerful 
 circulation, is the facility with which a tube, when 
 injured, may be removed and substituted by a new one. 
 It is stated that another advantage with the Field tubes 
 is the freedom from leakage, the pressure of the steam 
 
 J" 
 
 rfti ifh rf% „ rff-i 
 
 ( IWHH ) 
 
 lOnnrp — r rp 
 
 rrti mrHT,[f%ifWrfti ww 
 
 t — ) 
 
 DBUH ) 
 
 | Wr 
 
 \v-/ V_y w w \-/ <J W s/ \J 
 
 EQ 
 
 ILM\ I t 
 
 r ^ 
 
 Fig. 56. 
 
 shells are connected with stay-bolts, and at the fire-door 
 by a caulking ring. They are connected at the bottom 
 to a flat ring by angle iron rings. The firebox is divided 
 into two parts by the iron-plate partition, lined 
 with firebrick. This partition acts as a bridge. The 
 flues are arranged as a " down draught," and the 
 masonry is lined with firebrick. The crown of the 
 firebox is stayed to the crown of the boiler shell by 
 a number of suspension stays. The ashpit is formed 
 
 keeping the thin outer tube tight in the hole made in 
 the tube-plate. 
 
 Field tubes are also often adapted in existing boilers, 
 and can be introduced into the cylindrical flues of 
 Cornish and Lancashire boilers. Being only fixed 
 at one end, they do not assist in strengthening the 
 cylindrical flue, as is the case with the Galloway 
 tubes. 
 
 Vertical Boilers with Flue-Tubes. — The heating- 
 
100 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 surface of this class of boilers is increased by a number 
 of flue-tubes, similar to those used in locomotive and 
 other multitubular boilers. 
 
 Figs. P and Q are illustrations of Messrs. Davey, 
 Paxman, and Co.'s Essex boiler. The gases rising from 
 the grate strike the crown of the firebox and flow 
 through a cylindrical neck into the combustion chamber, 
 
 are made of wrought iron, are lap-welded, and ex- 
 panded in the tube-plate holes, thickness of material 
 being No. 11 B.W.G., and their diameter is 2£in. The 
 removal of a tube and the fixing of a new one is done 
 through the doorway to the combustion chamber. The 
 firebox is also made of steel plates fin. thick. The 
 material used for the smoke-box is cast iron, of which 
 
 Fig. 58. 
 
 where the combustion is completed. The combustion 
 chamber is shut from outside by a door lined with fire- 
 brick, and is partly bounded by the curved tube-plate 
 The gases pass from the combustion chamber through 
 38 flue-tubes into the smoke-box, whence they pass 
 through the uptake and escape by the chimney The 
 boiler shell is 8ft. 2in. high and 3ft. 4in. diameter. It is 
 made of steel, thickness of plate being fin. The tubes 
 
 the standpipe on the top of the boiler is also made. 
 The manhole ring as well as the manhole cover are 
 made of wrought iron. Steam-valve and safety-valve 
 are mounted on the standpipe. There are four mud- 
 holes equally divided round the bottom of the shell. 
 The tensile strength of the steel used for the shell is 
 28 tons, with about 24 per cent, elongation, and that 
 used for the firebox is 25 tons, with an elongation of 
 
SfBAM BOILERS. 
 
 101 
 
 28 per cent. The total heating-surface is 84'5 square 
 feet, and the grate-surface is 5 - 84 ,'square feet. The 
 working pressure is 751b. per square inch. 
 
 A vertical multitubular boiler made by Messrs. 
 Eansomes, Sims, and Jefferies, of Ipswich, is shown in 
 Figs. 58 and 59. The firebox is connected with the 
 combustion chamber by 20 short flue-tubes, and from 
 this 37 tubes lead the gases to the smoke-box. 
 
 The diameter of the tubes is 2&in. ; they are made of 
 
 different parts of the shell for the cleaning of the fire- 
 box shell. The outside plate of the combustion 
 chamber is lined with firebrick. The grate-surface is 
 6"8 square feet ; the total heating-surface is 111"5 
 square feet, the heating-surface of the tubes being 
 76"81 square feet. Eleven pounds of best Welsh coal 
 may be burnt per square foot of grate-surface per hour, 
 and 10" 21b. of water are supposed to be evaporated from 
 and at 212 deg. F, per pound of Welsh coal. 
 
 <&©Qi©©© 
 
 ©©©©©■©© 
 ©©©©©© 
 
 ©©©!©©© 
 
 ©©©©©©© 
 ©©©I©©© 
 
 Lwmmmmmm 
 
 ^QL 
 
 JOL 
 
 io. 59. 
 
 wrought iron and lap-welded. The material used for 
 the boiler shell, firebox, combustion chamber, and man- 
 hole ring is Siemens-Martin mild steel. The height of 
 the boiler is 8ft. 3in., and its outside diameter is 
 3ft. 5|in. The firebox and boiler shell are jointed at 
 the bottom by a Z iron ring, and at the fire-door by a 
 thick caulking ring. There is one manhole at the 
 crown of the boiler, and four handholes are placed at 
 
 Strength of Boilers, 
 
 The material which is used in making boiler shells 
 and internal flues is either wrought iron or mild steel. 
 Wrought iron combines great tensile strength and duc- 
 tility; it can be forged, welded, rolled, and flanged, 
 which are necessary properties in material to be used 
 for boilers. When wrought iron is hammered, flanged, 
 punched, or otherwise worked when cold, its tensile 
 
ioa 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 strength increases ; but at the same time its ductility 
 diminishes. Annealing, however, brings the material 
 back to its former state ; for this reason wrought-iron 
 plates and bars which have been worked cold should 
 be annealed before being used for boiler-making. The 
 tensile strength of wrought-iron plates is greatest when 
 the fracture is at right angles to the direction of rolling, 
 and least when the fracture is parallel to that direction. 
 The tensile strength of wrought-iron plates with the 
 fibres is about 21 tons per square inch, and when the 
 fracture is parallel to the fibres the tensile strength is 
 about 18 to 19 tons per square inch. The most suitable 
 wrought iron for boilers is that which in the testing 
 machine gives the greatest elongation before breaking, 
 combined with the greatest tensile strength. Great 
 elongation or great contraction of sectional area are 
 signs of great ductility. 
 
 The softer kinds of steels-so-called mild steel — ap- 
 proach wrought iron in character, and can be welded, 
 flanged) rolled, etc. Its tensile strength, however, is 
 greater than that of wrought iron. Steel conceals 
 faults more than wrought iron, and varies more in 
 quality; It is also more liable to be injured from over- 
 heating, and great care has to be taken when working 
 steel in the fire. 
 
 Hammering, flanging, and rolling are also more 
 injurious to steel than to iron. Annealing, however, 
 will remedy the faults. Steel plates are stronger with 
 the fibres than across the fibres. The tensile strength 
 of steel plates used for boilers is from 25 to 30 tons per 
 square inch, with an elongation of 20 to 28 per cent, in 
 test-strips of 8in. to lOin. long. 
 
 On account of the greater strength of the material, 
 steel shells can be made thinner than iron shells, thus 
 securing better riveting and less corrosion from the fire. 
 Cast iron is as strong as wrought iron when subject 
 to pressure, but is very much weaker under tension. It 
 is very brittle, and is liable to internal stress, in conse- 
 quence of its contraction while cooling. It is therefore 
 not fit to be used as boiler material. 
 
 Copper has sometimes been used for boiler construc- 
 tion. Its tensile strength is smaller than that of iron, 
 but it has a greater conductivity for heat, and can 
 therefore stand intense heat better than iron. For this 
 reason it is often used for making fireboxes for loco- 
 motive boilers, where the temperature of the fire is 
 greater than in ordinary boilers. Copper also resists 
 corrosion better than iron. 
 
 Boiler shells are built up of a number of cylindrical 
 rings, which fit alternately inside or outside into the 
 neighbouring ones, and are jointed by riveting. Each 
 ring is made of wrought-iron or steel plates, which 
 have been curved in the plate-bending machine to the 
 required radius. The plates are then joined, either by 
 welding, or more frequently by riveting. If the latter 
 process be used, it will be necessary to bring the edge 
 of the one plate to bed close against the other, in order 
 to make a tight joint. For this purpose the edges of 
 the plates are planed to a bevel before bending 
 When the joint is completed, the bevelled edge of the 
 
 plate is fullered with a broad fiat tool, whereby a tight 
 joint is secured. Before the planing of the edges was 
 introduced, it was customary to close the joint by 
 caulking, that is, to bring the edge against the plate by 
 
 Fig. 60. 
 
 means of a thin tool, like a chisel. By this process an 
 extra strain is liable to be thrown on the rivets. 
 
 ^a , 
 
 Fro. 61. 
 
 Blvet Joints. — Eivets used in boiler-making are made 
 either of best wrought iron or of best mild steel. When 
 
 f -*.&./:£ , 
 
 Fig. 62. 
 
 iron is used it should be best scrap iron, for which old 
 boiler-plates are generally used. The plates are cut 
 up in strips, which are well cleaned from oxide and 
 rust. The strips are then brought up to welding heat, 
 after which they are worked under a steam-hammer 
 
STEAM BOILERS. 
 
 m 
 
 into blooms ; they are then rolled down into rivet iron. 
 The rivet bars, whether made of wrought iron or mild 
 steel, should be tested for tensile strength and elonga- 
 tion. The tensile strength should be from 26 to 30 
 tons per square inch, and the elongation in lOin. 
 should not be less than 25 per cent. The iron is 
 then taken to the rivet-making machine, where it is 
 cut into the proper length, and each piece, when red- 
 
 Fig. 63. 
 
 hot, is placed in a die, where it is subjected to com- 
 pression by which the head is formed. The head of 
 rivets used for boilers is generally cup-shaped, as shown 
 in Fig. 60. The diameter of the rivet is made a little 
 smaller than that of the hole in the plate, but, after 
 riveting, the two diameters are supposed to be the 
 same, the body of the rivet having been compressed so 
 much that it fills up the entire hole. The process of 
 riveting consists in forming the second head while the 
 rivet is red-hot and is placed in the hole. This can be 
 done either by hand or by machinery. In the first 
 case the rivet must be held in the hole by letting it 
 rest on a heavy receiving die held by one man, while 
 two other men working with hammers form the second 
 
 Fig. 64. 
 
 head, the shape of which is shown in Fig. 61 ; it is often 
 finished by the use of a hollow die — the shape of the 
 head will then be more like a cup. Machine riveting, 
 however, gives a better result than hand riveting, as 
 greater pressure can be produced, and the rivets will 
 therefore fill up the holes more completely. The rivet 
 is held in the hole by a receiving die placed inside the 
 shell, whereas the heading die is moved backwards and 
 forwards by the machine, thus squeezing the rivet, 
 whereby the second head is formed. The pressure, 
 which must be adjusted according to the diameter of 
 the rivet, is produced either by a hydraulic head or by 
 
 Fig. 63. 
 
 Fig. 64. 
 
 an arrangement of levers and weights. Fig. 62 shows 
 the form of rivets used in machine riveting. 
 
104 
 
 PRACTICAL ELECTRICAL ENGINEERING-. 
 
 The rivet holes are either punched or drilled, but the 
 former process does not give such a good result as the 
 latter. In punching the plates, the metal surrounding 
 the hole will be compressed, and this weakens the 
 plates. Plates which are punched should therefore be 
 annealed, at any rate if they are of steel. Another 
 objection to punching is the difficulty in bringing the 
 holes which have to receive the same rivet to agree ; the 
 holes have therefore to be bored out by a rymer. The 
 old-fashioned way of bringing the holes opposite to each 
 other by means of a taper steel drift should be avoided, 
 as it often causes fractures in the plates which are 
 not easily discovered. Until a comparatively recent 
 date, the universal practice was to punch the plates, 
 
 Fig. 65. 
 
 and this before they had been bent to the proper 
 radius; but since the introduction of high pressure 
 steam, boiler-makers have found it necessary to improve 
 the manufacturing of the various parts which consti- 
 tute a boiler, and it is especially since steel plates have 
 come into use that punching has been entirely substi- 
 tuted by drilling in good boiler-making. In fact, the 
 Board of Trade require that steel plates should be 
 drilled ; assent may, however, be given for the plates 
 to be punched, but then the particulars of the punching 
 and annealing should be submitted to the Board for 
 consideration. 
 
 The Board of Trade rules only refer to marine boilers 
 and engines for passenger steamers, there being no 
 
 Fig. 66. 
 
 rules for land boilers; but there is no reason why the 
 same rules should not be adhered to in all cases 
 
 In boilers which are to be worked at low pressure 
 punching^ still used, as this process is much cheaper 
 than drilling. ^ 
 
 , n S id fL ma .y either be lap-joints, Figs. 63, 64, 
 
 wwf'°^ . Utt " J01 , ntS,KgS - 66 and 67 > ac ^ing to 
 whether the two plates lap over or meet abut 
 
 The following symbols will be used in dealing with 
 riveted joints : 
 
 d diameter of rivet, which is supposed to be the same 
 as that of the hole after riveting is completed. 
 
 iio. 65. 
 
 Fig. 66. 
 
 S thickness of plate. 
 
 Si thickness of strap or straps in butt-joints. 
 
 p x pitch of rivet— i.e., distance between centres of two 
 
STEAM BOILERS. 
 
 105 
 
 consecutive rivets in the same row parallel to the 
 
 edge of the plate. 
 p. 2 smallest distance between centres of two rivets 
 
 belonging to two consecutive rows of rivets. 
 p 3 distance between centre of rivet and edge of plate. 
 T original tensile strength of boiler-plate in pounds per 
 
 square inch — that is, the tensile strength of the 
 
 test-strip. 
 T x apparent tensile strength of boiler-plate in pounds 
 
 per square inch — that is, the tensile strength of 
 
 the plate as it appears to be when testing the com- 
 pleted joint in the testing machine. 
 S original shearing strength of rivet iron or rivet steel 
 
 in pounds per square inch. 
 S x apparent shearing strength of rivet iron or rivet steel 
 
 in pounds per square inch. 
 C bearing pressure — that is, the pressure on the rivet 
 
 tending to crush the rivet or that part of the plate 
 
 which surrounds the rivet hole. 
 r\ ratio of shearing strength of rivet to tensile strength 
 
 of plate. 
 <f> efficiency of joint — that is, the ratio of strength of 
 
 joint to that of the full boiler-plate. 
 n number of rivets in one row. 
 
 When, however, the joint is tested, the apparent tensile 
 strength of the plate is found to be less than that of 
 the drilled plate, the reason being that the pulling forces, 
 not acting in the same line, will bend the plates, whereby 
 a bending stress, in addition to the tensile stress, will 
 be thrown on the plates. 
 
 For the same reason the rivets are also subject to a 
 bending stress, with the result that the apparent shear- 
 ing strength of the rivets in a joint is smaller than the 
 original one. The diminution in the shearing strength 
 has further been found to be greater in joints with 
 drilled holes than with punched holes, the reason being 
 that the sharp edges of tbe drilled holes have a greater 
 shearing effect on the rivets than the more blunt edges 
 of the punched holes. By rounding the edges of 
 drilled holes the shearing effect is diminished. 
 
 Fig. 67. 
 
 The force which tends to burst the joint has to over- 
 come one of the following resistances : 
 
 (a) The resistance of the plate against being torn 
 along the centre line of the rivet holes belonging to the 
 same row, this line being the fracture line. 
 
 (b) The resistance of the rivets against shearing. 
 
 (c) The resistance offered by the rivet, or that part 
 of the plate which surrounds the rivet hole, to with- 
 stand crushing. 
 
 (d) The resistance of the plate against being torn 
 between the rivet hole and the edge of the plate. 
 
 Another resistance which has to be overcome, is 
 the friction between the plates. The rivet, being red- 
 hot while the second head is formed, will contract in 
 cooling, and thus press the two plates together. The 
 frictional resistance which is hereby produced cannot, 
 however, be relied upon, as it varies with the tempera- 
 ture of the rivet, when the riveting was performed, and 
 also upon vibrations. This resistance is therefore not 
 taken into account when calculating the strength of a 
 joint. 
 
 It has been found by experiment that the tensile 
 strength along the fracture line of a drilled plate is 
 greater than the original tensile strength of test-strips. 
 
 Fig. 67. 
 
 It is evident that in calculating the resistance of a 
 joint the strength of the material as it appears in an 
 actual joint should be taken. The values in the follow- 
 ing table are taken from Professor Unwin : 
 
 
 T 
 
 mean. 
 
 T 
 
 S 
 about 
 
 £ 
 
 i 
 
 Material. 
 
 Single 
 riveted. 
 
 Double 
 riveted. 
 
 are 
 
 
 A 
 
 B 
 
 A 
 
 B 
 
 A 
 
 B 
 
 Iron plates ... 
 Steel plates ... 
 Iron rivets ... 
 Steel rivets ... 
 
 46000 
 62000 
 62000 
 66000 
 
 0-88 
 1-00 
 
 0-77 
 0-90 
 
 0-95 
 1-06 
 
 0-85 
 1-00 
 
 0-8 T 
 0-8 T 
 0-8 T 
 0'8T 
 
 43000 
 49000 
 
 46000 
 53000 
 
 A. Drilled B. Punched. 
 
100 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 In lap-joints and butt-joints with single cover, the 
 rivets are sheared in one plane only ; such joints are 
 therefore exiled single-shear joints. In butt-joints with 
 double cover, the shearing takes place in two planes, 
 and the joints are therefore double-shear joints. In a 
 well-designed joint the shearing resistance of the rivets 
 should be equal to the tearing resistance of the plate 
 along the line of fracture. 
 
 The shearing resistance of the rivets in a single- 
 shear joint will be — for single-riveted joints 
 
 Table A. 
 
 .x^x Sl 
 
 for double-riveted joints 
 
 2 n ^-t- x Sj 
 
 (1) 
 
 (2) 
 
 The tearing resistance of the plates in single-riveted, 
 as well as in double-riveted joints, will be 
 
 n (p 1 - d) x S x Tj 
 
 (3) 
 
 By equating (1) and (3) we shall have for single- 
 riveted joints 
 
 ^ 4 ' S 
 
 (4) 
 
 by equating (2) and (3) we shall have for double-riveted 
 joints 
 
 ^i = ^+|x, x ^ 
 
 (5) 
 
 Q 
 
 where « =■ — L . 
 
 As the tearing resistance of the full plate is 
 n x Pl x S x T, it is evident that the efficiency of the 
 joint will be 
 
 f . n x fo - d) x S x T 2 ( 1 _d\T l ... 
 nxftxjxT -V pj "t" (b) 
 
 for both single and double riveted joints. 
 
 In the following tables, the holes are assumed to 
 have been drilled in the plates, and the values of 
 
 T 
 T, — x , and Sj are taken from the preceding table. 
 
 The table A shows that double-riveted joints are 
 stronger than single-riveted ones, and it will further 
 be seen that the pitch of the rivet and the efficiency of 
 
 the joint depend upon d, *?, and |; we must there- 
 fore have formulae by which these quantities can be 
 found. 
 
 _ In calculating the above table the possibility of the 
 rivet or the plate surrounding the rivet hole being 
 
 ZS ?T. n0t b ? n tak ? int ° acCOUnt > nor h ^e we which gives 
 examined the resistance of the plate against being torn 
 between the rivet bole and the edge of the plate The 
 latter however, will be great enough if W e mak 
 
 
 
 Single rivetec 
 
 joints. 
 
 
 1\ 
 
 Pi 
 
 4> 
 
 Iron plates 
 with y 
 
 106 
 1-21 
 
 0-69 
 0-79 
 
 0-98 
 1-12 
 0-65 
 0-75 
 
 d + 0-832 x- 
 8 
 
 d + 0-95x- 
 8 
 
 (^ + 0-542xl 2 
 8 
 
 d + 0-62x_ 
 8 
 
 Double riveted 
 
 d+l-5ix- 
 8 
 
 d+l-76xl 2 
 8 
 
 rf+l-02xl 2 
 8 
 
 rf + l-18xl 2 
 8 
 
 0-73x1 
 8 
 
 iron rivets. 
 
 Iron plates 7 
 with > 
 
 1+0-832x1 
 8 
 
 0-836x1 
 8 
 
 steel rivets. 
 
 Steel plates ] 
 with I 
 
 1+0-95x1 
 8 
 
 0-542x1 
 8 
 
 iron rivets. 
 
 Steel plates ] 
 
 with I 
 
 steel rivets. J 
 
 Iron plates 1 
 with l 
 
 1+0:542x1 
 8 
 
 0-62x1 
 8 
 
 1+0-62x1 
 8 
 
 joints. 
 
 146x1 
 8 
 
 iron rivets. 
 
 Iron plates ] 
 with L 
 
 1 + 1-54x1 
 
 1-67x1 
 8 
 
 steel rivets. 
 
 Steel plates ] 
 with i- 
 iron rivets. 
 
 Steel plates ] 
 with L 
 
 1 + 1-76x1 
 8 
 
 1-08x1 
 8 
 
 1 + 1-02x1 
 8 
 
 1-25x1 
 8 
 
 steel rivets. 
 
 1 + H8x£ 
 
 
 
 If the pressure — the bearing pressure— between the 
 rivet and the plate is too great, the effect will be to 
 make the joint loose. The resistance of the joint to 
 withstand this pressure depends upon whether the bear- 
 ing area, dxS, which supports the pressure is large 
 enough. We ought, therefore, to have, in a well- 
 designed joint, the maximum crushing force allowed 
 on the rivet equal to shearing resistance, or for single- 
 shear joints 
 
 d C 
 
 i-l"27x Fi - 
 
 As a result of experiments on riveted joints, Professor 
 Kennedy has found that the intensity of bearing 
 pressure on the rivets exercises a very iinportairt in- 
 
STEAM BOILERS. 
 
 107 
 
 fluence on their strength. So long as it does not much 
 exceed 40 tons per square inch on steel rivets it does 
 not seem to affect their strength, but pressures of 50 to 
 55 tons per square inch seem to cause the rivets to 
 shear at stresses varying from 16 to 18 tons per square 
 inch. 
 
 Q 
 
 We may therefore allow -q- = 2 in single-shear joints, 
 
 where Sj is taken from the table on page 105, so that we 
 have for steel as well as for iron rivets 
 
 d 
 
 2-54 
 
 (7) 
 
 The diameter of the rivet may be found from the 
 following rule, 
 
 d = 1-2 JJ (8 ) 
 
 which is due to Professor Unwin. The two formulae 
 (7) and (8) combined require S% 0'22in. This is, 
 however, within the limits used with boilers. 
 By eliminating S between (7) and (8) we have 
 
 d!> 0'57in.; or, say, fin. 
 Formula (8) may also be written thus : 
 
 d 2 
 
 = 1-44 
 
 (9) 
 
 (10) 
 
 We can now, by means of (7), (8), and (10) and the 
 above table, find^ and <f> of single-shear joints. 
 
 According to the Board of Trade rules we must 
 have — 
 
 p a ^2d (ID 
 
 for chain joints (see Figs. 74 and 75), and for zig-zag 
 joints (Fig. 76)— 
 
 = &p x + Ad _ (12) 
 
 Pi 
 
 10 
 
 It is necessary that p x should be greater than twice 
 the diameter of the rivet head, or, according to Figs. 71 
 and 73, we must have p x > 3'6 x d or 3'8 x d. It is 
 therefore possible that the table on page 106 will give us 
 values for p x which are too small. In such cases we 
 
 d 2 
 must make — > T44. Suppose we make it equal to a, 
 
 8 
 then as —.% 2*54 we must have 8- 
 
 6-45 
 
 Formula (6) shows that the strength of the joint in- 
 
 creases bv making — small. In boilers it is, however, 
 
 Pi 
 necessary that the joints should be water and steam 
 
 tight, and — cannot therefore be made so small as in 
 
 Pi 
 girders. 
 
 The shearing resistance of the rivets in double-shear 
 joints will be — for single-riveted joints 
 
 n x 2 x —r- x S, 
 
 (13) 
 
 and for double-riveted joints 
 
 4 
 
 2 n x 2 x '-^-^ x S x 
 
 . (14) 
 
 The tearing resistance of the plates in single-riveted, 
 as well as in double-riveted joints, will be 
 
 n (p x - d) S T a (15) 
 
 By equating (13) and (15) we shall have for single- 
 riveted joints 
 
 , t d 2 S, 
 
 • (16) 
 
 and by equating (14) and (15) we find that the pitch 
 for double 7 riv.eted joints will be 
 
 p 1= cZ +7r x|x|i 
 
 (17) 
 
 The efficiency of both kinds of joints will be 
 
 ^. XPi-flJT! A d\T± m , (1Q) 
 
 ip x 6 T v p x ■ 
 
 Table B. 
 
 
 Single-riveted double-shear joints. 
 
 
 Si 
 
 Pi 
 
 <£ 
 
 Iron plates ] 
 with L 
 iron rivets. 
 
 Steel plates ] 
 with \ 
 iron rivets. 
 
 Steel plates 1 
 with I 
 steel rivets. 
 
 0-82 
 0-53 
 
 061 
 
 Hi 
 d + V29x%. 
 
 
 
 rf + 0-83x^- 
 
 
 
 d + 0-96x1 
 
 
 
 1,4 4 
 
 1 + 1-29x1 
 
 
 
 0-83 x 1 
 
 1+0-83x1 
 
 
 
 0-96 x 1 
 8 
 
 1+0-96x1 
 S 
 
 Iron plates 
 
 with 
 iron rivets. 
 
 Steel plates 
 
 with 
 iron rivets. 
 
 Steel plates 
 
 with 
 steel rivets. 
 
 Double-riveted double-shear joints. 
 
 d 
 
 0-75 
 
 0-50 
 
 0-58 
 
 d + 2-36xl 
 
 d+l-57x 
 
 d+l-82x 
 
 d? 
 
 d* 
 
 2-24 x ' 
 
 l + 2-36x! 
 
 1-66 x 
 
 d 
 
 1 + 1-57x1 
 o 
 
 1-93 x' 
 
 l + l-82x!i 
 
 
 
 Professor Kennedy finds that in double-shear joints 
 it is impossible to develop the full shearing resistance of 
 the joint without getting excessive bearing pressure, 
 because the shearing area is doubled without increasing 
 the area, dxS, on which the pressure acts, 
 
108 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Let us assume a maxi mum bearing pressure of 50 tons A flat surface, subject to fluid pressure, 
 
 per square incb on steel rivets, tben the shearing will A circular cylinder, subject to fluid pressure acting on 
 probably take place at 17 tons per square inch. We its outside surface. 
 
 have thus increased the bearing pressure allowed in (i) j n the fi rs t case> w here we have a cylinder acted 
 
 formula (7) from about 43 tons to 50 tons, and at the U p on by a pressure from inside, the cylindrical form will 
 
 same time we have reduced the shearing strength from no t b e changed. The pressure, however, will tend to 
 
 22 tons to 17 tons per square inch, or to 77 per cent, fracture the shell, partly in a plane at right angles to 
 
 of the former value. We can, therefore, take ^ the axis of the shell, and partly in a plane through this 
 
 "i axis. 
 equal to three in double-shear joints. As the maximum The first-named fracture would be caused by the 
 
 crushing force should be equal to the shearing resist- total steam pressure acting upon the end-plates of the 
 
 ance of the rivets, we must have shell. This pressure is evidently equal to 
 
 (jxixC^x^xSj. . .(19) ^p**** (23) 
 
 d 2 C The resistance of the material which the pressure 
 
 or ' ~$ = ^ x Si' has to overcome is 
 
 c 7rD 6 Xc$xT (24) 
 
 and as S^ =3 ' where T is the tensile strength of the boiler plate 
 
 across the grains. The value of S, which we would 
 
 - <l - 9 (20) obtain by equating (1) and (2), would be tbat at which 
 
 the boiler would burst. It is therefore evident that we 
 
 Therefore, as long as (5^ fin., we may take must not take the whole value of T, but only a fraction 
 
 ^ 2 thereof, 
 
 ■y = 1'35 (21) We shall therefore have 
 
 By eliminating S between (12) and (13) we have —r~ X P" = "* ^ x ^ * ~~ • • • (25) 
 
 4 f s 
 
 d > - 71, or say fin (22) f s j s ca u e( j a sa f e ty factor, and is the pure number 
 
 p 2 should be taken from (11) and (12), on page 107. b y whi ° h we must divide the ^ tim ate strength of the 
 
 T ,, , , , -r, ,, , j. m j-i. material, m order to reduce the stress to a safe limit. 
 
 In the table B, the values of T x are the same as -n, ,. .„. . 
 
 in the table on page 106, but the shearing strength of ^ 
 
 the rivets has been reduced to 77 per cent, of the values § = x ^ e x fi- . (26) 
 
 given in tbat table, in accordance with what has been 4 T 
 
 explained above. The force which tendg to fcaoiliaie the snell ^ a plane 
 
 Prom a theoretical point of view, the thickness, S v through the axis of the shell will be 
 of butt straps should be for single-shear joints equal to ^ 
 
 j! H DiXLjXj), , . . . (27) 
 
 S, and for double-shear joints equal to~ 2 ; but it has where L& is the length of the shell. The resistance of 
 
 been found by experiment that this is not sufficient. the shdl to witnstand thi s force will be 
 
 In practice, therefore, the thickness of single straps is 2 S x L x — (28) 
 
 _ 9 f> 
 
 taken as <J X > g x <J, w here T is the tensile strength of the plate with the 
 
 grains. By equating (5) and (6) we obtain 
 
 and for double straps, <5, ^ - X S. r> ^ „ t 
 
 8 s= Ub Pe xA , (29) 
 
 2 T 
 
 Butt straps should be cut from plates, and not from A ., ., ,,.,-,, . 
 
 bars, and should be of as good a quality as the shell- the tenSlle stren g th Wlth the § rams ls & e&tex thm 
 
 plates, and for the longitudinal seams they should be that acr0SS the grams ' jt Wl11 be seen that S found from 
 
 cut across the fibre. ( 7 ) w * u ' 3e almost twice that found from (4), / s being 
 
 Plate Thickness '(S) of Boiler Shells and Cylindrical ^V?"? in b ° th **""**' ?' , ther f f • ff boiler 
 
 Flues.-For the purpose of calculating the strength of f^" strong enough to prevent a longitudinal fracture, 
 
 the various parts of the boiler we must consider the f, & } B ° h ° aW ? t0 Wlthstand the P ressure on the 
 
 following cases separately : end-plates. For the same reason the longitudinal joints 
 
 . . , .. , . . of a boiler shell should be double-riveted butt-joints, 
 
 ^M^rf' SU t0 Pr6SSUre a ° ting ° n With d0uble C0VerS ' Whereas the J' 0ints ' connecting two 
 
 . ° e " consecutive rings, will be strong enough if either single 
 
 A semi-sphere, subject to fluid pressure acting on its or double-riveted lap-joints, according to the stress, tq 
 
 inside surface, which they are subjected, 
 
S'fEAM BdlLERS. 
 
 109 
 
 It is evident that we must not have the longitudinal 
 joints of any two shell rings in the same line, as the 
 boiler would then be unnecessarily weakened. 
 
 Let <j>" be the efficiency of the longitudinal joint, 
 then we must have 
 
 When a larger plate of thickness S is supported by 
 stay-bolts or longitudinal stays, and the distance be- 
 tween centres of consecutive stays is a, then the 
 greatest stress on the plate will be 
 
 8 x L r x __ x 0" = i x D 6 x L r x p e . (43) 
 
 2 „ a 2 
 r = - x — 
 
 9 t 2 
 
 ■Pe 
 
 (50) 
 
 "Where L r is the length of a shell ring, (43) will give us — 
 
 20' T * 
 
 The lap-joint will be strong enough if 
 D 6 
 
 <S> 
 
 x P* x /« 
 4 T 
 
 (44) 
 
 (45) 
 
 The area supported by each stay will be a square 
 inches. Formulae (49) and (50) are due to G-rashof ; 
 the safe value of r depends upon the construction of 
 the stays, and whether the plates are exposed to the 
 impact of heat or flame. 
 
 Stays used for supporting flat surfaces may be : 
 
 where <j> is the efficiency of the joint 
 
 The safety factor used by the Board of Trade is 5 
 for boiler shells made of best material, rivet holes 
 drilled in place, and all longitudinal seams made with 
 double-riveted butt-joints with two straps, and alto- 
 gether of superior workmanship. 
 
 (2) The hemisphere is often used for end-plates for 
 boiler shells or domes. The shape of the sphere, like 
 that of a cylinder, is not altered by the inside pressure, 
 and the stress in the material is the same at every 
 point of the shell. Taking the diameter of the sphere 
 as equal to that of the boiler shell or dome whose ends 
 it is to close, we find that the force tending to fracture 
 it along a diametrical plane is 
 
 ttD 6 2 
 
 4 
 
 ■x p» 
 
 The resistance of the material will be 
 
 rD fc 
 
 T 
 
 where S 2 is the thickness of the material. 
 (46) and (47) we obtain 
 
 p e 
 
 4 
 
 T 
 
 • • (46) 
 
 . ■ (47) 
 By equating 
 
 • • (48) 
 
 The plate thickness of the sphere need not therefore 
 be more than half that of the cylinder-plate ; /„ might 
 be taken as 5 or 6. 
 
 Dished ends should, according to Board of Trade 
 rules, be stayed as .flat ends, but if they can be con- 
 sidered as portions of spheres, the stays may have a 
 stress exceeding stays for fiat ends. The ends of the 
 domes in the boilers, Figs. 32 and 36, are dished, and 
 supported by longitudinal stays. 
 
 (3) Fluid pressure acting upon flat surfaces tends to 
 bulge the plate. Such plates must therefore be of small 
 dimensions, or else they must be well stayed, so as to 
 leave small distances between the supporting stays. 
 
 When a circular plate, fastened along its periphery, 
 having a diameter D and a thickness 8, is acted upon 
 by the effective pressure, p„ then the greatest stress, 
 r, in the material will be 
 
 1 - D 2 
 
 t = — x — 
 
 6 S 2 
 
 X p e 
 
 (49) 
 
 Fig. 69. 
 
 (a) G-usset stays, Fig. 69, which support the flat 
 ends, B, of boilers by making connection with the 
 cylindrical shell, B P. They are usually made of a 
 single plate, Gr S, of iron or mild steel, the one end of 
 which is riveted to the end-plate, E, and the other end 
 to the boiler barrel, B P, by ni3ans of angle-iron. 
 Gusset stays are generally used in Cornish and Lanca- 
 shire boilers, see Figs. 26 and 33. 
 
 (6) Longitudinal stays, which are long iron or steel 
 rods reaching from the one end of the boiler to the 
 other, and supporting the flat surfaces of the end-plates 
 by balancing the pressures to which they are exposed. 
 
 Longitudinal stays are generally screwed at both ends 
 and fixed to the end-plates by nuts, as shown in Fig. 
 70, where E is the end-plate of the boiler, L S the 
 longitudinal stay ; C 1 and C 2 are copper washers. 
 
lio 
 
 'PRACTICAL ELECTRICAL ENGINEERING. 
 
 In multitubular boilers, some of the tubes are screwed 
 at both ends and are provided with nuts ; they act as 
 longitudinal stays, supporting the flat tube-plates, and 
 are therefore called stay-tubes. 
 
 The Board of Trade allow on solid steel screwed 
 stays, which have not been welded or otherwise worked 
 after heating, a working stress of 9,0001b. per square 
 inch of net section, provided the tensile stress is from 
 27 to 32 tons per square inch, and the elongation in 
 10in., about 25 per cent., and not less than 20 per 
 cent. This corresponds to a safety factor of between 
 6-7 and 8. 
 
 In the Garrett loco, boiler, Fig. 21, the longitudinal 
 stays do not go through the plates ; the end-plates 
 are therefore not weakened by making holes in them. 
 These connections are shown in detail in Figs. 71 and 72. 
 
 L.S. 
 
 Fig. 71. 
 
 In Fig. 71, e is an eye riveted to the end-plate, E ; the 
 longitudinal stay, L S, is fixed to the eye by the pin, p. 
 
 Fig. 72. 
 
 The stay, L S, in Fig. 72, is pinned to the two angle- 
 irons A 1 and A 2 , which at the same time stiffen the 
 end-plate, E. 
 
 (c) The flat crowns of loco, fireboxes and of the 
 combustion chambers of marine boilers are often stif- 
 fened by crown stays or girder stays, as will be seen in 
 Figs. 19 and 24. Girder stays are made of wrought iron or 
 mild steel; the girder must not rest directly on the roof- 
 plate, but the water must have free passage, so as to 
 prevent overheating. 
 
 Girder stays prevent, to a certain extent, the free 
 circulation of the water above the crown-plate, and for 
 
 this reason corrugated plates, as in the Garrett boiler, 
 Fig. 22, may be considered better. 
 
 The method of staying the firebox roof by long 
 bolts to the crown of the boiler, as in the Field boiler, 
 Fig. 56, is, no doubt, also better than girder stays. 
 
 (d) Stay-bolts are used for stiffening two surfaces, 
 generally flat surfaces, which are close together. 
 
 SB 
 
 Fig. 73. 
 
 The stay-bolt, shown in Fig. 73, is that which is 
 most commonly used. It is screwed at both ends, and 
 riveted over the plates, to which it acts as a stay. Stay- 
 bolts are made of wrought iron or steel; but in loco- 
 motive boilers, they are often made of copper, as this 
 metal stands the intense heat of the fire better than 
 the two other materials. In Fig. 73, B P stands for 
 boiler-plate, F P for firebox-plate, and S B for stay- 
 bolt. 
 
 (4) A circular cylinder, subjected to external pressure, 
 will remain unaltered in shape, if the cross-section is 
 a perfect circle, and there would be a uniform stress 
 throughout the material. But a small deviation from 
 the true circle will cause the cylinder to collapse at a 
 comparatively small pressure. 
 
 The determination of the plate thickness by theoretical 
 formulas would, therefore, be very complicated, and 
 empirical formulas based upon reliable experiments 
 must be resorted to. 
 
 The late Sir William Fairbairn made a series of 
 experiments with tubes closed at both ends and exposed 
 to external pressure, which latter was increased until 
 the tube broke. He found that the bursting pressure 
 did not only depend upon the diameter, D, of the tube 
 and the plate thickness, S, but also upon the length, I, 
 of the tube. The results of his experiments are 
 expressed in the following formula : 
 
 p e = 806,300 x 
 
 s*-™ 
 
 UB 
 
 ■ (51) 
 
 where p 6 is the effective collapsing pressure in pounds 
 per square inch, I is given in feet, whereas D and S are 
 in inches. 
 
 Fairbairn's formula can be applied for finding the 
 plate thickness of an internal circular flue. In order to 
 obtain a suitable 8, it is necessary to divide the flue into 
 short cylinders, joined together by strengthening rings 
 or the like. I, in formula (51), will then be the distance 
 between two consecutive stiffening rings, and not the 
 total length of the flue. We have seen that the plate 
 thickness of a flue must be small, say, ^in., in order to 
 
STEAM BOILERS. 
 
 ill ' 
 
 prevent the plate from being burnt by the heat of the 
 gases. We must therefore consider S as given, and find 
 I from formula (51), using a safety factor of 10, I to 
 be taken as the distance between the strengthening 
 rings. 
 
 The flue rings are generally made of mild steel 
 plates, but may also be made of best wrought iron. 
 The plates are curved in the bending rolls to the 
 required circle, and the two ends are bent into contact 
 for the purpose of being joined. As it is necessary that 
 the ring should be as near as possible a perfect circle, 
 it is evident that the joints must be very uniform ; they 
 should therefore either be welded or butt-jointed. If 
 butt-joints are used, they should either be double-riveted 
 with one strap, or single-riveted with double straps. 
 The riveted joints, however, will be struck by the flame, 
 and therefore have a tendency to be overheated ; for 
 this reason welded joints are preferable. 
 
 It is of great importance that the joints connecting 
 consecutive rings of the flue should be designed so as 
 to allow the flue to expand and contract longitudinally, 
 this having a great bearing upon the strength and 
 durability of the boiler. The joints should further be 
 removed from the direct line of draught, so as to 
 escape the scouring action and intense heat of the 
 gases as theyJpass]along-the flue. 
 
 and at the same time stiffening the flue, F P. The 
 rivets and laps, however, are in the direct line of the 
 draught. 
 
 The flue plates, F P, may also be flanged at both 
 ends, and then riveted together with a caulking ring, 
 C E, between them, as shown in Fig- 76. It is 
 evident that the joint in this case is outside the direct 
 action of the flame. 
 
 E.R. 
 
 A great improvement in expansion boiler-flues is the 
 Fox's corrugated furnace, which has already been shown 
 in several boilers, Figs. 24, 32, and 36. It is made of 
 Siemens mild steel, manufactured by the Leeds Forge 
 Company, Limited, andjis]machine-rolled,'so that the 
 
 Fig. 74. 
 
 As examples of expansion flues, the following have 
 been selected : 
 
 In Fig. 74 is shown the Paxman expansion boiler- 
 flue. It consists of a series of short lengths, which are 
 made of very ductile soft steel, the tensile strength being 
 25 tons, with about 28 per cent, elongation ; or it may 
 be made of best wrought iron. The rings are welded, 
 and the ends are then heated in a furnace and enlarged 
 in a flanging machine, in such a way that the end of 
 one fits exactly within or without that of the next ; the 
 holes are then drilled in position through both, and 
 the flue is completed by riveting. This construction 
 allows the flue to expand freely, the rivets and the laps 
 are outside the direct line of the flame, and the strength 
 of the flue is considerably increased, by the diameter of 
 the joint being greater than that of the flue rings. The 
 boilers illustrated in Figs. 26 and 33 are provided with 
 this expansion flue. . 
 
 Fig. 75 shows section of an expansion-joint which is 
 often used, the A-shaped ring, E E, acting as a spring 
 
 The advan- 
 
 cross-sections are almost perfect circles, 
 tages of this furnace are : 
 
 (a) The increased strength by the many corrugations 
 allowing furnaces of varying diameters to be made of 
 
 FP 
 
 FP 
 
 Ki«. 76. 
 
 material fin. thick, and subjected to external pressure 
 up to 2001b. per square inch. The Board of Trade 
 allow the working pressure on corrugated furnaces to 
 
lia 
 
 Practical electrical engineering. 
 
 be equal to — (where D is the mean diameter 
 
 of the furnace in inches) provided that the plain 
 parts at the ends do not exceed 6in. in length, and that 
 the plates are not less than fV™. thick. If the furnace 
 is riveted in two or more lengths, see Fig. 77, the case 
 should he submitted for consideration. 
 
 ■ ton fusible plus (if koiiikd) 
 
 (b) An increased heating-surface of about 50 per cent., 
 in addition to which the corrugated surface assists in 
 breaking up the gases, thus absorbing more heat than 
 flues with smooth surfaces. 
 
 (c) It possesses a higher degree of elasticity than 
 the previously mentioned flues, and yet at the same 
 time it is strong enough to act as a longitudinal stay. 
 It is claimed that the corrugated furnace has the advan- 
 tage of throwing off the scale and sediment due to its 
 great elasticity. 
 
 (d) Uniformity of thickness, except where two or 
 more lengths are jointed, as shown in Fig. 77. 
 
 When required, a flat space, see Fig. 77, may be in- 
 serted for a fusible plug. "When cross tubes are re- 
 quired to be fitted into the flues, flat spaces are rolled 
 on each ring for this purpose. 
 
 Another corrugated furnace, made by the Farnley 
 Iron Company, Limited, is shown in Fig. 78. It will be 
 
 seen that the corrugation follows a spiral. The appli- 
 cation of this furnace to marine boilers is illustrated 
 in Fig. 79. 
 
 Manholes and Mudholes. 
 
 The manhole is made in the boiler shell for the pur- 
 pose of allowing a man to get inside the boiler, in order 
 to inspect it. The hole is made oval for two reasons • 
 partly m order to make the orifice as small as possible 
 
 and partly because the door could not be passed through 
 a circular hole. The longest axis of the hole varies 
 from 18in. to about 14in., and the shortest axis from 
 13in. to lOin. In horizontal land boilers, the hole is 
 made in the crown of the boiler barrel, the longest axis 
 being parallel to the axis of the boiler. In marine 
 boilers, the manhole is generally made in the end-plates, 
 and its longest axis is then horizontal. For making it 
 convenient to get inside a vertical boiler, the shortest 
 axis of the hole is parallel to the axis of the boiler, the 
 hole being made in the barrel. 
 
 Fig. 79. 
 
 As the shell is weakened by the large hole, it is 
 evident that the part of the boiler-plate surrounding the 
 hole must be strengthened. This may be done in two 
 ways. In land boilers working with low steam pressure, 
 a flat wrought-iron or mild steel ring, the compensating 
 ring, riveted on the boiler-plate, may be considered 
 sufficient. This ring is oval, and is of the same shape 
 as the hole. It is thicker than the boiler-plate, and is 
 about 3in. wide. The manhole is then closed by a cast- 
 iron door, which fits inside the boiler-plate, against 
 which it is tightened by one or two studs, the nuts of 
 which rest on arched crossbars. 
 
 In marine boilers, that part of the end-plate, where 
 the hole is made, is further strengthened by the aid of 
 longitudinal stays, as shown in Fig. 23. 
 
 When the working pressure of the boiler is high, the 
 flat compensating ring will not be sufficient to stiffen 
 the plate surrounding the orifice ; it is therefore sub- 
 
 Fio. 80. 
 
 stituted by a flanged saddle of mild steel, which is 
 riveted on the boiler-plate, either inside or outside the 
 barrel. 
 
 In Fig. 80, is shown an inside flanged saddle with 
 an embossed door, made by Mr. M'Neil, of Glasgow. 
 Both saddle and door are machined on the faces which 
 meet, so that a good joint can be more easily obtained. 
 With this arrangement of saddle and door, admission 
 
STEAM BOILERS. 
 
 113 
 
 to the boiler can be obtained much more rapidly than 
 where outside saddles are used, with doors fixed by a 
 large number of bolts, passing through the flanges of 
 door and saddle, as shown in Fig. 81. A saving of 
 time is also effected in using inside saddles when the 
 door has to be closed, only two nuts having to be 
 screwed up to make a joint. The pressure of the steam 
 assists in keeping the joint tight, whereas in the outside 
 saddle the whole strain is thrown on the bolts. 
 
 boiler. The passages through the arms are liable to 
 get choked by impurities and dirt contained in the 
 
 Fig. 81." 
 
 Mudholes and their doors are made in the same way 
 as manholes, but as they are smaller they have only one 
 stud and one crossbar. In Lancashire and marine boilers, 
 the mudholes in the end-plates underneath the internal 
 flues may be as large as manholes. Mudholes are 
 sometimes made circular and threaded ; they are then 
 closed by screwed plugs. The diameter of such holes 
 is about 2£in. Mudholes are placed in those parts of 
 the boiler shell, where the removal of sediments would 
 otherwise be difficult. 
 
 Water-Level Indicators. 
 
 The height of the water level in a boiler may be 
 indicated by means of either water-gauges or test- 
 cocks. 
 
 The water-gauge shown in Figs. 82 and 83 is invented 
 and made by Messrs. Dewrance and Co., of London. 
 It has two horizontal arms, A and B, each provided 
 with an asbestos packed cock, which will be seen in 
 section in Fig. 83, and each bolted to the end-plate of 
 the boiler by means of three bolts (Jin. diam.) through 
 the flange. The nipple fits into the hole made in the 
 boiler, and an asbestos washer being placed between 
 the flange and the boiler-plate. Arm A must be fixed 
 somewhat above the highest, and B somewhat below 
 the lowest, water level ; when, therefore, the two arm 
 cocks are open, the hole in A will be passed by steam, 
 and that through B by water. A strong glass tube is 
 inserted between the two arms, the joints being 
 made steam-tight by means of packed stuffing- 
 boxes. When now the two arm cocks are opened, the 
 water will stand in the glass at the same level as in the 
 
 m a 
 
 Fig. 82. 
 
 water, and the apparatus would in such cases 
 not show the correct water level ; it is there- 
 fore necessary to blow frequently through these 
 
 Fig. 83. 
 
 passages. If, for instance, we want to try the pas- 
 sage through A, then we shut arm cock B and open the 
 
114 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 blow-through cock ; a steam jet will then pass througn 
 A and the glass, and thus clear the passage. In the 
 same way we test passage through B by shutting arm 
 cock A, and opening the blow-through cock. It may, 
 however, happen that one or both passages are found to 
 be choked, and thus require to be cleaned out. Let, tor 
 instance, the passage through A be stopped, then pro- 
 ceed as follows : (a) shut both arm cocks and open 
 blow-through cock ; (b) remove the screwed plug opposite 
 the passage ; (c) open arm cock A gradually while 
 pushing a wire through ; (d) when the passage is clear, 
 shut arm cock A and put plug back ; (e) shut blow- 
 through cock and open both arm cocks. 
 
 The clearing of the passages under steam should, 
 however, be avoided, and will never be necessary if 
 done frequently before steam is up. 
 
 When the apparatus is in working order, the handles 
 of all three cocks should be in the position shown in 
 
 Fig- 82. . . . . 
 
 So far as described, this apparatus is m principle like 
 other water-gauges, but differs from these by having 
 a valve at the bottom and a valve-plug with a hole at the 
 top of the glass. Should the glass break in an ordinary 
 water-gauge, a steam jet will issue from the top opening 
 and a water jet from the bottom, and it is the water 
 jet especially which makes it difficult to shut the two 
 arm cocks. If the glass breaks in the apparatus here 
 described, the water will lift the small ball- valve, which 
 will shut the opening, and prevent the water from rush- 
 ing out, thus making it safe to shut the arm cock B. 
 On the other hand, the steam jet issuing through the 
 valve-plug at the top will be harmless, on account of 
 the small opening. The two valves can easily be 
 taken out under steam, and inspected, by shutting the 
 arm cocks, opening the blow-through cock, and remov- 
 ing the corresponding screwed plugs. A new glass can 
 be inserted, by opening the two stuffing-boxes. 
 
 Most boilers are provided with two water-gauges, 
 so that if the glass of the one breaks, the other gauge 
 will indicate the water level. 
 
 Test-cocks are cocks fixed at various heights on the 
 front of the boiler. One can find between which two cocks 
 
 The cock shown in Fig. 84 is fixed by tightening 
 the nut against the boiler-plate, an asbestos washer 
 being inserted between the outside of the plate and 
 the shoulder of the cock. 
 
 Fig. 84. 
 
 the water level must be, by opening the cocks in suc- 
 cession, and observing which give water and which give 
 steam. The actual height of the water cannot there- 
 fore be determined in this way ; but test-cocks having 
 no glass tube cannot so easily get out of order, and if 
 they get choked it is detected at once. The Board of 
 Trade require that each boiler shall have at least three 
 test-cocks. 
 
 Fig. 85. 
 
 Fig. 85 shows another kind of test-cock, which is 
 fixed by three bolts to the boiler-plate. 
 
 The cleaning of the passage through the cocks is 
 donje in precisely the same way as has been explained 
 with the water-gauge. 
 
 PFessure-Gauges. 
 
 The effective steam pressure can be measured by 
 means of a column of mercury, the weight of which 
 balances the pressure. The height of such column 
 would, however, be too great with the present pressures 
 at which steam boilers are worked. 
 
 A more handy instrument can be constructed by 
 letting the steam pressure act on a spring, which will 
 be bent more or less according to the intensity of the 
 pressure ; the movement of the spring is then multiplied 
 and indicates the pressure. The two following instru- 
 ments are made on this principle. 
 
 Fig. 86. 
 
 (1) Bourdon's gauge is shown in Figs. 86 and 87. The 
 spring is a curved metallic tube, which is closed at the 
 one end and fixed to a tap at the other end. The tube is 
 of the flattened form, and its greatest width is in the 
 direction perpendicular to the plane in which the tube 
 is curved. The steam pressure which is communicated 
 through the tap to the inside of the tube will tend to 
 bulge the tube, whereby this will stretch, The motion 
 
STEAM BOILERS. 
 
 115 
 
 of the closed end of the tube is applied to turn a 
 toothed sector, which again turns a pinion whose 
 spindle carries a hand ; the latter, pointing on a scale 
 properly graduated, will indicate the effective pressure 
 to which the tube has been subjected. 
 
 The gauges illustrated in Figs. 86 and 87 are made 
 by the Crosby Steam Gauge and Valve Company. It 
 will be seen that Fig. 87 works with two tubes. 
 
 pressures below the atmosphere, and are then called 
 vacuum-gauges. 
 
 Safety Apparatus. 
 
 This class of boiler fittings is designed with the 
 object of preventing — (a) the steam pressure from ex- 
 ceeding a certain limit, (i) the water from sinking 
 below a certain level, and (c) the flues from being 
 damaged should the water fall too low. 
 
 Safety-Valves. — An apparatus of this kind consists 
 simply of a loaded valve, which will be lifted when the 
 pressure on the valve is high enough to overcome the 
 load ; the latter may be produced either by weights or 
 by springs. 
 
 (a) The valve may be loaded direct by a weight, and 
 is then called a dead-weight safety-valve. A valve of this 
 class, made by Messrs. Galloway, Limited, is illustrated 
 
 Fig. 87. 
 
 (2) The Schaffer gauge is shown in Fig. 88 ; the 
 pressure acts upon a corrugated diaphragm of steel, 
 which will be bent more or less, according to the in- 
 tensity of the pressure. The motion of the diaphragm 
 is communicated to the hand, as in the Bourdon's 
 gauge. The object of the watch-spring fixed to the 
 hand-spindle is to keep the gearing parts in tension. 
 
 Fig. 88. 
 
 The calibration of pressure-gauges >hould be done 
 under steam pressure by comparison with a standard 
 gauge, the scale of which is graduated by comparison 
 with a column of mercury. 
 
 Fig. 89. 
 
 in Fig. 89. B is the boiler shell, N the neck on the 
 boiler, to which the apparatus is fixed, N x is a cast-iron 
 pipe, which carries the gunmetal valve-seat, V S, on 
 which the valve, V, rests ; this latter forms part of a 
 
 TberiTu^erma7alsobe constructed to ideate sphere and is also of gunmetal, The load is composed 
 
116 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 of several cast-iron weights, W. When the steam enclosed in a case and locked up, as shown in Fig. 90 ; 
 pressure is sufficiently high, the valve will be lifted the lifting lever is outside the case, 
 together with the weights, and steam will escape 
 
 between the valve and the valve-seat, and then by the 
 
 JL 
 ■JMB 
 
 ft 
 
 _ sir la 
 
 Fig. 90. 
 
 Fig. 91. 
 
 openings at the top into the atmosphere. Dead- (b) Safety-valves may also be loaded direct by springs, 
 weighted valves should always be provided with a lift- as illustrated in Fig. 91 ; the lifting lever is shown on the 
 
 Fig. 92. 
 
 trteirS;vIlve hiChtheWeiglltSCanbe lift6d ln ° rder ri S ht - hand side at th « ^ of the drawing. Such 
 
 -RW ti, mim „'„ t x- „ valves with their springs may also be enclosed in a 
 
 H or the purpose of preventing the engineer or stoker case. 
 
 Irom tampering with the valve, safety-valves are often When valves are loaded with springs, their construe 
 
STEAM BOILERS. 
 
 117 
 
 tion should be such that the valve cannot fly off if the 
 spring should break. 
 
 (c) Bamsbottom's safety-valve for loco, boilers, and 
 made by Messrs. Bobey and Co., is shown partly in 
 section in Fig. 92. It will be seen that this is a double 
 safety-valve; the valve V x is shown in section, the 
 other, V 2 , is hidden inside the neck, N 2 . "When the 
 steam pressure is high enough, the two valves will be 
 lifted simultaneously, and the spring, Sg, will thereby 
 be compressed. By means of the lever, L, the valves 
 
 safety-valve ; the valves are spherical, and are placed 
 on the top of the seats, like in Fig. 89. The weights 
 of lever and valve are balanced by counter-weights 
 placed on the opposite end of the lever. 
 
 Let W denote the load on the end of the lever, 
 whether produced by weight or spring, L the length of 
 lever upon which the load acts, I the length of lever 
 upon which the valve acts, d the distance between 
 centre of gravity of lever and fulcrum, w the weight of 
 the valve, and q the weight of the lever, then the work- 
 
 -e 
 
 c 
 
 VP 
 
 -e 
 
 c 
 
 mmmwmrmm 
 
 Jr 
 
 *& 
 
 irnnimrnnnnnnnn 
 
 AU1UWUUUUUUUU 
 
 Fig. 93. 
 
 can be tested, and should the driver accidentally or in- 
 tentionally put a weight on the lever, L, the result will 
 be to relieve valve V x from some of its load, and the 
 steam will blow off through V x at a lower pressure than 
 intended. 
 
 Fig. 93 shows valve-pin, V P, for V 2 , and Fig. 94 
 represents the spring-bolt, S B. 
 
 The working pressure, p e , at which direct loaded 
 valves will be lifted, is given as follows : Let W be the 
 
 Fig. 94. 
 
 ing pressure, p e , at which the valve will be lifted, will 
 be for balanced levers 
 
 WxL .... (53) 
 
 Pe - 
 
 and for non-balanced levers 
 
 axl 
 
 Pe = 
 
 "Wx~L + qxd + wxl 
 
 axi 
 
 (54) 
 
 (e) Belief-valves are safety-valves used for preventing 
 
 load in pounds, including weight of valve, a the area 
 in square inches on which the pressure acts, then 
 
 Pe = 
 
 W 
 
 (52) 
 
 (d) The load may act indirectly on the valve by 
 means of a lever, In Fig. 95 is shown a double-lever 
 
 the bursting of feed-pipes, should the check-valve stick, 
 or for preventing the covers of steam-cylinders from 
 being blown out, should the cylinder contain too much 
 water. They are usually loaded direct by springs. In 
 Figs. 96, 97, and 98 are illustrated three forms of relief- 
 valves. 
 
 (/) Hopkinson's Compound Safety- Valve. This ap- 
 
118 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 paratus is invented and manufactured by Messrs. J. 
 Hopkinson and Co., of Huddersfield, and is illustrated 
 in Fig. 99; it comprises a lever safety-valve with a 
 dead-weight safety-valve, and it also acts in the dis- 
 charge of steam, should the water become low in the 
 boiler. 
 
 The low-water feature of this valve consists in the 
 suspension of a double lever, L L 1 , inside the boiler • 
 on the end of L 1 is placed a float, which, when in its 
 normal condition, is below the surface of the water, at 
 such a height as is deemed low- water mark. This float 
 is made of a special material, which has been found 
 
 Fig. 96. 
 
 Fig. 97. 
 
 Fig. 98. 
 
 On the valve-seat is placed a large flat valve, V 1 , 
 loaded by means of a ball, B, and a lever, I. Through 
 the centre of this valve is an orifice, which forms the 
 seat for another valve, Y\ of spherical or ball construe- 
 
 to be suitable to withstand high-pressure steam, and 
 the various conditions to which floats are subjected in 
 high-pressure steam-boilers. 
 
 On the end of L is placed a counter-weight, C W, 
 
 Fig. 99. 
 
 tion, and weighted inside the boiler on the dead-weight 
 principle, by means of iron plate weights W 
 
 On the steam increasing beyond the pressure to 
 which the valves are loaded, each valve will rise from 
 ito seat and discharge the excess of steam 
 
 to balance the float, and through an eye, E, on the lever, 
 the rod passes which suspends the dead weight, and on 
 this rod is fixed an adjustable collar, Cr. 
 
 Should the water in the boiler become low enough to 
 leaye the float, the latter falls and turns the lever, which 
 
STEAM BOILERS. 
 
 119 
 
 will then be brought in contact with the collar, Cr, 
 thereby raising the spherical valve from its seat and 
 discharging the steam. 
 
 The valves can be examined by removing the cover, 
 C. D P is a drain-pipe. 
 
 The safety-valve is one of the most important boiler 
 fittings, and should therefore have the greatest con- 
 sideration from engineers. Its duty is to discharge 
 from the boiler sufficient mass of steam at excess of pres- 
 sure, thereby preventing the pressure from rising to a 
 dangerous height. The valve must therefore be so 
 large that when open, it shall allow at least so much 
 steam to escape in unit of time, as is produced in excess 
 in unit of time. The valve opening must therefore be 
 a function of the following quantities : 
 
 (a) The size of the grate-surface. The greater this 
 part of the boiler is, the more coal can be burnt in unit 
 of time, and therefore the more steam can be produced. 
 But we have seen that the amount of coal which can 
 be burnt per hour per square foot of grate-surface, may 
 vary from less than 101b. to more than 401b., all accord- 
 ing to the force of the draught. 
 
 (|8) The size of the heating-surface. The amount of 
 steam which can be produced in a boiler per square foot 
 of heating-surface has been shown to depend upon the 
 force of the draught, and also upon the ratio of the 
 heating-surface to the amount of fuel burnt per hour. 
 
 (y) The force of the draught is an important factor 
 in the determination of the size of the safety-valve. 
 
 {S) The size of the water-room of the boiler ought also 
 to be taken into account in calculating the valve area. 
 It has been shown on page 68, that in taking steam out 
 of the boiler, and thereby lowering the pressure, the 
 water in virtue of its higher temperature will give off 
 steam, until the temperature corresponds with the 
 pressure. For this reason, the time required for lower- 
 ing the pressure will be increased with the size of the 
 water-room. 
 
 (e) The size of the steam-room will also have some 
 influence upon the time required for lowering the 
 pressure, but this is of minor importance. 
 
 (f) The pressure of the steam. The relation between 
 the absolute pressure of dry saturated steam and the 
 corresponding specific volume of the steam, is approxi- 
 mately given in the formula 
 
 p a x vH = 475 
 
 (55) 
 
 The volume of lib. of steam will therefore be the 
 smaller, the greater p a is, and consequently for high 
 pressures, the valve area need not be so large as with 
 low pressures, everything else remaining the same. The 
 velocity at which the steam will escape into the atmo- 
 sphere will also be greater at high pressures. 
 
 (rj) The temperature of the feed-water has, of course, 
 also a great influence upon the amount of steam which 
 can be produced in unit of time. 
 
 In the Board of Trade rules, which only refer to 
 marine boilers, the area of safety-valves is taken as a 
 function of the steam pressure and the grate-surface 
 
 only. The following table is taken from the Board of 
 Trade regulations : 
 
 Boiler pressure in pounds per Area of safety-valve in square inches 
 square inch. per square foot of grate-surface. 
 
 50 0-576 
 
 60 0-500 
 
 70 0-441 
 
 80 0-394 
 
 90 0357 
 
 100 0-326 
 
 110 0-300 
 
 120 0-277 
 
 130 0-258 
 
 140 0-241 
 
 150 0-227 
 
 160 0-214 
 
 170 0-202 
 
 180 0-192 
 
 190 0-182 
 
 200 0-174 
 
 It is not stated for what draught, nor for what ratio 
 of heating-surface to grate-surface, the above table is 
 calculated; but as boilers with forced draught may 
 require valves considerably larger than those found by 
 the table, the design of the valves proposed for such 
 boilers, together with the estimated coal consumption 
 per square foot of grate-surface, should be submitted to 
 the Board for consideration. 
 
 The lift of the valve should be at least one-fourth of 
 the diameter of the valve ; the area of the cylindrical 
 opening between the lifted valve and the valve-seat 
 will therefore be at least equal to the valve area. 
 
 The standpipe and neck, carrying the steam from the 
 boiler to the valve, should be as short as possible, so 
 that the steam pressure on the valve should be the same 
 as that in the boiler. 
 
 If the boiler be placed indoors, the valve ought to be 
 surrounded by a case, as shown in Fig. 99, from which 
 a pipe, intended to carry off the waste steam, leads 
 into the open air. If the boiler be placed out of doors, 
 such case and waste-pipe are not necessary, the steam 
 being allowed to escape straight into the atmosphere. 
 The area of the waste-pipe must not be less than the 
 valve area. The valve case should be fitted with a 
 drain-pipe at its lowest part, for the escape of condensed 
 steam. 
 
 Direct loaded valves should be fitted with a lifting 
 gear, for the purpose of testing the valve. Lever safety- 
 valves may be tried by lifting the lever by hand ; the 
 lever should not be allowed to drop back, but should 
 be put back gently. 
 
 The valve, as well as the valve-seat, should be made 
 of gunmetal, and in the case of lever valves all pinholes 
 should be bushed with gunmetal, or else the pins should 
 be of gunmetal ; iron working on iron ought not to be 
 allowed. 
 When the valve is resting on the seat, the steam 
 
 j 2 
 
 pressure acts on an area equal to ^j 3 -. where d x is the 
 diameter of the valve-chest; but when the valve is 
 
120 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 lifted, then the area upon which the pressure acts is 
 
 ^^ w here d 2 is the diameter of the valve. 
 
 4 ' 
 
 Let now W denote the total load in pounds on the 
 valve, then the valve will be lifted when the steam pres- 
 sure is 
 
 (56) 
 
 4 
 
 and the valve will go back again, when the steam 
 pressure has fallen to 
 
 W 
 
 Pe"=- 
 
 ■d<? 
 
 (57) 
 
 where 
 
 W = the load on the springs in pounds. 
 D = the diameter of the spring, from centre to centre 
 
 of wire, in inches. 
 § = the diameter or side of square of the wire in 
 
 inches. 
 C = 8,000 for round steel. 
 C = 11,000 for square steel. 
 The spring must have a sufficient number of coils, so 
 as to allow of a compression of at least one-fourth of 
 the diameter of the valve under the load at which it 
 ought to open. 
 
 The spring loaded safety-valve should be tested 
 under full steam and full firing, with feed-water shut 
 off and stop-valve closed. 
 
 Fig. 100. 
 
 we have therefore 
 
 PL 
 P." 
 
 dl_ 
 
 d* 
 
 (58) 
 
 As the valve must have some bearing surface, it is 
 impossible to have p e ' equal to p e ". For the same 
 reason, the load on the valve should be determined at 
 the hydraulic pressure test, and not by calculation only. 
 The Board of Trade rule for calculating the size of 
 springs, used for loading safety-valves, is given by the 
 following formula : 
 
 j= V 
 
 WxD 
 
 (59) 
 
 The springs should be protected from the waste 
 steam and from the impurities issuing from the valves. 
 
 Low-Water Alarms. — These apparatus are designed 
 for the purpose of giving the stoker due warning, when 
 the water is at or near the lowest level at which it is 
 safe to work the boiler. 
 
 Fig. 100 illustrates a low-water alarm ; as long as 
 the flat cylindrical weight is under water, the small 
 valve near the top of the apparatus will keep shut, 
 and thus prevent steam from entering the whistle. If, 
 however, the water level becomes so low that a suffi- 
 ciently large part of the flat weight remains out of the 
 water, then the valve will open and the signal will be. 
 
STEAM BOILERS. 
 
 121 
 
 given by the steam rushing through the whistle. "We 
 will assume that the counter- weight is placed at such a 
 distance from the fulcrum of the lever, that the whole 
 apparatus is balanced when there is no steam pressure, 
 and when the flat weight is completely immersed in the 
 water. Let now V denote the volume, in cubic inches, 
 of that part of the flat weight which must be out of the 
 water, in order to open the valve against the effective 
 steam pressure, p e , L the length of the lever upon which 
 the flat weight acts, d the diameter of the valve, I the 
 length of tbe lever upon which the valve acts, all in 
 inches, and y the weight of one cubic inch of water, 
 then we must have 
 
 ^X]),xl = VxyxL . . (60) 
 
 pipe leads from the steam-room into the flue, but is 
 shut by a valve until the water falls to a dangerously 
 low level, when part of the flat weight will become 
 bared of water, and the valve will open and discharge 
 steam into the flue, thus reducing the pressure and 
 extinguishing the fire. 
 
 (b) The double-cone fusible plug is shown in section, 
 in Fig. 102. Socket P is screwed into the crown of the 
 flue shell, at or near the firebox ; the space between the 
 two cones, C and D, is filled with an alloy, which melts 
 at a temperature which is somewhat higher than that 
 of the steam. Should the furnace-plates become bared 
 of water, the alloy will melt, and steam will thus be 
 discharged into the flue. The fusible metal should be 
 renewed at intervals, not exceeding two years. This 
 
 Fig. 101. 
 
 The valve will therefore open, and the signal will be 
 given, when 
 
 V = 21-8 x d * *P°* 1 cubic inches . (61) 
 Li 
 
 The temperature of the water at which the apparatus 
 is balanced has not been taken into account, but as 
 it would be lower than that corresponding top e , the 
 valve will open a little before V has reached the value 
 in (61). p e ought to be taken greater than that at which 
 the safety-valve blows off, especially as the apparatus 
 cannot be tested while steam is up. The cock leading 
 to the whistle is locked up, in order to prevent the appa- 
 ratus from being tampered with. 
 
 Flue Protectors are designed with the object of pre- 
 venting flue shells from being destroyed by overheating, 
 should the water level fall too low. 
 
 (a) A flue protector is illustrated in Fig. 101. A 
 
 Fig. 102. 
 
 plug is recommended by several boiler insurance com- 
 panies. 
 
 Blow-Off Apparatus. 
 The Blow-Off Pipe.— The object of the blow-off pipe 
 
 is ; 
 
 (a) To empty the boiler when it is to be cleaned. 
 For this purpose the pipe should be fixed at the lowest 
 point of the boiler. 
 
 (b) To diminish the formation of scale. It is well 
 known that the impurities contained in the feed-water 
 will be left in the boiler after the water has been 
 evaporated. Now, as long as steam is produced in the 
 boiler, the circulation of the water will to a great extent 
 keep the sediments suspended in the water, but when 
 the evaporation has ceased, the sediments begin to settle 
 down, and if we then blow off some of the water, we 
 shall at the same time relieve the boiler of part of this 
 
122 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 injurious matter. In boilers with mud-drums, as the 
 Babcock and Wilcox boiler, a large amount of the sus- 
 pended matter will settle down during evaporation, and 
 blowing off may therefore be done at any time with 
 advantage. 
 
 The pipe 
 shutting off 
 valve is fixed 
 
 Fig. 103. 
 
 is provided with a cock or a valve for 
 the boiler water. The blow-off cock or 
 on the boiler, and it should be made with 
 
 through blow-off cocks being left open, and the boiler 
 emptying itself of its water without the man in charge 
 being aware of it. 
 
 Figs. 103, 104, 105, and 106 are views of Hopkinson's 
 parallel slide blow-off valve, with locking gland. The 
 main feature of tbis blow-off valve consists in the 
 form of sliding discs, S V, see Fig. 103, which, by 
 means of a spring, Sg, between them, are kept against 
 their seats when sliding, and so remove any grit or dirt 
 which may have accumulated on the seats, V S. The 
 valve is opened or shut by turning the pinion, P, with 
 
 Fig. 105. 
 
 the key, and thereby lifting or lowering the rack, E. 
 Flange F is nearest to the boiler, and the connection 
 with the blow-off pipe is made through flange F x . 
 
 Fig. 104 shows the key way in the locking gland, and 
 Figs. 105 and 106 show another form of the valve, the 
 " Elbow pattern." 
 
 The Scum-Pipe.— When the boiler water contains 
 impurities which are lighter than the water, it is neces- 
 sary to blow off from the water surface. For this 
 purpose, a long narrow trough, the scum-pipe, is inserted 
 into the boiler. It runs parallel to the surface of the 
 water, and is fixed at the one end to the front of the 
 
 a locking gland-i.e., the key or spanner used for open- 
 ing and shutting the cock, cannot be removed unless 
 the cock is properly shut. Many accidents have arisen 
 
 boiler, where it opens into a cock, the scum-cock. It 
 is placed at such a height, that the opening is just a 
 little below the water surface. When the scum-cock is 
 
STEAM BOILEKS. 
 
 123 
 
 opened, the impurities floating on the water will be 
 carried into the trough and through the scum-cock into 
 a drain-pipe outside the boiler, whence they run into a 
 sink or the like. 
 
 box ; the cock is fixed to the neck on the boiler by 
 bolts through flange F 1; and the nipple, n, fits into 
 the neckhole, an asbestos washer being placed between 
 Fj and the neck flange. The drain-pipe is fixed to the 
 
 Fig. 107. 
 
 Figs. 107 and 108 show sections of Messrs. Dewrance 
 and Co.'s scum-cock, which is the Manchester Steam 
 Users' Association standard pattern. H is the handle 
 for turning the cock, Gl is the gland of the stuffing- 
 
 cock by bolts through flange F 2 . Cover C is taken off 
 when the scum-pipe is required to be cleaned. The 
 cock is made of gunmetal throughout and is packed with 
 asbestos. The cock as shown in the drawings is shut. 
 
t&4 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 The Steam-Pipe and its Fittings. 
 
 The duty of the steam-pipe is to convey the steam 
 from the boiler to the engine. Its sectional area, as 
 well as that of any other opening through which the 
 steam has to pass, should be of such size that the 
 velocity of the steam never exceeds 100ft. per 
 second. The pipe should, if possible, have a fall 
 towards the boiler, so as to allow condensed steam 
 to run back to the boiler instead of to the engine. fn 
 order to diminish condensation, the pipe should be 
 covered with a non-conducting material, and when the 
 pipe is long, the expansion of the pipe due to heating 
 should be taken up by special joints, expansion-joints, 
 
 P 2 , with flanges, ¥ 1 and F 2 , by which they are bolted to 
 the steam-pipe. The expansion of the latter is taken up 
 
 Fig. 109. 
 
 Fig. 111. 
 
 to be inserted at proper intervals of length. Fig. 109 by allowing P 2 to slide inside P r The joint between 
 Bhows Hopkinson's corrugated copper expansion-joint; Pj and P 2 must be steam-tight, and for this purpose T 1 
 
 t 
 
 E 
 
 ■> 
 
 Fro. 110. 
 
 ^"rf^ 180 ^^^^^- t ° eVMy26fl - t0 is pr0vided With an asbestos P acked stuffing-box, of 
 ^^^L^.Vk^ U " W T7 leWBOf,II,eipM1 - Which ' G1 ' is the S land ; the bushes, b x and ft* are 
 From he S£ \ T ^T?™ "* C °- made of g™ ta1 ' and the latter » ^ to the gland 
 
 £fl»Taft£ri£? T i , lg - 10, " win be seen by four screws - s - The ob J ect of the two lon S bolts > of 
 
 thatthejomtconsxstsoftwoshortcast-jronpipes.P.and which the one can be seen in Fig. Ill, is to prevent 
 
STEAM BOILERS. 
 
 12S 
 
 the joint from separating, and to denote the proper 
 position of the parts. 
 
 Connected with the steam-pipe, are several fittings 
 and apparatus, of which the following are the most 
 important : 
 
 (1) The Stop-Valve or Steam-Valve. — This is used for 
 shutting and opening the steam-passage, and consists of 
 a valve, which is moved by turning a screwed spindle. 
 The valve is placed in such a manner that it shuts 
 against the steam pressure. The spindle works in a 
 nut, and moves steam-tight through a stuffing-box. 
 The steam may leave the valve in the same direction 
 
 stop-valve, in order to be able to shut off the steam 
 from the engine, without having to get up on the boiler, 
 (a) Fig. 112 is a section of the stop-valve belonging 
 to the Eobey loco, boiler, illustrated on page 75. The 
 screwed part of spindle, Sp, works in a gunmetal nut, 
 n, which is fixed to the casing. By turning the hand 
 wheel, H, the valve, V, will be lifted or lowered accord- 
 ing to the direction in which we turn, but the valve 
 itself will not turn, as will be seen by looking at the 
 drawing. The direction in which the steam moves 
 through the valve is indicated by the two arrows. The 
 valve, V, valve-seat, V S, stuffing-box gland, G-l, and 
 
 Fig. 112. 
 
 as it enters, and is then a straight-through valve ; or 
 the steam may leave in a direction at right angles to 
 that in which it enters, and the valve is then a junction- 
 valve. Valves with screwed spindles are also called 
 screw-down valves. 
 
 The stop-valve should not be fixed direct on the 
 boiler shell, but should be bolted to a standpipe riveted 
 to the boiler. The stop-valve should be placed on the 
 boiler where the steam is driest; if, therefore, the 
 boiler is provided with a dome, the valve should be 
 fixed on the top of the dome. 
 
 The engine end of the steam-pipe is also closed by a 
 
 Fig. 113. 
 
 the nut, 11, are all of gunmetal ; the rest, except bolts 
 and spindles, is of cast iron. By removing the cover, 
 the valve and valve-seat can be inspected and ground. 
 
 (6) The valve shown in Fig. 113 is constructed by 
 Mr. Bhodes, of London. The valve face is not of 
 metal, but consists of a number of washers of an asbes- 
 tos and indiarubber compound. By unscrewing the 
 cover, the valve can be taken off and the washers 
 renewed. 
 
 (c) Hopkinson's parallel slide-stop valve, Fig. 114, is 
 made on the same principle as the blow-off valve, Fig. 
 103. By turning the hand wheel, we turn the nut, 
 
136 
 
 PRACTICAL ELECTRICAL ENGINEERING-. 
 
 but the screw simply moves up and down. The spiral reasonably hard. The under side of the seating is 
 spring on the top of the nut, N, is for keeping the grooved, so as to make a tight joint without packing 
 gland up to its seat ; no packing is therefore required. The loose valve, Fig. 118, is removed from the spindle 
 
 by unscrewing the nut which holds it. The casing and 
 everything else are made of gunmetal. 
 
 ,.. . ■ .,■ 
 
 
 j'hifijiUm 
 
 Fig. 118. 
 
 Fig. 114. 
 
 Ss« 
 
 Fig. 116. 
 
 Fig. 115. 
 
 Fig. 117. 
 
 {d) Dewrance's renewable stop-valves, so called The casing and cover of the stop-valve shown in 
 because the valve-seats can be taken out and renewed Fig. 119 are made of cast iron, and are Zl^ she 
 without having to remove the casing from its fixed 
 position, thus saving time and expense. fflBl 
 
 To renew a valve such as is illustrated in Fig. 
 115, we must first unscrew the cover, and then 
 
 Fig. 121. 
 
 iDiiiiuiif 
 
 Fig. 120. 
 
 Fig. 119. 
 
 Fig. 122. 
 
 unscrew the seating, either bv mpans nf „ • , 
 
 spanner, which follot with each valve or bvTT " th l illustrated in K §- ™- To renew a valve, 
 
 piece of iron. A new seating, Fig 116 ' cL I T rem ° ve * he cover a " d taI <e out the asbestos packing 
 
STEAM BOILERS. 
 
 127 
 
 bottom with an asbestos washer. Caulk down firmly (2) Steam Check-Valve. — When two or more boilers 
 the asbestos round the top of the seating-cage, and in are connected to the same steam-way, so as to supply 
 bolting down the cover, care must be taken to keep it the same engine or engines, it is necessary, or at any 
 
 Fig. 123. 
 
 Fig. 124. 
 
 level, so that the spindle may be vertical. Fig. 122 
 shows the empty casing. 
 
 The junction-valve illustrated in Fig. 123, is also 
 
 rate advisable, to insert a check-valve in the branch 
 steam-pipe, leading from the boiler to the main steam- 
 way w By such arrangement, steam from any one boiler 
 
 SIDE VIEW 
 
 Fig. 125. 
 
 constructed on the " renewable " principle. With the 
 exception of the shape of the casing, every detail is 
 precisely the same as in Fig. 119. 
 
 cannot enter any other boiler, should the stop-valve of 
 the latter not be shut. 
 
 Fig. 124 is a section of Hopkinson's steam check- 
 
128 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 valve. A is a gunmetal valve closing upwards, and B is an 
 iron float submerged in mercury, contained in the vessel 
 C ; the mercury is prevented from evaporating, by the 
 condensed steam floating on the top of the mercury. 
 The top side of the valve, A, is the inlet side, and is 
 also the side to which the steam comes, on leaving the 
 boiler. The under side of the valve is in connection 
 with the main steam-pipe. 
 
 On account of the buoyancy of the float, the pressure 
 on the inlet side must be about ilb. per square inch in 
 excess of the pressure on the outlet side, in order to 
 open the valve. The combination of the iron float and 
 the mercury not only acts in 'closing the valve should 
 
 K 
 
 Fig. 126. 
 
 the pressure be in equilibrium, but it also acts as a dash- 
 pot, because the float, although free to rise and fall in 
 the mercury, can only move at a comparatively slow 
 rate, as the mercury requires time to pass through the 
 thin annular space between the float and the vessel, C. 
 This prevents all objectionable oscillation of the valve 
 by the varying flow of steam, and obviates violent 
 knocking against the valve-seat. 
 
 Fig. 125 shows the most suitable position of the 
 check-valve on the boiler; it is placed next to the 
 iunction-valve, in the line of the radial pipe conveying 
 the steam to the main steam-pipe. 
 
 A steam check-valve, will be useful in many cases, of 
 which the following may be noticed : 
 
 {a) It will isolate one boiler of a series, and if such 
 boiler be empty, it will prevent accidents which other- 
 wise might happen should someone mischievously or 
 accidentally open the stop-valve. Of course, the right 
 
 [Fig. 127. 
 
 thing to do in such a case, would be to lock up the stop- 
 valve before emptying the boiler. 
 
 (6) When the boiler is provided with a low-water 
 safety-valve, a low-water alarm, or a flue protector, the 
 check- valve will prevent steam out of the other boilers 
 from replenishing the steam discharged through the 
 above-named safety apparatus. 
 
 Fig. 128. 
 
 (3) Steam Separators. — It has already been mentioned 
 that steam supplied by ordinary boilers is probably never 
 perfectly dry, and it has also been discussed several 
 times in this book how boilers ought to be constructed 
 in order to diminish the amount of water suspended in 
 the steam. 
 
STEAM BOILERS. 
 
 129 
 
 Where economy of ground space is an important 
 item, boilers of small bulk, and therefore also with 
 small water-surface, have to be resorted to ; in such 
 cases, and especially where water-tube boilers are 
 used, the steam-pipe must be provided with an 
 apparatus for separating the steam from the water 
 which has been carried along with it. All steam 
 separators are constructed in such a way as to cause 
 the wet steam to move in curved paths, whereby the 
 heavier water seeking straight paths will separate from 
 the steam. 
 
 provided with a drain-pipe, which open& into the boiler 
 water below the low-water level. 
 
 (b) Kg. 126 shows section of Messrs. Hydes and 
 Wigfull's steam separator. It is fixed outside the boiler 
 in an upright position, the steam inlet being at the base. 
 The fixed spiral, Sp, gives to the wet steam a rapid 
 circular motion inside the tube, T, which is pierced by 
 a number of narrow slits, S ; through these slits the 
 water is eliminated and drawn out through the drain- 
 valve. The apparatus is covered with non-conducting 
 material. 
 
 SECTIONAL ELEVATION 
 
 INLET 
 
 '//////////// ///////////////A 
 
 Fig. 129. 
 
 Amongst the numerous steam separators now in 
 existence, the following will be taken : 
 
 (a) The anti-priming pipe, which is often used 
 even in boilers which supply fairly dry steam. 
 It is a horizontal pipe, the top part of which is 
 perforated, and it is placed inside the boiler, see 
 Figs. 11 and 13, Sd., and is connected at the top 
 to the stop-valve standpipe. The steam will thus 
 be obliged to move up and down through the 
 pipe before it can enter the stop-valve casing, and 
 the greater part of the suspended water will therefore 
 remain in the boiler. To prevent water from accumu- 
 lating in the anti-priming pipe, the latter may be 
 
 OUTLET 
 
 PLAN 
 
 SECTION THROUGH A.B. 
 
 Fig. 130. 
 
 (c) Fig. 127 illustrates Messrs. J. Musgrave and 
 Son's "Globe" water separator. As shown in Fig. 
 128, the separator is fixed in the interior of the boiler 
 directly under the standpipe to which it is con- 
 nected, so that all steam leaving the boiler must first 
 pass through the separator. The steam enters the 
 apparatus by four inlets at the top, and these are so 
 placed and arranged, that the steam enters at a tangent 
 to the body of the separator, and in rushing in, is, by 
 its own velocity, given a rapid rotary motion ; the par- 
 ticles of water are thrown to the sides of the outer 
 casing, while the steam passes up through the pipe at 
 the centre, The water trickles down the sides of the 
 
130 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 casing, and accumulating at the bottom, it opens the 
 valve below, see Fig. 127, and passes back into the 
 boiler. The valve works on a knife edge ; to render it 
 as sensitive as possible, it is almost balanced, and is 
 so arranged that a small quantity of water inside the 
 separator will open it and allow the water to escape 
 into the boiler, but no water or steam can pass from 
 the boiler through the valve, 
 
 (d) Sectional elevation and plan of Savill's steam 
 separator are shown in Figs. 129 and 180. It consists 
 of a square casing of cast iron with an inlet and an out- 
 let for the steam. In the casing are placed three 
 vertical and angular plates, which are perforated in 
 such a manner, that the holes of the first and the last 
 are in the lower part of the plates, whereas the holes 
 in the centre plate are all in the top part. The steam 
 must therefore fall and rise twice before it reaches the 
 outlet ; the water and dirt originally suspended in the 
 steam, will impinge against the plates and drain down 
 
 Fig. 131. 
 
 the serrated grooves. The water outlet at the bottom 
 of the casing must either be trapped or syphoned under 
 water, to prevent air from getting into the separator. 
 
 (4) Beducing Valves.— The steam from the same 
 boiler may be used for various purposes, for some of 
 which the boiler pressure is too high. For instance, 
 the boiler may supply steam to several engines not 
 working at the same pressure, or some of the steam 
 may be used for heating purposes. In such cases, the 
 intensity of the steam pressure must be reduced before 
 the steam enters the apparatus working at the lower pres- 
 sure. Fig. 131 shows a section of Dewrance's reducing 
 valve. It is fixed to the branch steam-pipe leading 
 from the main steam-pipe to the apparatus for which 
 the valve is intended. 
 
 The reduction of the steam pressure is effected by 
 throttling the steam while passing through the reduced 
 openings left between valves V and V" and their 
 respective seats. These two valves form one " equili 
 brium valve," so that the steam pressure can neither 
 
 open nor shut them. The steam of reduced pressure 
 acts upon piston, P, which is loaded by weight, B, and 
 lever, L. If, therefore, this pressure is still too high, 
 the piston will be lifted, and thereby also the equili- 
 brium valve, which will thus leave still smaller openings 
 for the passage of the steam. For preventing violent 
 oscillations of the valve, a dash-pot, D P, is added. The 
 valve casing is of cast iron, and the working parts 
 are made of gunmetal. 
 
 Apparatus for Supplying Feed- Water. 
 
 Pumps. — The most common apparatus used for 
 forcing the water into the boiler is a pressure-pump— 
 the feed-pump — which may be driven by the engine or 
 by a separate motor. The tank is generally placed at 
 a lower level than the pump, which must therefore 
 suck the water up in the one stroke, and deliver it to 
 the boiler in the next stroke. To the pump belong 
 two pipes, one between the tank and pump — the inlet- 
 pipe or suction-pipe ; and another between the pump 
 and the boiler — the delivery-pipe or feed-pipe. 
 
 In order to compensate for possible leakage, blowing 
 off, and other waste, the pump should be made so large 
 that it can supply the boiler with at least twice the 
 amount of water which is to be evaporated. For this 
 reason, it will be necessary to have means by which the 
 supply can be regulated, so as to keep the water at a 
 constant level in the boiler. This can be done in 
 various ways — viz. : 
 
 (a) The pump can be disconnected from the engine. 
 By this arrangement, however, the water level in the 
 boiler would not be constant. 
 
 (6) By having a cock or a valve on the inlet-pipe, 
 which can be opened more or less, and thus regulate 
 the amount of water which enters the pump. 
 
 (c) By an overflow-pipe from the pump cylinder to 
 the tank, provided with a regulating cock or valve, 
 whereby some of the water in the pump will return to 
 the tank. This arrangement will also show whether 
 the pump is acting or not. 
 
 (d) By an overflow-pipe from the feed-pipe to the 
 tank, also provided with a regulating cock or valve, 
 whereby some of the water, which otherwise would 
 have entered the boiler, will be returned to the tank. 
 
 (e) By a regulating valve placed close to the boiler, 
 and which may also act as a check-valve. A relief- 
 valve, placed between the regulating valve and the 
 delivery valve of the pump, is in this case absolutely 
 necessary. An overflow-pipe should lead the rejected 
 water from the relief-valve to the tank. 
 
 if) If the pump be worked by a separate motor, the 
 number of strokes per unit of time may be regulated 
 to suit the supply which is wanted. 
 
 In order to prevent the water from running out of 
 the boiler, in case the feed-pipe bursts or the pump- 
 valves get out of order, the feed-pipe must be provided 
 with a valve— the check-valve— close to the boiler, and 
 placed in such a position that it is kept shut by the 
 boiler pressure. The check-valve, however, can also 
 get out of order, and it may be necessary to. take it out 
 
STEAM BOILERS. 
 
 131 
 
 while the boiler is at work ; a cock should therefore be 
 placed between the valve and the boiler, so as to be able 
 to shut off the boiler water. 
 
 To prevent the feed-pipe from being burst in case 
 the check-valve sticks, a safety-valve — relief-valve — 
 should be placed on the feed-pipe, or between the suc- 
 tion-valve and the delivery- valve of the pump. 
 
 If the pump be single-acting — that is, it only delivers 
 water every other stroke — then an air-vessel of suffi- 
 cient capacity should be placed on the feed-pipe, usually 
 close to the pump. This air-vessel will act as a cushion, 
 and prevent shocks due to the stopping and starting of 
 the water in the pipe. 
 
 To depend upon one pump only should not be 
 allowed, especially where the water-room of the boiler is 
 
 water returns by an overflow-pipe, provided with a 
 regulating feed-valve, to the feed-water heater. 
 
 Tor the purpose of giving access to the valves, 
 screwed plugs are placed above each valve. The plug 
 above the inlet-valve is further provided with a relief- 
 valve, E V, which is held down by a spring as long as 
 the pressure of the water in the valve-box is not in 
 excess. Should, however, the delivery-valve or the 
 check-valve stick, E V will discharge a sufficient quan- 
 tity of water to relieve the pressure. The lifts of the 
 three valves are limited by projections, p, on the plugs. 
 
 Donkey-pumps are pumps worked by separate small 
 steam-engines, which may either receive steam from 
 the main boiler or from a separate small boiler, the 
 donkey-boiler. The steam-piston and the pump piston 
 
 Fig. 132. 
 
 small ; besides the engine feed-pump, there should be 
 either a hand pump, a donkey-pump, or an injector. 
 
 In Fig. 132 is illustrated a single-acting feed-pump, 
 made by Messrs. Garrett and Sons, for portable engines. 
 The pump piston, P P— the plunger— is a hollow 
 cylinder of gunmetal, attached by the connecting-rod, 
 C E, to an eccentric on the engine shaft. The pump 
 cylinder, P C, is of cast iron, and ends in a packed 
 stuffing-box, so as to make the pump watertight. I V, 
 D V, and C V are inlet-valve, delivery-valve, and 
 check-valve respectively, which are all made of gun- 
 metal, as well as their valve-seats, V S. The pump 
 and the valve-box are fixed on the boiler ; no feed-pipe 
 is therefore wanted, except a very short piece between 
 the check-valve and the boiler, The excess of feed- 
 
 are connected to one common piston-rod, and the pump 
 with its motor is usually fixed on the wall of the 
 
 building. 
 
 A single-acting donkey-pump with steam-engine, and 
 made by Mr. Mumford, of Colchester, is illustrated in 
 Figs. 133 and 134, the latter showing longitudinal 
 section through steam-cylinder, S C, slide-valve, SV, 
 plunger, P P, pump cylinder, P C, air-vessel, A V, etc. 
 It will be seen that the steam piston-rod, P E, is con- 
 nected to the plunger by means of a pin, p ; this pin 
 need not be strong, as it has only to carry the weight 
 of the plunger while this is lifted, the pressure on the 
 plunger during the downward stroke being taken up by 
 the conical end of the piston-rod, which fits into the 
 plunger. 
 
132 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Fig. 133, 
 
STEAM BOILERS. 
 
 133 
 
 For the purpose of giving the plunger a uniform B, which latter is case-hardened. In order to prevent 
 motion, a craukshaft, C S, with a flywheel, F W, is shocks when the hole into which pin B fits is worn, a 
 
 _£. 
 
 Fig. 134. 
 
 Z^o'ZSSittXSZL'Z tz^^t^^^tz 
 
134 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 shaped bolt, WB. IP is the inlet-pipe, DP the 
 delivery-pipe, I V the inlet-valve, and D V the 
 delivery-valve. The valves can be taken out by remov- 
 ing the covers, VC. S P is the steam-pipe, E P the 
 exhaust-pipe, and Ex the eccentric which moves the 
 slide-valve, S V. As the steam-chest is small, it would 
 be difficult to tool the face if the cover were made in 
 the usual way ; the cover is therefore made slanting, as 
 shown in Fig. 134. 
 
 A double-acting donkey-pump by the same maker is 
 illustrated in Fig. 135. The pump piston is here a 
 gunmetal ring, and the pump delivers water in both 
 strokes ; there are, therefore, two suction- valves and 
 two delivery-valves, which are not shown. 
 
 Steam, tSjoi.nitl.l_e> 
 
 Fig. 136. 
 
 The lettering is the same as in Figs. 133 and 134. 
 No further explanation is therefore wanted. 
 
 The Steam Injector.— This apparatus was invented 
 by Giffard, and is based on the principle that a steam 
 jet escaping from a boiler under pressure, has a greater 
 velocity than a water jet escaping from the same boiler 
 For the same mass, the steam jet will therefore contain 
 more kinetic energy than the water jet. The steam 
 jet may therefore be so applied, that by condensing, it 
 transfers its kinetic energy to a water jet, thereby giving 
 to the latter a velocity which is greater than that at 
 which the water would escape from the boiler. If this 
 be the case, and the water jet be directed against a 
 
 proper opening in the water-room of the boiler, the 
 water jet, in virtue of its high velocity, will overcome 
 the boiler pressure and will enter the boiler. 
 
 (1) In order to understand fully the action of this 
 instrument, we will proceed to explain the injector shown 
 in Figs. 136 and 137, and made by Messrs. Gresham and 
 Craven, of Manchester. 
 
 It will be observed that the injector has four open- 
 ings or branches in its external casing. At the top of 
 the apparatus is a spindle — the steam spindle — with a 
 hand wheel ; the spindle ends in a cone which fits 
 accurately into a nozzle — the steam nozzle — and by 
 turning the hand wheel, the cone will be worked up or 
 down inside the steam nozzle, and thus regulate the 
 
 SCeanviSpirulLe 
 
 Vo^BoUer 
 ifio. 137. 
 
 admission of steam. Below the steam nozzle is another 
 nozzle — the combining or condensing nozzle— which 
 can be worked up and down by means of a rack and 
 pinion, the latter being attached to a graduated hand 
 wheel. ^ The water supply is regulated by increasing or 
 decreasing the annular space between the two nozzles. 
 The third nozzle near the bottom of the apparatus is 
 called the delivery nozzle. The injector just described 
 is a " lifting " injector— i.e., it can lift the supply water 
 from a tank below. The necessary manipulation in 
 order to start a lifting injector is as follows : Open the 
 water regulator about half way, raise the steam spindle 
 slightly so as to allow a small quantity of steam to rush 
 
STEAM BOILERS, 
 
 135 
 
 FW 
 
 <L^ 
 
 4 
 
 _L_ 
 
 ■ ■ ^ ■ 
 
 10 
 
 12 Inch;. 
 
 Fig. 135. 
 
STEAM BOILERS. 
 
 137 
 
 through the apparatus, whereby the nozzles will be 
 cleared, and air be drawn from the water supply pipe, 
 thus inducing a partial vacuum and lifting the water ; 
 then open the steam spindle fully, and the water will be 
 forced through the apparatus into the boiler. Should 
 the water supply be too great to pass through the feed- 
 
 The injector can only supply the boiler with com- 
 paratively cold water, as -the supply water must be cold 
 enough to condense the steam. The warmer the supply 
 water is, the more of it is required for condensing the 
 steam, and the mass of the water jet may thereby in- 
 crease so much, that its velocity will be too small to 
 
 Fig. 138. 
 
 valve at the boiler, some of the hot water will escape 
 through the overflow. The water regulator must there- 
 fore be adjusted, until the overflow of water ceases. 
 
 In Pig. 138 is shown the fixing of a lifting injector 
 feeding two boilers. 
 
 Fig. 139 is an external view of a non-lifting injector. 
 With the exception of having no steam spindle, it is 
 
 !llllilHIIUlIII:T r 
 
 ZoJioiter 
 
 \ 
 
 Fig. 139. 
 
 constructed precisely in the same way as the lifting 
 injector. The fixing of the injector is shown in Fig. 
 140. It will be seen that the tank is above the 
 apparatus, which will therefore start by simply opening 
 the two cocks shown in the diagram. 
 
 overcome the boiler pressure. The temperature of the 
 water should not exceed 120 deg. to 130 deg. F. 
 
 A great objection to the lifting injector is its want of 
 self-starting power. In the event of the jet being 
 broken by a sudden jolt or vibration, the same cycle of 
 manipulation as described above must again be gone 
 through, and there is a danger, if the attendant should 
 not be at hand when the injector " knocks off," that 
 the steam will in the meantime so heat the injector 
 that it will not readily start again. 
 
 (2) The exhaust-steam injector is an invention of 
 Messrs. Davies and Metcalfe, and is manufactured by the 
 Patent Exhaust-Steam Injector Company, Limited, of 
 Manchester. The injector described above is worked 
 with live steam having the same pressure as that against 
 which the apparatus works, whereas the injector which 
 we are about to describe is worked by the exhaust 
 steam of non-condensing engines, the pressure of which 
 is only a little above that of the atmosphere. 
 
 Fig. 141 is a section of the exhaust-steam injector, 
 which has three nozzles like the injector, Fig. 136. 
 The steam nozzle, S, is only remarkable in being of 
 very large bore to suit exhaust steam ; down its centre 
 passes the fixed cone, K, which serves to direct the 
 steam. The combining nozzle, D, is split from the 
 smallest bore or throat, E, for more than half of its 
 length, so that the " flap " piece, N, works freely on a 
 hinge, and by its movement enlarges or contracts the 
 throat of the combining nozzle. The delivery nozzle, 
 F, is connected to the combining nozzle, and by means 
 of a screw, E, the two nozzles can be lifted or lowered 
 together, in order to adjust the opening for the water. 
 
 The action of the apparatus may be described as 
 follows : 
 When not working, the flap, N, hangs open, so that 
 
138 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 there is the largest possible area for the flow of steam 
 and water. When the supply water is turned on, and 
 allowed to flow by its own gravity (the supply-tank 
 always being higher than the injector) into the combin- 
 ing nozzle and out • of the overflow, a partial vacuum 
 
 steam and grease separator, such as that described and 
 shown in Figs. 129 and 130, page 129. 
 
 The following are instructions for fixing the exhaust 
 steam injector : 
 
 (a) Position. The exhaust injector should be fixed 
 
 a. rALKHEJh ft PON 
 
 Fig. 140. 
 
 will be formed in the steam nozzle, whereby the steam 
 will be induced to enter ; the inrushing steam will then 
 be condensed, and thereby increase the vacuum. Th& 
 result of this is that more steam and water are drawn 
 in, until they mingle in sufficient quantities, and the 
 vacuum formed is so good that the rush of steam and 
 water into the combining cone generates a jet of such 
 velocity that, accelerated by the concentration given 
 it by the cone, it overcomes the boiler pressure and 
 enters the boiler. "When this vacuum is obtained, the 
 flap, N, is held close into the solid sides of the combin- 
 ing nozzle, forming practically a solid nozzle, through 
 which the water is delivered to the boiler exactly as in 
 the Giffard injector. 
 
 The automatic re-starting feature obtained by the use 
 of the flap nozzle, is the distinguishing feature in tbe 
 action of the exhaust injector. This principle is also 
 applied to lifting live-steam injectors, in order to make 
 them self-re-starting; but it is an actual necessity for 
 exhaust-steam injectors, which have to be applied not 
 only to constant but to intermittent engines, such as 
 winding-engines, etc. 
 
 The exhaust-steam injector can only force up to 
 about 751b. per square inch, but with the aid of live 
 steam, they can be made to feed against locomotive 
 pressures. 
 
 As the temperature of the water jet from this ap- 
 paratus is raised by applying the heat contained in the 
 ■ exhaust steam, it is evident that it will cause a con- 
 siderable saving in fuel consumption. 
 
 It is of great importance that the exhaust steam 
 should be dry and contain no grease, as the water 
 suspended m the steam is inactive, and the grease 
 would be carried into the boiler with the water jet 
 this reason, the injector should be placed near 
 
 vertically, sufficient space being left to admit of the 
 withdrawal of the nozzle. It may, however, be fixed 
 
 the engme where the steam is driest, and the steam, 
 before entering the injector, should pass through a 
 
 horizontally if more convenient, but in that case the 
 " flap " portion of the removable nozzle must be turned 
 
STEAM BOILERS. 
 
 139 
 
 round, so as to be on the top. This is done by turning 
 round the regulating nut with a spanner, the word 
 " top " being stamped on the nut. 
 
 (b) Exhaust steam. The branch exhaust-pipe to the 
 injector should be taken from the main exhaust-pipe 
 as far as possible from the open end, and should be 
 connected to the side if the pipe is vertical, as shown 
 in Fig. 142, or to the top if the pipe is horizontal, in 
 order to prevent grease and water from getting into 
 the injector. The main exhaust-pipe should not be 
 
 internal flues. It starts at the front of the boiler and 
 at the one side. In loco, boilers, the feed-water enters 
 at the side of the boiler, see Fig. 17, page 75. The 
 feed- water should never enter the boiler at the bottom, 
 as this part of the boiler is the coldest, and the cir- 
 culation of the water would therefore be diminished. 
 The check-valve or back-pressure valve, which is 
 
 WATER SUPPLY 
 
 Fig. 142. 
 
 throttled, but be as large as possible. Great care 
 must be taken that all joints between the cylinder 
 and injector are perfectly airtight, "as air is inactive. 
 
 (c) Water. The injector must be fixed below the level 
 of the feed-water; and in cases where the blast is 
 heavy, the injector must be fixed as low as possible. 
 The temperature of supply water should not exceed 
 90 deg. F. 
 
 (d) Overflow. In every case, this pipe must be led 
 downwards7and its end must dip in water two or three 
 inches, to prevent the admission of air. 
 
 (e) Delivery. There must be the usual check-valve 
 on the boiler ; and if the delivery-pipe is long, an 
 additional check-valve must be fixed two or three feet 
 away from the injector, to diminish friction between 
 water jet and pipe. 
 
 (/) Joints. These must not be made with white or 
 red lead ; thin sheet rubber or asbestos is preferred. 
 
 Fig. 144. 
 
 The Feed-Pipe.— The duty of this pipe is, as already 
 mentioned, to convey the feed-water from the pump 
 or the injector to the boiler. It is usually continued 
 inside the boiler by a long horizontal pipe, of which the 
 latter part is perforated for the purpose of producing an 
 effectual distribution of the water mside the boiler. 
 The position of this pipe is somewhat above the 
 
 Fig. 143. 
 
 placed in the feed-pipe close to the boiler, is of similar 
 construction as the relief-valve, Fig. 98 ; it may, how- 
 ever, be worked with or without a spring. 
 
 Another fitting belonging to the feed-pipe, and which 
 has already been mentioned, is the feed-valve or feed- 
 regulating valve. It consists, like the stop-valve, of a 
 
 Fig 
 
 valve and a screwed spindle with a hand wheel, but the 
 valve does not lift with the spindle. By turning the 
 hand wheel, we only regulate the lift of the valve which, 
 therefore, can act as a check-valve when placed direct on 
 the boiler. In the latter case, no other check-valve is 
 required, bub in both cases a shut-off cock must be 
 fitted between the valve and the boiler, in order to be 
 
140 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 able to remove the valve while steam is up or water is 
 in the boiler. 
 
 When one pump or one injector supplies water to 
 more than one boiler, branch pipes are led from the 
 main feed-pipe to each boiler. The supply of water to 
 each boiler is regulated by placing a feed- valve in each 
 branch pipe, as shown in Figs. 138 and 140. 
 
 is for assisting in removing the box from the casing. 
 The Ferranti feed-valve is made by Messrs. Dewrance 
 and Co. 
 
 The feed-valve shown by Figs. 146a and 146b 
 resembles the " renewable " stop-valves described on 
 page 126, but the valve is not attached to the spindle ; 
 it is therefore free to act as a check-valve, the spindle 
 
 Fig. 146a; 
 
 Fig. 146b. 
 
 In Ferranti's feed-valve, as illustrated in Fig. 143, 
 the ball-valve rests on a seating contained in a wedge- 
 shaped box, which is made to a standard size, and is 
 therefore interchangeable, so that the valve can be 
 taken out and replaced by a spare one, Fig. 144, in a 
 
 iiiiimiiiijB iiiiii 
 
 Fig. 147. 
 
 few minutes. The regulation of the supply of water is 
 effected by the hand wheel and screw spindle. 
 
 Fig. 145 is an external view of the valve without 
 shut-off cock ; but when connected direct to the boiler, 
 a cock may be formed in one piece with the body of 
 the valve, as shown in the section, Fig. 143. The 
 object of the two screws at the front of the valve-box 
 
 merely regulates the lift of the valve. By removing 
 the flange on the side branch, the passage to the boiler 
 can be easily cleaned. 
 
 Hopkinson's combined cock and feed-valve is shown 
 in Figs. 147 and 148. It will be seen that the screwed 
 
 ■mm 
 
 Fig. 148. 
 
 spindle works inside the cock-plug ; the valve-seat is 
 integral with the elbow, as shown in section. The 
 elbow, and with it the seat and the valve, may be 
 easily removed for examination and repairs. The lift 
 of the valve is regulated in the usual manner by 
 turning the hand wheel at the top, and the shut-off 
 cock is closed by turning the horizontal handle. 
 
STEAM Bdl'LE'kS. 
 
 141 
 
 Heating of Feed- Water. 
 
 Before steam can be generated in the boiler, we must 
 raise the temperature of the feed-water to that of the 
 steam. It is therefore evident that the feed-water 
 should enter the boiler at as high a temperature as pos- 
 sible. Suppose the absolute pressure of the steam is 
 1151b. per square inch, then the temperature of the 
 water is 338 deg. F.; and let us further suppose that 
 the temperature of the feed-water is 60 deg. F., we have 
 then to spend for each pound of feed- water 
 
 338 - 60 = 278 heat units 
 
 before evaporation can begin. The total heat required 
 for the evaporation of lib. of water at 338 deg. F. 
 from 60 deg. F. is, according to formula (2), page 62, 
 1,157 heat units ; there is therefore a loss of 24 per cent, 
 of fuel, due to the heating of the feed-water. The 
 
 partly because most of the foreign matters originally 
 contained in the water will remain in the heater. 
 
 For the purpose of heating the feed-water, we can 
 either utilise the heat contained in the exhaust steam 
 or tbat contained in the chimney gases. In the first 
 case, the apparatus is called a feed-water heater simply, 
 and the temperature to which the water can be raised, 
 cannot possibly be more than 212 deg. F. , as this tem- 
 perature is the boiling point under atmospheric pres- 
 sure ; in the second case, the apparatus used is called 
 an economiser, and it may raise the temperature of the 
 feed-water to almost that of tbe high-pressure water. 
 
 An arrangement by which exhaust steam is mixed 
 with the feed-water is shown in Fig. 149. It is em- 
 ployed by Messrs. Garrett and Sons, of Leiston, for 
 heating the feed-water for their portable engines. The 
 one end of the apparatus is connected to the exhaust- 
 
 Fig. 149. 
 
 economy in fuel will therefore be considerable, if the 
 feed-water, before it enters the boiler, be made to absorb 
 heat which would otherwise be wasted. In the table 
 below is shown the saving in fuel, by heating the feed- 
 water before it enters the boiler ; the absolute pressure 
 of the steam is taken as 1151b. per square inch. 
 
 Feed- Water Heated prom 60° F. to 
 
 100° 
 
 3 5 
 
 120" 
 
 140' 
 
 5-2 
 
 160< 
 
 6-9 
 
 180° 
 
 8-6 
 
 200° 220° 
 
 10-4 
 
 12-1 
 
 240° 
 
 13 8 
 
 260° 
 
 280° 300° 320° 
 
 15-6 1V3 190 
 
 20-7 
 
 338° 
 
 22-5 
 
 24 
 
 Saving of fuel in per cent. 
 The economy in fuel is, however, not the only result 
 obtained from heating the feed-water; the life of the 
 boiler is also greatly increased, partly because the tem- 
 perature will be more equal throughout the boiler, and 
 
 pipe of the engine, and the other end to the feed-pump, 
 Fig 132 F V is the regulating feed-valve; when 
 it is open, or partly open, the excess of water 
 from the feed-pump will be forced through the 
 nozzle, N, and the overflow-pipe into a tank below, 
 which is not shown. Tbe water flowing through the 
 nozzle, N, will create a partial vacuum in the space sur- 
 rounding the nozzle, and if now the regulating valve, 
 EV, be opened, the steam from the exhaust-pipe will 
 be invited to enter, and will mix with and heat the 
 water which will run into the tank. The pump takes 
 the water from the tank through a separate pipe. In 
 regular work, the feed-valve may be so set as to provide 
 the boiler with a constant and regular supply of water 
 and to return the residue to the water heater. By this 
 means, the feed-water is heated without causing a back 
 pressure on the steam piston. 
 
142 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 The exhaust steam always contains greasy matter, 
 which it carries from the steam-cylinder. It is there- 
 fore better to heat the water without mixing it with the 
 steam. Such an apparatus would, however, be incon- 
 veniently large for portable engines. 
 
 The ends of these tubes are securely fixed in tube-plates, 
 T P, formed in the case, which, while keeping the ends 
 tight, allow the body of the tubes to expand. The 
 feed-water is pumped through the heater, entering at 
 the bottom, F I, and passing out at the top, P 0, to the 
 
 Fig. 150. 
 
 Fib. 151. 
 
 The " Compactum " feed-water heater, manufactured 
 by Messrs. John Kirkaldy, Limited, of London, is shown 
 m Pigs. 150 and 151. The apparatus consists of a cast- 
 iron casing, C, with a number of solid drawn brass S 
 
 boiler, being heated on its passage through the appa- 
 ratus, and depositing its scale and grease in the heater. 
 The former falling down clear of the tubes, can be blown 
 or scraped out from time to time. The following 
 
 tubpa ST n,™^ \,- u i.t. "—«-««" ^a,B» o- or scrapea out rrom time to time. The follow: 
 tubes, b 1, through which the exhaust steam passes, fittings are provided, and form part of the apparatus 
 
STEAM BOILERS. 
 
 143 
 
 One drain and sludge valve, D V, at the bottom, for 
 emptying the heater and removing sediments ; 
 
 One cleaning-door, C D, also at the bottom of the 
 apparatus ; 
 
 A drain-pipe, D P, from the tubes, to take away the 
 condensed steam, thus relieving the back pressure ; 
 
 A scum-valve, SV, near the top, for taking away 
 grease and dirt from the feed-water^ 
 
 Exhaust 
 
 Outixt 
 
 approximately equal to | of the total capacity of the feed 
 per hour. The water will thus remain for about 1\ 
 minutes in the heater, and the temperature of the feed 
 is claimed to be about 10 per cent, below that of the 
 steam. 
 
 Messrs. Eobey and Co.'s multitubular heater is 
 illustrated in Figs. 152 and 153. It consists of three 
 cast-iron chambers, Ci, C 2 , and C 3 — the first and the 
 last being connected by a number of vertical brass tubes. 
 The water enters at the top, and passes through the tubes 
 into C 3 , whence it runs out of the heater. The cham- 
 ber C 2 receives the exhaust steam, which, by striking 
 the tubes, heats the water. The drain-cock at the 
 bottom is for blowing off sediments given off by the 
 water. Fig. 153 is a horizontal section of the heater, 
 and is drawn to double the scale of Fig. 154. In this 
 
 Fig. 152. 
 
 One soda and air cock, S C, at the top, and another, 
 S C, just above the exhaust inlet, B I, for the purpose 
 of passing soda in a liquid state into either the water 
 side or the steam side of the heater, thus cleansing the 
 inside of the tubes and softening the scale on the 
 
 outside. , 
 
 A pressure relief-valve, E V, on the top, for the teed 
 
 to escape should the check-valve be closed. 
 
 The volume of the water-room of the heater is 
 
 Fig. 153. 
 
 heater, as well as in the previous one, the water moves 
 in the opposite direction to that of the steam; the 
 efficiency of the heater is thereby increased, as the 
 hottest steam will meet the hottest water, and the 
 coldest steam the coldest water. 
 
 Another heater, the "Nozzle Feed-Water Heater," 
 of the same firm, is illustrated in Figs. 154, 155, and 156. 
 Fig. 154 shows the arrangement of the nozzle, N, for 
 producing a circulation of the exhaust steam through 
 the circulating and heating tubes of the heater. It will 
 be seen that the exhaust steam has a straight course 
 through the nozzle and bottom chamber, whereby the 
 back pressure on the engine piston is not increased. 
 
 The tubes are all made of solid drawn brass or 
 copper ; they are only fixed at one end, and are free to 
 expand or contract independent of each other at the 
 other end, thus avoiding the leakage which occurs when 
 both ends of straight tubes are fixed. The fixed ends 
 
144 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 of the tubes are expanded into bored holes in the tube- 
 plates. By the arrangement of the brass tubes, the 
 circulating steam is brought into intimate contact with 
 the heating-surface. 
 
 The object of using brass tubes for heaters is this, 
 that the expansion of brass is different to that of the 
 scale, whereby the latter will not adhere to the tubes". 
 
 The heater casing is made of boiler-plates sufficiently 
 strong to resist the boiler pressure. Most of the im- 
 purities in the feed-water are deposited in the bottom 
 of the casing, and can be blown off through the blow-off 
 cock, D C. 
 
 The condensed steam is gathered in the pipe, C P, 
 at the bottom of the heater. 
 
 Fig. 155 shows horizontal section of the heater, and 
 Fig. 156 gives an external view of the lower parts of 
 same. 
 
 WI is the water inlet, WO water outlet, mh 
 mudhole, and AV an air-vessel. The arrows show 
 
 number of brass tubes 2Jin. diameter, and then out of 
 the heater. The water enters at the bottom, W I, and 
 leaves at the top, W 0. The heater consists of two 
 cast-iron chambers, C^ and C 2 , and a water-room, W E. 
 The end-plates of C x and C 2 serve as tube-plates ; the 
 shell bounding the water-room is made of boiler-plates, 
 which must be able to stand the boiler pressure. 
 
 The heater is provided with the following fittings— 
 viz. : 
 
 Two blow-off valves, B V, one for each mud-leg ; one 
 blow-off cock, B C, which may be kept open ; one drain- 
 cock, D C, for letting the condensed steam out of 
 chamber C 2 ; two mudholes, m h, on the top of the 
 heater ; one scum blow-off valve, S B V ; one safety- 
 valve, S ; E I is the exhaust inlet ; E is the exhaust 
 outlet. 
 
 Economisers. — These apparatus consist of a system 
 of parallel water-tubes, and are designed to absorb the 
 last part of the heat which can be allowed to be taken 
 
 SBV 
 
 Fig. 157. 
 
 how the steam and the water circulate through the 
 heater. 
 
 A 400 h.p. feed-water heater with horizontal tubes 
 is illustrated in Fig. 157, and is constructed by 
 Messrs. Musgrave and Sons, Limited, of Bolton. A 
 horizontal heater will, of course, require more ground 
 space than a vertical one, but the former will have the 
 advantage of having a larger water-surface, whereby the 
 solid matters suspended in the water will have a better 
 chance of settling down before the water leaves the 
 heater. In a vertical heater with small water-surface, 
 the current of water passing through the heater has a 
 greater velocity, and will thus keep the sediments 
 stirred up. 
 
 Fig. 157 shows longitudinal section of the heater, 
 and it will be seen that it rests on two legs— mud- 
 legs, M L— into which the sediments fall, and whence 
 they can be removed by blowing off. The exhaust 
 steam enters at the one end and passes through a 
 
 from the chimney gases ; they may therefore be con- 
 sidered as a continuation of the boiler, but their heating- 
 surface is more effective than that of the latter, because 
 the water they contain is colder than the boiler water. 
 Suppose the working pressure of the boiler is 1001b. 
 per square inch, then the temperature of the steam 
 will be about 338 deg. F. ; and let the draught be -55in. 
 of water, then, according to diagram on page 67, the 
 temperature of the chimney gases should be 300 deg. F., 
 if the chimney height is 100ft. The gases will leave 
 the boiler at a temperature higher than 338 deg. F., 
 but how much higher, depends upon the size of the 
 heating-surface of the boiler. The temperature of the 
 water in the economiser must therefore be less than 
 300 deg. F., but how much lower, depends upon the 
 heating-surface of the economiser. 
 
 The feed- water to the economiser should not be 
 cold, but the chill should be taken off ; it is therefore 
 best to employ one of the above-described feed- water 
 
STEAM BOILERS. 
 
 145 
 
 Fig. 154 
 
 Fig. 156. 
 
STEAM BOILERS. 
 
 147 
 
 heaters, from which the water is taken into the econo- 
 miser instead of into the boiler ; but when the engine 
 has a condenser, the water is taken from the latter. 
 The best-known economise! - is that manufactured by 
 
 the cold water, and to another similar pipe through 
 which the hot water is let out of the apparatus; 
 both tubes are outside the masonry. Each vertical 
 tube is provided at the top with a conical cover, 
 
 Fig. 158. 
 
 Fig. 159. 
 
 Messrs. E. Green and Son, Limited, of Manchester, of which is removed when the tube^ is to be cleaned 
 which Eig. 158 shows a longitudinal section, and Figs, inside. For the purpose of keeping the outside of 
 159 and 160 an end view and plan respectively. It the tubes clean, each tube is provided with a soot- 
 will be seen that this special size of economise!- has cleaner, S C, which is continuously moved up and 
 
 Fig. 160. 
 
 j. i ™fv, QT1 PT+prnal down the soot-cleaners of two consecutive rows of tubes 
 
 24 rows of vertical cast -iron tubes w.tt: »n « terna] £°™; chain The md down 
 
 * T'Vb. t'rowT of tabes" corrnecfed »t be motion of tbe'cleaners is effected by .be tormog of . 
 
 bouom to one too™ boritl"nicb receives boricont,. sb,ft, H S. proved wtb .orro geanog, as 
 
148 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 shown in Pig. 158. H S is driven by gearing it to pulley, 
 A, which again is driven by belt from the engine ' or 
 other motor. The reversing of the motion is done 
 automatically. 
 
 The draught should be regulated by a damper between 
 the economiser and the chimney, whereby the gases are 
 held back between the tubes. The boiler dampers 
 should only be closed when the boiler is not working. 
 
 The sediments must be blown out every day through 
 the blow-off valve, so as to prevent them from being 
 carried into the boiler. The scum gathered at the top 
 should be blown out through the safety-valve. 
 
 The soot chamber underneath the apparatus serves 
 to gather the soot, and must be cleaned out at proper 
 intervals, and the horizontal tubes must at the same 
 time be cleaned by means of a brush. 
 
 A thermometer is usually placed in the top branch 
 pipe, or in the feed-pipe in front of the boiler. 
 
 As a rule, the repairing of the economiser does not 
 necessitate the stoppage of the boilers, as a bye-flue 
 may be built between the boilers and the chimney, to 
 be used only on that occasion. 
 
 Incrustation.— Cleaning. 
 
 Pure water is a chemical combination of hydrogen 
 and oxygen, and contains no other element. As water, 
 however, has the property of dissolving and suspending 
 to a less or greater extent almost all substances, it 
 follows that the water in our rivers, canals, lakes, and 
 wells cannot be pure. Even rain-water and snow, 
 while descending through the atmosphere, dissolve 
 some of the ingredients contained therein, and are 
 consequently not pure water. 
 
 The amount of impurities in the atmosphere is 
 smallest on the ocean and . greatest in manufacturing 
 towns. 
 
 The atmosphere always contains nitric acid, 
 ammonia, and chloride of sodium ; the two first com- 
 binations are formed by lightning, and the latter 
 originates from the ocean, from which it is carried 
 along with the evaporated water. The atmosphere in 
 towns contains also sulphuric acid from the smoke, and 
 this forms sulphates with the original impurities con- 
 tained in the ocean air. 
 
 Bain-Water 
 
 London. 
 Hydrochloric acid 
 Nitric acid . 
 Sulphuric acid 
 Sulphate of soda . 
 Sulphate of ammonia 
 
 Bain-Water. 
 
 Manchester. 
 Hydrochloric acid 
 Nitric acid 
 Sulphuric acid 
 Sulphate of soda . 
 Sulphate of ammonia 
 
 Grains per gallon. 
 0-087 
 0-069 
 0-063 
 1-431 
 0-939 
 
 2-589 
 
 Grains per gallon. 
 
 . 0-408 , 
 
 . 0-094 
 
 . 0-260 
 
 . 3452 
 
 . 1-602 . 
 
 . 6/816 
 
 is 
 
 It is not only solid and fluid matter, however, which 
 dissolved by water, but gases are also greatly 
 absorbed by water. The mass of a gas which can be 
 absorbed depends upon the temperature and the 
 pressure, but the volume of the gas depends upon the 
 temperature only. The gases will, however, escape by 
 boiling the water. 
 
 The dissolving properties of water are very much 
 increased when the water is acidulated; thus, for 
 instance, carbonate of lime is not soluble in pure water, 
 but is soluble to a great extent in water which contains 
 carbonic acid. The latter, however, escapes by boiling, 
 like other gases, and the carbonate of lime will then be 
 deposited in the vessel containing the water. 
 
 Water, however, can only dissolve a certain amount 
 of a given substance ; when this limit is reached, the 
 solution is said to be saturated. The saturation point 
 varies with the temperature. 
 
 If a saturated solution be heated, and the water be 
 allowed to evaporate, the solution will become beyond 
 saturation, and some of the dissolved matter will be 
 deposited — generally as crystals. By pouring water 
 into the solution, the latter will become weaker 
 and the deposited matter will dissolve again. 
 
 Water as used for steam boilers is drawn from rivers, 
 canals, lakes, and wells, and therefore contains, besides 
 the impurities of rain-water, also soluble matters from 
 the soil through which the water has percolated. 
 Eiver and canal water also contains matter which is 
 characteristic of the towns and works passed by the 
 water. The water may also contain suspended matter, 
 but this will settle down by letting the water rest for 
 a time, or it can be removed by filtering. 
 
 Biver Thames. 
 At low water at Greenwich. 
 
 Carbonate of lime 
 Sulphate of lime 
 Chloride of magnesia 
 Sulphate of magnesia 
 Nitrate of magnesia 
 Chloride of sodium 
 Silica . 
 
 Oxide of iron 
 Organic matter . 
 Suspended matter 
 
 Grains per gallon. 
 
 9-99 
 
 11-82 
 
 6-87 
 
 3-60 
 
 4-30 
 
 80-09 
 
 0-99 
 
 1-50 
 
 4-07 
 
 176-05 
 
 Biver Thames. 
 
 299-28 
 
 At high water at Greenwich. 
 
 Carbonate of lime 
 Sulphate of lime 
 Chloride of magnesia , 
 Sulphate of magnesia. 
 Nitrate of magnesia 
 Chloride of sodium 
 Chloride of potassium 
 Silica . 
 
 Oxide of iron 
 Organic matter . 
 Suspended matter 
 
 Grains per gallon. 
 14-72 
 10-05 
 32-43 
 15-40 
 3-93 
 244-89 
 545 
 0-34 
 0-53 
 37-24 
 469 
 
 369-67 
 
STEAM BOILERS. 
 
 149 
 
 Sulphate of lime 
 Sulphate of magnesia 
 Sulphate of alumina . 
 Chloride of sodium . 
 Silica 
 Oxide of iron 
 
 Watford Canal. 
 Near a Paper Mill. Grains per gallon. 
 
 123-81 
 
 45-86 
 
 14-88 
 
 6-15 
 
 8-48 
 
 0-52 
 
 199-70 
 The sulphate of alumina is due to refuse matter from 
 the paper mill. 
 
 The chief impurities contained in water are the car- 
 bonate and sulphate of lime, and carbonate, sulphate, 
 nitrate, and chloride of magnesia. These impurities 
 when contained in boiler water, will be deposited in 
 the boiler and form a scale. Carbonate of lime is the 
 first to settle down, but its scale is porous and soft, and 
 therefore does not burn on to the plates. The sulphate 
 of lime forms a hard, non-porous scale, which also acts 
 as a cementing medium on the carbonate of lime. The 
 sulphate scale will therefore easily cause the plates to 
 be overheated, and will burn on to the plates. 
 
 Sulphate Scale. Per cent. 
 
 Sulphate of lime 86-85 
 
 Hydrate of magnesia . . . . 11 "66 
 
 Silica 0-47 
 
 Oxide of iron 009 
 
 Organic matter 042 
 
 Water 0-21 
 
 Carbonate Scale. 
 Carbonate of lime 
 Carbonate of magnesia 
 Sulphate of lime . 
 Hydrate of magnesia 
 Oxide of iron 
 Silica . 
 Organic matter . 
 
 100-00 
 
 The carbonate of lime in the latter scale is cemented 
 
 together by the hydrate of magnesia, whereby a very 
 
 dense, hard, and non-porous scale is formed, which' is 
 
 easily burnt on to the plates. The hardest of all scales 
 
 are those which contain hydrate of magnesia as a 
 
 cementing medium. Thus the above-named sulphate 
 
 scale is as hard as porcelain, and is produced by using 
 
 Thames water. The following scale is the burnt 
 
 part of the carbonate scale given above : 
 
 Burnt Scale. 
 
 Oxide of iron 
 
 Protoxide of iron 
 
 with the other sediments forms a scale, grease scale, 
 which is soft, but is perhaps the most dangerous of all 
 scales, as it prevents the water from coming into con- 
 tact with the heating-surface ; the plates will therefore 
 be over-heated, and perhaps become red-hot and give 
 way to the steam pressure. 
 
 Great care should also be taken in selecting the oil 
 to be used for lubricating the steam-cylinder, as many 
 oils are acid, or are converted into acids when heated 
 by the high-pressure steam. These acids will attack 
 the boiler-plates, and will form a corrosion scale with 
 the other impurities contained in the water. 
 
 Corrosion Scale produced by Oil. 
 
 Per cent. 
 Oxide of iron 
 Protoxide of iron 
 
 100-00 
 
 50-59 
 
 6-47 
 
 8-77 
 
 2633 
 
 242 
 
 4-26 
 
 1-16 
 
 Sulphate of lime 
 Magnesia . 
 Silica . 
 Organic matter 
 
 Per cent. 
 69-60 
 20-16 
 8-89 
 0-50 
 0-12 
 0-73 
 
 100-00 
 
 Carbonate of iron 
 Sulphate of lime . 
 Oxide of copper . 
 Magnesia . 
 Silica . 
 Oily organic acid 
 
 72-26 
 926 
 2-00 
 1-75 
 0-84 
 7-68 
 0-26 
 5-95 
 
 Boiler water which has been heated by mixing it with 
 exhaust steam, contains grease which has been earned 
 along with the steam from the cylinder, The grease 
 
 100-00 
 For the purpose of reducing the formation of scale, 
 or even avoiding it altogether, we must use means 
 which will (1) prevent the injurious impurities from 
 getting into the boiler ; (2) prevent the impurities from 
 settling down in the boiler ; (3) prevent the impurities 
 from forming an injurious scale. 
 
 (1) The suspended matter can be removed, as already 
 mentioned, by filtering before the water is pumped into 
 the boiler. 
 
 The carbonate of lime can be removed by adding 
 limewater , whereby the latter combines with the carbonic 
 acid contained in the water, and forms more carbonate 
 of lime, which, together with the carbonate originally 
 contained in the water, will be deposited in the tank. 
 This softening process, however, requires large and 
 expensive tanks, and we must not add more limewater 
 than just what is required for binding the carbonic acid 
 
 in the water. ... 
 
 By the use of a feed-water heater, some of the impuri- 
 ties, such as carbonate of lime, will separate from the 
 water and be deposited in the heater. 
 
 (2) We may prevent the impurities from settling 
 down in the boiler by blowing-off before the water 
 becomes a saturated solution. The deposit of carbonate 
 of lime could not be prevented in this way, as carbonate 
 of lime is not soluble in water that is free from carbonic 
 
 Sulphate of lime is only slightly soluble in water, and 
 we should therefore have to blow-off a great deal of 
 water, in order to prevent the sulphate scale being 
 formed. This remedy would therefore be expensive, 
 and is consequently not practical. 
 
 We must therefore add to the boiler water such 
 chemical ingredients as, with the slightly soluble 
 impurities, will form soluble combinations. Such 
 chemical ingredients can only be prescribed when we 
 have an analysis of the water. 
 
150 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 In waters which contain sulphate of lime, a suitable 
 amount of soda added to the boiler water will soften it 
 by forming sulphate of sodium, which is soluble in 
 water; at the same time carbonate of lime will be 
 formed, but the scale of the latter is soft. 
 
 Soda alone would be of no use if the boiler water 
 contained magnesia salts, for then hydrate of magnesia 
 would be formed, and the latter would act as a binding 
 cement on the carbonate of lime, and thus form a hard 
 scale. If more soda be added than is required for 
 decomposing the sulphate of lime, caustic soda will be 
 formed, which will attack the fittings, for zinc and 
 copper are soluble in a strong solution of soda. 
 
 By boiling the water with salammoniac, the car- 
 bonate of lime will be decomposed, and the result 
 will be the formation of carbonate of ammonia 
 and chloride of calcium, which are all soluble in 
 water. But the carbonate of ammonia is volatile, 
 and is carried by steam into the engine, where it 
 attacks the metal. Salammoniac should therefore not 
 be used, although it has often been recommended. 
 
 The following is an example of what may take place 
 if blowing-off is neglected, and the boiler water thus 
 allowed to become a strong solution of injurious 
 impurities. The impurities contained in the feed- 
 water were : 
 
 
 Grains per gallon. 
 
 Carbonate of lime 
 
 . 11-91 
 
 Sulphate of lime 
 
 512 
 
 Chloride of calcium 
 
 . 7-18 
 
 Chloride of magnesium 
 
 . 1-99 
 
 Chloride of sodium 
 
 . 14-40 
 
 Silica .... 
 
 1-22 
 
 Oxide of iron 
 
 . 0-88 
 
 42-70 
 The following is the analysis of the boiler water after 
 one month's run without blowing-off : 
 
 Grains per gallon. 
 762-72 
 47-30 
 
 Chloride of calcium 
 Chloride of magnesium 
 
 Sulphate of soda 
 Chloride of sodium 
 
 On account of the strong solution 
 following corrosion scale was formed, 
 hard and perfectly black : 
 
 Oxide of iron 
 
 Hydrated protoxide of iron 
 Carbonate of iron 
 Alumina 
 Silica . 
 
 Sulphate of lime . 
 Carbonate of lime 
 Hydrate of magnesia 
 Chloride of sodium 
 Chloride of magnesium 
 Organic matter . 
 Moisture 
 
 53-32 
 . 4,010-92 
 
 4,874-26 
 of chlorides, the 
 This scale was 
 
 Per cent. 
 
 32-58 
 
 18-90 
 
 16-56 
 
 7-02 
 
 7-34 
 
 5-93 
 
 4-43 
 
 4-93 
 
 0-98 
 
 0-19 
 
 0-31 
 
 0-83 
 
 (3) We can prevent the scale from being hard by 
 prescribing proper boiler compositions, which will 
 prevent the deposits from sticking together. 
 
 Treacle and sugar are sometimes used for this pur- 
 pose, and there are impurities which are more easily 
 dissolved in a sugar solution than in water. 
 
 Tallow is often used for the purpose of making the 
 scale soft, but it forms a grease scale, which prevents 
 the water from coming in contact with the plates. 
 Fatty acids are also formed, which attack the iron and 
 form a corrosion scale. 
 
 The following deposit and scale are due to tallow and 
 grease. 
 
 The greasy deposit is floury, and although it does not 
 bake hard, it shows the fatty acids are attacking the 
 plates. 
 
 Greasy Deposit. 
 Silica ... 
 Oxide of iron 
 
 Carbonate of lime 
 Hydrate of magnesia 
 Stearate of magnesia 
 
 Per cent. 
 19-04 
 6-60 
 39-62 
 21-74 
 12-95 
 
 99-95 
 
 The tallow scale below is a more aggravated form. It 
 is a tough scale, and shows a much larger percentage of 
 
 Corrosion Scale. 
 
 Prom Tallow. 
 Silica . 
 Oxide of iron 
 
 Carbonate of lime 
 Sulphate of lime . 
 Hydrate of magnesia 
 Stearate of magnesia 
 Organic matter . 
 
 Per cent. 
 11-06 
 19-50 
 4-81 
 263 
 1-06 
 59-91 
 1-03 
 
 100-00 
 
 100-00 
 
 To prevent corrosion of the iron, metallic zinc is 
 sometimes soldered on the boiler-plates, for the corrosive 
 acids will attack the zinc before the iron. 
 
 Chloride of lead is also used for preventing the scale 
 from sticking to the iron. By this process a thin layer 
 of lead is deposited on the plates, but not without the 
 chlorine combining with an equivalent amount of iron. 
 The process is really a corrosive one, and the layer of 
 lead is soft, and is easily removed in scaling the boiler. 
 
 The author has hereby endeavoured to show that 
 steam users cannot be too careful in choosing and 
 using the boiler compositions which are recommended 
 to them. And it must be well understood that because 
 a composition has been found to answer well in one 
 boiler, it does not follow that it will answer with water 
 from another source. To choose the right boiler 
 composition is like choosing the right medicine for a 
 patient. 
 
 The logical way of solving the question is to pre- 
 scribe the boiler composition according to the chemical 
 analysis of the water which is to be used. It is on 
 this principle that Messrs. Sidney Minns and Co., of 
 London, make their " Scientific Boiler Fluids," and if 
 this principle is followed out, the result must be satis- 
 factory. 
 
STEAM BOILERS. 
 
 For the purpose of introducing the boiler com- 
 position into the boiler, a small force pump may be 
 used, or, what is more convenient, an injector, as 
 shown in Fig. 161. 
 
 The latter consists of a chamber, C, which is closed 
 at the top by means of the screwed plug, P, and at the 
 bottom by a cork, of which H is the handle. The 
 apparatus is fixed on the top of the boiler, and is 
 manipulated in the following way : (a) Shut the cock by 
 turning H into the horizontal position, (6) unscrew plug 
 P by means of a pin through the hole until holes h are 
 level with the cup-shaped top, A, of the injector, (c) pom- 
 boiler fluid into A, from which it will run into chamber 
 
 151 
 
 Fie. 161. 
 
 C, (d) when the latter is filled, screw plug, P, back, 
 (e) and open the cock, the fluid will then run into the 
 boiler, being displaced by the inrushing steam. 
 
 Cleaning of Boiler. — This operation consists partly 
 in removing the scale and other sediments deposited 
 inside the boiler, and partly in removing ashes and 
 soot which have gathered in the flues. 
 
 The first operation necessitates the boiler being 
 emptied, so as to allow a man to get inside it. The 
 scaling is done by the use of the chipping hammer, 
 this being a tool in shape of a hammer with sharp 
 edges, but in places where it is impossible to operate 
 with the chipping hammer, scrapers of various forms 
 are used. Thus, for scraping a Galloway tube a bar of 
 
 sufficient length and bent at the one end where it is 
 sharp is used. A cup-shaped scraper is employed for 
 scaling the Field tube. *■ 
 
 The cleaning of flues is done by sweeping with 
 brushes of a convenient shape. The brushes may 
 either be made of wire or badger's hair. The brush 
 shown m Fig. 162 is used for cleaning flue-tubes, and is 
 mounted on a long cylindrical handle. 
 
 The small space left between the tubes in water- 
 tube boilers, as Figs. 38 and 52, makes it difficult to 
 clean the tubes by brushing. In such boilers the 
 cleaning is done by employing steam jets, as explained 
 before. 
 
 Boiler-cleaning is a very important operation and 
 ought to be done frequently, but as it only requires 
 common sense it is unnecessary to comment further 
 on it. 
 
 Testing.— Explosions. 
 Testing.— Fox the purpose of testing a boiler, it is 
 completely filled with water at ordinary temperature. 
 By means of a force pump, provided with a safety-valve 
 and a standard pressure-gauge, the water inside the 
 boiler is brought to a pressure higher than the working 
 pressure allowed on the boiler. The testing pressure, 
 
 Fro. 162. 
 
 according to the Board of Trade rules, should not exceed 
 twice the working pressure, the latter being determined 
 by the surveyor, by calculating the strength of the various 
 parts of the boiler. As a rule, the testing pressure is 
 taken from l - 5 to If of the working pressure. 
 
 The test should not take place until all fittings are 
 placed on the boiler. 
 
 The objects for testing a boiler are : 
 
 (a) To see that the boiler is Btrong enough, by exam- 
 ining whether the boiler is safe at a pressure higher than 
 the working pressure, that at such pressure the 
 volume and shape of the boiler is not altered, and that 
 no permanent set takes place in any part of the boiler. 
 
 (b) That at the testing pressure, the boiler does not 
 leak — i.e., that water nowhere escapes from the boiler 
 except in drops — it being taken for granted that such 
 leakage will not occur when the boiler is warm. 
 
 (c) To determine the load to be placed on the 
 safety-valves. The pressure-gauge may at the same 
 time be tested by comparison with the standard-gauge. 
 
 "When the working pressure has been reached by 
 means of the force pump, and the loads on the safety- 
 valves have been determined, a further load is put on 
 the latter, in order to be able to increase the pressure 
 to the testing pressure. Care should be taken to shut 
 the pressure-gauge before the pressure becomes greater 
 than the gauge can indicate, 
 
152 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Explosions.— The principal causes of boiler explo- 
 sions are : 
 
 (a) Bad design— i.e., the shape of the boiler is weak, 
 the stays are not sufficient and not strong enough, the 
 joints and plates are too weak, the internal flues and 
 manholes are not stiffened enough, or openings for 
 inspection and cleaning are placed in wrong positions, 
 thus making it difficult to examine the boiler. 
 
 (6) Bad material and workmanship, defective fittings, 
 and leaky joints. 
 
 (c) The cleaning of the boiler has been neglected, 
 whereby a thick scale has been formed, causing the 
 plates to be overheated, and thus making them too soft 
 to withstand the pressure. A large piece of scale may 
 also suddenly break off, when the flue-plate is red-hot, 
 and thus cause an explosion. 
 
 (d) By the stoker having allowed the water level to 
 fall below the surface of the flues, whereby these have 
 become red-hot and have yielded to the pressure. 
 
 (e) By the stoker having overloaded the safety-valve 
 and allowed the pressure to increase beyond a safe 
 limit. 
 
 if) By the steam suddenly being allowed to escape 
 through a large opening — such as would take place if 
 the steam-pipe bursts, whereby a sudden and violent 
 evaporation takes place. 
 
 ig) The boiler being too old and worn to withstand 
 the pressure. 
 
 In order to prevent explosions as much as possible 
 the following rules should be adhered to : 
 
 (1) Buy your boiler from a firm of high standing, so 
 as to be sure of good design, material, and workman- 
 ship. Never buy a second-hand boiler unless it has 
 been thoroughly surveyed by a good engineer, who 
 should also determine the working pressure to be 
 allowed. 
 
 (2) Take as stoker a reliable man, see that the boiler- 
 room is always clean and tidy, and do not allow anybody 
 in the boiler-room except on business. 
 
 (3) Clean the boiler thoroughly at regular intervals. 
 Complete cleansing is of special importance previous to 
 thorough examination by a boiler inspector, for other- 
 wise the inspection cannot be satisfactory and reliable. 
 
 (4) The feed-water should be analysed, and boiler 
 composition, if necessary, should be prescribed 
 accordingly. 
 
 (5) Feed-valves, check-valves, stop-valves, blow-off 
 valves, and all other fittings should have constant 
 attention, and be taken to pieces and cleaned at proper 
 intervals. 
 
 (6) Before firing-up, see that there is sufficient water 
 in the boiler, and that the blow-off cock or valve is 
 properly shut. Try the test-cocks and water-gauges, 
 
 and if they act slowly, clear out the passages with a 
 wire. See that the stop-valve is. shut, as it might 
 happen that the engine stop-valve were open. "When 
 ready to start the engine, open stop-valves slowly. 
 Clean the grate thoroughly. 
 
 (7) Try water-gauges and test-cocks frequently when 
 boiler is at work, in order to keep the passages clear. 
 Also try safety-valves and pressure-gauge ; if the latter 
 acts slowly, clean it as soon as possible. 
 
 (8) Begulate the feed-valve so as to keep the water 
 level at the fixed mark. The water must never fall 
 below that level. 
 
 (9) Use the various feeding apparatus (pumps or 
 injectors) alternately, in order to be sure that they are 
 in working order. 
 
 (10) Never leave the fire-door wide open when the 
 boiler is hot, or the rivet-joints may be injured. 
 Eegulate the fire by dampers. The fresh charge of fuel 
 should be put on the fire as quickly as possible, and the 
 fire-door should be shut at once. 
 
 (11) The stoker should remain in the boiler-room as 
 long as there is a fire. 
 
 (12) The surface-water should be blown out regu- 
 larly through the scum-cock, in order to prevent 
 priming. 
 
 (13) The safety-valve must be out of the stoker's 
 control — i.e., it must be constructed in such a way that 
 no extra load can be put on. 
 
 (14) Should the water level fall low, and should it be 
 impossible to restore it to the proper level, the fire 
 should be drawn at once, and if the water level falls out 
 of sight, the fire should be damped immediately by 
 throwing on ashes or other non-combustible matter and 
 closing the dampers, as the act of drawing the fire 
 might cause serious increase of heat. In the latter case 
 the feeding must be stopped, and the engine be run 
 slowly so as to take the pressure off gradually. 
 
 (15) All low- water safeguards must be thoroughly 
 tested and examined every time the boiler is cleaned. 
 
 (16) Should the pressure increase beyond the allowed 
 limit the fire should be damped. 
 
 (17) The boiler should never be emptied under 
 pressure, but should be allowed to cool down. The 
 practice of emptying under steam pressure and filling 
 up immediately with cold water often occasions serious 
 fractures in the boiler and in the masonry. 
 
 (18) If the feed-water is heated by mixing it with 
 exhaust steam, the latter should pass through a 
 separator before mixing with the water, in order to 
 prevent grease from getting into the boiler. 
 
 (19) The boiler should be examined frequently, and 
 tested after each repair. When getting worn, the 
 working pressure should be reduced. 
 
STEAM BOILEkS. 
 
 153 
 
 CHAPTER IX. 
 
 PKINCIPLES OF DYNAMICS. 
 
 Mass.— Force.— Acceleration. 
 
 lSS. By matter we understand every- 
 thing which makes an impression upon 
 us through our senses. Thus all we 
 smell, feel, taste, and see is matter, and 
 it is only through vibration of matter 
 that sound can be produced and trans- 
 mitted to our ear ; in a perfect vacuum there would be 
 no sound. 
 
 A bounded part of matter is called a body, and the 
 part of space which a body occupies is called its volume. 
 The amount of matter contained in a body is called the 
 mass of the body. In practice, the mass of a body is 
 measured by comparison with the mass of a standard 
 body. 
 
 In Great Britain the standard mass is called a 
 pound, and it is the mass of a certain piece of platinum 
 kept in the Standard Office for comparison. 
 
 In France the standard mass is a kilogramme, and 
 it was originally the mass of a cubic decimetre of pure 
 water at its highest density, but now it is the mass of a 
 piece of platinum, which is as nearly as possible equal 
 to the original kilogramme. 
 
 For reasons which will be explained later on, the 
 engineering unit of mass is neither a pound nor a kilo- 
 gramme, but is in Great Britain the mass of 32-187 
 pounds, and in France the mass of 9*8087 kilogrammes. 
 
 The determination of the mass of a body is done by 
 means of a pair of scales, whereby is found the number 
 of times that the mass of the body is greater than the 
 standard mass. This is the only way in which the 
 mass of a body can be found direct. For very accurate 
 measurements it may be necessary to take into account 
 the mass of the air which is displaced by the body, but 
 this is of no consequence to the engineer. 
 
 The measurement of mass by means of a pair of 
 scales and a standard mass can be made at any place in 
 the world. 
 
 Length.— The British standard length is a yard, this 
 being the distance at 62 deg. F. between two marks on 
 a bar kept in the Standard Office of the Board of Trade. 
 The engineering unit of length, being one-third of the 
 standard, is called a foot. 
 
 The French standard length is a metre, and is also 
 the distance between two marks on a certain bar at a 
 temperature of deg. C. The French engineering 
 unit of length is also a metre. 
 
 Time.— Time is a consequence of motion; if there 
 were no motion, there would be no time. 
 
 Unit of time is a second, which is svhn of the time 
 it takes the earth to turn once round on its axis. 
 
 Density. — By the density of a substance, is under- 
 stood the mass contained in unit of volume of the 
 substance. In Britain, density is expressed in pounds 
 per cubic feet, and in France in kilogrammes per cubic 
 metre. The density of a substance varies with the 
 temperature and the pressure to which it is subjected. 
 
 Specific Gravity. — By relative density or specific 
 gravity of a substance, is understood the ratio of the 
 density of the substance to that of pure water at a 
 standard temperature. This temperature is often taken 
 as 39"1 deg. F., this being the temperature of the maxi- 
 mum density of water. The relative density of a gas 
 is usually taken as the ratio of the density of the gas to 
 that of air at 32 deg. F., and at a pressure of one 
 atmosphere. 
 
 Specific Volume. — By specific volume or bulkiness of 
 a substance, is understood the number of units of 
 volume occupied by a portion of the substance, of 
 which the mass is equal to the standard mass. There- 
 fore in Britain it is measured by the number of cubic 
 feet to the pound, and in France by the number of 
 cubic metres to the kilogramme. 
 
 Force. — A body which is at rest — i.e., remaining at 
 the same place in space — cannot change its position in 
 space without a cause. Nor can a moving body change 
 the conditions of its motion without a cause. Such 
 cause is called a force. 
 
 The pull due to the mutual attraction of the earth 
 and a body is called the weight of the body, and the 
 force of this attraction is called the force of the weight 
 of the body. 
 
 The engineering unit of force is taken as the force 
 which can balance the weight of the standard mass at 
 the sea level at a certain latitude. 
 
 In Britain, the engineering unit of force is the force 
 which can balance the weight of the mass of lib. at the 
 sea level at Greenwich. This force is called the force 
 of a pound, or the pound weight. 
 
 In France, the engineering unit of force is the force 
 which can balance the weight of the mass of one kilo- 
 gramme at the sea level at Paris, and is called the force 
 of a kilogramme, or the kilogramme weight. 
 
 A force which acts upon a moving body in the direc- 
 tion of the motion of the body is called an effort, and a 
 force which acts in the opposite direction of the motion 
 of the body is called a resistance. 
 
 The measurement of a force is done by means of a 
 spring balance accurately calibrated in pound weights at 
 the sea level at Greenwich, or in kilogramme weights at 
 the sea level at Paris, as the case may be. This is the 
 only way the magnitude of a force can be measured 
 
154 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 direct, due correction to be allowed for the temperature 
 of the spring. This measurement can be made at any 
 part of the world. 
 
 Velocity. — A moving body which is not acted upon 
 by any force, or is acted upon by forces which are 
 mutually balanced, will pass through equal lengths of 
 space in equal times. Such motion is. called uniform 
 motion, and the ratio of length of space passed over to 
 the corresponding length of time is called the velocity 
 of the motion, or the rate of motion. 
 
 If the moving body be acted upon by one force, or by 
 several forces which are not mutually balanced, the 
 length of space passed over by the body in equal times 
 will vary, and we call such motion variable motion. 
 We may have accelerated motion or retarded motion, 
 according to whether the force is an effort or a 
 resistance. By the velocity of such a body at any 
 given time, we understand the velocity the body 
 Would have if the force or forces suddenly ceased to act, 
 as the motion then would be uniform. 
 
 The unit of velocity is in Britain 1ft. per second, 
 and in France 1 metre per second. 
 
 Acceleration. — It is evident that the magnitude of a 
 force can be measured by the result which it can pro- 
 duce — i.e., by the amount of motion which it can produce 
 in unit of time. Let, therefore, the magnitude of the 
 force be/, the mass of the body which it moves be m, 
 and the velocity produced after the lapse of the first 
 second be X, then we must have, when using the corre- 
 sponding units, / = to x \, where X is called the 
 acceleration. Unit of force will therefore, when acting 
 upon unit of mass, produce unit of acceleration. 
 
 As the force produces a velocity X per second in the 
 first second of the motion, the velocity at the lapse of 
 two seconds must be 2 X per second, and after the lapse 
 of n seconds the velocity of the body will be n X per 
 second, provided that the force is still acting on the 
 body. For this reason we may define acceleration as 
 rate of change of velocity. 
 
 Acceleration is measured in Britain in feet per 
 second, per second, and in France in metres per second, 
 per second. 
 
 As the acceleration of gravity at the sea level at 
 Greenwich is 32187ft. per second, per second, then 
 lib. weight acting upon lib. mass will produce an 
 acceleration of 32187ft. per second, per second, but 
 when lib. weight acts upon engineering unit of mass, 
 then the acceleration should be 1ft. per second, per 
 second ; consequently engineering unit of mass is the 
 mass of 321871b. mass. In French measurements, the 
 acceleration of gravity at the sea level at Paris is 9-8087 
 metres per second, per second, and as unit of accelera- 
 tion is 1 metre per second per second, we find that 
 engineering unit of mass in French measures is equal 
 to 9 - 8087 kilogramme mass. 
 
 As the acceleration of gravity varies with the distance 
 from the centre of the earth, and also with the latitude 
 it follows that the weight of a body will also vary' 
 according to its position relative to the earth. For this 
 reason, by weighing a body on a pair of scales, we do 
 
 not determine the weight but the mass of the body. 
 
 We can therefore only find the weight of the body by 
 
 means of a spring balance. Suppose we thus find the 
 
 weight to be P pound, and the acceleration of gravity is 
 
 g feet, then the mass of the body in pounds will be 
 
 32187 xP P 
 
 , and in engineering units - . We may there- 
 
 9 9 
 
 fore say that the mass of a body is equal to the weight 
 
 of the body in pounds divided by the acceleration in 
 
 feet. In French measures, the mass of the body is 
 
 equal to the weight of the body in kilogrammes divided 
 
 by acceleration in metres. 
 
 If, for instance, the safety-valve of a boiler is loaded 
 
 by weights, the valve will open at a smaller effective 
 
 pressure at the equator than at the pole ; but if the 
 
 valve is loaded by a spring, then the effective pressure 
 
 which is required to open the valve will be the same at 
 
 any place in the world. 
 
 Work.— Power — Energy. 
 
 Work. — By doing work, we understand to produce 
 motion against a resistance. The product of the resist- 
 ance in terms of unit of force, and the distance in 
 terms of unit of length, through which it has been 
 moved, is, therefore, a measure for the work done upon 
 the resistance. 
 
 Unit of work is, in Britain, 1 foot-pound, and in 
 France, 1 kilogram-metre. 
 
 Power. — It is not only necessary to know what work 
 a machine has to do, but we must also know within 
 what time it must perform this work. Suppose we 
 have to pump a certain quantity of water into a tank, 
 and we are allowed eight hours for doing the work ; 
 then an engine of a certain size will do. But if we 
 are only allowed four hours, we must have an engine of 
 double the size, or, as we say, of double the power. 
 
 By the power of an engine, we, therefore, under- 
 stand the rate at which it can do work — i.e., the quan- 
 tity of work which it can do in a certain interval of 
 time. The unit of power is, in Britain, 1 foot- 
 pound-second, and in France, 1 kilogram-metre- 
 second. But as these units are small, it would be 
 more convenient, in practice, to use a derived unit 
 of power. The usual derived unit is 1 horse-power, 
 this being equivalent to the average work performed 
 by a horse, and is, in Britain — 
 
 550 foot-pounds per second ; 
 or, 33,000 ,, ,, minute. 
 
 In order to fit in with the metrical system, the 
 French horse-power is — 
 
 75 kilogram-metres per seoond, = 
 542'5 foot-pounds per second. 
 Energy. — By energy we understand the oapacity of 
 bodies for performing work, and a body is said to 
 contain the more energy the more work it is able to do, 
 regardless of time. 
 
 The terms energy and power must not be considered 
 to mean the same. Power has relation to time. Thus, 
 one ton of coal can, by being burnt in a sufficiently 
 
STEAM BOILERS. 
 
 155 
 
 large boiler, produce steam for developing, say, 750 
 horse-power for one hour ; whereas, when being burnt 
 in a small boiler, it can produce steam for 50 horse- 
 power for 15 hours. The efficiencies of the two 
 boilers and engines are assumed to be the same. 
 
 Energy appears under various forms— thus, mechanical 
 energy, heat, electric energy, etc.— but they are all con- 
 vertible. Energy cannot be created or destroyed ; this 
 statement is expressed in the principle of conservation 
 of energy, and also in the impossibility of perpetual 
 motion. 
 
 According to its definition, mechanical energy must 
 be expressed in the same units as work done— i.e., that 
 the unit of mechanical energy is a foot-pound, and in 
 French measures a kilogram-metre. 
 
 Mechanical energy appears in two forms — viz : 
 
 (a) Potential energy or energy of position ; this is 
 the product of an effort into the distance through 
 which it is capable of acting (Eankine) . 
 
 (6) Kinetic energy or energy of motion, which is half 
 the product of the mass of the body into the square of 
 the velocity of the body. 
 
 These two kinds of energy are exemplified by the 
 simple cases of a weight raised to a height, and a 
 rotating flywheel, both of which contain capacity for 
 doing work ; the former in virtue of its position, and 
 the latter in virtue of its mass and velocity. 
 
 Thus, a body placed on or above the surface of the 
 earth would, if it were possible, fall towards the centre 
 of the earth, or rather towards the common centre of 
 gravity of the earth and the body, and stop there. 
 Therefore, the potential energy of a body placed at a 
 certain distance from the centre of the earth is equal to 
 the energy exerted by gravity during the fall of the 
 body. 
 
 If a body, weighing Q pounds, falls from a point at a 
 height, hu above the sea level to another point at a 
 height, Jh, also above the sea level, then it will lose in 
 potential energy the amount (Aj-^xQ foot-pounds. 
 
 If the body, in falling through the distance Oh -hi), 
 is doing no work — i.e., is not overcoming a resistance — 
 then the difference of potential energy before and after 
 the fall will be converted into kinetic energy, and we 
 may say : 
 
 Energy exerted = kinetic energy accumulated in the 
 body. Or in symbols — 
 
 ff 2 
 
 (62) 
 
 Where V is the velocity of the body when reaching 
 the lower point, and g is the acceleration of gravity ; 
 
 therefore -^ is the mass of the body. 
 9 
 If the body, while falling through distance {h^hi), is 
 raising another body weighing P pounds, say, through 
 the same distance, then it will be doing work to the 
 amount Px Qh-ht), but as Q>P, a part of the energy 
 exerted will be converted into kinetic energy, and we 
 may say 
 
 Energy exerted = work done + kinetic energy accumu- 
 lated in the moving bodies. Or in symbols— 
 
 Q (h 1 -h 2 )=F(J h -h 2 )+^±^: 
 
 1L 
 
 2 
 
 (63) 
 
 where v is the velocity of the two bodies after having 
 traversed distance Qh-h 2 ), and it is evident that 
 v <s V in formula (63). 
 
 Moment of Inertia.— -The motion of the body has 
 hitherto been assumed to be such that every molecule 
 is moving at the same velocity; but if a body is 
 rotating round a fixed axis like a flywheel, then we can 
 no more talk about the velocity of the body, because the 
 velocity of the molecules varies with their distance from 
 the axis of rotation, and the problem of calculating the 
 kinetic energy accumulated in the flywheel becomes 
 more complicated. 
 
 Let the molecules of the rotating body have distances, 
 x l ,x 2 ,x 3 . . . , from the axis of rotation, and have masses, 
 m u m 2 , to, ... , and let the body make n revolutions 
 per minute, then the velocity per second of the mass m^ 
 
 will be — x , and the kinetic energy accumulated 
 in the mass m^ will be i (-^) «i 2 wh, and the total 
 kinetic energy stored in the body will be 
 
 =KW Xmx2 ' 
 
 (64) 
 
 where Soto; 2 is called the moment of inertia of the 
 rotating body. 
 
 Imagine, now, that the total mass, M, of the rotating 
 body be concentrated into one point at a distance, 8, 
 from the axis of rotation, so as to have the same 
 moment of inertia as the rotating body itself, then we 
 must have 
 
 M.S 2 = 2mx 2 .... (65) 
 and 
 
 -V 
 
 2 to a; 2 
 
 (66) 
 
 <5 is called radius of gyration or radius of rotation. 
 
 Example. — A cast-iron flywheel is making 100 revo- 
 lutions per minute. The cros^section of the rim is 
 a circle with a radius equal to 3in., and the distance 
 between centre of circle and axis of rotation is 3ft. 
 What will be the kinetic energy in foot-pounds stored 
 in the flywheel rim? Take density of cast iron as 
 4501b. 
 
 The moment of inertia of the arms and boss is only 
 small compared with that of the rim, and may therefore 
 be neglected. Eadius of rotation of the rim is approxi- 
 mately equal to the distance between centre of gravity 
 of cross-section and axis of rotation ; we may therefore 
 take S = 3ft. The volume of the rim is 
 
 V = 7r(0 ' 5)2 x2ttx3 = 3-7 cubic feet. 
 4 
 
 The mass of the rim is, therefore, 450 x V pounds, or 
 
 expressed in engineering unit of mass 
 
 M = 
 
 450 xV 
 = 32-187 
 
 = 52. 
 
156 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 The moment of inertia of the rim is 
 
 I = Mx(5 2 = 52x9 = 468. 
 Kinetic energy stored in the rim = £ x f^— — ) x 468 
 
 = 25,670 foot-pounds. 
 
 Work by Tangential Forces. — A force which produces 
 or resists rotary motion round a fixed axis is called a 
 tangential force, because the straight line in which the 
 force acts is a tangent to a circle with centre in the 
 fixed axis, and whose radius is equal to the perpendi- 
 cular distance between the force and the axis. This 
 radius is called the arm of the force, and the product 
 of the force into the arm is called the statical moment 
 of the force. 
 
 If the tangential force be an effort, then the energy 
 exerted, while the rotating masses make one revolution 
 round the axis, will be 2 irr x E, where r is the arm of 
 the effort, E. 
 
 If the tangential force, E, be a resistance, then 
 2 x r x E will be the energy consumed by E during 
 one revolution of the masses round the axis. 
 
 As an example, we may take a horizontal shaft on 
 which two pulleys are fixed. The one pulley, the 
 "driver," receives energy from a prime mover by 
 means of a belt ; the other pulley, the " driven," drives 
 a dynamo machine, also by belt. Starting from rest, 
 the speed of rotation of the shaft will continue to 
 increase until a maximum — n revolutions per second 
 — is reached. Let t denote the time which elapses till 
 the speed of rotation has reached n revolutions per 
 second, then we must have : 
 
 Energy exerted on the " driver" during the interval 
 of time, t = work done on the " driven " during the 
 same period of time + kinetic energy accumulated in 
 the rotating masses. 
 
 The latter quantity is according to formula (64) 
 = i(27rw) 2 2ma; 2 . 
 In 2 ma; 2 is included all the rotating masses. 
 
 As the speed of rotation is to be kept constant at n 
 revolutions per second, it is evident that after that 
 speed has been reached, no more kinetic energy must 
 be added to the rotating masses, and consequently the 
 energy exerted must be spent in doing work only; 
 thus we must have 
 
 Energy exerted on the " driver " = work done on the 
 " driven." 
 
 Let E denote the tangential effort, E the resistance, 
 D the diameter of the "driver," and d that of the 
 " driven," then the energy exerted during one revolu- 
 tion, when the motion is uniform, will be j-DxE 
 and the work consumed by the resistance during the 
 same period will be ird x E. According to the above, 
 we shall have 
 
 wDxE = 7rixE, 
 or D x E = d x E, 
 
 that is " The condition of uniform motion of a rotating 
 axis is, that the statical moment of the effort must be 
 equal to the statical moment of the resistance, or the 
 effort must balance the resistance." 
 
 There are cases, such as in the steam-engine, where 
 the efforts vary in magnitude during each revolution. 
 In such cases we can consider that we have three 
 periods during a revolution — viz. : 
 
 First period, during which the statical moments of 
 the efforts are greater than those of the resistances ; we 
 may therefore say 
 
 Energy exerted = kinetic energy added to the moving 
 masses during this period + work done. 
 
 Second period, during which the statical moments 
 of the efforts are equal to those of the resistances ; we 
 may therefore say 
 
 Energy exerted = work done. 
 
 Third period, during which the statical moments of 
 the efforts are smaller than those of the resistances ; 
 we may therefore say 
 
 Energy exerted = work done - kinetic energy added 
 to the moving masses during the first period. 
 
 By adding all three periods we shall have 
 
 Energy exerted during a revolution = work done 
 during a revolution. 
 
 Work in Terms of Pressure and Volume. — The resist- 
 ance may be a pressure uniformly distributed on the 
 area of a piston, such as the back pressure of water in 
 a pump ; the total resistance in such a case will be the 
 intensity, E, of the pressure, multiplied by the area, A, 
 of the piston. Let now the distance traversed by the 
 piston be L, then the work consumed by the resistance 
 will be 
 
 ExAxL = ExV .... (67) 
 where V is the volume weight by the piston. If E is 
 given in pounds per square foot, A in square feet, and 
 L in feet, then (67) will be expressed in foot-pounds. 
 
 If the effort be a pressure of intensity, P, acting on a 
 piston, such as steam in a steam-cylinder, then the 
 energy exerted by the effort would be 
 
 PxAxL = PxV .... (68) 
 Both effort and resistance may be pressures acting 
 on the same piston, or there may be two pistons 
 connected by a common piston-rod, the effort acting 
 on the one, and the resistance on the other piston. As 
 an example of the latter case, we may take the donkey- 
 pump, Pig. 134, where the steam acts as an effort on 
 the piston, S P, while the pressure of the water acts as 
 a resistance on the pump-piston, P P. The water is 
 forced into the boiler during the downward stroke, and 
 the back pressure of the water is then greater than the 
 mean effective pressure on the steam-piston ; whereas 
 during the upward stroke the opposite is the case. For 
 this reason the crank-shaft, C S, with flywheel, F W, 
 are added, whereby sufficient kinetic energy can be 
 stored, during the upward stroke, to assist the steam 
 pressure in overcoming the resistances during the 
 downward stroke. 
 
 Work represented by an Area. — As work and energy 
 exerted are products of two quantities, force and length, 
 they can be represented by the area of a plane figure, 
 which is also the product of two quantities. 
 
PRINCIPLES OF DYNAMICS. 
 
 157 
 
 Let the abscissae in Fig. 163 represent distances 
 traversed by a constant effort, and let tbe latter be 
 represented by the ordinate o m, then it is evident that 
 the area omn~L will represent the energy exerted by the 
 effort in traversing through the distance represented 
 by oL. 
 
 If the effort is variable, then let distances traversed 
 and corresponding magnitudes of the effort be repre- 
 
 vn 
 
 Fig. 163. 
 
 sented respectively by abscissae and corresponding 
 ordinates in Fig. 164. Let, for instance, the distances 
 traversed be equal to d times length of abscissae, and 
 let magnitudes of effort be equal to b times length of 
 ordinates, then, when the variable effort has traversed 
 a distance equal to a x S, the value of the effort is 
 £ = b x y. Let now the effort move through a small 
 distance S = a x h, then the effort will be equal to E = b 
 (y + k). Let w denote the energy exerted by the variable 
 effort in traversing distance S, then we shall have 
 
 we please, by making h small ; consequently we must 
 have 
 
 w = a bxq. 
 
 Now the total energy exerted by the variable effort 
 in traversing through the distance represented by L 
 is W = 2 w, and the area, o toFBwLO, isQ = 2g; 
 therefore 
 
 W = a6xQ, 
 
 or, in words, area Q represents the energy exerted by 
 the variable effort. 
 
 Instead of being an effort, the force might have been 
 a resistance, and area Q would then represent the 
 energy consumed by the variable resistance, or the work 
 done by the effort on the variable resistance. 
 
 The mean ordinate of the curve, m F B n, Fig. 164, 
 is the height of the rectangle, whose base is L, and 
 whose area is equal to Q. The mean ordinate, there- 
 fore, represents the constant force, which by traversing 
 the distance represented by OL, would give the same 
 result as the forces represented by the ordinates of 
 
 E x S> w> i x S 
 
 or 
 
 <» + *) 
 
 h\ 
 
 w , 
 
 i 
 
 ab 
 
 (69) 
 
 Fig. 164. 
 
 Let, further, q denote the area of S F B D S, then we 
 shall have, according to Fig. 164, 
 
 {y +k)h>q> y x h. . ■ • ( 7 °) 
 
 (69) and (70) show that q and !L lie between the same 
 
 limits, the difference of which may be made as small as 
 
 
 m, 
 
 oLL- 
 
 *3I 
 
 *! 
 
 Fig. 165. 
 
 the curve, m'FBn, give, in moving through the same 
 distance. 
 
 This force is called the mean effort, the mean 
 pressure, or the mean resistance, as the case may be. 
 
 Determination of Areas of Plane Figures— -If the 
 problem be to determine the area enclosed by the curve, 
 I m, Fig. 165, the base line, L, and the two ordinates, 
 ol and wL, then divide the base line, OL, into a 
 convenient number of equal parts, h, and draw the 
 corresponding ordinates, y v y 2 , etc. The area is now 
 divided by these ordinates into eight parts, of which each 
 may be considered to approximate to a trapesoid. The 
 area of a trapesoid (which is a plane figure bounded by 
 a pair of parallel straight lines, and by a pair of straight 
 lines not parallel) is equal to half the sum of the two 
 parallel sides', multiplied by the perpendicular distance 
 between them. 
 
 The true area of Fig. 165 approximates, therefore, to 
 
 q = h p5-±l? + Vl + y 2 + y s +y i + y,+ y.+ tfrj 
 
 and the mean ordinate of the curve will approximately 
 
 be </ m =£ [^±^ + y l +y i +ys+y i + y^y^ + y^\ 
 
158 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 If we had divided L into n equal parts, we should 
 have approximately 
 
 Q= oj,n^ +2/i+2/2 
 
 n 
 
 Vm 
 
 nL 2 
 
 + 2/1 + 2/2 
 
 . +^-i] . (71) 
 . +y*-i] . (72) 
 
 (71) and (72) are the closer approximations, the greater 
 n is. In fact, we must divide the curve into so many 
 parts, that each part can, without appreciable error, be 
 considered a straight line. 
 
 Planimeters. — When the areas of a number of irregu- 
 lar figures are to be determined, considerable time and 
 labour may be saved by the use of a planimeter, which 
 is an instrument worked by moving a tracer over the 
 line enclosing the area to be measured ; the latter will 
 be registered on a small graduated wheel. 
 
 (1) Amsler's Polar Planimeter is illustrated in Fig. 
 166; it consists of a measuring wheel, C, and two 
 
 Fig. 166. 
 
 arms, of which the one carries the tracer and is called 
 the pole-arm ; the other arm — the radius-arm — carries 
 a needle-point, which is pressed into the board and 
 serves as a centre round which the apparatus is turned. 
 The tracing point is placed on the line of the figure, 
 and a slight indentation is made in the paper to serve as 
 a starting point. The graduated wheel, C, is set at the 
 zero mark. The tracer is then moved clockwise over the 
 boundary of the figure, making a complete circuit and 
 returning to the starting point. The number of divisions 
 and fractions of a division, shown on the wheel at the 
 point opposite the stationary zero mark, indicates the 
 area enclosed by the curve traced. The wheel has 10 
 main graduations, each of which represents 1 square inch. 
 Each main division is subdivided 10 times. A stationary 
 vernier, E, is placed beside the graduated wheel, 
 and serves to indicate the smaller fractions— viz., 
 hundredths. 
 
 That the integrating wheel really registers the area 
 
 enclosed by the path moved over by the tracer, can be 
 seen in the following way : In Fig. 167, is the centre, 
 round which the apparatus is turned, Fj is the radius- 
 rod, and a F x the pole-rod. Suppose we want to move 
 the tracer to point b, the rods would then have to be 
 
 Fio. 167. 
 
 moved to position O F 2 b ; in order to do this we will 
 first turn the pole-rod round point F t until it is in the 
 position Fj a x parallel to F 2 b, and then slide the pole- 
 rod parallel to itself until it reaches position F 2 b. The 
 
 Fig. 168. 
 
 area moved over by the pole-rod is therefore F x a a Y b F 3 Fi, 
 which consists of the sector ¥ 1 aa x Fj and the curvi- 
 linear parallelogram F 1 a 1 6F 2 F 1 . Let the length, of 
 the pole-rod be I, then the area of the sector will be 
 1 1 arc a a v 
 
PRINCIPLES OF DYNAMICS, 
 
 159 
 
 Let us assume that the graduated wheel is placed in 
 the middle of the pole-rod, so that a c 1 — c 1 J? lt then the 
 wheel will roll over arc c x c 2 , which is equal to J arc 
 aa\. The area of the sector will, therefore, be b x arc 
 Cj c 2 , and the graduated wheel will thus register the 
 area of the sector. The area of the curvilinear paral- 
 lelogram is equal to b x c 2 c 3 . The wheel will be moved 
 over the curve c 2 d, but this motion can be resolved into 
 dc 3 and e 2 c 3 . The former will cause the wheel to slide 
 only, and the latter motion will make the wheel turn. 
 We see, thus, that the area of the curvilinear parallelo- 
 gram is also registered by the wheel. 
 
 Let, now, the problem be to measure area, Z, which 
 is enclosed by the curve, a lt a 2 , a 3 , a 4 , a lt in Fig. 168 ; 
 
 therefore nought, and the wheel can consequently be 
 placed at any convenient distance from the tracer. 
 
 2. If we imagine the radius-rod, Fig. 167, infinitely 
 long, so that point F of the pole-rod moves in a straight 
 line instead of in a circle, then we have the Coffin plani- 
 meter illustrated in Fig. 169. 
 
 The one end of the bar of this instrument rests in a 
 groove made in a metal plate, which is fixed on the board 
 of the apparatus ; the other end of the bar carries the 
 tracer. In using the instrument, we must place the 
 card containing the area to be measured under the 
 clamps, C and K, which may be sprung away from the 
 board sufficiently to allow the card to be introduced, 
 and the card is then moved towards the left into such 
 a position that the base line of the curve is near to and 
 parallel with the lower edge of the stationary clamp, 
 C— i.e., the base line is placed at right angles to the 
 
 Fig. 169. 
 
 then place the instrument in position, F x a v and let 
 the tracer move from a x over the curve a v a 2 , a v the 
 area F, a, a 2 a 3 F 3 F i; call it B, will be registered 
 by the wheel : then let the tracer move over curve fl ? , 
 o 4 , a-i, and the wheel will measure Z+B. Area B is 
 thus described twice, but in opposite directions ; the 
 area registered by the wheel will therefore be Z + B ±J 
 = Z. We have thus proved that the graduated wheel, 
 when properly calibrated, will measure the area 
 enclosed by the path of the tracer. 
 
 It has hitherto been assumed that the rolling wheel 
 should be placed in the middle of the pole-rod but this 
 is not necessary, for the angular motion of the rod in 
 moving from position a, F a to a s F 3 is the same as when 
 moving from a, F 3 to a, F x , only in an opposite direc- 
 tion; the resultant angular motion of the pole-rod is 
 
 Fig. 170. 
 
 gr00V e_while the left-hand - end of the curve is made to 
 touch the other edge of the clamp. The movable clamp, 
 K, which can slide at the bottom in a groove, is then 
 moved towards the left until its edge touches the right- 
 hand end of the curve. The tracer is then moved to 
 the point D, at which the edge of clamp K touches 
 the curve. Here a slight indentation is made in the 
 paper by pressing the finger on the top of the tracer. 
 The graduated wheel is next turned, until its zero 
 mark is opposite the zero mark of the vernier. The 
 tracer is then moved carefully over the curve, in the 
 direction of the hand of a watch, until it has made a 
 complete circuit, returning to point D. 
 
 If it be required to determine the mean ordinate ot 
 the curve, then move the tracer— after it has been 
 moved over the curve-from D, along the edge of 
 
160 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 clamp K ; as this motion is made in an anti-clockwise 
 direction, the wheel will gradually turn round to zero 
 again ; at that moment another slight indentation is 
 made in the paper, in order to mark the new position 
 of the tracer. This point is represented by A in Fig. 
 169. The distance, D A, will then be equal to the 
 mean ordinate of the curve. That this is right, can 
 be seen in the following way. In Fig. 170, F x D is the 
 position of the bar after the tracer has been moved over 
 the curve. F 2 A, parallel to Fj D, is the position of the 
 bar when the graduated wheel has returned to zero. 
 The bar has thus described the parallelogram F X D AF 2 , 
 whose area must be equal to that enclosed by the curve ; 
 but the area of Fj D A F 2 is equal to the area of rect- 
 angle Lj D A L 2 ; and as D L x is the base line of the 
 curve, A D must be the mean ordinate. Cj C 2 is equal 
 to the length of the arc turned over by the graduated 
 wheel. 
 
 Useful Work and Waste Work. — The total work done 
 by the effort or efforts in moving a machine is partly 
 spent in overcoming useful resistances for the purpose 
 of which the machine is designed, and partly in over- 
 coming resistances within the machine. The total 
 work is therefore equal to the useful work plus the 
 waste work. 
 
 By the efficiency of a machine is understood the 
 ratio of the useful wqrk to the total work. The greater 
 this ratio is, the better is the machine for transmitting 
 energy, but it can never reach unity, which is the 
 efficiency of a perfect machine. 
 
 In practice, a machine usually drives another 
 machine, which again drives a third, etc. 
 
 The efficiency of any machine of such a series of 
 machines will be equal to the work done by the 
 machine on the next one, divided by the work done on 
 this machine by the one in front. Suppose, for instance, 
 we have a series of four machines having efficiencies 
 iv %, %, and >/ 4 respectively, and let w denote the 
 work consumed by No. 1 ; then the work done by the 
 latter will be j/ 1( w, and that done by No. 2 will be 
 %, m, w > therefore the work done by No. 4 will be 
 Vv m> iv l v w. The efficiency of the whole system of 
 the four machines will evidently be 
 
 *li< Is, %, m > w 
 w 
 
 = It, %> 12< Vl 
 
 (73) 
 
 or in words : The efficiency of a series of machines is 
 equal to the product of the efficiencies of the single 
 members of the series. 
 
 Work Consumed by Friction.— The greater part of 
 the energy lost in a machine is spent in overcoming 
 the force of friction, which resists the sliding of two 
 bodies on each other at their surface of contact— i.e., 
 their bearing surface. 
 
 The surface of a body is never perfectly smooth 
 although it may appear to be so, but, by examining it 
 with a magnifying glass, it will be found to be more or 
 less undulated. The friction is thus caused by moving 
 the sliding body over these undulations. The degree of 
 smoothness which can be given to the surface of a body 
 
 depends upon the material ; thus, the surface of metal 
 can be made smoother than that of leather, but by 
 using proper substances which will adhere to the 
 bodies, and which are soft enough to be squeezed into 
 and fill up the undulations, the smoothness of the bear- 
 ing surfaces can be greatly increased. Such substances 
 are called unguents or lubricants, and may be oil, 
 tallow, or soap and water, and for some materials even 
 water alone. 
 
 To overcome friction, therefore, consists in lifting 
 the sliding body over the undulations. The friction 
 between two sliding bodies must therefore be propor- 
 tional to the force, P, by which they are pressed 
 together, P being estimated at right angles to the 
 •bearing surfaces. If P is so small that it does not 
 cause any indentations in the surfaces of the bodies 
 but leaves them in their original condition, then the 
 friction must be independent of the area of the bearing 
 surface. In this case, the force of friction, F, will be 
 F - / x P (74) 
 
 where / is a factor called the coefficient of friction, 
 whose value depends on the state of the bearing 
 surfaces as to smoothness and lubrication. 
 
 If P is so great that it causes the bodies to grind into 
 each other, then / will increase with P. In construct- 
 ing machinery it is of great importance to make the 
 bearing surfaces large enough, so as to prevent P from 
 producing excessive friction and injuring the bearing 
 surfaces. 
 
 The work consumed by friction is evidently equal to 
 the force of friction, F, multiplied by the distance, d, 
 traversed by the sliding body. This energy is converted 
 into heat, to the amount of one British heat unit for 
 each 772 foot-pounds exerted in overcoming the 
 friction. The heat thus produced may be made useful 
 in softening the lubricant, as when tallow is used, but 
 when excessive it will decompose the lubricant and 
 soften the sliding bodies, whereby these will grind into 
 each other and increase the friction. To prevent the 
 temperature from rising beyond a certain limit, the 
 cooling surface of the sliding bodies must be so large 
 that at this limit of temperature, the heat'dissipated is 
 equal to the heat produced during unit of time. 
 
 Deviating Force.— Centrifugal Foree. 
 
 A body which is acted upon by no force, or by 
 balanced forces, will continue to move in a straight 
 line ; but if the moving body be continually acted upon 
 by an unbalanced force in the direction at right angles 
 to that of the motion; then the path of the body will be 
 a curve, and the force is called the " deviating force." 
 
 Thus, for example, if a train moves on a curve, the 
 deviating force is supplied by the rails being strong 
 enough to deviate the motion of the train from the 
 straight line. The train will, on account of its inertia, 
 produce a pressure on the rails, which is equal to the 
 deviating force. This pressure is commonly called 
 "centrifugal force," but as it disappears with the 
 deviating force, the centrifugal force is only a reaction, 
 and can produce no motion. 
 
HEAT. 
 
 161 
 
 If there were no friction or other resistance, a 
 passenger would be thrown out of the train in the 
 direction of the tangent to the curve, in virtue of his 
 inertia, but not, as is commonly understood, in virtue of 
 the centrifugal force ; the latter does not exist unless 
 there is a deviating force, and when it does exist, it will 
 be balanced by the deviating force. 
 
 The acceleration of the deviating force is 
 
 v 
 
 (75) 
 
 where v is the tangential velocity of the body, and r is 
 the radius of curvature. If v be given in feet per 
 
 Fig. 171. 
 
 acceleration of gravity at the place where the weight is 
 taken, then the deviating force would be 
 
 second and r in feet, then p would be expressed in feet 
 per second per second. Let to denote the mass of the 
 body in pounds, then the deviating force will be 
 
 TO 
 
 x — pounds 
 
 (76) 
 
 32.187 
 If w be the weight of the body in pounds, and g the 
 
 - x - pounds 
 9 r 
 
 (77) 
 
 In rotating parts of a machine, such as a flywheel, 
 the deviating force is supplied by the force of cohesion. 
 If at a certain speed of rotation, the required deviating 
 force is greater than the cohesion, then, at such speed, 
 the molecules will separate and the wheel will fly to 
 pieces. 
 
 It is of great importance in machinery that the 
 rotating masses should be so distributed round the axis 
 of rotation as to produce no pressure on the bearings ; 
 or, in other words, that the deviating forces acting on 
 the molecules should mutually balance. Fig. 171 
 represents a thin slice of a substance rotating round 
 axis AB ; let r lt r 2 , r 3 , etc., denote the distances of the 
 molecules from B A, and to^ to 2 , m 3 , etc., their respec- 
 tive masses, then the deviating force required for m x , 
 when the speed of rotation is n revolution per second, 
 will be 
 
 ^ZSLl£ mi ^(2^n) i m 1 r 1 
 
 (78) 
 
 n 
 
 and as the deviating forces of all the molecules are 
 parallel, the resultant deviating force will be 
 
 (27rra) 2 2mr .... (79) 
 
 (79) will be nought if 2 to r = 0, which requires that the 
 axis of rotation should pass through the centre of 
 gravity of the slice. 
 
 If the rotating body be of any shape, then we can 
 imagine the body being cut up in thin slices passing 
 through the axis of rotation, and consequently, the 
 condition for no pressure on the bearings due to rota- 
 tion will be that the axis of rotation shall pass 
 through the centre of gravity of the body. 
 
 CHAPTER X. 
 
 HEAT. 
 
 Temperature— Seales of Temperature. 
 
 | EAT is the cause of certain phenomena which 
 take place in material bodies. By means of 
 our sense of feeling, we term a body warm or 
 cold according to the condition of the body 
 brought on by heat. Heat has been proved 
 by Joule to be a special form of energy. The 
 thermal state of a body A, is higher than that of a 
 body B, if A is able to give off heat to B, this thermal 
 state is termed " temperature," A body is, therefore, 
 
 termed warm, when its temperature is higher than that 
 of our own body, in which case we receive heat ; the 
 body is cold, if its temperature is lower than that of our 
 body, as it then receives heat from us. 
 
 Temperature is measured, by means of the ther- 
 mometer, by observing the expansion of a fluid when 
 heated from a certain standard temperature to the 
 temperature to be measured. Let V denote the volume 
 of the fluid at the standard temperature, k Vthe expan- 
 sion of the fluid when heated up to another and higher 
 standard temperature, and h the quantity of heat which 
 
162 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 is required for producing expansion kV ; then the 
 difference of temperature produced by the quantity of 
 
 heat, -, where n is a pure number, is called a degree 
 
 n 
 of temperature, and if the fluidbeair — air thermometer — 
 
 then we shall find that the expansion produced by - 
 
 n 
 
 k V 
 will be — , consequently the latter will be a represen- 
 n 
 
 tation of a degree of temperature. 
 
 The lower standard temperature is usually taken as 
 that at which pure water freezes, and the higher 
 standard is the temperature of pure water when the 
 pressure of its vapour is one atmosphere, or, as it is 
 commonly called, the boiling point of pure water. 
 
 In the centigrade scale, n is taken as 100, and one 
 
 k V 
 centigrade is therefore represented by — -. In this 
 
 scale the temperatures are reckoned from the freezing- 
 point as zero-point. The temperature of the boiling- 
 point is therefore 100 deg. 
 
 British engineers use the Fahrenheit scale, in which 
 w is taken as 180; 1 deg. F. is therefore equal to f deg. 
 centigrade. The temperatures are reckoned from a 
 temperature 32 deg. below the freezing-point; the 
 boiling-point will therefore be 212 deg. 
 
 In practice, the thermometer fluid is mercury, but 
 
 by using the latter, — will at certain temperatures pro- 
 
 k V 
 duce an expansion greater than —-, and at others an 
 
 n 
 
 A mercury thermometer 
 
 k V 
 expansion less than — - . 
 n 
 
 should therefore be calibrated by comparison with an 
 air thermometer. 
 
 In measuring the temperature of a body, the thermo- 
 meter is brought into contact with the body until its 
 temperature is the same as that of the body ; hereby 
 the body will either give off heat to the thermometer 
 whereby its temperature will fall, or the body will 
 receive heat from the thermometer and its temperature 
 will rise. The error hereby incurred can be made 
 inappreciable by making the mass of the thermometer 
 small. 
 
 Heat Units. -Thermal Capacity. -Specific Heat. 
 
 Heat Units.— For the purpose of measuring the heat 
 required for raising the temperature of a body from one 
 temperature to another, we must .choose an arbitrary 
 quantity of heat, which we call a heat unit. 
 
 In Britain, the heat unit is the quantity of heat which 
 is required for raising the temperature of the mass of 
 lib. of pure water from 32deg. F to 33 deg. F. at a pres- 
 sure of one atmosphere. 
 
 In France, the heat unit is called a calorie major, and 
 is the quantity of heat required for raising the tempera- 
 ture of the mass of 1 kilogramme of pure water from 
 deg. C. to 1 deg. C, at the pressure of one atmosphere. 
 
 In the C.G.S. system, the heat unit is called a calorie 
 minor, and is the quantity of heat required for raising 
 
 the temperature of the mass of 1 gramme of pure water 
 from deg. C. to 1 deg. C. at the pressure of one atmo- 
 sphere — the calorie minor is therefore equal to O'OOl 
 calorie major. 
 
 Thermal Capacity. — Specific Heat. — The thermal 
 capacity of a body, of mass to, is the quantity of heat 
 required for raising the temperature of the body 1 deg., 
 and will be 
 
 S x to heat units 
 
 where S is a coefficient which depends upon the sub- 
 stance, and must be determined by experiment. The 
 heat required for raising the same mass of pure water 
 1 deg. of temperature from the ice-point, will be m 
 heat units. Consequently, S will be the ratio of the 
 heat required for raising a certain mass of a substance 
 through 1 deg. of temperature, to the heat required 
 for raising the temperature of the same mass of pure 
 water 1 deg. from the ice-point. S, which is a pure 
 number, is called the specific heat of the substance, but 
 as it is neither specific nor heat, the term is utterly 
 absurd, and S ought to be termed thermal coefficient. 
 
 The specific heat varies with the temperature of the 
 substance, but if we have found, for instance, in French 
 measures that the specific heat for a given substance is 
 S at a temperature of t deg. centigrade, then the 
 specific heat for the same substance, using English 
 measurements, will also be S at a temperature 
 of (I t + 32) deg. F. This, however, is only 
 true if the British heat unit is denned as above. In 
 some text-books it will be found that the British heat 
 unit is denned as the quantity of heat required for 
 raising the temperature of one pound of pure water 
 from 50 deg. F. to 51 deg. F. ; in other books the tem- 
 peratures are 39 deg. F. arid 40 deg. F. In either case 
 the specific heat in the British system will not be the 
 same as in the French system at the same temperature, 
 because the specific heat of water at 50 deg. F. or 
 39 deg. F. is not the same as at 32 deg. F. 
 
 The heat required for raising the temperature of a 
 body, of mass to, from ^ deg. to t 2 deg., will be 
 
 S,„ x to x (£ 2 - tj heat units. . . (80) 
 
 where S is the average specific heat of the substance 
 between the above two temperatures. 
 
 Transfer of Heat. 
 
 The transfer of heat from one body, A, at a higher 
 temperature, to another body, B, at a lower tempera- 
 ture, takes place through the surfaces of the bodies. 
 The surface through which heat is given off or received 
 is called cooling-surface or heating-surface respectively. 
 The temperature of the cooling-surface will be denoted 
 by t Ci and that of the heating-surface by th . 
 
 The transfer of heat can take place by three 
 processes— viz., "conduction," "radiation," and 
 " convection." 
 
 (1) Conduction.— When the two bodies, A and B, of 
 given shapes are in contact with each other, then the 
 transfer of heat from A to B will be done by conduction 
 through their surfa.ce of contact, which will be cooling- 
 
HEAT. 
 
 163 
 
 surface with regard to A, and heating-surface with 
 regard to B. The quantity of heat transferred at any 
 moment will be proportional to {t c - t 1 ), and to the 
 area of the surface of contact, where t c and tf 1 *, refer to 
 A and B respectively. 
 
 As A loses heat, t c will fall at a rate which will be 
 directly proportional to (t c -t l h ), and inversely to the 
 thermal capacity and conductivity for heat of A. 
 
 This will go on until t e has fallen so much, and 
 th has risen so much, that t c = fi h ; i.e., when the two 
 bodies are at the same temperature. 
 
 As B receives heat, t l h will rise at any moment, at a 
 rate which will be inversely proportional to the thermal 
 capacity aud conductivity of B, and directly propor- 
 tional to ( t - Ph). 
 
 Let A be entirely surrounded by B, whose thickness, 
 estimated at right angles to the surface of contact, is 8; 
 and let C denote the conductivity of B, C„ and C 1 , the 
 cooling surfaces of A and B respectively, and t e the 
 temperature of the cooling-surface of B ; then the rate 
 at which heat is conducted through B at any moment 
 will be — 
 
 H = 
 
 C X Ct + C1 * x ( t e - t e ) 
 
 8 
 
 (81) 
 
 For this reason, steam-boilers, steam-pipes, etc., 
 should be covered with non-conducting material of 
 
 such a thickness that - is small enough to prevent a 
 o 
 
 superfluous loss of heat. 
 
 For a cylinder of diameter d, length I, and covered 
 
 with a mantle of thickness 8, the rate of heat lost will 
 
 be from the cylindrical surface — 
 
 H= • 
 
 (82) 
 
 and if the ends of the cylinder are closed and also 
 covered, then the rate of heat lost from each end will be 
 
 n Ct(P (fc - t e ) 
 
 H = 48 • 
 
 (8b) 
 
 Formulae (82) and (83) can be applied to boilers, steam- 
 pipes, steam-cylinders, etc. 
 
 For steam boilers set in brickwork, formulas (82) and 
 (83) are to be used ; but for boilers which are not set in 
 brickwork, 8 will be small compared with d, and formula 
 (82) may then be reduced to 
 
 tt _ C IT d I (t e -t e ) (g4) 
 
 8 
 
 In steam-cylinders 8 would also be small compared 
 with d, and formula (84) will, therefore, also hold good 
 in this case. 
 
 For steam-pipes formula (82) must be used. 
 
 The above formulas show that if the heat lost is to be 
 the same for cylinders of various diameters, 8 must be 
 taken in proportion to the diameter of the cylinder, and 
 that for the ends, 8 must be increased with the square 
 of the diameter. Thus, if we are satisfied with 8 equal 
 to lin, for a steam-pipe of 2in. diameter, then for a 
 
 steam-pipe of 12in. diameter, we must take 8 = 6in., 
 using the same non-conducting material in both cases. 
 
 It would, however, be inconvenient in practice to 
 increase 8 with d. Suppose we choose 8, so that the 
 ratio of the heat lost to the quantity of steam carried 
 by the pipe in unit of time, is the same for all sizes of 
 pipes. 
 
 Then for the same velocity and temperature of the 
 steam and for the same non-conducting material, we 
 should have the heat lost and the quantity of steam 
 
 passing through the pipe proportional to and d 2 
 
 8 
 
 respectively, and the ratio of these two quantities 
 
 should be constant for all pipes, or 
 
 d + 8 
 
 Sd 1 
 
 = k 
 
 (85) 
 
 where k is a constant. 
 
 If we are, as above, satisfied with lin. covering for 
 a 2in. pipe, then k will be equal to - 75. According to 
 formula (83) we have 
 
 d 
 
 8=- 
 
 kd 2 -l 
 
 (86) 
 
 If we want to find 8 for a 6in. pipe, we must insert 
 d = 6 and k = 0"75 in formula (86), and we then find 
 8 = 0"-23. 
 
 In practice, however, 8 is made the same for all sizes 
 of pipes — viz., about 2in. Consequently the efficiency 
 of a covered pipe of large diameter is greater than that 
 of a small pipe. 
 
 Non-conducting materials used for lagging boilers, 
 steam-pipes, etc., are generally composed of sawdust, 
 charcoal, asbestos, hair felt, silicate cotton, paper fibre, 
 etc. 
 
 The same formulae may be applied for calculating 
 the heat transferred through the heating-surface, 
 H s of a boiler. The rate of heat transferred will be 
 
 H _ C x H. x (fr - t w ) heat nnitB> 
 8 
 
 (87) 
 
 Where t f and t a are the temperatures of the fire 
 and the water respectively, and 8 the thickness of the 
 flue-plates. The relative conductivity of copper and iron 
 is 100 and 16, and for this reason copper is often used 
 for fireboxes in locomotive boilers where t f is very high. 
 
 (2) Radiation. — The transfer of heat from one body 
 to another at any distance apart, follows the laws for 
 radiation of light. 
 
 To illustrate radiation of heat, place a thermometer 
 in front of a fire, and we shall soon find that its tem- 
 perature will rise ; if we now suddenly screen the ther- 
 mometer from the fire, its temperature will fall to that 
 of the surrounding air, which shows that the heat 
 cannot have been conducted through the air to the 
 thermometer. 
 
 The rate at which a hotter body radiates heat, and a 
 colder body absorbs heat, depends upon the state of the 
 surfaces of the bodies as well as on their temperatures. 
 Bough and dark surfaces radiate and absorb heat better 
 than bright and smooth surfaces. For this reason, the 
 
164 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 covering of steam-pipes and boilers should be smooth 
 and of a light colour; uncovered pipes and steam- 
 cylinder covers should be polished. 
 
 (3) Convection.- — The density of fluid is generally 
 diminished by the rise of temperature. If a fluid con- 
 tained in a vessel be heated, the portion of the fluid 
 next to the heat source will be less dense than the 
 other parts of the fluid, and will therefore tend to rise 
 to the surface ; but this rise must evidently be accom- 
 panied by a fall of the colder parts of the fluid, and thus 
 currents — convection currents — will be set up in the 
 fluid, whereby the latter will be heated. It is also 
 evident that the heat source ought to act on the fluid 
 at the lowest point of the vessel, so as to produce con- 
 vection currents right through the fluid. If the heat 
 source were placed at the surface, no convection 
 currents could be produced, and the heating of the fluid 
 would then be done by conduction, which would require 
 very much more time. 
 
 Boiler-flues should therefore be placed as close as 
 possible to the bottom of the boiler. 
 
 Expansion of Gases. 
 
 The volume occupied by a certain portion of a gas 
 depends upon its temperature and the pressure to 
 which it is subjected. If the pressure of the gas be 
 diminished, the volume will increase — i.e., the gas will 
 expand ; the same effect is caused by raising the 
 temperature of the gas. 
 
 The First Law of Expansion of Gases. — Assume that 
 we have a certain portion of a gas enclosed within a 
 cylinder, with an air-tight fitting piston. We may then 
 compress the gas by placing weights on the top of the 
 piston, in which case we shall find that the tempera- 
 ture of the gas will rise, but still by conveying a 
 sufficient amount of heat from the gas, we can keep the 
 temperature constant. We could otherwise diminish 
 the weights on the piston, and the gas would then 
 expand again and at the same time the temperature 
 would fall, but by adding the necessary quantity of 
 heat to the gas, we can keep the temperature constant 
 in this case also. 
 
 Let now V , V 1( V 2 , V 3 . . . denote the volumes 
 occupied by the gas, corresponding to the pressures per 
 unit of area P , P 1; P 2 , P 3 ; then we shall find by 
 keeping the temperature of the gas constant that 
 
 Yi-li. Yi-Zi. v 2_p 3 , m 
 
 V P = V 1 P 1 = V 2 P 2 = V 3 P 3 .= C . . (89) 
 
 or 
 
 The first law of expansion of gases may therefore be 
 stated as follows : " The product of the volume occu- 
 pied by a portion of a gas at constant temperature into 
 its pressure is constant." 
 
 This law was first discovered by Eobert Boyle in 
 1662, and verified by Mariotte in 1715 ; hence it is 
 called Boyle's law or Mariotte's law. 
 
 The constant, C, varies, of course, with the tempe- 
 rature, everything else remaining the same, 
 
 The Second Law of Expansion of Gases states that 
 " The increase of volume of a gas at constant pressure 
 is proportional to the temperature." 
 
 It has been found that the increase of volume of a 
 gas by heating it from the ice-point to the boiling-point 
 is 0'3665 of its volume at the ice-point. The expansion 
 of a gas by increase of temperature is therefore for 
 
 1 deg. F. equal to ^|§^- = 0-002036 ; this number is 
 180 
 
 called the " coefficient of expansion " of gases. 
 
 The second law of expansion of gases can now be 
 expressed in symbols, as follows : 
 
 Let V 82 , V 1( and V 2 be the volumes of a gas at 
 32 deg. F., ^deg. F., and t 2 deg. P. respectively, then 
 
 Vi= [1 + 0-002036(^-32)] V 32 . . (90) 
 
 V 2 = [1 + 0-002036 (t 2 - 32)] V 32 . . (91) 
 
 and by dividing (90) by (91) 
 
 Vi _ 1 + 0-002036 (fr- 32) , q9 . 
 
 V 2 1 + 0-002086 (*j- 32) " " [ ' 
 The second law of expansion of gases is also called 
 the law of Charles or the law of Gay-Lussac. 
 
 The two laws of expansion can be joined together in 
 one formula, stating the relation between volume, tem- 
 perature, and pressure of a gas. 
 
 Let V and V 1 be volumes of the same portion of a 
 gas at 32 deg. F., and under pressures of P and P^ then 
 
 Io_Pi 
 
 yi-p n • • ' 
 
 (93) 
 
 and let Vj be the volume of the gas at ^ deg. F. and at a 
 pressure of Pj, then 
 
 Y l _ 1 
 
 V; = l + 0-002036 fo-32) ' ' (94) 
 now multiply (93) by (94) then 
 
 Vi = V [1 + 0-002036 {t x - 32)] ij> . (95) 
 
 Pi 
 
 If the temperature and the pressure be changed to 
 t 2 and P 2 respectively, then the volume of the gas will 
 be 
 
 V 2 = V [1 + 0-002036 &- 82)]; 
 
 Now divide (95) by (96) then 
 
 V 9 P„ 
 
 (96) 
 
 -=E (97) 
 
 1 + 0-002036 (Cj - 32) 1 + 0-002036 (t 2 - 32) 
 where E is constant, depending upon the nature and 
 portion of the gas. Formula (97) expresses in symbols 
 the two laws of expansion of gases. 
 
 Absolute Temperature of Gases. — Imaginary Gases. — 
 The absolute temperature of a gas is a theoretical con- 
 sequence of the second law of expansion, by assuming 
 that it is possible to continue the cooling of a gas until 
 its volume is diminished to nought. The temperature, 
 T , at which the volume of the gas should be nought, 
 must satisfy the following equation : 
 
 = 1+ 0-002036 (T - 32)) . . (98) 
 and we find 
 
 T„ =. - 460 deg. F (99) 
 
tt EAT. 
 
 168 
 
 If gases really followed the second law of expansion 
 strictly, as assumed in formula (98), then all heat would 
 have been extracted from the gas at the moment its 
 temperature fell to T , and the temperature would 
 therefore be the absolute zero of temperature. But 
 all substances hitherto known have been found to have 
 three states of aggregation — solid, fluid, and gaseous — 
 besides which it has been proved by accurate experi- 
 ments that gases only follow the two laws of expansion 
 within certain limits of temperature and pressure, and 
 it has also been found that the further the temperature 
 of a gas is from its temperature of liquefaction, the closer 
 does it follow the laws of expansion. For example : 
 atmospheric air, hydrogen and oxygen, at temperatures 
 and pressures within practical limits, are far from their 
 fluid state of aggregation, and follow the two laws of 
 expansion fairly well ; but carbonic acid, which can be 
 condensed to a fluid at a temperature of minus 108 deg. 
 F., or at a pressure of 36 atmospheres, does not follow 
 the two laws so well, especially at high pressures, as it 
 is then approaching its fluid state of aggregation. 
 
 It is, however, convenient in theory to distinguish 
 between " imaginary gases " and "actual gases," the 
 former following the two laws of expansion absolutely, 
 and remaining gases at all temperatures and pressures, 
 whereas actual gases only follow the two said laws 
 approximately within certain limits. 
 
 The temperature, T„, given by (99) will be the 
 absolute zero for imaginary gases, and the temperature 
 of the gas reckoned from this zero will be the absolute 
 temperature of the gas. Thus, if the temperature of 
 an imaginary gas is t deg. F., its absolute temperature 
 is T = 460 + 1. One degree of absolute temperature is, of 
 course, equal to 1 deg. F.; we have only lowered zero 
 460 deg. F. 
 
 The temperature of an imaginary gas will hereafter 
 be denoted by T in the absolute scale, and by t in the 
 ordinary scale. 
 
 Formulse (98) and (99) gives _ ( JL_ = 492, which 
 
 makes 1 + 0-002036 (^ - 32) equal to 460 + f; we can, 
 consequently, write formula (97) as 
 
 v 1 Pi_v 1 p 2=e . . ( ioo) 
 
 T x ~ T 2 
 
 Formula (100) expresses the two laws of expansion 
 for imaginary gases. 
 
 Mechanical Theory of Heat. 
 
 When Joule had proved by well-known experiments 
 tbat heat is a special form of energy, a theory was 
 introduced, according to which the effects of heat upon 
 substances are explained by the laws of dynamics 
 This theory is called " thermo-dynamics," and its first 
 law states "that heat and mechanical energy are 
 mutually convertible ; a British unit of heat corre- 
 sponding to a certain fixed amount of mechanica 
 energy (equal to 772 foot-pounds) , called the mechanical 
 equivalent of heat." The latter will always be denoted 
 by J. 
 
 A body is supposed to consist of molecules, which 
 are tied together by the force of cohesion. By a 
 molecule we understand the limit of division of a body 
 effected by mechanical means. A molecule, again, 
 consists of atoms, which are tied together by chemical 
 affinity. An atom is, therefore, the liinit of division of 
 a body effected by chemical means. According to the 
 above-named theory, the molecules are in a state of 
 agitation — i.e., the molecules are flying about within 
 the space occupied by the body in all directions, at a 
 velocity which increases with the temperature. 
 
 We may conveniently compare the molecules of a 
 body to a number of billiard-balls moving to and fro 
 on a billiard-table. In their motion the balls will hit 
 each other, and, being elastic bodies, they will part in 
 opposite directions if their collision is straight ; but it 
 will also happen that the balls meet at an oblique angle, 
 in which case they will set each other spinning. The 
 kinetic energy, K lt of the balls consists therefore partly 
 of energy of translation and partly of energy of rotation, 
 and will be 
 
 K^Pm^ + ^IWi 2 • • (101) 
 
 Where m is the mass, «x the velocity of translation, 
 I the moment of inertia, and w x the velocity of rotation 
 of a ball. 
 
 The energy K x evidently corresponds to a quantity 
 
 of heat, Hj = -j- 
 
 Suppose we could somehow or other increase the 
 velocity of translation of the balls to v 2 , then by the 
 hitting of the balls against each other, their velocity of 
 rotation would also be increased, say, to w 2 , and their 
 kinetic energy to 
 
 K 2 = i2TOi; 2 2 + i2lM> 2 2 . . . (102) 
 The energy imparted to the balls in increasing their 
 velocity from v 1 to v t must be (E^-K^), which evidently 
 corresponds to a quantity of heat 
 
 -g. _ K 2 - K t 
 
 (103) 
 
 In heating a body we have not only to increase the 
 kinetic energy of the molecules, but as the body gene- 
 rally expands, we have also to overcome the force of 
 cohesion, and the external pressure, which resist the 
 expansion. 
 
 We may, therefore, say that the heat is partly spent 
 in exerting 
 
 Internal energy = work done in overcoming 
 internal resistance + kinetic energy added to 
 
 the molecules • • (104) 
 
 and partly in exerting 
 
 External energy = external work done in over- 
 coming external resistances .... (105) 
 Latent Heat.— If a body be moved against gravity at 
 a constant velocity, then its potential energy will be 
 increased, but its kinetic energy will remain un- 
 altered, or 
 
 Energy exerted = work done in increasing the 
 potential energy of the body .... (106) 
 
166 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 To apply (106) on (104) would mean to spend heat in 
 overcoming internal resistances only, and thus to 
 increase the potential energy of the molecules by 
 bringing them further apart, and allowing them to move 
 more freely, but the temperature would remain con- 
 stant. Now this is precisely what takes place when a 
 body is changing from one state of aggregation to 
 another. Thus, when melting a piece of ice, we may 
 continue to add heat, but as long as any ice is left, the 
 temperature of the ice- water will remain constant at 
 32 deg. F. The temperature of water while being 
 evaporated at a constant pressure will also remain 
 constant. 
 
 The heat which is spent in increasing the potential 
 energy of the molecules is called "latent heat," in 
 opposition to " sensible heat," which is spent in 
 raising the temperature only ; instead of the two latter 
 terms, " potential heat " and " kinetic heat " might be 
 used with advantage. 
 
 Ther-mo-dynamics of Imaginary Gases. 
 
 It has been shown by experiments that actual gases, 
 when far from their point of liquefaction, show hardly 
 any sign of internal resistances, for if such a gas be 
 allowed to escape into a vacuum, it fills up all parts of 
 the space, and its temperature does not fall appreciably. 
 If the gas had absolutely no internal resistances to 
 overcome by expanding, its temperature would not fall 
 at all. We may, therefore, assume that the reason 
 why no actual gas follows the laws of expansion is due 
 to internal resistances, whereby the nature of the gas 
 will be altered when we alter its temperature, pressure, 
 and volume. Imaginary gases can consequently have 
 absolutely no internal resistances to overcome at any 
 temperature and pressure, and the heat spent on such 
 a gas will partly be exerted as 
 
 Internal energy = kinetic energy added to the 
 molecules (107) 
 
 and partly as 
 
 External energy = external work done by the 
 expansion of the gas (108) 
 
 If by /3 we denote the ratio of the total molecular 
 energy to that of translation of an imaginary gas, 
 then the internal energy of the gas will be 
 
 i/3 2m« 2 (109) 
 
 and as the nature of the gas does not alter with the 
 temperature and pressure, we must conclude that /3 
 remains constant. 
 
 In consequence of the molecular agitation and the 
 property of gases to expand without limit, the molecules 
 must strike the boundary of the vessel in which the gas 
 is enclosed, with a force equal to their momentum, 
 mv. But the greater the velocity, v, is, the more fre- 
 quently will the vessel be struck by each molecule, so 
 that the pressure exerted by the gas must be propor- 
 tional to m v 2 . By compressing the gas we bring the 
 molecules closer together, and the intensity of the 
 bombardment on the vessel will thus be increased, or, 
 
 in other words, the pressure of the gas must be, if V is 
 
 the volume of the gas, 
 
 ^ mv 2 «,„. 
 
 P oc -y- (HO) 
 
 By comparing (110) with (100), it will be apparent 
 that temperature is measured by m v 2 . If, therefore, 
 two gases have the same molecular energy of transla- 
 tion, they will also have the same temperature. 
 
 Specific Heat of Imaginary Gases. — When the tem- 
 perature of a gas is zero, then the quantity of heat con- 
 tained in the gas will also be zero. If, therefore, the 
 mass of a portion of a gas is M, then the heat required 
 for raising its temperature from absolute zero to T x 
 will be 
 
 Hi J^mV =gmixMxTi| (111) 
 
 and if the temperature be raised to T 2 , instead of T x , 
 then the heat will be 
 
 H. J^" 8 ' = S m2 xMxT 2 . (112) 
 
 where S mi and S m2 are the mean specific heats of the 
 gas between zero and T x and zero and T 2 respectively. 
 Now divide (111) by (112), we shall then have 
 
 m v x 2 _ Smi x Tj 
 
 ^m 2 X 1 2 
 
 but according to our theory 
 
 Consequently 
 
 mv 9 
 
 5a 
 T 2 
 
 C>mi — C3m2 
 
 (113) 
 
 (114) 
 
 (115) 
 
 or, in words, the specific heat of an imaginary gas is 
 constant for all temperatures and pressures, and only 
 varies with the substance of the gas. Actual gases 
 follow this law approximately. The specific heat of a 
 gas will hereafter be denoted by S„. 
 
 The heat contained in the mass of q pounds of an 
 imaginary gas at an absolute temperature, T v in degrees 
 Fahrenheit will therefore be 
 
 Hj = S„ x q x Tj British heat units . . (116) 
 
 The equivalent mechanical energy will be 
 
 772 x S„ x q x T x = K„ x q x T x foot-pounds . . (117) 
 
 where K„ is the quantity of mechanical energy in foot- 
 pounds which is required to raise the temperature of lib. 
 of the gas 1 deg. F. 
 
 Molecular Heat. — Assume that we have two portions 
 of two different gases each enclosed in a vessel, and 
 both occupying the same volume and having the same 
 pressure and temperature. 
 
 The pressure exerted by the first gas must be pro- 
 portional to the molecular energy of translation and to 
 the number, n lt of molecules contained in the gas ; or, 
 in symbols, 
 
 Pj oc m x v 1 2 X n x . . . . (118) 
 
 Similarly the intensity of the pressure of the second 
 gas will be 
 
 P 2 oc m 2 v 2 2 x n 2 . . . . (119) 
 
HEAT. 
 
 167 
 
 But as P 1 is equal to P„, we shall have 
 
 m x «i 2 xm^Mjflj 2 x»i 2 . . . (120) 
 
 We have, further, m 1 v-? equal to m 2 v 2 , as the gases 
 have the same temperature, and consequently also n x 
 equal to n 2 — i.e., " the two gases contain the same 
 number of molecules." 
 
 The heat contained in the first gas will be 
 
 i fan xm, Vi 
 
 and that contained in the other will be 
 -n- ifan x m 2 v 2 2 
 
 ■C-!! = = • 
 
 (121) 
 
 (122) 
 
 Let us now for a moment assume that fa is equal to 
 fa, then H, would be equal to H 2 — i.e., " the two gases 
 contain the same amount of heat." 
 
 We could also have written 
 
 H 1 = S 1 „xM 1 xT .... (123) 
 
 where M 1 is the mass of the portion of the first gas. 
 Similarly, we have for the second gas 
 
 H 2 =S"„xM 2 xT .... 
 Therefore 
 
 H x S\ x M 1= S 1 ,x)txm 1 = S\ x m 1 
 
 (124) 
 
 (125) 
 
 S"„ x M 2 S"„ xnxm 2 S"„ x m 2 
 
 From (125) we learn that " the molecular thermal 
 capacity, S» x m, of an imaginary gas is the same for 
 all imaginary gases." Actual gases have been found to 
 follow this law, and our assumption that /3 is the same 
 for all gases is therefore right. 
 
 Specific Seat at Constant Volume and Constant 
 Pressure. — When heating a certain portion of an 
 imaginary gas from a temperature T 1: in a cylinder 
 with air-tight fitting piston, we can do one of two 
 things — viz. : 
 
 (1) Prevent the gas from doing external work, by 
 increasing the pressure on the piston, and thus keep the 
 volume of the gas constant. 
 
 (2) Or we can let the pressure on the piston remain 
 constant. The gas will then expand and do work in 
 raising the weight. The gas is therefore heated at a 
 constant pressure. 
 
 In the first case, the heat is spent in doing internal 
 work only, and we will have 
 
 H = S,xjx (T 2 - Tj) British heat units (126) 
 
 where q is the mass of the gas in pounds, and T 2 the final 
 temperature. The energy exerted in mechanical energy 
 units will be 
 
 772 x H = K v x q x (T 2 - T a ) foot-pounds (127) 
 
 As the volume of the gas has been kept constant, 
 S„ is called "the specific heat of the gas at constant 
 volume." 
 
 In the second case, the heat will be spent partly in 
 doing internal work and partly in doing external work, 
 and we will have 
 
 H = [S„ x q x (T 3 - T x ) + q x H s ] heat units (128) 
 
 Where IL is the heat spent in doing external work 
 per pound of gas and T 3 is the final temperature, we 
 will also have 
 772 x H = [K„ xgx (T 3 - T x ) + E w X q] foot-pounds (129) 
 
 Where E w is the external work done per pound of gas. 
 We might also have written (128) as 
 
 H = S p X q X (T 3 - TO heat units . . . (130) 
 
 Where S^, is called " the specific heat of the gas at 
 constant pressure," instead of (129) we can write 
 
 772 H - K p xq* (T 3 -Tj) foot-pounds. . (131) 
 
 where K p is the mechanical energy in foot-pounds 
 required for raising the temperature of lib. of gas 
 1 deg. P. at constant pressure. 
 
 It is evident that T 3 < T , for we have spent the 
 same amount of energy in both cases ; but in the first 
 case, we have only raised the temperature of the gas, 
 whereas in the second case, a part of the heat has been 
 spent in doing external work.- 
 
 We have therefore 
 
 S p > S„ and K p > K v ; 
 
 or in words, " The specific heat at constant volume of 
 a gas is smaller than the specific heat at constant 
 pressure of the same gas." 
 
 It remains now to find an expression for the external 
 work, E w . If P be the constant pressure in pounds per 
 square foot, V; the initial volume of the gas in cubic 
 feet at temperature T v and V the final volume of the 
 gas. then the piston will have swept through volume 
 (V - Vi ) cubic feet, while the temperature of the gas 
 has been raised from T 1 to T 3 . According to (100), page 
 165, PV = BT 3 = r?T 3 and PV» = BT X = rqT^ we 
 have therefore 
 
 E M 2 = P (V- Vi ) = r x q (T 3 -T x ) foot-pounds (132) 
 
 where r is the external work done by lib. of gas, raised 
 1 deg. E. at constant pressure. 
 
 By the combination of (129), (131), and (132) we 
 obtain 
 
 K p = K v + r . . . . (133) 
 
 Example. — Let the problem be to find K Pi K„, and r 
 for atmospheric air, taking the values S p = 0'2375 and 
 S» = 0'1685 as found for actual air. 
 
 (1) For Pound-Mass and Foot-Pounds. — The amount 
 of heat required for raising the temperature of lib. of 
 air 1 deg. E. at constant pressure will be 
 
 Up = 0-2375 British heat units, 
 and at constant volume 
 
 H„ =0'1685 British heat units. 
 We shall therefore have 
 
 Kp =0-2375x772 = 183-35 foot-pounds, 
 
 K„ =0-1685 x 772 = 130-08 foot-pounds, 
 andr = Kp - K„ =53'27 foot-pounds. 
 
 (2) For Kilogramme-Mass and Kilogram-metres. — 
 The amount of heat required for raising the tempera- 
 
PRACTICAL ELECTRICAL ENGINEERING. 
 
 ture of 1 kilogramme of air 1 cleg. C. at constant pres- 
 sure, will be 
 
 Up = 0'2375 calories (major), 
 and at constant volume 
 
 EL = 0*1635 calories (major). 
 As J =424 kgm, we shall have 
 
 Kj, = 0-2375x424 = 1007 kilogram-metres. 
 
 K„ =0-1685 x 424 = 71-44 kilogram-metres. 
 r =K P -K„ =29'26 kilogram-metres. 
 
 (3) For Gramme-Mass and Ergs. — The amount of heat 
 required for raising the temperature of 1 gramme of 
 air 1 deg. C. at constant pressure will be 
 
 Up =0'2375 calories (minor), 
 and at constant volume 
 
 H„ =0'1685 calories (minor). 
 As J is 41,550,000 ergs, we shall have 
 K p =0-2375 x 41,550,000 = 9,868,000 ergs. 
 K, =0-1685x41,550,000 = 7,001,200 ergs. 
 r =K P -K v =2,866,800 ergs. 
 
 "We have seen that two imaginary gases which occupy 
 the same volume, V, exert the same pressure, P, and 
 have the same temperature, T, contain also the same 
 quantity of heat. Let us now heat both gases up to 
 the same temperature, Tj, by adding so much heat to 
 each that they occupy the same volume, V 1; »nd exert 
 the same pressure, P 1( then they must again contain tbe 
 same heat, and they have, therefore, also received the 
 same quantity of heat, and they have done the same 
 amount of external work. 
 
 "We must therefore have 
 
 K i) xMx(T 1 -T) = K 1 „xM 1 x(T 1 -T) . (134) 
 
 rxMx(T 1 -T) = r 1 xM 1 x(T 1 v-T) . (135) 
 
 K„xMx(T 1 -T) = K 1 „xM 1 x(T 1 -T) . (136) 
 
 By combination of (133), (134), (135), and (136), we 
 obtain 
 
 -Kg == ^P - S P _ &P - -y /ions. 
 
 where y is a constant, or in words, " The ratio of 
 specific heat at constant pressure to specific heat at 
 constant volume is the same for all imaginary gases." 
 Actual gases also follow this law, and y has been found 
 to be equal to about 1 - 41. 
 
 Adiabatic and Isothermal Expansion and Compression. 
 
 When a gas expands in a cylinder doing work by 
 overcoming the resistance of a piston, its temperature 
 will fall, as some of the molecular energy will be trans- 
 formed into external work. 
 
 Let us assume that the cylinder and piston, Fig. 172, 
 are perfect insulators for heat, so that they can neither 
 absorb nor conduct heat. Let the cylinder contain 
 0"841b. of imaginary air, of which the volume will 
 be 1 cubic foot when the piston is in position BK, 
 and the temperature is 60 deg. P.; the pressure of the 
 air will then be 11 atmospheres, or 23,'2871b. per 
 square foot. If we now let the air expand, driving the 
 
 piston before it, then the volume will increase, and 
 the pressure will diminish with the ordinates to curve 
 A. The work done by the gas in expanding from 1 
 cubic foot to 11 cubic feet will therefore be represented 
 by the area B D E KB. The same area would also be 
 proportional to the heat lost by the gas. Let T x be 
 the final temperature of the gas, the initial one being 
 T 2 = 520 deg. F. ; then the heat lost by the gas 
 would be 
 
 S„ x 0-84 x (520 - Tj) British heat units ; 
 and the external work done would be 
 
 K„ x 0-84 (520 - Tj) foot-pounds. 
 
 As S„ = 0-1685, and K v = 130-08 for air, it only 
 remains to determine T x . It can be proved that when 
 an imaginary gas, by expanding, is doing work and 
 does not receive any heat nor gives off heat to surround- 
 ing bodies, the relation between corresponding tempe- 
 ratures, volumes, and pressures will be 
 
 T, 
 
 $r •§-&)'■ •<>» 
 
 where y is 
 
 Op 
 
 In our case we have therefore 
 
 T x = 520 (A) '" = 194-5. . . (139) 
 The heat- given off by the gas will be 
 0-1685 x 0-84 (520 -194-5) = 46-07 British heat units, 
 and the work done by the gas will be 
 
 130-08 x 0-84 (520 - 194-5) = 35,566 foot-pounds. 
 
 The curve A, Fig. 172, is called " adiabatic " curve, 
 and the expansion is termed "adiabatic expansion. - " 
 
 The ratio of the final volume to the initial volume of 
 the gas is called the " ratio of expansion," and will be 
 hereafter denoted by r e . 
 
 "We might now move tbe piston towards the bottom 
 of the cylinder, and thereby compress the gas, whereby 
 the pressure will increase with the ordinates of the 
 curve A. When we have moved the piston to its 
 initial position, then the volume of the gas will again 
 be 1 cubic foot, the pressure 11 atmospheres, and 
 the temperature 520. The compression, however, 
 requires that we should do work on the gas to an 
 amount of 35,566 foot-pounds, whereby the gas receives 
 46 '07 British heat units, the same quantity which it 
 lost during expansion. The gas has thus been returned 
 to its original thermal state. As neither heat has been 
 added from external sources, nor heat been dissipated 
 while moving the piston, the compression is termed 
 " adiabatic compression." 
 
 If the cylinder, Fig. 172, be a good conductor, then 
 we may add heat to the gas during expansion, at such 
 a rate that the temperature of the gas remains constant, 
 while the volume of the gas increases from 1 cubic 
 foot to 11 cubic feet. In this case, the gas will follow 
 the first law of expansion, and we shall have 
 
 As P 2 = 11 atmospheres, it is evident that Pj will be 
 one atmosphere or 2,1171b. per square foot, and the 
 
HEAT. 
 
 169 
 
 intermediate pressures will vary as the ordinates to 
 curve I. As the thermal state of the gas remains the 
 
 remains constant, and its asymptotes are the axes of 
 ordinates and abscissae. Any point of the curve may 
 
 same during expansion, it must be the heat added 
 which is transformed into external work ; the latter is 
 represented by the area BCEKB. 
 
 be constructed in the following way : Draw, for instance, 
 the ordinate, G P, then draw F and L N, the latter 
 to be parallel to the axis of abscissae ; N G will be the 
 
 /M5;//)///MS')'''''S »"''"'''/"""""'77fr77?- 
 
 Fig. 173. 
 
 The curve I is evidently a rectangular hyperbola, 
 since the product of the abscissa and the ordinate 
 
 ordinate corresponding to G as abscissa. This 
 construction is right, because LK ; FG = OK; G, 
 
170 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Take KB=i/ 1) CE = y 2 , OK=a; 1 , and OE=% then 
 area BCEKB will be 
 
 C x log*. 
 
 (140) 
 
 and the work done by the expansion of the gas will be 
 
 P 2 V 2 x log, — i = P 2 V 2 x log* r e 
 
 *2 
 
 (141) 
 
 where log e means the hyperbolic logarithm. 
 
 Expansion of a gas at constant temperature is called 
 " isothermal " expansion, and also " hyperbolic " expan- 
 sion, because the curve I is a hyperbola. 
 
 In our particular case we have P 2 = 23,2871b. per 
 square foot, and V 2 = l cubic foot, and r e = ll. The 
 work done by the isothermal expansion of 0'841b. of 
 air at an absolute temperature of 520 deg. F., will be 
 
 23,287 x 1 x log e 11 = 55,839 foot-pounds, 
 
 and the heat added to the gas during expansion will be 
 
 55,839 =72 . 33 British heat units. 
 772 
 
 The curves I and A in Fig. 172 show that the work 
 done in isothermal expansion is greater than that done 
 in adiabatic expansion. 
 
 In Fig. 173 the adiabatic as well as the isothermal 
 compression begin at the same temperature, T x = 520, 
 and at the same pressure, P x = 2,1171b. per square foot. 
 
 The adiabatic compression will take place if the 
 cylinder as well as the piston are absolute insulators for 
 heat. The pressure will increase with the ordinates of 
 the curve A. Let D be the point at which this curve 
 intersects ordinate CK — i.e., when the gas has been 
 compressed to 1 cubic foot, then area BDKEB will 
 represent the work done on the gas during compression. 
 According to (138) we will have 
 
 T 2 = 52o(^V 41 =1390 
 
 and 
 
 P 2 = 2117 x ( y V ^ = 62 > 2 411b. per square foot. 
 
 The work done on the gas will therefore be 
 
 K„ X 0-84 (T s - TO = 130-8 x 0"84 (1390 - 520) = 95,061 
 foot-pounds, 
 
 and the heat added to the gas will be 
 
 B, x 0-84 (T, - T,) = 0-1685 x 0-84 (1390 - 520) = 123"1 
 British heat units. 
 
 The isothermal compression will take place if we 
 extract heat from the gas during the motion of the 
 piston, at such a rate that the temperature of the gas 
 remains constant. The pressure will in this case 
 increase with the ordinates of curve I, which is exactly 
 the same as curve I in Fig, 172. The work done on the 
 gas will be 55,839 foot-pounds, the same amount as was 
 given off by the gas during expansion, Fig. 172. The 
 heat rejected will be 72'33 British heat units, the same 
 as that added to the gas during expansion, 
 
 Heat Engines. 
 
 By a heat engine is understood "a machine which 
 transforms heat energy into mechanical energy, or vice 
 versa." 
 
 The action of a heat engine requires a working 
 substance which shall receive heat from a source, and 
 again give it off in doing external work. If the per- 
 formance of the engine is to be continuous, the opera- 
 tions of its working parts must repeat themselves 
 within a certain interval of time, during which the state 
 of the working substance with regard to pressure, 
 volume, and temperature must also be repeated. Such 
 a complete set of repeated operations is called a 
 " cycle." The temperature, volume, and pressure of 
 the working substance must therefore be exactly the 
 same at the end as at the beginning of the cycle. 
 
 Let us assume that the working substance is an 
 imaginary gas, acting upon a piston in a cylinder, as in 
 Fig. 172, the cycle being performed while the piston 
 moves through distance K E, and back through dis- 
 tance E K. Let the initial pressure be represented by 
 KB, the initial volume by OK, and let the tempera- 
 ture be T 2 . We may now apply the source during the 
 forward stroke in such a way that the temperature of 
 the gas remains constant at T 2 ; the expansion curve 
 will therefore be I, and the heat, H, taken from the 
 source will be represented by area K B C E K, which 
 area will also represent the work done by the gas on 
 the piston. The gas can evidently do no work on the 
 piston during the return stroke, but must be com- 
 pressed by absorbing energy, and as the gas must be 
 brought back at the end of the cycle to its initial state, 
 the compression curve must pass through point B ; we 
 can satisfy the latter condition by choosing the 
 adiabatic curve A as back-pressure curve. This, how- 
 ever, requires us to reduce the pressure of the gas to 
 D E, by applying a refrigerator whose temperature is 
 Tj, and which will absorb from the gas a quantity of 
 heat, h, which is represented by area DEKBD; when 
 this is done, the communication with the refrigerator 
 is shut off, and the piston can then move backwards, 
 compressing the gas adiabatically and leaving the gas 
 at the end of the stroke in its initial state. As the 
 work done by the gas is represented by area BCEKB, 
 and the work done on the gas by area BDE KB, it is 
 evident that the work done in driving the piston 
 through a cycle must be proportional to area BCDB; 
 this work is called the " indicated" work, as it can be 
 determined by the "indicator," generally called the 
 steam indicator, and which will be described later on. 
 The indicated work is evidently equivalent to H - h, 
 and as the gas receives a quantity of heat, H, from the 
 source during the forward stroke, the efficiency of the 
 engine with regard to converting heat into mechanical 
 energy must be 
 
 9 = ^ (142) 
 
 Camot's Imaginary Perfect Heat Engine. — An 
 imaginary heat engine working between two tempera- 
 tures T a and T ; can be made more efficient than the 
 
HEAT. 
 
 171 
 
 above described, by a cycle of operations known as 
 Carnot's cycle. In Fig. 174 the cycle begins with 
 applying the source, of which the temperature is T 2 , 
 and allowing the gas to expand isothermally at tempe- 
 rature T 2 , exactly as in Fig. 172, but only while the 
 piston moves through distance K N ; during the rest of 
 the forward stroke the source is disconnected from the 
 engine and the gas expands adiabatically, the expansion 
 curve being G L, whereby the temperature falls to T lt 
 the temperature of the refrigerator. The refrige- 
 rator is now connected to the engine, and the backward 
 stroke begins by compressing the gas isothermally at T lt 
 and along the hyperbola L Q, until the piston has moved 
 through distance EF. From this position of the 
 piston until the end of the stroke the corapression 
 follows the adiabatic curve Q B through point B, the 
 refrigerator being disconnected. 
 
 and similarly as L and Q lie on hyperbola L Q 
 
 P 1 xV 1 =.p 1 x»i-BT 1 . . (144) 
 As G and L lie on curve G L we have 
 
 i>2 _ /V \y 
 
 ■w) 1 
 
 And as Q and B lie on curve Q B we have 
 
 P, 
 
 According to (141) and (143) we have 
 AreaBGNKB = RT 2 x log. 
 
 (145) 
 
 (146) 
 
 P? 
 
 And, similarly, (141) and (144) will give us 
 AreaQLBFQ = ET,x log, |i. 
 
 By this cycle, the work done by the gas on the piston 
 during the forward stroke will be represented by area 
 BGLEEB, and the work done on the gas during the 
 backward stroke by B Q L B K B, and the useful work 
 by area B G L Q B. As the two areas GLENG and 
 B Q F K B both represent S„ (T a - Tj), it follows that 
 area BGL QB = area BGNKB minus area QLEFQ, 
 
 or 
 
 n = 
 
 H- A _ BGNKB-QLEFQ 
 BGNKB 
 
 Let the pressures of the gas corresponding to points 
 B, G, L, and Q be denoted by P 2 , p 2 , P 1( and p v and 
 the corresponding volumes occupied by the gas V 2 , v 2 , 
 V and v v then, as B and G lie on hyperbola B G, we 
 
 must have 
 
 P ? xV ? =^2 X «2 = BT 2 
 
 (143) 
 
 A proper combination of the four formulae (143), 
 (144), (145), and (146), will show us that 
 
 P* Pi' 
 
 H T s * 
 
 B-h = T,-T 1 
 H T 2 ' ' 
 
 Formula (147) shows that the efficiency of the 
 Carnot's engine only depends upon the temperatures of 
 the source and the refrigerator, but not upon the working 
 substance, as long as it only follows the laws of 
 imaginary gases. 
 
 The engine is perfectly reversible — i.e., when driving 
 the engine in the reverse direction by another engine, 
 
 and therefore 
 
 and consequently t\ = 
 
 (147) 
 
172 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 the temperatures of the source and refrigerator will be 
 reversed. The machine would in this case convert 
 mechanical energy into that of heat. The expansion 
 will first take place along the adiabatic curve BQ, until 
 the piston has moved through distance KF, when the 
 temperature will have fallen to T 1 ; for the rest of the 
 stroke the temperature of the gas remains constant 
 while expanding along curve QL, and receiving a 
 quantity of heat, h, from a source, of which the 
 temperature is T r The first part of the compression 
 curve is L G-, while the piston moves through distance 
 E N ; the gas is then compressed at constant tempera- 
 ture T 2 , it being all the time in connection with a 
 refrigerator whose temperature is also T 2 . The amount 
 of heat generated during this reversed cycle is evidently 
 H - h, this being the same quantity as is required for 
 driving the machine as a motor. 
 
 We can now prove that no heat engine (whether 
 imaginary or actual), receiving heat from a source at 
 temperature T 2 , and giving off heat to a refrigerator at 
 a temperature Tj, can have an efficiency higher than 
 
 T - T 
 2 — -. For suppose an engine did exist which had a 
 
 2 
 
 higher efficiency, then we might couple it with a 
 
 Carnot's engine and drive the latter as a heat generator. 
 Let the generator during one cycle give off a quantity 
 of heat, H, to the source, and receive h from the 
 refrigerator, and let the motor, during the same 
 interval of time, receive H from the source and give 
 off h to the refrigerator. The heat account would 
 thus be balanced, but as the efficiency of the motor is 
 greater than that of the Carnot's engine, the motor 
 would evidently be able to do more work than just 
 driving the generator. But this work cannot be 
 accounted for, and must therefore have been created 
 from nothing, which is absurd. The efficiency of the 
 
 motor can therefore not be greater than — 2 -1. 
 
 J-2 
 
 The practical conclusion to be derived from the 
 above theoretical statement is this : That we must 
 expect to find the efficiency of our actual heat engines 
 to be less than that of Carnot's ideal engine, working 
 between the same limits of temperature. 
 
 CHAPTER XL 
 
 THBOBY OF STEAM ENGINES. 
 
 Single-Acting Engine Without Expansion. 
 
 'HE definition and properties of dry saturated 
 as well as of superheated steam have been 
 given on page 61, and the generation of 
 steam in a boiler has been thoroughly dis- 
 cussed in Chapter VIII. We may therefore 
 proceed to investigate the behaviour of the 
 steam in a cylinder when driving a piston before it. 
 The diagram, Fig. 175, represents a boiler with a 
 cylinder on the top, which is connected to the 
 boiler by a short steam-pipe, S P. The line a b repre- 
 sents the water-level, when the piston is at the 
 bottom of the cylinder; the steam-room of the 
 boiler is filled with dry saturated steam at an absolute 
 pressure p per square inch, or P B per square foot. 
 Suppose now we have applied so much heat that lib. 
 of water is evaporated, then the water-level will have 
 fallen to cd, and the piston will be in position "2," 
 having swept through a volume equal to (v - s) cubic 
 feet, where v is the specific volume of steam at a 
 pressure p a , and s the volume of lib. of water at that 
 pressure, this volume being equal to the space between 
 levels a b and c d. Let us now pump lib. of water into 
 the boiler, whereby the water-level will again be a b ; in 
 order to do this we must overcome the pressure of the 
 steam and move the piston through a volume, s, to posi- 
 
 tion "3." We will now shut inlet-valve, IV, whereby 
 no more steam can get into the cylinder, which at that 
 moment contains lib. of dry saturated steam at a 
 pressure T f = P a (where T f denotes that the pressure 
 is pushing the piston forward), assuming that no con- 
 densation has taken place. The energy which has been 
 exerted by the lib. of steam is evidently equal to 
 P/ (v —s), of which P/ x s is the work done in pumping 
 the water into the boiler. We will now open the 
 exhaust valve, E V, whereby the steam is admitted 
 through the exhaust-pipe, E P, to the refrigerator, or, 
 as it is called, the condenser, in which the absolute 
 pressure is P 6 per square foot, corresponding to a tem- 
 perature, t 1 . The piston will now move downwards, 
 overcoming pressure P 6) and sweeping through 
 volume, v. The indicated work of this engine will 
 evidently be 
 
 IW=(P, -V b )xv .... (148) 
 
 It would be interesting to see how I W would vary 
 with the pressure of the steam ; for this purpose we 
 must refer to the table on page 61. To make the 
 subject clearer, the two curves, Fig. 176, have been 
 drawn, of which the one represents the relation between 
 the steam pressure in pounds per square inch and the 
 temperature in degrees of Fahrenheit, while the other 
 illustrates the variation of the specific volume of steam 
 
THEORY OF STEAM ENGINES. 
 
 173 
 
 in cubic feet with the temperature. p a , v, and t trate the indicated work of lib. of dry saturated 
 are drawn to the same scale, and t begins with steam, see formula (148), it is necessary to fix the 
 
 
 ""WiWVO?? 
 
 IV 
 
 TspI (ep 
 
 JUy 
 
 JTio. 175. 
 
 Fig. 178. 
 
 100 deg. F., instead of 32 deg. F., in order to save value of P& — i.e., the pressure in pounds per square 
 space. By glancing at the curves, Fig. 176, it will be foot of the steam in the condenser. The ordinates to 
 
 ~t 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■ 
 
 
 
 1 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 v 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 \ 
 
 < 
 
 
 
 
 
 
 
 
 
 \* 
 
 / 
 
 
 
 
 
 
 
 \£ 
 
 
 
 
 
 
 
 
 1 V i 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 /] 
 
 : 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 j 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 1 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 W 
 
 /' 
 
 
 
 I 
 
 St 
 
 ■j. 
 
 u 
 
 ■it 
 
 So 
 
 3C 
 
 v> 
 
 3 
 
 a 
 
 3 
 
 id 
 
 4- 
 
 LO t- 
 
 Fig. 176. 
 
 clear that the product, p a x v, will only vary a little 
 with the temperature ; this is made more apparent in 
 Fig. 177, where the ordinates to the curve A represent 
 P r x v, and the corresponding abscissae, p a . To illus- 
 
 curves B, C, and D represent P& x v for P& equal to 
 2,160, 1,080, and 360 pounds per square foot respec- 
 tively, and consequently the difference between the 
 ordinates of curves A and B will give the indicated 
 
174 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 work of lib. of steam if the pressure in the con- 
 denser is 151b. per square inch and the temperature 
 213 deg. F.; similarly, the difference between the ordi- 
 nates of curves A and C will give the indicated work 
 of lib. of steam, when the pressure in the condenser 
 
 Example 1. — Take p a = 301b. ; curve A, Fig. 177, 
 will give us P/ x v = 58,280 foot-pounds. 
 
 (a) If the pressure in the condenser bepb = 2' 51b. 
 per square inch, then P& x v will be equal to 4,856 
 foot-pounds, and the indicated work done by lib. of 
 
 Pfao 
 
 i 
 
 Vjmo 
 Siwo 
 iicoo . 
 
 1fw 
 
 >io>c 
 
 GdOD 
 
 6«Kj 
 
 looo 
 
 IS 
 
 3° 
 
 no ns ' iso 
 Fig. 177. 
 
 isf 
 
 180 
 
 iW zn zzs 
 
 z+o 
 
 [It will be noticed that there is a break in this diagram between the points 30,000 and 57,000, 
 because none of the curves referred to would appear on this portion, and therefore to 
 insert the full abscissae and ordinates would take up space without any corresponding 
 advantage.] 
 
 is 7'51b. per square inch, and the temperature 
 180 deg. F. If the pressure in the condenser is 251b. 
 per square inch, and the temperature is 134'6deg. F., 
 then the difference between the ordinates of curves A 
 and D will represent the indicated work of lib. of 
 steam. 
 
 steam will be (P/ - P 6 ) x v = 53,424 foot-pounds. 
 Assuming the temperature of the feed-water to be that 
 of the condenser, then the quantity of heat required for 
 turning lib. of water at 134'6 deg. F. into dry satu- 
 rated steam at the pressure required for the engine, will 
 be, according to formula (2) page 62 ; 
 
THEORY OF STEAM ENGINES. 
 
 175 
 
 H = 1,114 + 0-305 x 250-2 - 134-6 = 1,055 heat units. 
 
 The efficiency of the engine, as a heat engine, is 
 evidently 
 
 53,424 
 
 = 0-0656. 
 
 1,055 x 772 
 
 (b) If we have pb = 7'51b. per square inch, then 
 (P - Pi ) x v = 43,708 foot-pounds. The tempera- 
 ture of the feed- water will be 180 deg. F., and we will 
 therefore have 
 
 H = 1,114 + 0-305 x 2502 - 180 = 1,010 heat units. 
 
 Therefore, 
 
 43708 n-nKRK 
 
 1= - ~-~ — „„~ = 0565. 
 1,010 x 772 
 
 (c) Let pbbe equal to 151b. per square inch, then 
 (P/ - Pj ) x v = 29,142 foot-pounds. 
 
 The temperature of the feed- water is 213 deg. F., and 
 therefore 
 H = 1,114 + 0-305 x 250-2 - 213 = 977 heat units. 
 
 and 
 
 = 29,142 = . Q386 
 
 ' 977 x 772 
 
 Example 2. — Take p a = 2401b. ; curve A, Kg. 177, 
 will give us P/ x v = 65,730 foot-pounds. 
 
 (a) Take p b =2-5, then (P, - P 6 ) x v = 60,874 foot- 
 pounds and H= 1,055 ; therefore 
 
 60,874 =0 . Q747 
 ' 1,055x772 
 
 (b) pb =7-5, then (P/ ~P 6 ) x © = 63,676 foot-pounds, 
 and H= 1,010 ; therefore 
 
 63,676 
 
 n=- 
 
 = 0-0817. 
 
 1,010 x 772 
 
 (c) p b =15, then (P/ - P 6 ) x v = 61,622 foot-pounds, 
 and H = 977; therefore 
 
 61,622 =0 . 0817- 
 ' 977 x 772 
 
 These examples clearly show the advantage of the 
 condenser with low-pressure steam. The advantage 
 of high-pressure steam is also evident ; for not only is 
 the efficiency higher, but we can, at any rate when 
 the pressure is very high, dispense with the condenser 
 altogether, and thus free the engine from doing work 
 in pumping the condensing water through the con- 
 denser. 
 
 If we compare the above efficiencies with those of 
 Carnot's imaginary engine, working between the same 
 temperatures, we shall find that our engine is far from 
 being a perfect one. But by investigating the matter, 
 we shall find that we have not entirely followed the 
 principle laid down in the imaginary engine, for we 
 nave allowed the temperature of the steam to fall 
 suddenly, by heating a large mass of water which 
 is of no use to us, instead of which we ought to let the 
 steam lose heat in doing work — that is, allowing the 
 steam to expand while pushing the piston before it. 
 
 Ideal Single-Acting Engine with Expansion. 
 Suppose that the admission valve, Fig. 178, be shut 
 
 at the moment lib. of steam has been admitted into 
 the cylinder, and that the volume through which the 
 piston can sweep in the cylinder is r* x v, the steam 
 would then be able to move the piston from position 
 "1" to position "2" while expanding, and thus do an 
 additional amount of work by the expenditure of a 
 quantity of heat, which otherwise would be wasted in 
 the cod denser. The work done during expansion 
 can be determined if we know the form of the expansion 
 curve, which, experience tells us, does not differ much 
 from a rectangular hyperbola. We can now construct 
 the expansion curve in the manner shown in Fig. 172, 
 and the work done during expansion, being represented 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 uo 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ;tn 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 , 5 , 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 m- 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 / 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 / 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 / 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 'f 
 
 
 
 
 / 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 / 
 
 
 
 
 
 
 
 ' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 i 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 t 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 *v 
 
 
 // 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 „ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 * 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 | 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 P 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 1 
 
 )•> 
 
 i 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 
 
 
 t — i 
 
 r-- 
 
 » 
 
 • 
 
 
 ■>■ . 
 
 1 ■ 
 
 — 1 
 
 fcrr; 
 
 e— 7 
 
 c / 
 
 ■' -* 
 
 > / 
 
 1 
 
 Fig. 179. 
 
 by area B C D E B, can be calculated by means of 
 formula (141). Let now P/j denote the initial forward 
 pressure, which is represented by O A and E B, then 
 the total energy exerted by the lib. of steam during 
 the forward stroke will be 
 
 ~Pji x-v + Pftxvx log e r e =P/i x v 
 
 [l + log e r e ] foot-pounds . . . (149) 
 
 The work to be overcome by the piston during the 
 backward stroke will be 
 
 P& x r e x v foot-pounds. . . . (150) 
 
176 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Table A. 
 
 No. 
 
 Pa 
 
 * 
 
 Pi 
 
 £ 
 
 r e 
 
 IW 
 
 H 
 
 S 
 
 n 
 
 T 2 -T\ 
 
 T 2 
 
 1 
 
 50 
 
 280'9 
 
 2 
 
 126-3 
 
 1 
 
 57694 
 
 1073 
 
 34-3 
 
 0-069 
 
 0-208 
 
 2 
 
 JJ 
 
 J J 
 
 15 
 
 213 
 
 1 
 
 42066 
 
 987 
 
 47-1 
 
 0-055 
 
 0-091 
 
 3 
 
 JJ 
 
 jj 
 
 2 
 
 126-3 
 
 3 
 
 118909 
 
 1073 
 
 16-7 
 
 0-143 
 
 0-208 
 
 4 
 
 JJ 
 
 jj 
 
 15 
 
 213 
 
 3 
 
 72022 
 
 987 
 
 27-5 
 
 0-094 
 
 0-091 
 
 5 
 
 J J 
 
 jj 
 
 2 
 
 126-3 
 
 5 
 
 144714 
 
 1073 
 
 13-7 
 
 0-174 
 
 0-208 
 
 6 
 
 JJ 
 
 j j 
 
 10 
 
 193-2 
 
 5 
 
 96634 
 
 1006 
 
 20-5 
 
 0-124 
 
 0-118 
 
 7 
 
 JJ 
 
 jj 
 
 2 
 
 126-3 
 
 9 
 
 170510 
 
 1073 
 
 11-6 
 
 0-206 
 
 0-208 
 
 8 
 
 JJ 
 
 j j 
 
 2 
 
 126-3 
 
 25 
 
 193449 
 
 1073 
 
 10-2 
 
 0-233 
 
 0-208 
 
 9 
 
 200 
 
 381-6 
 
 2 
 
 126:3 
 
 1 
 
 64722 
 
 1104 
 
 30-6 
 
 0-076 
 
 0-303 
 
 10 
 
 JJ 
 
 JJ 
 
 15 
 
 213 
 
 1 
 
 60473 
 
 1017 
 
 32-7 
 
 0-077 
 
 0-200 
 
 11 
 
 JJ 
 
 JJ 
 
 2 
 
 126-3 
 
 3 
 
 135237 
 
 1104 
 
 14-6 
 
 0-158 
 
 0-303 
 
 12 
 
 JJ 
 
 J J 
 
 15 
 
 213 
 
 3 
 
 122488 
 
 1017 
 
 16-2 
 
 0-156 
 
 0-200 
 
 13 
 
 JJ 
 
 JJ 
 
 2 
 
 126-3 
 
 5 
 
 167323 
 
 1104 
 
 11-8 
 
 0-196 
 
 0-303 
 
 14 
 
 JJ 
 
 JJ 
 
 15 
 
 213 
 
 5 
 
 146076 
 
 1017 
 
 13-6 
 
 0-186 
 
 0-200 
 
 15 
 
 JJ 
 
 J J 
 
 2 
 
 126-3 
 
 9 
 
 203137 
 
 1104 
 
 9-75 
 
 0-238 
 
 0-303 
 
 16 
 
 JJ 
 
 )) 
 
 15 
 
 213 
 
 9 
 
 164890 
 
 1017 
 
 12 
 
 0-210 
 
 0-200 
 
 17 
 
 J> 
 
 JJ 
 
 2 
 
 126-3 
 
 25 
 
 259452 
 
 1104 
 
 7-6 
 
 0-304 
 
 0-303 
 
 18 
 
 JJ 
 
 JJ 
 
 8 
 
 182-9 
 
 25 
 
 210420 
 
 1047 
 
 9-4 
 
 0-260 
 
 0-236 
 
 19 
 
 JJ 
 
 JJ 
 
 2 
 
 126-3 
 
 27 
 
 263189 
 
 1104 
 
 7-5 
 
 0-308 
 
 0-303 
 
 20 
 
 JJ 
 
 JJ 
 
 1-6 
 
 116-6 
 
 125 
 
 315640 
 
 1114 
 
 6-3 
 
 0-367 
 
 0-315 
 
 Consequently, the indicated work done during a 
 double stroke of the piston will be 
 
 I W - T fi x v [1 + log, r e J - P 6 xvxr e foot-pounds (1 51) 
 
 We can also express the indicated work by the mean 
 forward pressure, which is 
 
 p ^x.tl + log ^ x (1 + log.r.) 
 
 r e x v r e 
 
 and we obtain 
 
 I W = [P/m-Ps] x r e x v foot-pounds . (153) 
 
 To illustrate the advantage of working the engine by 
 expansion, the two curves A and B, Pig. 179, have 
 been plotted. The ordinates represent values of 
 formula (149) for various values of r e , which latter are 
 represented by the abscissa. Curve A is calculated for 
 p a , equal to 2001b. per square inch, whereas in the case of 
 curve B, p a is only 501b. per square inch. As the 
 ratio of expansion cannot be less than one— i.e., steam 
 working without expansion as in the engine, Fig. 175— 
 the curves start at this value of r e . To save space, zero 
 of ordinates is 60,000 foot-pounds. The curves clearly 
 show the great advantage gained by the expansive 
 working of the steam. 
 
 We may now calculate the amount of steam, S, in 
 pounds, which is required for producing one indicated 
 horse-power hour, knowing that the indicated work 
 produced by lib. of steam in the ideal engine, Fig. 178, 
 can be found by formula (153). It is evident that we* 
 must have 
 
 c, 60 x 33,000 
 
 b = — j^= pounds 
 
 •(154) 
 
 Table A (given above) has been calculated for the 
 purpose of showing how the performance of an ideal 
 single-acting steam engine varies with the ratio of 
 
 expansion, temperature of the condenser, and initial 
 pressure. 
 
 In calculating the quantity of heat, H, required for 
 turning lib. of water into steam, the temperature of 
 the feed- water has been taken to be the same as that of 
 the condenser. The table also enables the reader to 
 compare the efficiency, tj, of the engine with that of a 
 Carnot's imaginary engine. 
 
 The discussion of the figures given in the table will, 
 however, be deferred to a more convenient place in the 
 book. 
 
 Compound Engines. 
 
 The temperature of the cylinder, Fig. 178, will 
 undergo a series of changes during a stroke, on account 
 of the temperature of the steam falling while expanding. 
 The cylinder will therefore be colder than the entering 
 steam, which will cause some of the latter to condense 
 during admission, and so much steam will be condensed 
 that the latent heat given off by condensation will be 
 sufficient to raise the temperature of the walls of the 
 cylinder to that of steam. The heat thus received will 
 again be given off by the cylinder during expansion, 
 and it is evident that the greater the ratio of expansion 
 is, the lower will the temperature of the cylinder be at 
 the end of the stroke. For this reason it is not prac- 
 tical, or rather not economical, to drive the ratio of 
 expansion too far in one cylinder ; the expansion may, 
 however, be continued in two, three, or four cylinders ; 
 the temperature of each cylinder will be lower than that 
 of the one in front, but the variation of the temperature 
 of any one cylinder during a stroke will be comparatively 
 small. 
 
 By a " Compound Engine " is understood an engine 
 which has two cylinders. The steam is passed direct 
 from the boiler into the first cylinder, where it is 
 
THEORY OF STEAM ENGINES. 
 
 177 
 
 moderately expanded, say, three to five times, and is 
 then exhausted into the second cylinder : here it is 
 further expanded three to five times, and is finally 
 exhausted into the condenser. 
 
 Such an engine may either be a " Woolf s Engine," 
 in which the second cylinder may be considered as a 
 simple continuation of the first one, or a " Eeceiver 
 Engine," in which case the steam from the first cylinder 
 is exhausted into a reservoir, called the " Eeceiver," 
 from which again the second cylinder receives steam 
 when convenient. 
 
 Woolf s Engine. — Eig. 180 represents an ideal engine 
 of this class, the cylinders are placed side by side, and 
 the angle between the cranks is 180 deg. ; the piston 
 of the one cylinder will therefore be at the crank end 
 of the cylinder at the same time as the piston of the 
 other cylinder is at the bottom. The length of stroke 
 is the same for both cylinders, but the diameter of the 
 second must evidently be larger than that of the first, 
 as the volume of the steam contained in the first 
 cylinder must be augmented in the second before 
 being exhausted into the condenser. 
 
 H PC 
 
 L P C 
 
 Fig. 180. 
 
 The first cylinder, HPC, called the high-pressure 
 cylinder, receives the steam which is cut off when the 
 piston has moved through a portion of the stroke ; it 
 then expands during the rest of the stroke, the ratio 
 of expansion being r e 1 . The ordinates of the diagram 
 oabcf represent the absolute pressure exerted by the 
 steam during the stroke; a b being the admission line, b c 
 the expansion curve, and of being the absolute vacuum 
 line, or the line of no pressure. The final pressure in 
 the high-pressure cylinder is cf, at which pressure the 
 steam enters the second cylinder, L P C, which is called 
 the low-pressure cylinder. The ordinates to curve h d 
 represent the pressures of the steam during the stroke 
 of the low-pressure piston ; o 1 e is the absolute vacuum 
 line, and the ordinate to the straight line m n, which is 
 parallel to o 1 e, represents the pressure in the con- 
 denser, which will also be the pressure on the low- 
 pressure piston, during the return stroke. Area 
 mhdnm will therefore represent the indicated work 
 done in the low-pressure cylinder during a double 
 
 stroke. 
 
 As the intensity of the pressure on the high-pressure 
 piston, during the return stroke of this piston, must be 
 exactly the same as that on the low-pressure piston at 
 the same moment, it is evident that the back-pressure 
 curve for the high-pressure cylinder must be the same 
 as the expansion curve for low-pressure cylinder — that 
 
 is, curve c g is the same as curve h d, h o 1 being equal 
 to cf, and d e likewise being equal to o g. The indi- 
 cated work done in the high-pressure cylinder will 
 therefore be represented by area g abc g. 
 
 Let the problem be to determine the indicated work 
 done by lib. of steam in this engine at an initial 
 pressure of P«, the pressure in the condenser being 
 P& per square foot. The ratio of expansion in the 
 first cylinder is r e l , and that in the second cylinder is 
 r e ". Let, further, the common stroke of the two 
 pistons be denoted by , I, the diameter of the small 
 cylinder by d, and that of the large cylinder by D, then 
 it is evident that the volume swept by the small piston 
 in each stroke will be 
 
 4 
 
 v x r e * = I x 
 
 and that swept by the large piston will be 
 
 v x r e 1 x r e " = I x 
 
 D 
 
 D 2 
 
 from which follows, that — is equal to *jr& 
 
 0/ 
 
 .i.e., "the 
 
 ratio of the diameter of the large cylinder to that of 
 the small cylinder, is equal to the square root of the 
 ratio of expansion in the large cylinder." 
 
 In order to understand clearly the behaviour of the 
 steam in the engine, we will imagine that we have a 
 low-pressure cylinder with a diameter, d, instead of 
 large D, the stroke of the piston being L instead of I, 
 where the relation between the two latter must be 
 
 Lx: 
 
 ■d? 7 ttD 2 L D 2 
 
 = l x — _ or — = — = r 
 4 4 I d 2 
 
 Place now the long cylinder on the top of the 
 small cylinder, as shown in Fig. 181. The dis- 
 tance between the high-pressure piston, " 1 ", and 
 the low-pressure piston, "2," will be I at the moment 
 that the steam begins to enter the low-pressure 
 cylinder ; the steam is thus confined between the two 
 pistons, which will now begin to move in the same 
 direction ; but the low-pressure piston will move with 
 a velocity which bears to the velocity of the high- 
 pressure piston as — , but relative to the high-pressure 
 
 b 
 
 piston, the velocity of the low-pressure piston is equal 
 to the difference of the actual velocities of the two 
 pistons. At the end of the stroke, the high-pressure 
 piston has moved through a distance I, and the low- 
 pressure piston will be on position "4", the steam still 
 being confined between the two pistons, but its volume 
 having been increased to v x r t x x r e " . As far as the 
 effect on the steam is concerned, it would be precisely 
 the same as if the high-pressure piston had not moved 
 at all, and the low-pressure piston had traversed the 
 distance L from position " 1 " to position "3"; or, in 
 other words, " the expansion of steam in a Woolf's 
 engine takes place as if there were only one cylinder, 
 of diameter d, the stroke being L ; or, with a diameter 
 D, the stroke being I." 
 
178 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 In Fig. 181 a b is the admission line, the same as in 
 Fig. 180; the curve bed, being a hyperbola, is the 
 expansion curve of the steam, the ratio of expansion 
 being r e 1 x r„ ". 
 
 Let A 1 denote area o a befo, and B, A", C 1 and C" 
 areas fc def,ohdeo, og cfo, and o m n e o respectively, 
 then we find that the indicated work done in the first 
 cylinder is 
 
 PW^AJ-C 1 , 
 
 and that done in the second cylinder is 
 
 I"W" = A"-C", 
 and consequently the total indicated work done during 
 a double stroke of the pistons, if the engine is single 
 acting, will be 
 
 I W = A 1 - C 1 + A" - C" 
 
 But for lib. of steam we have evidently 
 
 A 1 = P/i x v x (1 + loge r e *) 
 
 B = P/i x v x log e r e " 
 
 C"=Pb * v x r e l x r e " 
 As curve h d is curve c d stretched out, we must have 
 
 and as curve g c is c d compressed, we must have 
 
 I 
 
 C' = Bx 
 
 L-r 
 
 The indicated work done by lib. of steam in Woolf s 
 engine will therefore be 
 
 I W = P/» x v x (1 + log« f.'xr.T-Pj xvxr e 1 xr e ", 
 
 which is precisely the same as in an engine with 
 one. cylinder of the size of the low-pressure cylinder. 
 
 Fig. 181. 
 
 Receiver Engine.— -When the angle between the 
 cranks of the two cylinders has to be different to deg. 
 or 180 deg., it is evident that the steam from the H.P. 
 cylinder could not be exhausted direct into the L.P. 
 cylinder, as in the Woolf's engine, but that the two 
 cylinders in this respect must be independent of each 
 other. The H.P. cylinder must therefore exhaust into 
 a chamber, the receiver, in which the pressure should 
 be constant and equal to the final pressure in the H.P. 
 cylinder. The L.P. cylinder receives steam from the 
 receiver at the beginning of the forward stroke, and the 
 cut-off takes place when the cylinder has received 
 exactly the same quantity of steam as the H.P. 
 cylinder exhausted into the receiver in one stroke. At 
 
 this moment, the expansion of the steam begins in the 
 L.P. cylinder, and is continued until the end of the 
 stroke. In the return stroke, the L.P. piston expels 
 the steam into the condenser. 
 
 In Fig. 182 are illustrated the two cylinders of an ideal 
 receiver engine. The H.P. piston has just finished a 
 stroke, whereas the L.P. piston has only moved through 
 about half its stroke, the cranks being about 90 deg. 
 apart. Using the same letters as before, the volume 
 swept by the H.P. piston in one stroke will be 
 
 , 7T # 
 
 I x — — = r e 
 
 xv, and that swept by the L.P. piston 
 
 will be lx* 
 
 D 
 
 r e ' xr e " xv ox -= Vr„ " just the same 
 4 d e ' 
 
 as for the Woolf's engine. In the diagram, o a is the 
 
 H PC 
 
 L 
 
 p c 
 
 ! t J\ 
 
 X///////////: 
 
 ; 
 
 s 
 s 
 
 
 
 assssssv: 
 
 Fig. 182. 
 
 initial pressure of the steam, V f i, a b the steam line, and 
 be the expansion curve for the steam, while in the H.P. 
 
 P 
 
 cylinder, and cf represents the final pressure, — t -L i in 
 
 this cylinder. The performance of the steam in the 
 H.P. cylinder is so far the same as in Fig. 180, but as the 
 pressure in the receiver is constant, equal to cf, the 
 back-pressure curve will be the straight line, eg, 
 parallel to the absolute vacuum line, of. The indicated 
 work done in the single-acting H.P. cylinder during 
 a double stroke will therefore be represented by area 
 gab eg. 
 
 The performance of the steam in the L.P. cylinder is 
 somewhat different to what we saw took place in the 
 Woolf's engine. The steam is admitted from the 
 receiver at a pressure o' g' = og = cf, and is cut off at the 
 moment the L.P. piston has described a volume equal 
 to r e ' xv; the length of the steam line, g' c', must 
 
 therefore be equal to — -; for the remainder of the stroke 
 
 the steam expands, the pressure being proportional to 
 the ordinates of curve c' h. The pressure in the con- 
 denser is represented by en = dm, and the back- 
 pressure line will therefore be the straight line mn 
 parallel to o' e, the absolute vacuum line. The indi- 
 cated work done during one double stroke of the L.P. 
 piston will therefore be represented by area m g' h n m. 
 A clearer understanding of the performance of a 
 receiver engine may be gathered from Fig. 183, in 
 which E is the receiver on which the two cylinders are 
 placed. The H.P. cylinder is precisely the same as in 
 Fig. 182, but the diameter of the L.P. cylinder is d, 
 the same as that of the H.P. cylinder ; the stroke, L, 
 must therefore be equal to r c " x I. 
 
THEORY OF STEAM ENGINES. 
 
 179 
 
 The steam from the H.P. cylinder is exhausted into 
 the receiver, whereas communication between the latter 
 aDd the L.P. cylinder is cut off by means of the valve 
 
 LPC 
 
 Fig. 183. 
 
 V. The steam line, g' c', is equal to the line g c. As 
 the pressure c'/' i s * ne same as cf, it is evident that 
 
 were admitted direct into the L.P. cylinder and there 
 expanded at the ratio r e ' x r e ". The total indicated 
 work done by lib. of steam in this engine will be 
 IW=P /i X'ux(l + log,, r«'xf/)-PjXDxr f 'xr/ 
 which is the same as in the Woolfs engine. 
 Hyperbolic Expansion of Steam. 
 
 In working out the figures given in Table A, on page 
 176, for the ideal steam engine with expansion, Fig. 178, 
 it has been considered, as is usually done in text-books, 
 that the total heat spent for the performance of lib. of 
 steam in the cylinder, is that required for turning lib. 
 of water from the temperature of the condenser into 
 dry saturated steam. In cases where the expansion of 
 the steam has been carried on until the final pressure 
 of the steam was equal to, or very nearly equal to, that 
 of the condenser, the table shows that the efficiency of 
 the engine is greater than that of the Carnot's engine, 
 working between the same temperatures. This, how- 
 ever, is absolutely impossible, as proved at the end of last 
 chapter. It is true that if the final pressure of the 
 steam is equal to that in the condenser, we are then 
 approaching the imaginary perfect engine, but the effi- 
 ciency of the steam engine could never be higher than 
 that of Carnot's engine. The table therefore clearly 
 shows that we have not accounted for all the heat 
 which is necessary for performing hyperbolic expansion 
 of steam. 
 
 So — To — i so* Jo i' l ' °___']f_ ^° ,jlf_„'L_flL. -° J°^^jj^^S S 
 
 Fig. 184. 
 
 expansion curve c' h is simply a continuation of the Let us first examine the case when the steam remains 
 hyperbola be; the performance of the steam in a dry saturated steam during expansion ; in this case 
 receiver engine is therefore precisely the same as if it the relation between volume, pressure, and tempera- 
 
180 
 
 PRACTICAL ELECTRICAL ENGINEERWG. 
 
 ture is given by Eegnault's table. It has been shown 
 that the latent heat of lib. of dry saturated steam, at a 
 lower temperature, is greater than that at a higher tem- 
 perature, and consequently, by increasing the volume 
 of the steam from v to v x and still keeping it dry and 
 saturated, we must add a quantity of heat which is 
 equal to the difference of the latent heat at tempera- 
 ture t v corresponding to v x , and the latent heat at 
 temperature t, corresponding to v. 
 
 This condition is clearly shown in Pig. 184, where 
 pressures in pounds per square inch are measured along 
 the axis of abscissae, and British heat units along the 
 axis of ordinates. The curve is drawn so as to show 
 the increase of latent heat of lib. of dry saturated 
 steam from p a = 200, to any lower pressure above 
 
 point b ; and if the initial pressure had been 501b., and 
 the final volume of the steam were 120 cubic feet, then 
 c d would be the curve of expansion. 
 
 Curve B in the same drawing is part of the rect- 
 angular hyperbola for p a = 200, and would be the expan- 
 sion curve, if the initial pressure were 2001b., and the 
 expansion were hyperbolic, as assumed in the ideal 
 engine, Fig. 178. As the two curves A and B run 
 very close together at high pressures, they have not 
 been continued further than shown in the diagram. 
 
 It will be seen that the ordinates to the hyperbola 
 are longer than the corresponding ones of the satura- 
 tion curve ; therefore the energy exerted by the 
 steam during hyperbolic expansion must be greater 
 than when the steam remains saturated. For this 
 
 Fig. 185. 
 
 p a = 50. If, for example, the initial pressure be 
 p a = 140, and the steam expands until the pressure 
 has fallen to p a = 60, then if the steam is to remain 
 dry saturated, a quantity of heat equal to about 43 
 British heat units must be added gradually to the 
 steam during expansion. 
 
 In Fig. 185, the curve A is plotted by taking the 
 ordinates and abscissae proportional to p a and the corre- 
 sponding specific volume of dry saturated steam ; this 
 curve is called the saturation curve of steam, and will 
 evidently be the expansion curve, when sufficient heat 
 is added during expansion so as to keep the steam at 
 the point of saturation. If, for example, the steam 
 be expanded to a volume of 100 cubic feet before being 
 exhausted, then the expansion curve would end in 
 
 reason, hyperbolic expansion requires an additional 
 quantity of heat, which must at least be equivalent to 
 the area between the two curves B and A. 
 
 If the initial pressure had been 501b., the hyperbola 
 should be constructed through point c, and the area 
 between the latter hyperbola and curve A would repre- 
 sent the quantity of heat to be added to the saturated 
 steam. 
 
 As the equation for the saturation curve is approxi- 
 mate^ 
 
 V a x v Tw = 68,400 
 
 (155) 
 
 where P 0) as before, is the pressure in pounds per 
 square foot, it can be proved that the energy exerted 
 on the piston by the expansion of lib. of dry saturated 
 
THEORY OP STEAM ENGINES. 
 
 181 
 
 Table B. 
 
 No. 
 
 Pa 
 
 Pb 
 
 r 6 
 
 H x 
 
 H + Hj 
 
 IW 
 * 772(H + H!) 
 
 * ♦ 1) 
 
 T 2 -T t 
 T 2 
 
 3 
 
 50 
 
 2 
 
 3 
 
 50-1 
 
 1123 
 
 0-137 
 
 17-43 
 
 0-208 
 
 4 
 
 
 15 
 
 3 
 
 501 
 
 1037 
 
 0-090 
 
 28-88 
 
 0-091 
 
 5 
 
 
 2 
 
 5 
 
 ■ 71-5 
 
 1145 
 
 0-164 
 
 14-58 
 
 0-208 
 
 6 
 
 
 10 
 
 5 
 
 71-5 
 
 1078 
 
 0-116 
 
 21-92 
 
 0-118 
 
 7 
 
 
 2 
 
 9 
 
 96-7 
 
 1170 
 
 0-189 
 
 12-66 
 
 0-208 
 
 8 
 
 
 2 
 
 25 
 
 138-3 
 
 1212 
 
 0-207 
 
 11-54 
 
 0-208 
 
 11 
 
 200 
 
 2 
 
 3 
 
 64-6 
 
 1169 
 
 0-150 
 
 15-50 
 
 0-303 
 
 12 
 
 )> 
 
 15 
 
 3 
 
 64-6 
 
 1082 
 
 0-147 
 
 17-18 
 
 0-200 
 
 13 
 
 
 2 
 
 5 
 
 92-4 
 
 1196 
 
 0-181 
 
 12-82 
 
 0-303 
 
 14 
 
 
 15 
 
 5 
 
 92-4 
 
 1130 
 
 0-170 
 
 14-78 
 
 0-200 
 
 15 
 
 
 2 
 
 9 
 
 123-2 
 
 1227 
 
 0-214 
 
 10-83 
 
 0-303 
 
 16 
 
 
 15 
 
 9 
 
 123-2 
 
 1141 
 
 0-187 
 
 1345 
 
 0-200 
 
 17 
 
 
 2 
 
 25 
 
 174-3 
 
 1278 
 
 0-263 
 
 8-83 
 
 0-303 
 
 18 
 
 
 8 
 
 25 
 
 174-3 
 
 1222 
 
 0-223 
 
 1096 
 
 0-236 
 
 19 
 
 
 2 
 
 27 
 
 177-4 
 
 1281 
 
 0-265 
 
 8-73 
 
 0-303 
 
 20 
 
 ) j 
 
 16 
 
 125 
 
 251 
 
 1365 
 
 0-299 
 
 7-68 
 
 0-314 
 
 steam, remaining dry saturated, will be 
 
 foot-pounds . (156) 
 
 the piston 
 
 1,094,400 16 Jr~ e - 1 
 
 hour would not be S pounds as given in Table A, but 
 H, 
 
 16 / 
 
 vr. 
 energy exerted c 
 
 The corresponding 
 during hyperbolic expansion will be 
 
 P/i x v x loge r e . . . . (157) 
 
 where P# is the initial forward pressure of the steam 
 in pounds per square foot. 
 
 We can now determine the quantity of heat, H 1( in 
 British heat units, which must be added to H in Table 
 A, in order to produce hyperbolic expansion. It is evi- 
 dent that we must have H a equal to the difference of 
 latent heat of lib. of dry saturated steam at volumes 
 r e x v and v plus 
 
 / 1,094,000 "V7, -1 \ 
 rM T fi xvxlog e r e w j- x w j- ) 
 
 The efficiency of the engine as a heat engine will 
 therefore be 
 
 *l = , 
 
 IW 
 
 (158) 
 
 772(H + Hj) 
 
 Let us assume that the wall thickness of the steam- 
 cylinder is very small indeed, and therefore has a very 
 small heat capacity, which may be neglected, and also 
 that the material of which the cylinder is made is a 
 perfect conductor for heat. Let, further, the cylinder 
 be surrounded with a steam-jacket — i.e., a ring-forming 
 space filled with live steam from the boiler. This 
 steam-jacket should be bounded externally by a cylinder 
 concentric to the steam-cylinder and made of a perfect 
 non-conducting material. The jacket should extend 
 over the cylinder-covers. For each pound of steam 
 admitted into the steam-cylinder, the jacket should 
 receive so much steam, that, by the condensation of the 
 latter, the quantity of heat, H l7 would be given off. If, 
 now, the condensed steam leaves the jacket at the same 
 temperature as that of the condenser, then the total 
 quantity of steam required per indicated horse-power 
 
 would be S x 1 + _ J Jpounds. 
 
 The annexed table, B, may be considered as a con- 
 tinuation of Table A, but the quantity of heat, Hj, has 
 been taken into account in working out the efficiency of 
 the engine, as well as the amount of feed-water required 
 per indicated horse-power hour. The two tables should 
 be carefully compared. 
 
 Steam Engines with Clearance. 
 
 In the engines hitherto discussed, it has been assumed 
 that the volume occupied by the steam at the end of 
 a stroke was equal to the volume swept by the piston ; 
 or, in other words, we have assumed that the volume 
 of the steam passages from the regulating valves to the 
 cylinder could be neglected, and we have also assumed 
 that the piston came right against the cover at the end 
 of its journey, so as to leave no space between itself and 
 the cover. 
 
 In actual engines it is, however, necessary to give 
 the steam passages a certain length, and, therefore, also 
 a sufficient cross-section in order to diminish the inter- 
 mediate fall of pressure between the steam-chest and 
 the cylinder, and for the safety of the cylinder, we must 
 allow a play between the piston and the cylinder-cover. 
 
 For these reasons the final volume of the steam in the 
 cylinder will be greater than the volume described by 
 the piston. Let I denote the length of the stroke, and 
 d the diameter of the cylinder, then the volume 
 occupied by the steam at the end of a stroke will be 
 
 V/ = 
 
 ■d? 
 
 1 + 
 
 X 
 
 (159) 
 
 where 
 
 ■d* 
 
 x A is called the " clearance volume " and 
 
 X the " clearance length." 
 
 Single-Cylinder Engine with Clearance. — For the 
 purpose of discussing the performance of the "steam in 
 
182 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 a cylinder with clearance, we can consider, as is done in 
 formula (159), that the length of the cylinder is (Z + X), 
 instead of I. 
 
 The volume (I 4- X) is called the total capacity 
 
 of the cylinder, and volume —r~ x I the effective 
 
 capacity. 
 
 In Fig. 186, the line of is the absolute vacuum line, 
 oh = \, hf= I; the initial volume of the steam before 
 being cut off will therefore be proportional to o g, and the 
 
 l- c 
 
 ■■'/A 
 
 If X 
 
 
 
 Fin. 186. 
 
 hyperbola c d must be constructed as if the piston 
 started at point o. The actual ratio of expansion 
 will be 
 
 IL = 
 
 = °f 
 
 og 
 
 1 + T xr, 
 
 (160) 
 
 because the apparent ratio of expansion, r e> is equal to 
 
 --, and therefore hg = — . The energy exerted by 
 h g r e 
 
 the steam on the piston is represented by area hbc dfh, 
 
 the work done on the back pressure by area h n mfh, 
 
 and the indicated work by area nbcdmn, 
 
 For lib. of steam we would have 
 
 area o ac g o =P/;x v 
 
 area g cdfg = P/,- x v x log e B,, 
 
 and as — =- is equal to 
 
 og 
 
 1 + 7 X ^ 
 
 area hb eg h 
 
 l + 7 xr. 
 
 we will have 
 
 x P /; x v, 
 
 and consequently the energy exerted by lib. of steam on 
 the piston will be 
 
 area hbc df /* = P/j x v 
 
 1 +— x r. 
 
 + loff e 
 
 I 
 
 i + *V 
 
 1 + - x ?\. 
 
 I 
 
 The final volume of the steam is 
 
 V,- =-,- - x (l + \) =K, x = r x — '— 
 
 1 + — x /•,. 
 I 
 
 be 
 
 The effective capacity of the cylinder will therefore 
 
 4 l + \xr e 
 
 consequently 
 
 area h n mfh = P& x vx 
 
 X 
 1 x _ x r e 
 
 v 
 
 The indicated work done by lib. of steam in this engine 
 will thus be 
 
 TW P / 1 , Xljt\ 
 I W =P /; x v x 1 + log c I 
 
 \l + 4 x r e 1 + 4 x r e / 
 
 I 
 minus Pi, x« x 
 
 1 +jxr e 
 
 if 
 
 (161) 
 
 The fraction — may on an average be taken from 0"05 
 
 b 
 
 to 0-07. 
 
 Example 1. — Take —j- = 0'07 and the apparentratio 
 
 of expansion r e = 3 ; the indicated work per pound of 
 steam in the cylinder will then be, according to (161), 
 
 IW= (1-802 x F fi - 2-479 x T b ) x v foot-pounds. 
 
 The actual ratio of expansion will be 
 
 E e = 2-65. 
 
 Example 2. — Take -y- = - 07 as before, but make 
 
 r e = 5 ; the indicated work per pound of steam in the 
 cylinder will be 
 
 I W= (2-118 x P,j - 3-704 x P 6 ) x o fgot-pounds, 
 and we will have E e = 3 '96. 
 
 Woolf's Engine with Clearance. — "We have seen that 
 the result of the performance of the steam in a "Woolf's 
 engine without clearance is the same as if the steam were 
 admitted direct into the L.P. cylinder, and there ex- 
 panded. But if the cylinders have clearance, then the 
 final volume of the steam in the L.P. cylinder will be 
 equal to the total capacity of the L.P cylinder, 
 including the clearance of the intermediate space 
 between the cylinders plus the clearance of the H.P. 
 cylinder. The initial pressure in the L.P. cylinder 
 will therefore not be equal to the final pressure in the 
 H.P. cylinder, but there will be a fall of pressure, due to 
 the expansion of the steam, in the clearance of the 
 L.P. cylinder and the intermediate space between the 
 cylinders. The back-pressure curve in the diagram of 
 the H.P. cylinder will start at a point somewhat lower 
 than point c, Fig. 180, and will also end at a point 
 lower than g. The expansion curve in the L.P. 
 cylinder will be the same as the compression curve in 
 H.P. cylinder. 
 
 If we now imagine the L.P. cylinder to be of 
 diameter d, and the stroke to be L = I x ; precisely 
 
THEORY OF STEAM ENGINES. 
 
 183 
 
 as in Fig. 181, then the final volume of the steam 
 will be 
 
 we have also 
 
 V, = ?4- [V + * + L ] 
 
 (162) 
 
 where X' is the clearance length of the small cylinder, 
 and X" that of the large cylinder, including the inter- 
 mediate space between the cylinders. 
 
 Fig. 187. 
 
 In diagram Fig. 187, the line of is the absolute 
 vacuum line, and the length of of is equal to 
 (X' + X" + L), the diameter of the cylinder being d. 
 We have, further, oh = X', hf = I, fti = X", and 
 h'f - L - I ; the line o a is, therefore, the clearance 
 line. 
 
 It can be proved, by a process of calculations similar to 
 that used with diagram Fig. 181, that the total indicated 
 work of the engine will be the same as if the steam were 
 admitted at once into the L.P. cylinder, filling 
 
 up the space X' x — — and then pushing the piston 
 
 through distance h g, the steam being cut off when the 
 piston arrives at g. From this position of the piston 
 and until it arrives at/, the steam expands hyperbolically, 
 
 the actual ratio of expansion being -£. A further 
 
 ex- 
 
 °9 
 
 pansion of the steam takes place in the space 
 X" x - — without doing work on the piston, the pres- 
 sure falling from fd to hi c'. During the remainder of 
 the stroke the energy exerted upon the piston will be 
 measured by area h' c' 61 f hi, the ratio of expansion 
 
 being ^,. The total indicated work during one 
 oh 
 
 stroke will thus be measured by area nb c dm n plus 
 
 area nl c' 61 ml ri. 
 
 For lib. of steam we will have 
 
 area o a c 61 f o = P/; x v x (1 + log,,^- 
 
 V off 
 
 ri 2 
 
 = ¥fi x v x 
 
 X' 
 T *r e 
 
 area o a bho = P/i x X' x = ±Vi - 
 
 4 -1+^xr/ 
 
 I 
 
 and 
 
 area / d c' h' f = P/i 
 
 oh' 
 
 i on, -p. 
 v x log c —- =V/i xv 
 
 x log. 
 
 1 + il. 
 
 The work done on the back pressure during the 
 return stroke will be 
 
 v x r e ' 
 4 
 
 I 
 
 P„ x 
 
 I + X' x n ' 
 
 = P;, x v x 
 
 1 + ~ x r e • 
 
 
 
 The indicated work done by lib. of steam will there- 
 fore be expressed by 
 
 I W = P /f x v x 
 
 1 
 
 , . X' 
 
 1 + T X '' 
 
 (t+T + tH 1 *! 1 )**' 
 
 + l0g e - -^ 
 
 I~*')"( 1+ T + t) 
 
 - P(, X V X 
 
 T x r ° 
 
 1 ^ * 
 
 1 + T x r e 
 
 foot-pounds. (163) 
 
 P/i x v x ( 1 + log 
 
 (L+X' + X")xr/ 
 
 I + X' x r, 
 
 P/i X D X ( 1 + log e 
 
 /■L X 
 
 X' X" 
 
 7 + 7 
 
 xn. 
 
 1 + ~xr« 
 
 6 
 
 Example 1. — Take ^ = Jl = 0-07, r e ' = 3, and— = J^ 
 I L Z cZ 2 
 
 = 3, whereby the effective capacity of the second 
 
 cylinder will be equal to three times that of the first 
 
 one. The indicated work per pound of steam will be 
 
 I W=.(2'744 x P^ - 7-438 x V b ) x v foot-pounds. 
 
 The total actual ratio of expansion will be 
 
 E c =813. 
 
 Example 2. — Take T = — = 0-07 as before, but make 
 
 L D 2 
 
 r e = 5, and-- = __=5. We shall then have 
 I d l 
 
 I W= (3'456 x P/i - 18-518 x P 6 ) x v foot-pounds, 
 and E g = 20-81. 
 
 Receiver Engine with Clearance. — Fig. 188 is a 
 diagram of a receiver engine with clearance, the cylin- 
 ders being placed on the top of the receiver, E, as in 
 Fig. 183, and to make the performance of the steam 
 clear, the diameter of the second cylinder is made 
 equal to that of the first. 
 
184 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 1. First cylinder. 
 
 Trie stroke is hf = I ; oh = X'; r e '= -^ ; the actual 
 
 kg 
 
 , of_\ lj_ 
 
 the pressure of the steam in the second cylinder is to 
 remain equal to df during admission, the steam must 
 be cut off when the L.P. piston has moved through 
 a distance h' g' = I - X". It is also evident that it is 
 
 ■a 2 
 
 ■6? 
 
 ratio of expansion is E e ' = — — 
 
 t 
 
 The final pressure in the cylinder will be df ■■ 
 
 E<. ' 
 
 volume I x 1LH. and not volume (I + X') x ?-%- which 
 4 4 
 
 will be augmented in the second cylinder, and the 
 
 hyperbola c' d' can, consequently, not be a continuation 
 
 of curve c d. 
 
 and this pressure being the same as that of the receiver, 
 the line h d, parallel to the absolute vacuum line of, 
 will be the back-pressure line. The indicated work 
 will thus be represented by the area kbcdk. The 
 piston will expel into the receiver, during the backward 
 
 stroke, a volume of steam equal to I x _ — , at a pres- 
 
 4 
 
 Pfi 
 
 sure ft , ; the clearance will thus retain a volume 
 
 He 
 
 ,72 
 
 X ' x — _ of steam at the same pressure. This quan- 
 tity of steam will be considered as wasted in the 
 following calculations. 
 For lib. of steam we have 
 
 / ( 1 + T>''' 
 
 oacdfo<=~Pfi x v x I 1 + log„ 
 
 \ 1 + -. 
 
 area 
 
 I 
 
 x r. 
 
 area oabho = \'x — - — x Py £ = ~P/i x v x 
 
 
 For lib. of steam admitted into the H.P. cyhnder, 
 we shall have 
 
 area o'a'c'd'fo' =Sl*xZx^x(l+ log e E/) = 
 
 Ee 4 
 
 p/i 
 
 area h k dfh = -~^- x I x 
 
 J E«' 4 
 
 - P/i x u x . 
 
 1 + 
 
 1 / 
 
 """IT* \ 
 
 i + 
 
 l + log* — y 
 
 '"> 
 
 Z 
 
 The indicated work per pound of steam will con- 
 sequently be 
 
 1 1 
 
 area o' a! V h' o' = X" x JLOl. x^; = p fi x „ x 
 
 4 Ee' 
 
 X" 
 
 I 
 
 1 + 
 
 X" 
 
 rw=p«x»x 
 
 i + *»,.■ "i + 4 + 
 
 ( 1 + t)-' \ 
 
 2. Second cylinder. The stroke is 
 
 A'/ = L; o'h' = \"; r e ° = %£,= l „ 
 
 L X r' 
 
 area &' ra ra /' A' = P 6 x ^ — x L = R x » x 
 
 4 , X' 
 
 l + y xr e 
 
 The indicated work will thus be 
 
 I" W" = P,* x v x _L_ x ( 1 - *" + l ogc 
 
 . X \ I I 
 
 lx' 
 
 Z 
 
 _X" 
 1 + L 
 
 *'?' 
 
 P& x v x 
 
 x r c 
 
 and the actual ratio of expansion is 
 E, 
 
 1 + ^ 
 o' /'_ L 
 
 1 + T X *' 
 
 The final pressure of the steam in the second cylinder 
 will be 
 
 o 9 
 
 _Z_ 
 L 
 
 X' 
 
 As the H.P. piston expels during each stroke a 
 r_d 
 "4 
 
 i^x.^' = P„x Lll 
 
 E «' °' f (i + x T y 
 
 L 
 
 1 + 
 
 X" 
 
 volume I x "?—— cubic feet of steam, it follows that if 
 
 P/.- 
 
 E« ' x E e 
 
THEORY OF STEAM ENGINES. 
 
 185 
 
 Table C. 
 
 — 
 
 No. 
 
 Pa 
 
 i?6 
 
 ?V or r e ' 
 
 lie 
 
 IW 
 
 Hi 
 
 s (^l) 
 
 IW 
 n 772(H + H X ) 
 
 T 2 
 
 
 1 
 
 50 
 
 2 
 
 3 
 
 2-65 
 
 102338 
 
 45 
 
 20-15 
 
 0-118 
 
 0-208 
 
 <D 
 
 2 
 
 
 15 
 
 3 
 
 2-65 
 
 63602 
 
 45 
 
 32-55 
 
 0-080 
 
 0-091 
 
 £ 6 
 
 3 
 
 
 2 
 
 5 
 
 3-96 
 
 118384 
 
 64-5 
 
 17-73 
 
 0-135 
 
 0-208 
 
 %,.s 
 
 4 
 
 
 10 
 
 5 
 
 3-96 
 
 82768 
 
 64-5 
 
 25-46 
 
 o-ioo 
 
 0118 
 
 ° be 
 
 5 
 
 200 
 
 >2 
 
 3 
 
 2-65 
 
 116187 
 
 59-6 
 
 18-34 
 
 0-129 
 
 0-303 
 
 U^ 
 
 6 
 
 
 15 
 
 3 
 
 2-65 
 
 105653 
 
 59-6 
 
 19-84 
 
 0-127 
 
 0-200 
 
 a 
 
 7 
 
 
 2 
 
 5 
 
 3-96 
 
 136045 
 
 82-4 
 
 15-64 
 
 0-149 
 
 0-303 
 
 
 8 
 
 }t 
 
 15 
 
 5 
 
 3-96 
 
 120305 
 
 82-4 
 
 17-78 
 
 0-142 
 
 0-200 
 
 
 9 
 
 50 
 
 2 
 
 3 
 
 8-13 
 
 147030 
 
 93-7 
 
 14-64 
 
 0-163 
 
 0-208 
 
 DO • 
 
 10 
 
 it 
 
 2 
 
 5 
 
 20-81 
 
 163184 
 
 134-4 
 
 13-65 
 
 0-175 
 
 0-208 
 
 o •- 
 
 11 
 
 200 
 
 2 
 
 3 
 
 8-13 
 
 174529 
 
 119-2 
 
 12-57 
 
 0-185 
 
 0-303 
 
 O M 
 
 12 
 
 
 15 
 
 3 
 
 8-13 
 
 142922 
 
 119-2 
 
 15-47 
 
 0-163 
 
 0-200 
 
 13 
 
 
 2 
 
 5 
 
 20-81 
 
 213833 
 
 168-3 
 
 10-61 
 
 0-218 
 
 0-303 
 
 
 14 
 
 JJ 
 
 8 
 
 5 
 
 20-81 
 
 177514 
 
 168-3 
 
 12-95 
 
 0-189 
 
 0-236 
 
 
 15 
 
 50 
 
 2 
 
 3 
 
 8-52 
 
 144127 
 
 95-7 
 
 14-96 
 
 0-160 
 
 0-208 
 
 
 16 
 
 > i 
 
 2 
 
 5 
 
 21-2 
 
 157294 
 
 135-4 
 
 14-17 
 
 0-168 
 
 0-208 
 
 •£ a 
 
 17 
 
 200 
 
 •2 
 
 3 
 
 8-52 
 
 171371 
 
 122-2 
 
 12-83 
 
 0-181 
 
 0-303 
 
 
 18 
 
 >> 
 
 15 
 
 3 
 
 8-52 
 
 139764 
 
 122-2 
 
 15-87 
 
 0-159 
 
 0-200 
 
 Ph m 
 
 19 
 
 j ) 
 
 2 
 
 5 
 
 21-2 
 
 207427 
 
 169-3 
 
 11-00 
 
 0-211 
 
 0-303 
 
 
 20 
 
 )) 
 
 8 
 
 5 
 
 21-2 
 
 171107 
 
 169-3 
 
 1344 
 
 0-182 
 
 0-236 
 
 3. The two cylinders taken together. 
 The total indicated work in the engine per pound of 
 steam will be 
 
 I W = r W + I" W" 
 
 The total actual ratio of expansion, H e , will be equal 
 to the initial pressure of the steam in the first cylinder, 
 divided by the final pressure of the steam in the second 
 cylinder, or 
 
 Re = R c ' x Re " 
 
 Example 1. — Take 
 
 I 
 
 ¥■ =007, r e 
 Li 
 
 :3,and - = — - 
 I a 2 
 
 = 3. The indicated work in the engine per pound of 
 steam will be 
 
 I W= [2-696 x T fi - 7-438 x P 6 ] xv foot-pounds 
 
 The total actual ratio of expansion will be 
 
 R e =8-52 
 
 Example 2. — Take again — - = — = 0'07, but r e ' = 5, and 
 
 I L 
 
 = 5. We shall have 
 
 I ~ d 2 '' 
 
 I W = [3-358 x V fi - 18-518 xP,j x v foot-pounds, 
 and Re = 21'2. 
 
 For the purpose of showing the influence of clearance 
 upon the performance of steam engines, Table C has 
 been added and should be compared with Table B. 
 
 The Steam Indicator. 
 
 We have several times used the expression " indi- 
 cated work," by which we understand the energy 
 
 exerted by the working substance during the forward 
 stroke of the piston in a heat engine, minus the work 
 done on the back pressure on the same side of the 
 piston during the return stroke. In double-acting 
 engines there must be indicated work done on both 
 sides of the piston, and by adding the two we obtain 
 the total indicated work done in the cylinder. 
 
 If the engine makes n revolutions per minute, and 
 the total indicated work be denoted by I W, then the 
 indicated work per minute will be I W x » ; and if 
 I W be given in foot-pounds, then the indicated power 
 expressed in horse-power will be 
 
 IHp = IWxn 
 
 33,000 K ' 
 
 For the purpose of measuring the indicated work of 
 an engine, and at the same time to obtain a picture of 
 the performance of the working substance in the 
 cylinder, we use the " steam indicator." It is evident 
 that this instrument must be able to produce two 
 motions at right angles to one another ; the one motion 
 must be up and down, and must be proportional to the 
 pressure of the wprking substance at each position of 
 the piston ; the other motion must be to and fro, and be 
 proportional to the travelling of the engine piston. 
 These two motions, being performed simultane- 
 ously, will produce the indicator diagram, and 
 the area enclosed within the diagram, will be a 
 measure of the indicated work on the one side of the 
 piston. If the engine be double acting, two indicators 
 are required, one for each side of the engine piston. 
 
186 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 The two essential parts of a steam indicator are the 
 cylinder and the paper-drum ; the former is in direct 
 communication with the engine cylinder, and is fitted 
 with a piston, H, Fig. 192, which is pressed from below 
 by the working substance, and from above by the 
 atmosphere. There is also a spring, N, the one end of 
 which is fixed to the piston, and the other end to the 
 cover of the cylinder. "When the pressures on both 
 sides of the piston are the same, the spring will not be 
 in tension, and the position of the piston will indicate 
 the pressure of the atmosphere. When the pressure 
 of the working substance exceeds that of the atmo- 
 sphere, the piston is driven upwards, and the spring 
 
 way that, within the range of the instrument, the 
 pencil must move in a vertical straight line, and it is also 
 essential that the movement of the pencil should always 
 be proportional to the movement of the indicator piston. 
 In some indicators the latter is not the case, and the 
 pencil therefore does not indicate the true value of the 
 pressure. 
 
 The drum, U, when about to be used, is covered with 
 a piece of paper, called the card, which is held on the 
 drum by two flat steel clips. The drum oscillates 
 about its axis by alternately being pulled in one direc- 
 tion by a cord wrapped round a pulley on the lower 
 part of the drum, and in the other direction by the 
 
 Fig. 189. 
 
 will be compressed in proportion to the difference of 
 the pressures below and above the piston ; the position 
 of the piston will thus indicate the pressure of the 
 working substance above the atmospheric pressure. 
 The pressure of the working substance may be less 
 than that of the atmosphere, in which case the piston 
 will move downwards, thus stretching the spring, but 
 always in proportion to the effective pressure of the 
 atmosphere. 
 
 As the motion of the indicator piston is but small, 
 a set of multiplying levers are attached to the piston 
 rod, the last lever, 8, carrying a pencil at its extreme 
 end. The system of levers must be designed in such a 
 
 tension of a spring, E. The cord is attached to the 
 cross-head of the engine, and the forward and backward 
 motions of the drum will therefore represent the forward 
 and backward strokes of the engine piston. 
 
 When the pencil is brought against the card on the 
 paper-drum, a diagram will be produced, which presents 
 a record of the pressure of the working substance in the 
 engine cylinder at every point of the stroke. 
 
 There are two points of great importance to be con- 
 sidered in designing an indicator which is to give a 
 true diagram. 
 
 The first of these points refers to the inertia of the 
 moving masses of the indicator, in virtue of which the 
 
THEORY OF STEAM ENGINES. 
 
 187 
 
 motion of these parts would be continued beyond what 
 is desired ; thus the position of the pencil will not 
 indicate the true pressure while the pencil is moving. 
 In order to reduce this error, the velocity and the 
 masses of the piston, piston rod, spring, multiplying 
 levers and pencil must be diminished as much as is 
 practicable. The effect of the inertia of the paper- 
 drum and the other parts accompanying the drum in 
 its movements, such as the drum carriage and the 
 drum spring, is to lengthen the diagram, whereby also 
 the spring will be overwound, and in the return stroke 
 the cord may be broken. For slow-speed engines, 
 these defects are not very difficult to overcome, but in 
 order to indicate a high-speed engine with success, only 
 the most improved forms of indicators can be used. 
 
 The other point to be considered is the friction 
 caused by the motion of the various parts. The effect 
 of friction is in the opposite direction to that of 
 inertia, as friction tends to prevent motion. 
 
 Having given a general idea of the construction of 
 the steam indicator, we will proceed to describe some 
 of the most important types which are at present in 
 use. 
 
 The Tabor Indicator. — This instrument is one of the 
 latest and most improved forms of steam indicators, 
 and has been designed for the purpose of indicating 
 high-speed as well as low-speed engines. 
 
 The following description of the Tabor indicator is 
 taken from a paper read by Mr. A. G. Brown before the 
 Sheffield Society of Engineers, on the 15th of March, 
 1890. 
 
 The special peculiarity of the Tabor indicator, Fig. 189, 
 is the means employed to communicate a straight line 
 movement to the pencil. A plate containing a curved 
 slot is fixed in an upright position, and secured to a 
 swivel plate on the cover of the indicator cylinder. This 
 slot serves as a guide, and controls the motion of the 
 pencil bar. A pin is fixed on one side of the pencil 
 bar, which carries a roller, and this is fitted so as to 
 roll freely from end to end of the slot. The position 
 of the slot is so adjusted, and the pin attached to such 
 a point on the pencil bar, that the curve of the slot 
 compensates the tendency of the bar to move in a 
 circular arc, and the end of the bar, which carries the 
 pencil, moves up and down in a straight line, when the 
 roller is moved from one end of the slot to the other. 
 There is thus very little chance for friction in this 
 movement, and the bar and connections are very light, 
 though strong enough for the purpose. It will be 
 noticed that the base of the paper-drum and the steam- 
 cylinder jacket are made in one piece. The steam- 
 cylinder is a straight tube inside the jacket, with an 
 air space around the sides, and attached to the jacket 
 by means of thread cut on the bottom of the cylinder. 
 The cylinder is thus left free to expand or contract 
 without affecting other parts of the instrument. Slots 
 are cut in the top of the cylinder for the insertion of a 
 key for screwing it in or out. Openings through the 
 side of the jacket allow the steam which leaks past the 
 piston to escape. 
 
 The pencil mechanism is carried by a swivel plate 
 fitted to the cylinder-cover, on which it can be freely 
 moved. The pencil movement consists of three pieces — 
 the pencil bar, the back link, and the piston rod link. 
 The two links are parallel to each other in every posi- 
 tion they may assume. The lower pivots of these links 
 and the pencil point are always in the same straight 
 line. If an imaginary link parallel with the pencil bar 
 be supposed to connect the two, the combination 
 would form an exact pantagraph, and would serve the 
 purpose of making the pencil point move in a straight 
 line, but the friction and wear of the pencil movement 
 would be greatly increased. The slot and roller serve 
 the purpose of this imaginary link to much better 
 advantage. 
 
 Fig. 190. 
 
 The connection between the piston and pencil 
 mechanism is by means of a steel piston rod, hollow at 
 the upper end where it passes through the cylinder- 
 cover, but solid below with a reduced diameter, and 
 having a ball formed on the lower end. 
 
 Fig. 191. 
 
 A ball and socket joint forms the connection of this 
 rod with the piston. This prevents the tendency to 
 bind either rod or piston while working, a fault which 
 causes considerable error where solid connections are 
 used. The socket is an independent piece which fits 
 into a square hole in the piston, and is fastened with a 
 thumb nut below. The piston is made very light, and 
 has a number of shallow grooves cut upon the outside 
 to serve as water-packing. 
 
 The springs, Fig. 190, used in this indicator are of the 
 duplex type, being made of two coils of wire fastened 
 
188 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 exactly opposite each other on the fittings. This 
 arrangement equalises the side strain on the spring, 
 and keeps the piston central in the cylinder, avoiding 
 the excessive friction caused by a single-coil spring 
 forcing the piston against the side of the cylinder. 
 
 The maximum pressures to which the springs for this 
 indicator should be subjected, to give good results, are 
 given in the following table : 
 
 English Springs. 
 
 
 Revolutions of Engine. 
 
 Scale 
 
 
 
 
 
 of Spring. 
 
 up to 200 
 
 200 to 400 
 
 400 to 600 
 
 
 
 
 
 per inch. 
 
 Maximum Pressures in 
 
 
 pounds per square inch above atmosphere. 
 
 10 
 
 15 
 
 10 
 
 6 
 
 12 
 
 20 
 
 15 
 
 12 
 
 16 
 
 24 
 
 20 
 
 18 
 
 20 
 
 36 
 
 30 
 
 25 
 
 24 
 
 48 
 
 40 
 
 35 
 
 30 
 
 70 
 
 55 
 
 45 
 
 40 
 
 100 
 
 80 
 
 60 
 
 50 
 
 125 
 
 100 
 
 75 
 
 60 
 
 150 
 
 120 
 
 90 
 
 80 
 
 200 
 
 160 
 
 120 
 
 100 
 
 250 
 
 200 
 
 150 
 
 12*0 
 
 300 
 
 240 
 
 180 
 
 150 
 
 350 
 
 300 
 
 225 
 
 Metric Springs. 
 
 
 Revolutiot s of Engine. 
 
 Scale 
 of Spring. 
 
 up to 200 
 
 200 to 4C0 
 
 400 to 600 
 
 mm. per kilog. 
 
 Maximum Pressures in 
 
 
 kilogs. per square cm above atmosphere. 
 
 3mm. 
 
 20 
 
 17 
 
 13 
 
 4 „ 
 
 16 
 
 13 
 
 10 
 
 5 „ 
 
 13 
 
 10 
 
 8 
 
 6 „ 
 
 10 
 
 8 
 
 6 
 
 8 „ 
 
 8-5 
 
 7 
 
 5 
 
 10 „ 
 
 6-5 
 
 5 
 
 3-5 
 
 15 „ 
 
 3-25 
 
 2-75 
 
 2-3 
 
 20 „ 
 
 2-75 
 
 2- 
 
 1-75 
 
 30 „ 
 
 1-5 
 
 1- 
 
 0-75 
 
 The paper -drum turns on a vertical steel pin secured 
 to the frame of the indicator. The drum, which is very 
 light, has a closed top, with an inside sleeve fitting the 
 steel pin and serving to guide the top. The bottom 
 rests on and is guided by the drum carriage, which has 
 a long bearing on the central pin. Stops are provided on 
 the drum and corresponding openings for them on the 
 carriage, arranged so that tbe position of the drum maybe 
 reversed, in order to take diagrams from the opposite 
 side. The pencil movement is made to correspond, so 
 that in two or three minutes' time the indicator may be 
 changed from a right to a left hand instrument. Steel 
 clips are attached to the drum for holding the paper. 
 
 The one of these is shorter than the other, which 
 makes it more convenient for putting on the paper. 
 
 The drum spring, which furnishes the returning force 
 for the drum, consists of a fiat spiral spring placed in a 
 cavity underneath the drum carriage and encircling the 
 bearing. One end is attached to the frame below, and 
 the other to the carriage. A stop is provided on the 
 frame to prevent the carriage unwinding the spring 
 when released. The tension of the spring may be regu- 
 lated by unscrewing the knurled nut above, which holds 
 the carriage in place, lifting the carriage clear of the 
 stop, and winding and unwinding the springs as may be 
 desired. 
 
 A carrier-pulley at the end of a swinging arm placed 
 below the paper-drum, serves to guide the indicator 
 cord [in any direction. A circular plate, carrying a 
 single-grooved pulley, is mounted in a clamp at the end 
 of the swinging arm, and this plate and clamp are so 
 arranged as to swivel to any position ; while the groove 
 on the pulley, being central with the driving cord at all 
 times, leads the cord in any desired position. 
 
 A ratchet is cut on the edge of the drum carriage, 
 and a pawl is so arranged as to engage in it whenever it 
 is desired to stop the motion of the drum without un- 
 hooking the driving cord. This is a convenience when 
 indicating slow-speed engines, but is of no use at high 
 speeds. 
 
 A simple form of a cord-adjuster is illustrated in 
 Fig. 191, for adjusting the length of the driving cord. 
 The form and method of connecting this attachment to 
 the cord is shown in the figure. A hook on the 
 indicator cord connects with the ring on the end. 
 
 A cock is screwed into the engine cylinder, and the 
 indicator is attached to this cock by means of a coupling 
 having a single thread, see Pig. 189. The indicator is 
 easily connected and fixed in any desired position. 
 
 The pressure of the pencil on the paper is regulated 
 by a screw, which passes through a projection on the 
 guide-plate and strikes against a pivot on the frame. 
 
 The length of the diagrams with slow speeds may be 
 taken as 4in. to 4§in., and will show well-proportioned 
 diagrams ; but as the speeds increase, the length of the 
 diagram should be shortened, to avoid the effects of 
 inertia of the paper-drum. As a guide in this matter, 
 the following lengths of diagrams are recommended 
 when using the Tabor indicator : 
 
 Speeds up 
 
 to 200 revs, per minute 
 
 4 in. long. 
 
 j ) 
 
 300 
 
 rU 
 
 n 
 
 850 
 
 3 
 
 
 400 
 
 • 2J „ 
 
 , , 
 
 450 
 
 ■• 2± „ 
 
 
 500 
 
 2 
 
 
 550 
 
 
 ) ) 
 
 600 
 
 With these lengths, the drum springs will not require 
 alteration or increase of tension, and the diagrams will 
 be well-proportioned ones, as a stronger piston spring 
 must be used with the higher rotative speeds to give 
 accurate cards, therefore both length and height of dia- 
 gram will decrease in about the same proportion. 
 
THEOR Y OF STEAM ENGINES 
 
 189 
 
 The Crosby Indicator. — This instrument, which is 
 illustrated in Pig. 192, is also designed to meet the 
 
 forming the cylinder-head and holding all attachments 
 in place. The sleeve, B B, surrounds the upper part of 
 
 Fig. 192. 
 
 requirements of high-speed engines. The cylinder- 
 cover or cap, A, has a hollow downward projection, 
 through which the piston rod, C, is guided. Its lower 
 
 the cylinder and carries the pencil movement ; it turns 
 freely, and is held in place by the cover. The adjusting 
 
 Fig. 193. 
 
 part is threaded to screw into the spring-Lead, D, and its 
 under part is threaded to screw into the cylinder, thus 
 
 Fig. 194. 
 
 screw, 5, is threaded through B, and when it is in 
 contact with the stop, 6, the pencil point maybe delicately 
 
190 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 adjusted to the paper on the drum. The swivel-head, E, 
 is threaded on its lower half to screw into the piston rod 
 more or less, according to the required height of the 
 atmospheric line on the diagram. Its head is joined to 
 the long link of the pencil movement. 
 
 The piston rod, C, is made of steel, and is hollow and 
 threaded inside to receive the swivel-head, B. Its 
 movement is guided through the hole in the cover, A. 
 Near the middle is set a transverse steel pin, I, with 
 ends projecting slightly to engage the slots of a hollow 
 
 end of the spring with its head to drop to its bearing 
 on the steel screw, G, which is closely threaded into 
 the lower part of the socket. 
 
 The spring, N, Fig. 193, is made of a single piece of 
 steel wire, wound from the middle into a double coil, the 
 ends of which are screwed into a head, D, with four radial 
 wings, having spirally drilled holes to receive and hold 
 them securely in place. The makers adjust the strength 
 of the spring, by screwing the spring in or out of the 
 head until it is of the right strength ; it is then securely 
 fastened in the head. The head is threaded inside to 
 screw on to the lower projection of the cover, and up 
 firmly against it. The foot of the spring, where the 
 motion is greatest, and therefore lightness is essential, is 
 a small steel bead, 4, fastened on thq transverse portion 
 
 Fig 195. 
 
 wrench and to screw the piston rod into the piston, H. 
 Near the lower end is a shoulder, 2 ; in its under side 
 is cut an annular channel, to receive the top edge of the 
 slotted socket of the piston, when the piston rod is 
 screwed down to the shoulder. Its lower end is concave 
 to fit the head, 4, of the spring on which it rests. 
 
 The piston, H, has a transverse web, supporting a 
 central socket, which projects both above and below. 
 The upper part is threaded to receive the lower end of 
 the piston rod ; it is also slotted, 3, to permit the lower 
 
 of the wire from which the coils proceed ; the bead, in 
 connection with its bearings in the piston rod, C, and 
 the steel screw, G, forms a ball and socket joint, which 
 allows the spring to yield to pressure from any direc- 
 tion without causing the piston to bend. 
 
 The indicator is fixed to the cock on the engine in the 
 same way as the Tabor indicator, by means of nozzle, 
 11, and the coupling, 12, which latter, however, is 
 provided with a double set of thread. The paper-drum, 
 U, on the base, Q, rotates on the spindle, P. Its 
 squared section fits into the spring-head, S, and is held 
 in place by the nut, 10. Y is the guide-pulley for the 
 cord, X is the drum stop, and Z is one of the clips for 
 fastening the cord. The drum springs E, is a short 
 spiral spring. Fig. 194 gives an external view of the 
 Crosby indicator, 
 
THEORY OF STEAM ENGINES. 
 
 191 
 
 Bichards' Indicator. — This instrument, of which an 
 external view is given in Fig. 195, is one of the oldest 
 types, and is therefore also better known to engineers 
 than the two former ones. The peculiarity of Bichards' 
 indicator is tlie parallel motion by which the piston 
 movement is multiplied, in order to allow the pencil to 
 describe a diagram of convenient size. It will be noticed, 
 however, that the multiplying gear is somewhat heavier 
 than in the two former indicators, and therefore 
 renders the instrument less fit for high speed work. 
 
 the paper-drum can be stopped and started, without 
 disconnecting the driving cord. 
 
 The indicator, as shown in Fig. 195, is mounted on 
 the cock, which serves for making connection between 
 the engine cylinder and the indicator. By means of the 
 cord-adjuster, 2, the length of the driving cord can be 
 adjusted and readjusted by forming a loop on the cord, 
 which is held tight by the thumb-screw. 
 
 Darke's Indicator, Fig. 196, is perhaps the earliest 
 form of high-speed indicators. The pencil motion is 
 
 Fin. 197. 
 
 But for slow speed, the apparatus works admirably, 
 and being made strong, can stand a great deal of 
 knocking about, which is not of little importance. 
 
 The cylinder is jacketed like that of the Tabor 
 indicator, but the piston springs, 6, are single coiled ; 
 the one end is screwed to the top of the piston, and 
 the other end to the cover, the piston rod moving inside 
 the coil. 
 
 The drum spring is a flat coiled one, like a watch 
 ^spring, but of course very much stronger. 
 
 This indicator, as made by Messrs. Elliott Bros., is 
 provided with Darke's detent, described below, whereby 
 
 formed by a single light steel lever, carrying at the one 
 end a cross-head moving upon steel centres; to this lever 
 the motion of the piston is communicated by a jaw, 
 fitted on the piston-rod head, which jaw supports, 
 between centres, a sleeve through which the lever slides 
 with the varying angle of the motion of the lever. 
 
 The pencil is carried by a block sliding upon the other 
 end of the lever, the pencil being kept in a straight line, 
 parallel to the axis of the paper-drum, by a slot guide 
 placed between the sliding block and the paper-drum, 
 in which slot the pencil moves. The pencil is kept 
 against the paper by the elasticity of the lever, 
 
192 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Table D. 
 
 
 Weight in Grammes. 
 
 Scale of 
 Spring. 
 
 Millimetres. 
 
 Ratio of Pencil 
 
 Movement to 
 
 that of Piston. 
 
 Name of 
 Indicator. 
 
 Piston. 
 
 Drum. 
 
 Spring. 
 
 Diameter 
 of Piston. 
 
 Diameter 
 of Drum. 
 
 Distance between 
 axis of Drum and 
 that of Cylinder. 
 
 Richards 
 
 Crosby 
 
 38-5 
 14-4 
 13-4 
 
 172-3 
 .45-7 
 100-4 
 
 27-1 
 12-6 
 12 
 
 i 
 
 1 
 1 
 
 20 
 
 hardly 20 
 
 20 
 
 §2 
 
 38-5 
 
 45 
 
 62 
 78 
 55 
 
 4 
 
 about 6 
 
 
 6 
 
 The paper-drum is smaller and lighter than that of 
 the Bichards' indicator. Darke's detent is also attached 
 to this instrument. It consists of a pawl, which is 
 made to fall or rise by the movement of a flat spring 
 fixed on the outside of the indicator cylinder, so as to 
 engage or liberate the paper-drum, by means of a small 
 segment of a ratchet placed at the base of the drum. 
 
 M'Innes' Indicator. — This instrument is made by 
 Messrs. T. S, M'Innes and Co., of Glasgow, and is 
 illustrated in Fig. 197. 
 
 The drum spring is of the spiral form, like that used 
 in the Crosby indicator ; it can be adjusted to increase 
 or diminish the tension of the drum to suit speed of 
 engine. The multiplying levers are strong but light. 
 The piston rod is hollow, and threaded inside at the 
 top to receive a swivel-head, like in Fig. 192, whereby 
 the height of the atmospheric line can be adjusted. 
 This adjustment is done by turning the milled head, 
 shown in Fig. 197, on the top of the piston rod, and a 
 lock-nut prevents the swivel-head from turning back 
 when the adjustment is finished. A clip cord adjuster 
 is fitted on the cord, and by pressing the tails the cord 
 may be lengthened or shortened instantaneously. The 
 cylinder is open at the foot, and the part of the cylinder 
 in which the piston moves is of a smaller bore than the 
 rest, so as to allow of its being easily cleaned. One of 
 the peculiarities of this instrument is that the cylinder, 
 cylinder cover, and coupling ring are sheathed with 
 vulcanite, to enable the operator to handle the instru- 
 ment more comfortably without burning his fingers. 
 An improved cord is sent out with this indicator ; it 
 has a wire core, and this diminishes the stretching of 
 the cord. 
 
 The springs are similar to those used in the Bichards' 
 indicator. 
 
 Table D (above) gives the weights and dimensions 
 of parts of indicators which the author has been using 
 for some time. 
 
 Driving Gear for Indicators. 
 
 The motion of the paper-drum should be that of the 
 engine piston on a reduced scale. The drum can there- 
 fore not be driven directfrom the cross-head of the engine, 
 but a driving gear must be applied by which the motion of 
 the cross-head can be reduced, in a proper proportion. 
 The driving gear must be accurate if we want to pro- 
 duce accurate diagrams, and it must reduce the 
 
 travelling of the cross-head at any position of the latter 
 at a constant ratio. 
 
 Pantagraph motions have been successfully used for 
 driving gear, and when well made and kept in good 
 order will reproduce the motion of the cross-head with 
 great accuracy. In Fig. 198 is shown a form of panta- 
 
 graph often called " Lazy Tongs" ; it consists of strips 
 of wood put together with rivets. The end B is 
 attached to the stationary post by means of the screw 
 
THEORY OP STEAM ENGINES. 
 
 193 
 
 and thumb nut, and the end A is attached to the cross- 
 head, and therefore moves in a straight line. The cord 
 which drives the indicator is attached to a pin, E, on 
 
 Messrs. Musgrave and Sons, of Bolton, to their large 
 engines,' either horizontal or vertical. Two pins are 
 fixed, one at each end of the guide-bars, on each of 
 
 Fig. 199. 
 
 the cross-bar, C D, which may be moved in different 
 positions with relation to the centre, B. This pin must 
 always be placed in the line A B, as any point in this 
 
 which a grooved pulley is placed, which can turn freely 
 on the pins. The centres of the grooves on the pulleys 
 are in line with a pin fixed on the cross-head. A 
 
 Iig. 201. 
 
 line will have a motion exactly corresponding with the 
 movement of A. 
 
 Fig. 199 shows " Lazy Tongs" attached to a hori- 
 zontal engine and driving two Tabor indicators. 
 
 cotton rope of \m. or T^-in. diameter is passed tightly 
 round these pulleys, and fastened to the pin in the 
 cross-head. As the cross-head moves back and forth 
 it carries this rope along with it, which, passing over 
 
 Fig. 202. 
 
 Another form of pantagraph is shown in Fig. 200, 
 and its application to indicator driving gear is illustrated 
 in Fig. 201. 
 
 Fig. 202 shows the reducing motion usually applied by 
 
 the grooved pulleys, turns them correspondingly. To 
 the hub of one pulley is fastened a small sheave, around 
 which the indicator cord is wound, and which winds 
 up and upwards with the movement of the engine, 
 
194 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 giving the correct motion to the indicator. The cord 
 may be led at any angle from this sheave to the 
 indicator, without affecting the movement. 
 
 In Fig. 203 is illustrated a form of driving gear, called 
 the " Brumbo pulley "; it is a plain reducing lever, con- 
 nected to the cross-head by a short link, and having a 
 
 The hook, A, on the indicator cord connects the 
 loop, B, on the driving cord when the paper-drum is 
 running ; the loop is made long enough to hold with 
 the hand, when hooking on the indicator. To disconnect, 
 merely catch the hook and hold it stationary for a few 
 seconds, and the loop will come off. The rubber 
 
 Fig. 203. 
 
 segment of a grooved pulley for the indicator cord. 
 The whole apparatus should be made very light. 
 
 Mr. A. G. Brown recommends for high-speed engines, 
 that the driving cord should be continued beyond the 
 loop for hooking on the indicator, and fastened to a 
 rubber band attached to a carrier-pulley, so that it 
 
 band then takes the strain, and keeps the driving cord 
 tight. 
 
 The Management of the Indicator. 
 The indicator must be kept clean and in good order, 
 the piston should be frequently taken out, and the 
 inside of the cylinder and the outside of the piston 
 
 m <ojg^ ^f — j_ (/ i 1 lite 
 
 
 • 
 
 o ' - 
 
 wan 
 
 Fig. 204. 
 
 always keeps a tension on the driving cord, whether the 
 indicator is running or not, and prevents entanglement 
 or breakage of the cord when the indicator is unhooked. 
 This method, as applied to the " Globe " compound 
 high-speed engines, is shown in Fig. 204 ; diagrams can 
 be taken with this arrangement as quickly as on slow- 
 speed engines. 
 
 should be wiped carefully ; a little dirt in the cylinder 
 may produce friction and cause distortions of the 
 diagram. When the indicator is in use, the cylinder 
 should be lubricated with best cylinder oil, and the 
 pivots at the joints in the pencil movement should be 
 lubricated with watch oil. The movements of the 
 piston rod and pencil levers must be perfectly 
 
THEORY Ob STEAM ENGINES. 
 
 195 
 
 free ; this can be tested by taking the piston spring off, 
 then raise the pencil lever to its highest position, when 
 it should drop to its lowest position with perfect 
 freedom. 
 
 It is also of importance that the paper-drum should 
 be kept in good order, and that the bearings should be 
 clean and properly lubricated. The strength of spring 
 to be used, depends upon the boiler pressure and the 
 speed of the engine. The lighter the spring is, the 
 higher will the diagram be, and the more accurate the 
 measurement ; but care should be taken that the diagram 
 is not distorted by the inertia of the moving parts. 
 Springs used for indicating condensing engines will be 
 compressed while subjected to the pressure of the 
 steam, and will be stretched during the exhaust of the 
 steam into the condenser; these springs are, therefore, 
 shorter than those used for indicating non-condensing 
 engines." 
 
 The indicator-cock, to which the indicator is fixed, 
 is screwed into a short steam-pipe, which, again, is 
 screwed into a hole made in the engine cylinder. In 
 horizontal engines, this hole should be at the end of 
 the cylinder barrel, and in such a position that the 
 engine piston never covers the opening of the hole. 
 In vertical engines, the hole for the indicator-cock 
 is made in the cover. The steam connections between 
 the engine cylinder and the indicator should be as 
 short as possible, and with easy bends. 
 
 Fig. 205. 
 
 Accurate work requires that diagrams should be taken 
 from both ends of the engine cylinder, and an indicator 
 should therefore be used at each end. To make sure 
 that both indicators are alike, they should be inter- 
 changed, and the spririgs should be frequently tested 
 under steam. 
 
 If only one indicator is available, it should be used 
 alternately, first at one end and then at the other end 
 of the engine cylinder. 
 
 "When the engine cylinder is short, the indicator may 
 be placed in the centre, and fixed on a three-way cock, 
 Fig. 205, with steam connections to both ends of the 
 engine cylinder. In order to reduce the errors due to 
 friction of the steam, the passages should be large and 
 have easy bends. 
 
 Before attaching the indicator to the cock, the latter 
 should be opened, in order to clear the passages from 
 any dirt which may have accumulated within, them. 
 The indicator and its carrier-pulley must be adjusted in 
 
 such a manner that the driving cord does not bear 
 against the sides of the pulleys, but runs central. The 
 length of the cord must be adjusted so as to prevent the 
 paper-drum from touching either stop when running. 
 The cock is then opened to admit steam into the 
 indicator cylinder to heat the piston, spring, and 
 cylinder. 
 
 Taking the Diagram. — The drum is now taken off 
 by unhooking the driving cord from the indicator cord, 
 and a card is placed smoothly on the drum. The card 
 is a blank piece of paper which is chemically prepared, 
 so as to enable the brass pin, which acts as a pencil, to 
 draw the diagram on the card. A lead pencil marking 
 on ordinary paper might be used, but the frequent 
 sharpening of such a pencil would be inconvenient. 
 The card is fixed on the drum by bending the two ends 
 over the clips. The drum with the card attached is 
 then put back on the drum carriage, and the cords 
 hooked together. 
 
 When the indicator is provided with a detent, the 
 drum is taken off by letting the pawl fall into the 
 ratchet ; and in order to prevent entanglement of the 
 cord, the operator must keep the cord in tension by 
 letting his other hand touch it, and follow the motion. 
 To start the drum it is necessary to lift the cord a 
 little by the hand, so as to turn the drum a little, 
 whereby the pawl will be disengaged. 
 
 The indicator-cock is now turned in such a position 
 that it allows the steam from the engine cylinder 
 to escape into the atmosphere, whereby water which 
 may have accumulated in the passages, will be blown 
 out. The cock is then turned so as to allow the air to 
 enter underneath the indicator piston ; the pressure on 
 both sides of the latter will thus be the same — viz., 
 that of the atmosphere. The pencil is now brought 
 lightly into contact with the card, and traces a line 
 which, when the card is unfolded, will appear to be a 
 straight line. This line is called the "atmospheric 
 line." The cock is again opened to admit steam under 
 the piston, and after a few revolutions of the engine the 
 pencil is again made to touch the card, this time tracing 
 the indicator diagram. Before taking the card off, the 
 pencil should again be allowed to trace the atmospheric 
 line, to ascertain whether the condition of the line is 
 the same. If a new line be traced, the diagram is 
 useless. 
 
 The pressure of the pencil on the card should be as 
 small as possible ; just sufficient to make a legible 
 diagram. A greater pressure will only create friction, 
 and, consequently, inaccuracy in the diagram. 
 
 On each diagram should be marked boiler pressure, 
 number of revolutions the engine is making per minute, 
 apparent ratio of expansion, diameter of engine piston, 
 length of stroke, the end of the cylinder at which the 
 diagram was taken, the date at which the diagram was 
 taken, the number of the indicator spring, etc. 
 
 "When making accurate tests of the engine, the 
 absolute vacuum line should be drawn on the card. 
 This line should be parallel to the atmospheric line, 
 and should be drawn below the latter ; the distance 
 
196 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 between the two lines should correspond to the atmo- 
 spheric pressure at the time when the diagram was 
 taken. 
 
 The Shape of the Indicator Diagram. 
 
 The diagram of an ideal engine, such as that shown 
 in Fig. 186, is composed of four straight lines and a 
 curve. The line n b is called the admission line, and 
 represents the pressure of the steam entering 
 the cylinder, and, at the same time, it shows that 
 the pressure on the piston is suddenly changed 
 from that of the condenser to that of live steam ; 
 line b c is the steam line, and represents the distance 
 travelled by the piston froth the point of admission to 
 that of cut-off, and it also shows that no throttling of 
 the steam has taken place, the pressure of the steam 
 remaining constant, and equal to that of the steam 
 during admission. The curve, c d, is the expansion 
 line, and this line meeting the steam line in a 
 sharp point, c, shows that the cut-off valve has acted 
 suddenly. The release takes place at point d, and as 
 the line dm, is at right angles to the motion of the 
 piston, the exhaust-valve must have been opened sud- 
 denly, presenting at once a sufficiently large opening 
 for the steam to allow the pressure to fall at once 
 to that of the condenser. The line m n is called the 
 exhaust line, and as it is parallel to the vacuum line, 
 the steam has been able to escape freely during the 
 
 does not act suddenly, but gradually closes the opening 
 through which the live steam is admitted to the 
 cylinder. The release begins at point E, but owing to 
 the slow shutting of the exhaust-valve,|the release is 
 not completed^till point F. 
 
 Fig. 206. 
 
 The exhaust-valve begins to shut at point G, and is 
 completely shut at H. The remaining steam is now 
 compressed in the clearance space, whereby its pres- 
 sure is increased before the fresh charge of live steam 
 is admitted. The effect of the slow action of the valves 
 in opening and shutting the steam-ports is called 
 " wire-drawing," and is represented on the diagram by 
 
 Fio. 2C7. 
 
 backward stroke of the piston ; and it also shows that 
 the condenser has been able to keep the pressure 
 constant. 
 
 The essential features of a diagram from an actual 
 engine, resemble that of the ideal engine with some 
 few modifications. In Fig. 206 is shown a diagram of 
 a non-condensing engine. A Bis the admission line, 
 B C the steam line, D E the expansion line, F G the 
 exhaust line or back-pressure line, H A is called the 
 compression line, and J I is the atmospheric line. 
 
 The short curve, C D, shows that the cut-off valve 
 
 curved corners instead of the sharp corners on the ideal 
 diagram. 
 
 Fig. 207 represents indicator diagrams taken from 
 the two ends of the cylinder of a steam engine, and re- 
 produced on one card. The straight line across the 
 diagram is the atmospheric line. The diagrams bear 
 evidence of having been taken from a condensing 
 engine, as the exhaust lines fall below the atmospheric 
 line. The admission lines are very straight and are 
 perpendicular on the atmospheric line, and this bears 
 evidence of a prompt admission. The ratio of expan- 
 
THEORY OF STEAM ENGINES. 
 
 197 
 
 sion is very high, which shows that the engine had 
 very little work to do when the diagrams were taken. 
 The diagrams are practically equal, and the exhaust 
 lines are parallel to the atmospheric line. The expan- 
 sion curve of the left-hand diagram is slightly wavy near 
 the cut-off, which is due to the inertia of the indicator. 
 
 The engine from which the diagrams were taken is a 
 Musgrave horizontal engine with Corliss valves. The 
 diameter of cylinder is 38in. ; stroke 6ft. ; boiler pressure 
 791b. ; the engine made 51 revolutions per minute ; 
 vacuum was 13Jlb., and the scale of the indicator 
 spring was ^. 
 
 Figs. 208 and 209 are diagrams taken from a 
 
 Determination of Clearance. 
 
 The clearance of a cylinder can be measured by filling 
 the clearance volume with water, the quantity of which 
 must be known. But in many cases this is impossible, 
 and the clearance must then be determined approxi- 
 mately by the indicator diagram. 
 
 Figs. 210 and 211 are diagrams taken from the high 
 and low pressure cylinders of a compound engine. Line 
 A B is the atmospheric line, and D C the absolute 
 vacuum line. Curves C E are the compression lines, 
 and assuming that these curves are hyperbolas, we can 
 determine the clearance line, remembering that the 
 latter line must be the axis of ordinates for the hyper- 
 
 Fro. £08 
 
 Fig. 209. 
 
 Musgrave tandem compound engine without receiver. 
 The H.P. diagrams from both ends of the cylinder are 
 shown in Fig. 208, the ratio of expansion being about 
 2. The diagrams show that the compression was 
 driven so far, that the end pressure in the clearance 
 space before admission was greater than the initial 
 pressure of the steam, and thus caused the indicator 
 pencil to draw the loop at the top. The scale of the 
 spring used for taking these diagrams was -^. 
 
 The L.P. diagrams are given in Fig. 209, and show 
 the presence of a condenser into which the steam was 
 exhausted from the L.P. cylinder. The scale of the 
 indicator spring used for taking these diagrams was ^ T> . 
 
 bola. The process of construction is the reverse of that 
 shown in Fig. 172, for determining the hyperbola I. In 
 Fig. 172 we constructed the curve, knowing the axes of 
 ordinates and abscissae, whereas in the case of diagrams 
 Figs. 210 and 211, we have given the curve and the axis 
 of abscissae, and shall now determine the axis of ordi- 
 nates. For this purpose we choose two points on the 
 compression curve, E and C, and draw the parallelogram, 
 EFCHE, and the straight line, F H, and extend the 
 latter until it intersects the vacuum line at D, which point 
 will correspond to point O in Fig. 172. The straight 
 line, D J, at right angles to CD, must therefore be the 
 clearance line ; the distance between D J and the 
 
198 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 admission line of the diagram must be the clearance 
 length, X. 
 
 As the length of the diagram represents the stroke, I, 
 the distance between the clearance line and the admis- 
 sion line, divided by the length of the diagram, will 
 give the percentage of clearance of the engine. 
 
 The two points C and E must be selected as far apart 
 as possible, but within the limits of the true curve. 
 Combining the Diagrams from Compound Engines. 
 
 The method of combining the diagrams from an ideal 
 compound engine has been explained for a Woolf's 
 
 introduce in turn other losses which occur between the 
 two cylinders. 
 
 These losses are due to condensation and friction in 
 the pipes and steam passages between the two cylin- 
 ders, and also to the expansion of the steam in these 
 passages while doing no useful work. There is, there- 
 fore, a fall of pressure of the steam between the two 
 cylinders, the extent of which can be shown and 
 measured by combining the diagrams from the actual 
 engine. 
 
 The following diagrams have been taken with the 
 
 Fig. 210. 
 
 Fio. 211. 
 
 engine in Fig. 181, where the H.P. diagram is g ab eg, 
 and the L.P. diagram mhdnm; and for a receiver 
 engine in Fig. 183, where g a b c g is the H.P. diagram 
 and mg' c' hnm is L.P. diagram. The diagrams from 
 actual engines will, however, be somewhat modified, 
 due to various losses. The loss due to clearance has 
 been considered, and must occur in any actual cylinder 
 to a smaller or greater extent, according to the size of 
 the clearance. 
 
 Although compound engines have the advantage over 
 non-compounding engines of reducing the amount of 
 initial condensation at high degrees, of expansion, they 
 
 Tabor indicator from engines built by Messrs. Musgrave, 
 of Bolton, to whom the author is indebted for the 
 blocks. 
 
 To correctly combine the- two diagiaras of a 
 compound engine, the total capacities of the two cylin- 
 ders must be known and accounted for, as done in Figs. 
 187 and 188. The next is to decide on the total length 
 of the combined diagram. It is best to decide on the 
 total length of the L.P. diagram first, and a length which 
 can be easily divided into 100 parts will be found the 
 most convenient, as percentages of this length can then 
 be easily measured ; we may, for instance, take the 
 
THEORY OF STEAM ENGINES. 
 
 199 
 
 length 12|in., a hundredth part of which will then be 
 one-eighth of an inch. The combined diagrams, 
 Figs. 214, 217, and 219, were each 12£in. long, and the 
 photo was reduced to the length shown. 
 
 The next is to decide on the scale of pressure to 
 
 the clearance line. All measurements of distance 
 should be made from the clearance line, and all 
 measurements of pressure from the atmospheric line. 
 
 We now divide the two original diagrams into 10 parts, 
 as show in Fig. 218. 
 
 Fig. 212 
 
 Fis. 213. 
 
 Fig. 214. 
 
 which the two diagrams are to be plotted — usually it 
 may be most convenient to take the scale of the original 
 L.P. diagram for this. Now draw in the atmospheric 
 and vacuum lines, and erect perpendiculars at the two 
 extremes of the combination diagram, one of which is 
 
 (1) Low-pressure Diagram. — Find the effective 
 capacity of the L.P. cylinder, and add to it the volume 
 of the clearance; the total length of the combined 
 diagram will represent the total capacity of the L.P. 
 cylinder. Divide the clearance of the cylinder by the 
 
200 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 total capacity of the cylinder, and the quotient will 
 give the percentage this clearance bears to the whole 
 length. Set off this distance from the clearance line, 
 
 and 219. If the pressure scale selected is the same as 
 that of the original diagram, then transfer the pressures 
 directly, with a pair of dividers, from the lines on the 
 
 Fig. 215. 
 
 Fig. 216. 
 
 Fig. 217. 
 
 and divide the remainder, representing the effective 
 capacity of the cylinder, into the same number of parts 
 that the original diagram is divided into, see Figs. 218 
 
 original diagram, to the corresponding lines on the 
 combination ; draw in the connecting portion of the 
 diagram, and the result will be an elongated diagram 
 
THEORY OP STEAM ENGINES. 
 
 2C1 
 
 from the L.P. cylinder, similar to those shown in Figs. 
 181 and 183. 
 
 of theL.P. cylinder; the quotient is the percentage which 
 the length of the H.P. diagram on the combination 
 
 Fin. 218. 
 
 Fig. 219. 
 
 (2) High-pressure Cylinder. — Find the total capacity bears to the length of the corresponding L.P. diagram 
 of the H.P. cylinder, and divide it by the total capacity both measured from the clearance line. Divide the 
 
202 
 
 Practical electrical E'tiGWEERiNG. 
 
 clearance volume of the H.P. cylinder by the total 
 capacity of the L.P. cylinder, and the quotient will be 
 the length of the clearance, B, Fig. 219, on the com- 
 bination measured on the scale of the combination. 
 Divide the remaining length of the H.P. diagram, C, 
 Fig. 219 (representing the effective capacity of the 
 cylinder), into the same number of parts as the original 
 diagram, and transfer the pressures from the lines on 
 the original diagram, Fig. 218, to the corresponding lines 
 on the combination, and to the new scale of pressures. 
 If the original H.P. diagram were taken with a 40 
 spring, see table on page 188, and the combination 
 diagram made to a scale of 101b. per inch, the new H.P. 
 diagram will be four times as high as the original one. 
 
 The combination diagram is finished by drawing the 
 hyperbola which touches the expansion line of the H.P. 
 diagram at the point of release. If there were no losses 
 between the two cylinders, this hyperbola ought to 
 coincide with the expansion line of the L.P. expansion 
 line. The space between the hyperbola and both 
 diagrams below the point of release of the H.P. diagram, 
 and also the space between the two diagrams, represent 
 the losses between the two cylinders. 
 
 Example 1. — The engine from which diagrams Figs. 
 212 and 213 were taken is a side-by-side compound 
 (receiver) engine with Corliss valves. The diameter of 
 the H.P. cylinder is 26in., and that of the L.P. cylinder 
 48in.; the strokeis6ft. for both cylinders. The boiler pres- 
 sure was 691b. above the atmosphere. The engine made 
 50 revolutions per minute, and the vacuum was 13'751b., 
 i.e., the pressure in pounds per square inch below the 
 atmosphere. The H.P. diagrams, Fig. 2.12, were taken 
 with a spring ^ — i.e., the resistance of the spring 
 allowed the pencil to move lin. per 301b. of effective 
 pressure per square inch on the indicator piston. The 
 straight line which is almost touching the diagrams, is 
 the atmospheric line. 
 
 The L.P. diagrams were taken with a spring ^ ; the 
 atmospheric line appears at the top of the diagrams ; the 
 vacuum line is not shown. 
 
 Fig. 214 represents the diagrams combined. A scale 
 is added at the bottom of the diagram whereby all 
 volumes can be measured in percentage of the total 
 capacity of the L.P. cylinder. The dotted curve is the 
 hyperbola through the point of release of the H.P. 
 diagram, and has been constructed as shown in Fig. 172, 
 the initial pressure line corresponding to line B F in 
 Fig. 172. 
 
 Example 2.— Diagrams Figs. 215 and 216 are also 
 taken f i om a side-by-side co mpound (receiver) engine with 
 Corliss valves. The diameters of the two cylinders are 
 35in. and 57in. respectively, and the stroke is 7ft. for 
 both. The boiler pressure was 651b., the engine made 
 33£ revolutions per minute, and the vacuum was 14£lb. 
 
 The H.P. diagrams were taken with spring 7J 1 Tr , and 
 they show that the pressure in the receiver was 
 not constant while the steam was exhausted from the 
 H.P. cylinder. 
 
 The L.P. diagrams were taken with spring ^\. The 
 
 straight lines drawn in both diagrams are the 
 atmospheric lines. 
 
 Fig. 217 is the combination of the diagrams by which 
 the losses between the cylinders can be measured. 
 
 Example 3. — Diagrams Fig. 218 were taken from a 
 tandem compound engine. Diameters of cylinders are 
 30in. and 50in. respectively, the stroke is 6ft., and the 
 diameter of the piston rod in both cylinders is 6^in. 
 The boiler pressure was 821b., the engine made 56£ 
 revolutions per minute, and the vacuum was 121b. 
 
 The effective capacity of either cylinder must be 
 equal to the area of the piston multiplied by the stroke, 
 minus the volume occupied by the piston rod. 
 
 The effective capacity of the H.P. cylinder must 
 therefore be 
 
 x 30 2 
 
 6-25 
 
 x 72 = 48,685 cubic inches. 
 
 V 4 4 
 
 Clearance volume in H.P. cylinder is 2,545 cubic 
 inches, so that the total capacity of the cylinder is 
 51,230 cubic inches. 
 
 The effective capacity of the L.P. cylinder will be 
 (1963-5 - 30-68) x 72 = 139,163 cubic inches, and the 
 clearance volume is 7,673 cubic inches. The total 
 capacity of the L.P. cylinder is therefore 146,836 
 cubic inches. 
 
 Determination of Indicated Horse-Power. 
 
 The area enclosed by the line traced by the indicator 
 pencil, measuresjthe indicated work done on the one side 
 of the engine piston during one stroke. 
 
 Fig. 220. 
 
 The length of the diagram, measured parallel to the 
 atmospheric line, represents the stroke of the engine 
 piston and is therefore given in feet ; the ordinates to 
 the diagram line, measured in a direction at right angles 
 to the atmospheric line, represent pressures in pounds 
 per square inch. The area enclosed by the diagram 
 line will therefore measure the indicated work in foot- 
 pounds per square inch of the engine piston. 
 
 It has been lully explained on page 157, and shown 
 in Fig. 165, how the area enclosed by a curve can be 
 determined. 
 
 In Fig. 220 is shown the method adopted for the 
 purpose of calculating the area of an indicator diagram. 
 
THEORY OF STEAM ENGINES. 
 
 203 
 
 The base line is here either the atmospheric line or the 
 vacuum line. The two extreme ordinates are C D and 
 A B. The base line, B D, is divided into ten equal parts, 
 and the corresponding ordinates are drawn. In order 
 to obtain an accurate result, the divisions of the curves 
 should be approximately straight lines. This is not the 
 case with FE, EG, the release curve and the compres- 
 sion curve. In the cases of FE and EG, we may 
 consider that the areas enclosed between the straight 
 lines FE and EG, and the curves EE and EG, are 
 equal. But the release curve and the compression 
 curve should be further divided. By using formula 
 (71), on page 158, the area enclosed by the diagram 
 will be obtained. 
 
 Let A! denote the foot-pounds represented by the 
 area of the diagram taken at the crank end of the 
 cylinder, and A, the same for the diagram taken at the 
 other end of the cylinder ; and, further, let D be the 
 diameter of the piston, d the diameter of the piston 
 rod, and S the diameter of the tail rod (if any), all in 
 inches; then the indicated work on the crank side of 
 piston will be 
 
 A 1 (1L&-JL*} foot-pounds, 
 
 and the indicated work on the other side of the piston 
 will be 
 
 A 2 (-X^-Jrii) foot-pounds. 
 
 The indicated horse-power of the engine, when mak- 
 ing n revolutions per minute, will be 
 
 I.H.P. = 
 
 { A ,(^!-^) + A^--i) } , 
 
 33,000 
 
 (165) 
 
 We might also have determined the energy exerted 
 by the steam during the forward stroke, and likewise 
 the work done on the back pressure during the return 
 stroke. For this purpose, we require the absolute 
 vacuum line to be drawn on the card. Let us imagine 
 that B D is this line. Area DEABD would thus 
 represent the energy exerted by the steam, and the area 
 between the back-pressure curve and the line B D would 
 represent the work done on the back pressure. These 
 two areas can be calculated in the same manner as 
 described above. 
 
 For the purpose of designing new engines, it is desir- 
 able to know the mean forward pressure, p/m, which will 
 be exerted on the piston of the new engine, and also the 
 mean backward pressure, pb m , which the piston has to 
 overcome. p fm will vary with the initial pressure of the 
 steam, the ratio of expansion, and the general construc- 
 tion of the eDgine, and p^m will depend on the construc- 
 tion of the condenser as well as on that of the engine. 
 
 These two pressures can be calculated by predeter- 
 mining the indicated diagrams of the new engine, 
 taking into consideration the losses which are likely to 
 occur in the new engine. If diagrams taken from 
 actual engines of the same class as that of the engine to 
 be designed are available, then the mean pressures calcu- 
 
 lated from such diagrams would give the designer more 
 reliable data than the theoretical diagrams could give. 
 
 The two mean pressures are to be calculated by means 
 of formula (72), page 158, the pressures being measured 
 from the absolute vacuum line. 
 
 If we know the mean pressures, the indicated work 
 on the crank side of the piston will be 
 
 (p/m - Pbm ) x ( —f ~ —£- ) x I foot-pounas, 
 
 and that on the other side of the piston will be 
 
 /ttD 2 ttS\ ,, 
 (P/m " - Pbm ) x l—£ Y) x * foot-pounds, 
 
 where I is the stroke in feet, the other dimensions being 
 in inches. 
 
 The engine when running at n revolutions per minute 
 will indicate 
 
 ' (Pfm -Pbm ) ^— - 
 
 \ 
 
 rrd? 
 
 4 
 
 J + (Pfm" -Pbm") (^— - ~ J }' X I X n 
 
 33,000 
 horse-power. 
 
 The pressure expressed iij P/m - Pbm,' is called the 
 mean effective pressure, but is nothing of the kind unless 
 the diagrams taken from both ends of the cylinder are 
 exactly the same. As this is hardly ever the case, the 
 usual meaning of the mean effective pressure is 
 erroneous. It is evident that the actual mean effective 
 pressures, in pounds per square inch, are for the two 
 sides of the piston (p /m ' - pi„ a ") and (pf m " - pbm')- The 
 two total mean effective pressures, in pounds, on the 
 piston are, for the stroke of the piston towards the 
 bottom of the cylinder, 
 
 , / TT D 2 7T d 2 \ m „ ( 7T D 2 7T <S 2 \ , 
 
 p fm x [— r - — r ) - Pbm (—j- - —£_) pounds, 
 and for the stroke of the piston towards the crank, 
 
 Pfm 
 
 11 ( 7T . 
 
 , /ttD 2 t*\ , 
 
 Pbm x ( — _ - — _ _ 1 pounds. 
 
 The areas enclosed by the indicator diagrams, as well 
 as the mean pressures, can also be determined by the 
 use of planimeters, as fully described on pages 158 and 
 159. 
 
 The planimeter, however, would give the area in 
 square inches, and we must therefore find the number 
 for each case by which the area must be multiplied to 
 obtain, the corresponding foot-pounds. 
 
 Suppose the scale of the spring used in taking the 
 diagram was ^, i.e., each inch measured at right angles 
 to the vacuum line corresponds to 401b. pressure per 
 square inch, and as the distances measured along the 
 vacuum line represent feet, it is evident that each 
 square inch measured by the planimeter must mean 
 
 — = 3 J foot-pounds. 
 
 In non-condensing engines carrying a light load, the 
 expansion line often extends below the back pressure 
 line and forms a loop in the diagram as shown in 
 
204 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Fig. 221. The indicated work represented by this This expression must be equal to the mean forward 
 
 diagram would be measured by the area on the left- pressure, multiplied by the volume swept by the piston ; 
 
 hand side of the loop minus the area of the loop. It is so we must have 
 easy to see that when using a planimeter, the tracer 
 
 P/m x » x 
 
 (167) 
 
 l +1 *r e 
 
 By equating (166) and (167) we obtain 
 
 1 + (l + -^ J x r e x log. 
 
 ( 1 + t) 
 
 x?-, 
 
 1+ ] Xr ' 
 
 Fig. 221. 
 
 Pm = P«x 
 
 ' e 
 
 pounds per square foot. 
 
 The mean forward pressure in pounds per square inch 
 
 will be p/m = 
 
 _ P/m 
 
 144 
 
 would move clockwise over the first part of the diagram 
 
 and anti-clockwise over the loop. The following table, B, has been calculated by Prof. 
 
 For the purpose of showing how diagrams, taken Haedicke, and is derived from actual engines : 
 
 from the same engine, vary with the cut-off, Figs. 222 Table E 
 
 1 
 
 01 
 
 0-2 
 
 0-25 
 
 0-3 
 
 0-33 
 
 0-4 
 
 0'5 
 
 0-6 
 
 0-625 
 
 0-7 
 
 Pfm_ 
 Pfi 
 
 036 
 
 053 
 
 0-60 
 
 0-66 
 
 0-67 
 
 0-76 
 
 0-84 
 
 0-90 
 
 0-92 
 
 095 
 
 Fig. 222. 
 
 and 223 have been added, The former is from a non- 
 condensing engine, and the latter from a condensing 
 one. 
 
 We can calculate the mean forward pressure by 
 means of the formula found for engines with clearance. 
 
 Fig. 223. 
 
 For example, for the single cylinder, Fig. 186, we found 
 that the energy exerted by lib. of steam was 
 
 P/i x v x 
 
 l 1 + r) xr - 
 — — + 10&— ^ — £ 
 
 The average mean back pressure of actual engines 
 may be taken as 161b. per square inch for non-condens- 
 ing engines, and as 31b. per square inch for condensing 
 engines. By assuming that the indicator diagrams 
 from both ends of the cylinder are equal, the mean 
 effective pressure will be — 
 
 (a) For non-condensing engines 
 
 Pem = (pfm - 16) pounds per square inch. 
 
 (b) For condensing engines 
 
 Pem = (p/m —3) pounds per square inch. 
 In using Table E, we must remember that p fi is not 
 equal to the absolute boiler pressure, but we may take 
 on an average : 
 
 (a) For engines working with throttle-valves 
 P/i = 0'8 times the absolute boiler pressure. 
 
 (b) For engines working with automatic expansion 
 gear 
 
 P/i = 0'9 times the absolute boiler pressure. 
 Actual Horse-Power of an Engine. 
 The actual energy given off by an engine is the energy 
 available for doing useful work. The actual horse-power of 
 an engine is therefore the actual energy in foot-pounds 
 given off per minute, divided by 33,000. The number 
 of A.H.P. of an engine is less than the number of 
 I.H.P. of the engine, by the amount required for over- 
 coming the resistances in the engine itself. The ratio 
 of A.H.P. and I.H.P. is called the efficiency of mechan- 
 ism of the engine, and must necessarily always be 
 smaller than unity. The efficiency of mechanism, « u 
 can be obtained approximately from the indicator by 
 
THEORY OF STEAM ENGINES, 
 
 205 
 
 determining the number of I.H.P., n x and n. 2 , when the 
 engine is running empty, and when it is running with 
 a certain load at the same speed. The efficiency of 
 mechanism would be 
 
 ^1 = 
 
 «, 
 
 (168) 
 
 As the resistances in the engine increase with the 
 load, formula (168) will only give ^ approximately. 
 
 The efficiency of mechanism is determined accurately 
 by measuring the A.H.P. by a dynamometer applied 
 on the engine shaft, and at the same time measuring 
 the I.H.P. by the indicator. 
 
 The efficiency of mechanism may be estimated from 
 the following empirical formulas due to Prof. Hrabak : 
 
 (a) For engines giving off less than 5 A.H.P. , 
 
 A.H.P. 
 
 >7i = 
 
 + 5 
 A.H.P. + 10 
 
 (169) 
 
 (b) For engines of more than 5, and less than 50 
 A.H.P. 
 
 A.H.P. + 20 ,_,., 
 
 ^ A.H.P. + 32-5 ■ • • (1 ' 0) 
 
 (c) For engines of more than 50 A.H.P. 
 A.H.P. + 300 
 
 1\ = 
 
 A.H.P. + 366 
 
 (171) 
 
 An engine making n revolutions per minute, with a 
 stroke equal to I, and delivering the same indicator 
 diagrams from both sides of the piston, will indicate 
 
 (P/m-Pbm) x A X I X 2 TC -rj p 
 
 33,000 ' ' 
 
 where A is the effective area of the piston in square 
 inches, and will be 
 
 (a) For engines with tail rod ; 
 
 -=»(' 
 
 jD 2 irj? ttD 2 jt, 
 
 1~~ 4 + _ 4~ 
 and in case where d = S 
 
 TT S 2 \ - D 2 , flT d* IT S 2 
 
 ttD 2 
 
 ird 1 
 4 ' 
 
 (b) For engines with no tail rod : 
 
 , ttD 2 
 4 
 
 d* 
 
 The A.H.P. given off by this engine will be 
 
 Vl (Pfm—Pbm) x A X Z X 2 ^ 
 
 33,000 
 
 If we denote the pressure ^ (p/ m — ptm ) by p„ , then 
 we shall have 
 
 p n x A x I x 2ra 
 
 A.H.P. = J 
 
 (172) 
 
 33,000 
 
 p n is evidently the mean effective pressure which would 
 produce the same effect as (j?f m —pt, m ) , if there were no 
 resistances to overcome in the engine itself; p n is called 
 the mean actual pressure. 
 
 (a) For non-condensing engines, we have 
 pn = >h (pjm - 16) pounds per square inch (173) 
 
 (6) For condensing engines 
 Pn =m (Pm —3) pounds per square inch 
 
 (174) 
 
 Steam Accounted for by the Indicator. 
 
 The connection between pressure and volume of lib. 
 of dry saturated steam is given in the table on page 61. 
 By the same table and a slight calculation we can 
 evidently determine the mass of steam contained in a 
 steam-cylinder at any point of the stroke, if we know 
 the corresponding pressure of the steam and volume of 
 the cylinder, and if the steam contained in the cylinder 
 is dry saturated steam. The indicator diagram will give 
 us this volume and pressure, if we only draw in the 
 clearance line and the vacuum line. The indicator 
 therefore is also useful in connection with a feed-water 
 test in determining the amount of steam consumed per 
 I.H.P. hour. 
 
 The indicator diagram does not only give an account 
 of the steam which is present in the cylinder during the 
 forward stroke of the piston, but the amount of steam 
 retained in the cylinder during compression can also be 
 measured in the same manner ; the difference of these 
 two quantities will be the amount of steam accounted 
 for per stroke. 
 
 Fig. 223* clearly shows that if the expansion of the 
 steam is hyperbolic, the steam in the cylinder will be 
 superheated, because the ordinates to the hyperbola are 
 greater than the corresponding ones to the saturation 
 curve. In diagrams taken from jacketed engines with 
 an early cut-off, it will always be found that the steam is 
 superheated at points near the end of the stroke. Such 
 points on the expansion curve should therefore not be 
 selected for determining the quantity of steam contained 
 in the cylinder, but points near the cut-off will give 
 more accurate results. 
 
 The data given in Table B, on page 181, show that, in 
 order to get the full advantage of the steam in the 
 cylinder, the jacket must be filled with steam, the 
 amount of which varies with the ratio of expansion ; in 
 example No. 20, the steam required in the jacket is 
 about 22 • 5 per cent, of the steam contained in the 
 cylinder — i.e., for each 1001b. of steam admitted to the 
 cylinder, 22"51b. of steam must be admitted to the 
 jacket. In calculating the latter' quantity of steam, it 
 was assumed that there was no radiation, and that the 
 material of which the cylinder-barrel was made was a 
 perfect conductor for heat ; this, of course, is not the case 
 in actual engines, and the quantity of steam condensed in 
 the cylinder and in the jacket must therefore be con- 
 siderable in actual engines, and it will increase with 
 the ratio of expansion. 
 
 The indicator diagram can therefore only show the 
 quantity of steam which is actually contained in the 
 cylinder, but not the total quantity consumed by the 
 engine. If we measure the quantity of feed- water 
 pumped into the boiler, and from that subtract the 
 quantity accounted for by the indicator diagram, we 
 obtain the extent of losses due to leakage and con- 
 densation. 
 
 It would not, however, be fair to the engine to debit 
 
206 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 it with the amount of steam escaping through leakages 
 in the boiler, boiler fittings, and steam-pipes, nor with 
 the steam which is condensed in the steam-pipes. The 
 rate of leakage can be determined by measuring the 
 quantity of feed- water required for keeping the water 
 level in the boiler constant, while the stop-valve of the 
 ' engine is shut ; and the extent of condensation can be 
 measured by applying a steam-separator close to the 
 engine stop-valve, and for accurate tests, an analysis of 
 the wetness of the steam should be made by means of a 
 calorimeter, which will be explained below. 
 
 Leakage may also occur at the valves and at the 
 piston of the engine, but this can be detected by a trial 
 
 When testing engines to determine their economy, 
 careful tests should be made of the quality of the steam 
 entering the engines, as in many cases water is carried 
 over from the boilers, and condensation in the pipes adds 
 to the amount ; sometimes apparatus is provided for 
 separating the water from the steam before the latter 
 enters the engine, but it is always advisable to make 
 these tests, if for no other reason than to know how 
 efficient the separator is. 
 
 No calorimeter should be used that cannot be worked 
 continuously over an extended period, and also, no 
 calorimeter should be used where allowance must be 
 made for the specific heat of the materials in the 
 
 Fi 223*. 
 
 under boiler pressure while the engine is at rest ; the 
 leakage will be shown by steam escaping through the 
 indicator cock. 
 
 The steam admitted to the jacket can be measured 
 by collecting the water formed in the jacket by con- 
 densation. 
 
 The total quantity of steam, S 2 , lost by leakage and 
 condensation in engines which are in average working 
 order, may be approximately estimated from the follow- 
 ing formula due to Professor Voleker : 
 
 /ah P 
 
 S 2 = (0-0527 x y' — ^— + 0-0121) lb. per second (175) 
 
 where C is the piston speed in feet per second. 
 
 The following is a description of the calorimeter test 
 referred to above, and is due to Messrs. Musgrave and 
 Sons, of Bolton : 
 
 instrument, as there is too great a liability to error in 
 both cases. 
 
 Fig. 224 shows a simple form of apparatus for the 
 purpose. It consists of a small tank, A, carefully 
 covered with non-conducting material, B, and encased 
 in tin, C, to protect the covering. The tank contains a 
 worm, D, the upper portion of which is connected as 
 closely as possible with the steam-pipe, and the connec- 
 tions carefully covered with non-conducting material. 
 The connecting-pipe should be carried to the centre of 
 the steam-pipe, and the end bent, as shown at E, to face 
 the current of steam. The lower end of the worm, D, 
 is connected to a thermometer, G, and from this a pipe 
 is led to a tank or barrel. The condensing water first 
 passes the thermometer F, and enters the bottom of 
 the tank, A, then passing up around the coils, it passes 
 out at the top, and is led away to a tank. The tempera- 
 
THEORY OF STEAM ENGINES. 
 
 207 
 
 ture is indicated by the thermometer H just as the 
 water leaves the tank. 
 
 In operation, the tank is first filled with condensing 
 water, and a small stream allowed to run through. 
 Steam is then admitted to the room, and the Yalve 
 adjusted to allow the desired amount to pass through 
 the worm. The quantity of condensing water is then 
 regulated to give the desired temperature of the over- 
 flow at H (about 160 deg. is a suitable temperature). 
 When the temperatures at all the thermometers have 
 become constant, start the test by simultaneously 
 turning the overflows of condensing water and con- 
 densed steam into their respective tanks. Note the 
 pressure of steam with the temperatures at all thermo- 
 meters. (These should be recorded every five minutes 
 during the test.) In ending the test at the same 
 instant, turn both overflows out of the tanks, and 
 then shut off the steam and condensing water. 
 
 Fig. 224. 
 
 The apparatus will be at the same temperature at 
 the end as at the beginning of the test ; therefore, no 
 account need be taken of its specific heat, as no change 
 has taken place, and from the low temperature and 
 careful protection, the radiation from the instrument 
 will be inappreciable. 
 
 Weigh the condensed steam and the condensing 
 water separately. Multiply the weight of condensing 
 water by the units of heat each pound has received, 
 and the product is the quantity of heat the steam has 
 lost. Multiply the weight of condensed steam by the 
 units of heat each pound lost after it was condensed to 
 water, subtract this product from the first one, and 
 divide the remainder by the latent heat of the steam. 
 If the quotient is less than the weight of condensed 
 steam, the difference between the two must have entered 
 the calorimeter as water. If the quotient is more than 
 the weight of condensed steam, the steam has been 
 superheated, and to find the number of degrees of super- 
 heat, multiply the weight of condensed steam by the 
 number of units of heat each pound has lost, both as 
 
 steam and water, and subtract this product from the 
 quantity of heat received by the condensing water ; 
 divide the remainder by the weight of the steam con- 
 densed, and by the specific heat of steam under constant 
 pressure; the quotient will be the degrees of superheat, 
 and this amount added to the temperature of saturated 
 steam in the tables, should correspond with the readings 
 of the thermometer I. 
 
 Example. — Take^„ = 1001b., to which corresponds a 
 temperature of 327 "9 deg. F. 
 
 Weight of condensed steam 901b., and temperature 
 70 deg. F. 
 
 Weight of condensing water, 1,0001b. ; temperature 
 at inlet 60 deg. F., and at outlet 160 deg. F. 
 
 Latent heat contained in lib. of dry saturated steam 
 at 327-9 deg. F. from 32 deg. F. is 883' 1 heat units. 
 
 Sensible heat contained in the latter is 298'3 heat 
 units. 
 
 Total heat contained in lib. of steam at 327'9 deg. 
 F. from 32 deg. F. is 1,181-4 heat units. 
 
 Difference of heat contained in lib. of water at 160 
 deg. F. and at 60 deg. F. is 100-365 heat units. 
 
 Difference of heat contained in lib. of water at 70 
 deg. F. and at 32 deg. F. is 3302 heat units. 
 
 Heat units received by condensing water : 
 100-365 x 1,000 = 100,365. 
 
 Heat units lost by the water to which the steam was 
 condensed : 
 
 (298-3 - 3802) x 90 = 23,425-2. 
 Weight of steam condensed to water : 
 100,365 - 23,4252 
 
 883-1 
 
 = 87-lllb. 
 
 Percentage of water contained in the steam : 
 
 (90 - 8711) x 100 _ 01 , 
 — = 3 21 per cent. 
 
 90 
 
 Had it required 1,0401b. of condensing water to con- 
 dense the steam, other conditions being the same, the 
 result would have been : 
 
 Heat units received by condensing water : 
 
 100-365 x 1,040 = 104,379-6. 
 Heat units lost by saturated steam : 
 
 (1,181-4 - 38-02) x 90 = 102,904-2. 
 Amount of superheat : 
 
 104,3796 - 102,9042 
 
 90 x 0475 
 
 = 34-5 deg. F. 
 
 where - 475 is specific heat of steam at constant pres- 
 sure. 
 
 Temperature of steam : 
 
 327-9 + 34-5 = 362"4 deg. F. 
 Losses in Steam Engines Summarised. 
 In summarising the losses which occur in a steam 
 engine while performing its duty, we must assume that 
 the steam in entering the engine stop-valve is perfectly 
 dry, either in the form of dry saturated steam or as 
 slightly superheated steam. It is unfair to the engine 
 
208 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 to test it with wet steam, as the suspended water is 
 inactive, and will produce losses which are not due to 
 faults in the engine. In testing the engine, the pressure 
 of the steam should be measured at the engine stop- 
 valve, and not at the boiler, as fall of pressure between 
 boiler and engine is not due to faults in the engine. 
 
 The losses which must occur more or less necessarily 
 in a steam engine may be tabulated as follows : 
 
 1. Throttling of the Steam before entering the Steam- 
 Chest. 
 
 (a) This may be caused by the stop-valve not being 
 fully opened, and also by the passages through the stop- 
 valve and the steam-pipe, between the stop-valve and 
 the steam-chest, not being wide enough. 
 
 (6) It will occur when the quantity of steam admitted 
 to the engine is controlled by means of a throttle-valve 
 worked by the governor. 
 
 2. In the Steam-Chest. 
 
 (a) Eadiation of heat from the external surfaces of 
 steam-chest. 
 
 (b) Heat conducted through eccentric rods. 
 
 (c) Steam required to compensate heat radiated from 
 slide-valve into exhaust-pipe. 
 
 (d) Leakage of steam through stuffing-boxes and 
 joints. 
 
 (e) Leakage of steam between valves and wearing 
 surfaces into exhaust-pipe. This can be tested as 
 described above. 
 
 3. In the Cylinder. 
 
 (a) Loss due to wire-drawing caused by the valves 
 opening and shutting gradually. The wire-drawing 
 will be augmented if the steam passages between steam 
 chest and cylinder are narrow. 
 
 (b) Condensation of part of the high-pressure steam 
 during admission, caused by the steam passages, 
 cylinder- walls, covers, piston, piston rod, etc., having 
 been cooled during the periods of expansion and 
 exhaust. Also condensation of some of the steam 
 during the early part of expansion ; this is due to the 
 cylinder not being warm enough to keep the steam at 
 the point of saturation. Some of the condensed steam 
 will be re-evaporated at a low pressure during the 
 latter part of expansion, and will do work on the 
 piston. But this re-evaporation requires heat, which 
 must be extracted from the cylinder walls, piston, etc., 
 whereby the temperature of these parts will fall. 
 
 (c) If the re-evaporation of the condensed steam has 
 not been completed before the release takes place, the 
 heat required for re-evaporating the remaining water 
 will be carried into the exhaust and thus wasted. 
 
 (d) Eadiation of heat from external surfaces of 
 cylinders which are not jacketed. 
 
 (e) Leakage through stuffing-boxes and joints. 
 (/) Leakage between cylinder walls and piston. 
 
 (g) Waste of steam in blowing through drain-cocks. 
 
 (h) Waste of steam in filling clearance volume at 
 every stroke. This loss can be diminished by a proper 
 amount of compression. It is evident that for economy 
 in compression, it is necessary that the work done 
 on the compression steam should be smaller than the 
 energy required for producing the steam saved by com- 
 pression. 
 
 (i) As the exhaust steam must pass through passages 
 of limited cross-sections, it follows that the pressure of 
 the exhaust steam will be greater than that of the 
 atmosphere or that of the condenser. 
 
 4. In the Jacket. — The object of the jacket is to 
 diminish the loss of heat due to condensation in the 
 cylinder. In an ideal jacket, such as assumed in 
 Tables B and C, the heat given off by the condensation 
 of the jacket steam is spent in doing work on the piston 
 by heating the cylinder steam and preventing con- 
 densation of the latter. Of course the ideal jacket can 
 not be carried out in practice, and therefore loss of 
 heat is the consequence. 
 
 (a) As the material of the jacket cylinder and covers 
 is not a perfect non-conductor, a certain quantity of 
 heat is wasted by radiation. 
 
 (6) As the quantity of heat which will be given off 
 by the jacket steam to the cylinder steam depends on 
 the conductivity of the material of which the cylinder- 
 barrel is made, it follows that an actual steam-jacket 
 will be less effective than an ideal one. 
 
 (c) If the temperature of the jacket is less than that 
 of the steam-cylinder, heat will be extracted from the 
 latter tore-evaporate condensed jacket steam accumu- 
 lated on the external surface of the steam-cylinder. 
 The jacket should therefore be well drained. 
 
 (d) Leakage of jacket steam into the exhaust-pipe 
 through joints between jacket-cylinder and steam- 
 cylinder. 
 
 5. Fall of Pressure between the Cylinders of Compound 
 Engines. 
 
 (a) Condensation of steam in pipe connections and 
 receivers. 
 
 (b) Expansion of steam in the latter parts while 
 doing no useful work. 
 
 (c) Friction between steam and walls of passages. 
 
 (d) Leakage through joints. 
 
 6. Mechanical Losses. 
 
 (a) Friction between the working parts of the 
 engine. 
 
 (b) Energy spent in driving feed pumps, air pumps, 
 and circulating pumps. 
 
 (c) Energy spent in driving the governor. This loss 
 may be considerable in engines working with automatic 
 expansion gear. 
 
DETAILS OF STEAM ENGINES. 
 
 209 
 
 CHAPTER XII. 
 
 DETAILS OF STEAM ENGINES. 
 
 The Cylinder. 
 
 \ HE steam-cylinder is a cast-iron pipe accu- 
 rately bored inside, and ending in faced 
 flanges, to which the covers are fixed by 
 studs. The cylinder barrel has to withstand 
 internal fluid pressure, like a steam boiler, 
 and its thickness might therefore be calcu- 
 lated in the same way as that of a boiler shell. The 
 thickness of the steam-cylinder barrel is, however, 
 always in excess with regard to strength, partly for the 
 purpose of allowing it to be rebored when worn, and 
 partly for making it convenient to cast. 
 
 the piston rod of the air pump, which latter is placed 
 behind and in line with the steam-cylinder. Gd are 
 guards to prevent accidents which might otherwise be 
 caused by the reciprocating movement of the tail rod. 
 E is the exhaust-channel. S C is the steam-chest 
 which is bounded by a cast-iron box fixed by bolts to 
 the main casting, and shut from outside by cover C. 
 S F is a faced surface on which the slide-valve works. 
 The slide-valve rod slides in stuffing-box S V, and the 
 cut-off valve rod, which passes right through the steam- 
 chest, slides in the two stuffing-boxes C V and V T. 
 To diminish loss of heat by radiation, the cylinder 
 
 i i i i ; i i i i i i 
 
 Fig. 225. 
 
 Non^Jacketed Cylinders. — Fig. 225 illustrates the 
 steam-cylinder, with steam passages, and steam-chest, 
 of a simple double-acting horizontal engine. The main 
 casting, containing the cylinder barrel, with steam 
 passages, P lt rests on four legs on the foundation frame, 
 to which it is fixed by eight bolts, F B. 
 
 The cylinder covers are turned to fit into the 
 cylinder ends, and are fixed by studs, Sd, through the 
 flanges. P B and T R are stuffing-boxes for the piston 
 rod and the tail rod respectively. Tail rods are 
 sometimes used in horizontal engines to diminish the 
 wearing of the cylinder caused by the weight of the 
 piston and piston rod. 
 
 In horizontal condensing engines, the tail rod forms 
 
 barrel is covered with non-conducting material, such 
 as felt, held in position by a sheet of steel laid on the top. 
 
 Jacketed Cylinders. — The cylinders of a compound 
 (receiver) engine, made by Messrs. Eansomes, Sims, 
 and Jefferies, of Ipswich, are illustrated in Figs. 226 
 and 227, the former being a section through the centre 
 line A B, and the latter a section through hue C D. 
 
 The main casting consists of the jacket cylinders, 
 J C, steam-chests, S x G 1 and S 2 Co, and receiver, E. 
 Cylinders J C are bored to receive the working cylinder 
 barrels, C B, which are cast separately and turned on 
 the outside to fit into the jacket cylinders. After 
 having been bored, the working cylinder barrels are 
 forced into the main casting by hydraulic pressure, and 
 
210 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 when in their proper position' a finishing cut is given 
 to the bore, so as to ensure perfect accuracy with the 
 cylinder faces. The space surrounding the barrels, 
 C B, when filled with steam, forms the steam-jackets, 
 S J. The jackets are tested for leakage under water 
 and steam pressure before the cylinder-covers are 
 put on. 
 
 H P C is the high-pressure cylinder. LPC is the 
 low-pressure cylinder. S I is the steam-inlet, the 
 steam passing, as shown by arrows, into the high- 
 pressure steam-chest, S x C lt which is shut from outside 
 by cover G x G v P x are high-pressure steam passages, 
 which alternately takes live steam from steam-chest 
 S x C x into the cylinder, and exhaust steam into exhaust 
 
 are 
 and is 
 
 slide valves to work on. hj are holes for slide-valve 
 rods ; the stuffing-boxes are not shown. dr 
 cylinder-drains. B dr is the receiver-drain 
 carried to the back of the cylinders. 
 
 J P is the jacket pocket, into which the condensed 
 jacket steam collects. The jacket should always be 
 kept free of water, as explained at the end of last 
 chapter ; the water is therefore continuously drained 
 off through a trap, which prevents the escape of steam. 
 A C is a facing for a steam-cock, by which high-pressure 
 steam is taken to the receiver and the low-pressure 
 steam-chest when starting the engine. F W is a facing 
 for a cock, by which a portion of the exhaust steam can 
 be taken to heat feed-water. L B are facings for 
 
 HPC 
 
 €-' 
 
 D 
 
 Fig. 226. 
 
 channel, E x G v B is the receiver, into which the 
 exhaust channel, E x C 1; opens. 
 
 The steam flows through the receiver, as shown by 
 arrows, into the low-pressure steam-chest, S, C 2 , 
 which latter is shut from outside by cover, C 2 C 2 ." P„ 
 are the low-pressure steam passages, which will be 
 seen to have a larger sectional area than passages V lt 
 the reason being that the low-pressure steam is not so 
 dense as the high-pressure steam. B 2 C, is the exhaust 
 channel, through which the exhaust steam from the 
 low-pressure cylinder is carried into the condenser or 
 into the atmosphere. 
 
 C E are the crank end cylinder covers provided with 
 stuffing-boxes, S B, for the piston rods. Holes h are bolt- 
 holes for attaching guide blades. B C are the bottom 
 end cylinder-covers, SF are faced surfaces for the 
 
 attaching lubricators. Cv is the cylinder covering, 
 and consists of asbestos non-conducting composition, 
 covered over with steel lagging sheets, cold rolled and 
 close annealed, presenting a very smooth surface, 
 which is painted and varnished. F F is the wrought- 
 iron foundation frame. 
 
 Steam-Trap. — The object of a steam-trap is to allow 
 the condensed water in jackets or steam-pipes to pass 
 out without permitting the steam to follow. Fig. 228 
 is an illustration of Hopkinson's self-acting steam-trap. 
 The condensed water will soon fill up the space between 
 the inner movable cylinder and the cylindrical casting, 
 and thus lift the former against the end of the outlet 
 pipe. Steam and water is thus prevented from passing 
 through the outlet, until the weight of the water accu- 
 mulated in the inner cylinder is sufficient to overcome 
 
DETAILS OF STEAM ENGINES. 
 
 211 
 
 the buoyancy of this cylinder. At that moment, the 
 lower end of the outlet pipe will be free and a small 
 quantity of water will be discharged, but no steam will 
 
 ones. The elasticity of spring rings is often increased 
 by hammering. 
 
 Davey, Paxman, and Co.'s Pistons.— Figs. 229 and 230 
 
 i i I i i l i i i 
 
 2 Fech 
 
 Fig. 227. 
 
 follow. The object of the valve at the top is to permit 
 
 air to escape. 
 
 The Piston. 
 
 The piston should fit steam-tight in the cylinder, to 
 prevent leakage of steam from the steam end into the 
 exhaust end of the cylinder. This could be done by 
 turning the piston to fit accurately into the cylinder, 
 but as both cylinder and piston wear in working, 
 leakage must occur after a time. 
 
 The piston is made steam-tight by employing elastic 
 metallic spring rings made of cast iron, steel, or gun- 
 metal. The spring ring is split at one place so as to 
 allow its diameter to be varied. The initial diameter 
 of the ring is greater than that of the cylinder. The 
 ring will therefore exert a pressure against the cylinder 
 barrel, which should be sufficient to keep the piston 
 steam-tight. 
 
 To prevent steam from leaking through the split of 
 the ring, two or more rings are placed side by side and 
 close together in such a manner that the splits of any 
 two rings are not in the same line. The thickness of 
 the ring may be uniform, or it may be thickest in the 
 middle and thinnest at the ends where it is split. 
 
 For small pistons one set of rings may be sufficient, 
 but with larger pistons two sets of rings are employed, 
 the inside ring or rings pressing against the outside 
 
 show a piston designed by this firm for their horizontal 
 engines. 
 
 ■III, 1 :' ■'11 
 
 INLET 
 
 =^-- "-' UTfutlet 
 
 Fig. 228. 
 
 The piston body, P, is made of cast iron, and is 
 hollow in order to diminish the momentum of the 
 piston. To prevent the clearance being increased by 
 
212 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 the hollow piston, the latter is closed by a cast-iron 
 disc, J E, the junk ring, which is fastened to P by six 
 steel studs, St, as shown. 
 
 The piston rod, P E, is of steel, and is fastened to 
 the piston by tightening the nut up against the junk 
 ring. This nut fits into a hollow in the cylinder-cover 
 in order to diminish the clearance ; the nut is prevented 
 from working loose by a pin, as shown. The spring 
 rings S 1 are of cast iron, and their wearing surfaces are 
 case-hardened. The inner spring ring, S 2 , is of steel. 
 
 Part of the projections at the two ends of the piston 
 are cut away for the steam-ports, so as to allow of an 
 easy admission of steam. 
 
 In the low-pressure piston, the spring rings; S 1 , are 
 made of cast iron, but in the high-pressure piston it is 
 preferred to make these rings of phosphor bronze, to 
 prevent scurring in case lubrication should fail. The 
 spring S 2 is a spiral steel spring with two turns. 
 
 Piston Speed. — In actual engines the velocity of the 
 crank-pin is controlled by the flywheel, and may be 
 assumed to be uniform. The motion of the piston can 
 therefore be determined ; but the simplest case will be 
 that where the connecting rod is very long compared 
 with the length, r, of the crank. This particular case 
 is illustrated in Fig. 233, where C P is the crank-pin 
 moving in a circle, with as centre and r as radius. 
 
 Fig. 229. 
 
 Fig. 230. 
 
 Fig. 231 is a drawing of the low-pressure piston of 
 the Windsor high-speed engine, made by the same firm. 
 This piston is made of cast steel, and is very light in 
 order to balance the smaller high-pressure piston. This 
 precaution is of great importance in high speed work. 
 The spring rings, S, are of steel, and are sprung over 
 the piston when brought into position. The piston 
 rod is made of steel. 
 
 Bobey and Go.'s Piston. — Fig. 232 shows section of a 
 piston designed by this firm. The junk ring is not 
 fastened to the piston body by studs, but is held in 
 position by the nut, N. The piston rod, P E, as well 
 as the nut, N, are of steel. 
 
 Assuming the crankshaft to make n revolutions per 
 second, the velocity of the crank-pin will be 2 ir r n feet 
 per second. At the moment the crank has turned 
 through an angle <£, reckoned from the dead-point a, 
 the piston must have traversed the distance af=x, and 
 it is evident that we must have 
 
 x = r (1 - cos <j>) . . . . (176) 
 
 The velocity of the piston will be 
 
 2 it rn x sin $ feet per second . . (177) 
 
 The velocity of the piston, therefore, is not uniform, 
 but increases from zero to a maximum during the first 
 
DETAILS OF STEAM ENGINES. 
 
 213 
 
 half part of the stroke, and then diminishes until it 
 becomes zero again at the end of the stroke. 
 
 The acceleration or rate of change of piston velocity 
 will be 
 
 4 tt 2 n 2 r x cos feet per second per second . (178) 
 
 and if by M we denote the mass in pounds of the 
 piston, piston rod, cross-head, and connecting rod, the 
 pressure on the crank-pin due to the variable motion of 
 the piston will be 
 M 
 
 Schmidt, gives a fair idea of the relation between the 
 piston speed, C, and the size of engines. 
 
 G = a (10+ VAHP) feet per minute (180) 
 The values of a are given in the following table : 
 
 Values of Speed of piston may be 
 
 u.. considered as 
 
 32-187 
 
 x 4 7T 2 n 2 r x cos pounds. 
 
 (179) 
 
 Fig. 231. 
 
 This pressure is maximum for = and = 180, and is 
 zero at mid-stroke. As the direction of the pressure is 
 positive between = and = 90, and negative between 
 = 90 and = 180, it will diminish the crank effort during 
 the first half part of the stroke and increase it during 
 the latter part of the stroke. In engines working with 
 expansion, the inertia of the moving masses will there- 
 fore nave the effect of making the crank effort more 
 uniform. 
 
 Engineers understand by the piston speed the mean 
 velocity of the piston in feet per minute. The piston 
 speed will therefore be 2 I n feet, where I is the stroke 
 in feet and n the number of revolutions per minute ; it 
 varies with the different classes of engines, and for 
 the same class it increases with the size of the engine. 
 The following formula, which is due to Professor G. 
 
 9-84 
 
 Very, slow 
 
 1378 
 
 Slow 
 
 17-72 
 
 Normal 
 
 21-65 
 
 High 
 
 25-59—39-37 
 
 Very high 
 
 The Piston Bod is subjected 
 
 to compression while the 
 
 Fig. 232. 
 
 piston moves towards the crank, and to tension during 
 the return stroke. As the length of the rod is great 
 
 compared with its diameter, it is liable to bend if not 
 sufficiently strong. The diameter, d, of the piston rod 
 
214 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Fig. 234. 
 
 45- 
 
 a* 
 
 :lo 
 
 \i 
 
 ^ 
 
 ^1 
 
 ^ 
 
 S,B 
 
 CHP 
 
 T 
 
 I 
 
 Fig. 235. 
 Fig. 237. 
 
 Fig. 236. 
 
 DP 
 
 Fig. 238. 
 
 should therefore be calculated with due consideration to where K is a constant whose value depends upon the 
 
 its length. Using the same letters of reference as on material of the rod and the factor of safety. The latter 
 
 page 203 we must have varies immensely in practice, but if the rod is made of 
 
 .J— steel, we may take, according to Eeuleaux, the average 
 
 ■ valine of K as - 03, and formula (181) will therefore be. 
 
 d 
 D 
 
 - K ^iV p ' 
 
DETAILS Oh STEAM ENGINES. 
 
 215 
 
 S - °V:e> V* 
 
 (182) 
 
 where p e is the maximum effective pressure on the 
 piston in pounds per square inch. 
 
 It is evident that every part of the rod must be able 
 to withstand the tensile stress to which it is subjected, 
 and for that reason the sectional area of the rod must 
 
 nowhere be smaller than — — , where d can be found by 
 
 d. 
 T> 
 
 57 + 0-5 p e 
 
 1,000 
 This formula is also due to Eeuleaux. 
 
 (183) 
 
 Fig. 239. 
 
 SB 
 
 4 6 8 10 12 INCHES 
 
 Fig. 240. 
 
 The Cross Head. 
 
 The object of the cross-head is to connect the 
 piston rod with the connecting rod ; the linear re- 
 ciprocating motion of the former is thereby trans- 
 formed into that of rotation. The piston rod is rigidly 
 fastened to the cross-head by a cotter, and in order 
 to prevent the rod being bent, the motion of the cross- 
 
 head is guided by making it slide between the parallel 
 surfaces of the guide blades. 
 
 The connection between cross-head and connecting 
 rod is made through the cross-head pin, which is usually 
 fastened to the cross-head and fits into a hole in the 
 connecting rod end. The connecting rod will thus 
 oscillate on this pin. 
 
 Single-acting engines have neither piston rod nor 
 cross-head, the connecting rod being attached direct to 
 the piston by a pin on which it oscillates. 
 
 In engines with oscillating cylinders, the cross-head 
 as well as the connecting rod are dispensed with, the 
 piston rod being connected direct to the crank-pin. 
 
 Paxman's Cross-heads. — 1. The cross-head shown in 
 Figs. 234 and 235, is forged out of one piece of steel. 
 The piston rod, P E, fits into a hole in the socket, S O, 
 and is fastened by a steel cotter, C O, which is further 
 secured by a pin, as shown. The other end of the 
 cross-head is forked, and embraces the connecting rod. 
 The cross-head pin, CHP, is fixed by a pin, p, and 
 ends in two journals, see Fig. 236, which fit into holes 
 in the slide blocks, S B. 
 
 CHP 
 
 Fig. 241. 
 
 Figs. 237 and 238 show the guide blades, GB, for 
 this cross-head. The slide blocks slide in grooves cut 
 in the guide blades. The height of a slide block is h, 
 and its width is w. The guides are fixed by bolts to 
 the foundation-frame ; each bolt has a shoulder, which 
 bears against the bottom guide blade, the latter will 
 therefore not get loose when the nut, which keeDS the 
 top guide blade in position, is unscrewed. Distance- 
 pieces, D P, are inserted to keep the guides the right 
 distance apart. The guide blades, as well as the slide 
 blocks are made of hard cast iron. 
 
 2. The cross-head shown in Figs. 239, 240, and 241 
 is made of cast steel, and is designed for girder engines. 
 The piston rod, P B, is cottered to the socket, S O. The 
 forked end of the cross-head embraces the connecting rod, 
 and the cross-head pin, C H P, is held by the two caps, 
 
216 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Cp, the grip of which can be adjusted by the bolts, AB. 
 The slide blocks, S B, are made of hard cast iron, and 
 are fixed to the cross-head by bolts as shown. The 
 position of the slide blocks is adjusted by screws, Sc. 
 The guide blades are replaced by the bored-out upper 
 and under surfaces of the girder trunk. On each side 
 
 are made to bear against the cross-head pin by cap, Cp, 
 and bolts AB. This cross-head is adopted for the 
 Windsor high-speed vertical engine, and the slipper 
 block, S B, slides on the standard which carries the 
 steam-cylinder, the slipper slides being bolted to the 
 standard and embracing the slipper block. 
 
 Fig. 242. 
 
 of the cross-head is an oil-box, as shown, from which 
 the oil is sucked up by a wig and made to drop on the 
 connecting rod end ; and through a hole in the latter 
 the oil is carried to the cross-head pin. Which of the 
 two oil-boxes is to be used depends upon whether the 
 girder is left-handed or right-handed. 
 
 Fig. 243. 
 
 The Bobey Gross-head for girder engines is illustrated 
 in Pigs. 244 and 245. To facilitate the removal of the 
 piston rod, P B, from the cross-head, the former is 
 screwed and furnished with a large nut, as shown in 
 Fig. 245. By tightening the nut the rod is easily with- 
 drawn, when the cotter, C 0, is taken out. The cross- 
 
 Fig. 244. 
 
 3. The cross-head, Figs. 242 and 243, is designed for 
 slipper slides. The piston rod, P B, the cross-head, 
 C H, and slide block, S B, are forged from one piece of 
 steel. The connecting rod end is forked and embraces 
 the cross-head. The cross-head pin, C H P, is firmly 
 shrunk into the eyes of the connecting rod end, and 
 oscillates jn the adjustable gunmetal steps, St, which 
 
 Fig. 245. 
 
 head itself is made of malleable iron, the box end, B, 
 receives the connecting rod end, which it embraces. 
 The slide blocks, S B, are of hard cast iron, and are 
 fastened by screws to the cross-head. Should it be 
 necessary to adjust the slide blocks, Messrs. Kobey and 
 Co. prefer to insert thin metal liners between the blocks 
 and the cross-head body, instead of employing adjusting 
 
DETAILS OE STEAM ENGINES. 
 
 217 
 
 screws. The cross-head pin, which is embraced by the 
 connecting rod end, is prevented from working loose in 
 the sides of the cross-head by the following arrange- 
 ment : 
 
 The holes in both sides of the cross-head, see Fig. 
 244, are bored conically, and the pin is correspondingly 
 tapered under the head. A key, K, which fits into a 
 slot, prevents the pin from turning. The other end of 
 the pin is fitted with a loose conical spring ring, S B, 
 which is split to allow it to be compressed. On tighten- 
 ing the nuts on the outer end of the pin, both cones are 
 pressed well home, and the centrality and tightness of 
 the pin is secured. 
 
 The Connecting Rod. 
 
 The motion of the connecting rod is that of rotation 
 in a plane at right angles to the crankshaft. Like the 
 piston rod, the connecting rod is alternately subjected 
 to compression and tension, and its diameter must 
 therefore be calculated with due regard to its length. 
 If by P we denote the force transmitted through the 
 cross-head to the connecting rod, then Q = P x sec /?, 
 see Fig. 246, is the force by which the rod is corn- 
 
 factor. In practice, the safety factor varies consider- 
 ably, but on an average we may take it as 25, and the 
 diameter of the rod will then be 
 
 d = 0-0363 V Q V L inches 
 
 (185) 
 
 The connecting rod is made thicker at the centre, 
 where the bending moment is greatest. The ends of 
 the rod contain bearings for the crank-pin and cross- 
 head pin, except when the cross-head end is forked 
 in which case the cross-head pin is fastened to the 
 connecting rod end, and finds a bearing in the cross- 
 head, as in Figs. 242 and 243. 
 
 Bobey's Connecting Bod with Box Ends. — The con- 
 necting rod, shown in Figs. 247-252, is forged from 
 steel in one piece with the crank-pin end, Figs. 247 
 and 248. The cross-head end is cottered to the rod, 
 as shown in Figs. 249 and 250. The brasses or gun- 
 metal steps, St, at the crank-pin end are adjusted by 
 a screw and wedge, C 0, and embrace the crank-pin, 
 C P. By this adjustment, the connecting rod will be 
 lengthened when the brasses are brought together to 
 compensate the wear. Fig. 251 shows the shape of the 
 
 Fig. 246. 
 
 pressed. P is partly due to the total effective steam 
 pressure, Pj, on the piston, and partly to the inertia of 
 the moving masses, which force, P 2 , is given in formula 
 (179). During the earlier part of the stroke, until the 
 piston speed has reached its maximum velocity, P is 
 equal to ~P 1 - P 2 , and during the remainder of the stroke 
 we have P = P\ + P,. Let now L denote the length 
 
 maximum 
 then the 
 
 of the connecting rod in inches, 
 P x sec j8, and / the factor of 
 diameter of steel rods must be 
 
 d 
 
 Q the 
 safety, 
 
 (184) 
 
 0-0162 V/TVQ s/lT inches . 
 
 Besides the compressive stress, the connecting rod is 
 also subjected to bending action at right angles to the 
 rod, due to the rotary motion of its particles. The 
 curve A, Fig. 246, is that described by a particle, a, in 
 the middle of the rod, and it will be seen that when the 
 rod is nearly at right angles to the crank, the deviating 
 force will be at right angles to the rod. This bending 
 action may be considerable in high-speed engines. 
 
 It is, however, usual to take Q as maximum P x 
 sec fl only, and to allow for the straining actions due 
 to the inertia of the moving masses in the safety 
 
 steps ; they are kept in position by a cap (not shown) 
 screwed on the crank-pin. The adjustment of the steps 
 at the cross-head end is done by tightening up the 
 cotter, whereby the connecting rod, C B, will be 
 pushed further into the socket, S O, and the box, B, 
 and will thus be shortened. The adjustments at the two 
 ends of the connecting rod will therefore compensate 
 each other, so that the rod remains always the same 
 length. Fig. 252 is an end view of Fig. 250. The 
 diameter of the connecting rod at the middle is 5£in. 
 All parts, except the brasses and the lubricator, L, are 
 made of steel. 
 
 Paxman's Connecting Bod with Strap Ends. — (a) 
 Both ends of the connecting rod, C B, Figs. 253-256, 
 are made flat to receive straps, Sp, which hold the 
 steps, St. The straps are bolted to the connecting rod 
 ends as shown, and the steps are adjusted by tighten- 
 ing the wedges, CO. The connecting rod is 5ft. 
 between centres of crank-pin and cross-head pin, and 
 its diameter at the centre is 2|in. 
 
 (b) The crank-pin end of the connecting rod, Figs. 
 257 and 258, is strapped like the one just described. 
 The cross-head end, Figs. 259 and 260, is in one solid 
 
'218 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Fig. 247. 
 
 r.v r 
 
 A 
 
 Fig. 251. 
 
 St. 
 
 2 4 6 8 10 12 INCHES 
 
 jj i 1 I I i i i i i i i i 
 
 Fig. 248. 
 
 Fig. 250. 
 
 Fig. 252. 
 
 piece with the rod a hole being cut out to receive the The length of this rod is 10ft. between centres of C V 
 steps, winch are adjusted by tightening the wedge, C 0. and C H P, the diameter at the centre being 4f in 
 
DETAILS OF STEAM ENGINES. 
 
 219 
 
 3 
 
 CO 
 
 *4 
 
 >-* 
 
 P 
 
 CO 
 
 The material used for all parts of these two connect- Connecting Bod Ends with Cotter and Gib. — In 
 ing rods is steel, except for the steps. Figs. 261-264, are shown connecting rod ends with 
 
220 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Fig. 257. 
 
 Fig. 259. 
 
 Fig. 258. 
 
 Fig. 260. 
 
 Si— : 
 
 Fig. 263. 
 
 Cfc. 
 
 E 
 
 Co 
 
 Fig. 261. 
 
 Sr 
 
 i^ 
 
 Fig. 262. 
 
DETAILS OF STEAM ENGINES. 
 
 221 
 
 straps fastened to the rod by cotters, C 0, and gibs, 
 
 Gb. The rod is only 2ft. 2in. between centres, and 
 
 its largest diameter is lfin. 
 
 Marine Connecting Bod End. — The crank-pin end, 
 
 Figs. 265 and 266, is formed of two steps, St, which 
 
 are held together by the two adjusting bolts, A B. 
 
 The cap, Cp, is of steel, and the rod is forged in one 
 
 piece from steel or wrought iron ; the cross-head end 
 
 is forked, and a steel cross-head pin, CHP, is firmly 
 
 shrunk into the eyes. B is an oil-box, the oil being 
 
 carried through tube, T, to the crank-pin, C P. 
 
 Fig. 267 is an end view of the crank-pin end. This 
 
 connecting rod is designed for the Windsor high-speed 
 
 engine. 
 
 The Crankshaft.— The Bearings. 
 
 The energy exerted by the crank effort is transmitted 
 through the crankshaft in order to drive the machinery 
 
 arms and are afterwards riveted ; their dimensions are 
 7|in. long by 6in. diameter. 
 
 The shaft, which is of steel, is supported by two 
 beariDgs, and the journals, Jl, are 13in. long by 8^in. 
 diameter. In the middle, where the flywheel is fixed, 
 the diameter of the shaft is lljin. ; the drawing shows 
 the section of this part of the shaft with the key-way 
 for fixing the flywheel. 
 
 The two lines a, b, are centre lines of the girder 
 heads. The positions of the eccentric are marked by 
 crossed lines, and the letters HPS, AEG, and LPS 
 refer to the eccentrics for the high-pressure slide-valve, 
 automatic expansion gear, and low-pressure slide-valve 
 respectively. 
 
 End cranks are often made in the form of flat 
 cylindrical discs, which are shrunk on the shaft, and 
 thus have the advantage of being balanced. 
 
 Fig. 267. 
 
 Fig. 265. 
 
 Ir- 
 
 AB 
 
 t 
 
 CP 
 
 CP 
 
 
 S* 
 
 CR 
 
 OT 
 
 Fig. 266. 
 
 for the purpose of which the engine is working. As 
 the position of the axis of rotation must remain 
 unaltered, the shaft must be supported by proper 
 bearings. 
 
 To the crankshaft are attached the necessary eccen- 
 trics for working slide-valves and feed pumps, the gover- 
 nor driving gear, flywheels, and transmission pulleys. 
 
 Shaft with End Cranks. — Fig. 268 illustrates shaft 
 and arms of the Paxman 100 nominal H.P. girder com- 
 pound engine. The two crank arms, which are forged 
 from the best quality of Martin-Siemens steel, are set 
 at right angles and shrunk on the ends of the shaft. 
 H P Cr is the high-pressure crank, and LPCr the low- 
 pressure one. The length of each arm — i.e., the dis- 
 tance between centre of crank-pin and centre of shaft — 
 is 2ft., thus making the stroke of the engine 4ft. long. 
 The steel crank-pins, C P, are shrunk into eyes in the 
 
 Cranked Shafts. — (a) Figs. 269 and 270 are drawings 
 of the shaft of the Paxman 40 nominal H.P. horizontal 
 compound engine with cranks at right angles. The 
 cranks and shaft are forged from one piece of steel, and 
 the gaps between the arms are slotted out of the solid 
 forging. The crank journals, C P, are 5in. long by 6in. 
 diameter ; the crank arms are 12in. long, and the stroke 
 of the engine is therefore 2ft. The width of an arm, 
 measured parallel to the shaft, is 4Jin. ; and the thick- 
 ness, at right angles to the shaft, is 6|in. 
 
 The diameter of the shaft is 6in., and its length is 
 9ft. 2in., there being 3ft. 6in. between left-hand end of 
 shaft and centre line of high-pressure crank, 2ft. 7in. 
 between centre lines of cranks, and 3ft. lin. between 
 centre line of low-pressure crank and other end of shaft. 
 The shaft is supported by three bearings, the journals 
 being Jl 1 , Jl 2 (both 8fin. long), and Jl 3 , the length of 
 
222 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 which is llin. ; the diameters of all three bearings are 
 the same as that of the shaft. 
 
 between Jl 1 and HPCr; the position of the eccentric 
 for the low-pressure slide-valve is between HP Cr 
 and Jl 2 . 
 
 (6) The cranked shaft of the 25 nominal H.P. 
 Windsor compound engine is illustrated in Figs. 271 
 and 272. The shaft and cranks are forged in one from 
 
 ( 
 
 
 Nf 
 
 >" n 
 
 
 f 
 
 \ 
 
 v.. 
 
 yk 
 
 Jk. 
 
 *r 
 
 
 * 
 
 
 
 a. 
 
 1 
 
 
 z 
 
 \ 
 / 
 
 / 
 \ 
 
 
 
 l_ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 / 
 
 / 
 \ 
 
 trf— 
 
 
 w 
 
 > 
 
 < 
 
 
 
 a 
 
 
 
 3* 
 
 
 
 
 
 
 o 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 Oi — 
 
 
 
 
 
 u 
 a — 
 
 
 
 J 
 
 
 * — 
 
 
 
 
 
 
 a. 
 
 
 
 
 X 
 
 ■> 
 
 < 
 
 r — 
 
 
 o 
 
 
 
 
 
 UJ 
 
 
 
 
 
 «* 
 
 
 
 
 
 
 7 
 
 \ 
 
 o — 
 
 
 
 
 
 l 
 
 C3 
 
 
 
 
 
 2 
 
 The eccentrics for working the high-pressure slide- 
 valve and the automatic expansion gear are placed 
 
 steel, and the cranks are afterwards slotted. The shaft 
 is 5in. in diameter, it is 3ft. 5in. between left-hand end 
 and centre of high-pressure crank, 2ft. 4in. between 
 centres of cranks, and 2ft. 7in. between centre of low- 
 pressure crank and other end, The crank journals are 
 Sin. in diameter by 7iin. long, and the arms are Gin. 
 thick and 3in. wide, the latter dimensions being 
 estimated parallel to the shaft. 
 
DETAILS OF STEAM ENGINES. 
 
 223 
 
 The shaft is supported by three bearings ; the 
 journals, Jl 1 and Jl 3 , are 12in. long by 5in. in diameter, 
 and the centre journal, Jl 2 , is 7Jin. long by 5in. 
 diameter. The eccentrics are lettered as on Fig. 268, 
 and G W is the position of wheel for driving governor. 
 
 Bearings. — The function of the main bearings of the 
 engine is to furnish the reactions which are required 
 
 Fig. 273. 
 
 however, vary during one revolution as well as with 
 the load, and the tensions of the belts increase with 
 the load. For these reasons the bearings must be 
 designed so as to allow of adjustment in more than one 
 direction. 
 
 The main bearing of the Eobey compound girder 
 engine is illustrated in Figs. 273, 274, and 275. The 
 
 Fig. 275. 
 
 F B 
 
 CO — | M^'CO 
 
 Fig. 274. 
 
 for balancing the forces acting on the crankshaft. 
 These forces are partly the weights of the shaft, pulleys, 
 flywheels, and other masses carried by the shaft, and 
 partly pressures on the cranks and tensions of belts. 
 If the direction of the resultant of these forces remained 
 the same, then an adjustment of the bearings in that 
 direction would be sufficient to compensate for wear. 
 The directions and magnitudes of the crank pressures, 
 
 cast-iron foundation-plate, F F, has four feet, which 
 are planed underneath and fastened to the foundation 
 by four bolts, F B. The foundation-plate carries the 
 plummer block or pedestal, Pd, which is cast in one 
 with the girder, G-. The bearing itself has four cast- 
 iron steps, St, lined with anti-friction metal, M. Ex- 
 ternally the steps are cylindrical where they fit the 
 gap in the plummer block ; but the lateral steps are 
 
•2-24 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 provided with wedge-shaped projections to fit the 
 adjusting wedges, C ; the top step is held down by 
 the cap, Cp. 
 
 "When it becomes necessary to take up the steps to 
 compensate for wear, they can be tightened sideways 
 by drawing up the wedges, C 0, and in a vertical 
 direction by screwing down the nuts of CB. But 
 before this can be done the lateral steps must be taken 
 out and filed a little to enable them to come closer 
 together. All four steps can be taken out for examina- 
 tion without removing the crankshaft, which only 
 requires to be raised a little so as to allow the bottom 
 step to be drawn upwards round the shaft. 
 
 Bearing Surface. — The heat produced by friction 
 between journals and their bearings often causes great 
 trouble in machinery, the cooling surface being too 
 small for the heat to dissipate. When the journal fits 
 accurately in the bearing, so that the pressure is equally 
 distributed over the bearing surface, the friction caused 
 
 by the pressure, P, will be ^ P/, where / is the co- 
 
 efficient of friction. 
 
 Let P be given in pounds and the diameter of the 
 journal be d inches, then the work consumed by the 
 friction while the journal makes one revolution will be 
 
 V 12 I P ^ = -5 dP ^ foot -P omids - 
 
 If the journal makes n revolutions per minute, then 
 the rate of consuming work, expressed in horse-power, 
 
 wmbe^ o P -^H.P. 
 24 x 33,000 
 
 and the heat generated per minute will be 
 
 Ts*d~P fn BHI j 
 24x33,000x772 ' ' ' 
 
 / is usually taken as 0*05 for journals which are con- 
 tinually lubricated, but it may sometimes be very much 
 less. 
 
 The cooling surface is estimated as equal to the sur- 
 face of the journal, or w d I, where I is the length of the 
 journal. In order to keep the bearings cool, we must 
 have 
 
 ■a a I oc ■L 
 
 24 x 33,000 x 772 
 
 or, say, 
 
 i=£Pn (1S6) 
 
 where K is a factor to be determined by experience. 
 The friction between a journal and its bearing will 
 
 be smaller than -^ P/ when the bearing is worn. It is, 
 
 however, difficult to estimate the exact value of the 
 friction in this case, and it is also doubtful whether we 
 can consider the cooling surface as being proportional 
 to the surface of the journal. For these reasons it is 
 better to adopt empirical formulfe which are based on 
 actual practice. 
 
 According to Yon Eeiche, the length of a journal 
 may be taken as 
 
 Z-0-0010-2 fs*^ . . (187) 
 
 By comparing the latter expression with (186), it 
 will be seen that the length of journals, working under 
 ordinary conditions, may be increased less rapidly with 
 the pressure and the speed of rotation than journals 
 which fit accurately in their bearings. 
 
 Overhung Journals. — The diameter of a journal, 
 which is supported at one end only, such as the crank- 
 pin, Pig. 268, and subjected to a pressure, P, uniformly 
 distributed over its length, will be 
 
 where S is the greatest stress to be allowed in the 
 material. If the journal be made of steel, S may be 
 taken as 14,0001b. per square inch, and 
 
 d = 0-019^/T ^ inches . . (188) 
 
 Ob 
 
 and by eliminating I between (187) and (188) 
 
 d = 0-00719 s /p * ^ inches . . (189) 
 Dividing (187) by (189), we have 
 I 
 
 d 
 
 = 0-142 
 
 v n 
 
 (190) 
 
 Neck Journals. — For journals supported at both ends 
 and only subjected to a pressure, P, uniformly dis- 
 tributed over the length of the journal, the diameter 
 will be 
 
 and if we take S as 14,000 for steel 
 
 d = 0-00952 . H k A inches 
 
 W 
 
 E limin ating I between (1ST) and (191) 
 
 d = 0-00452 s fP~* Jn inches 
 and by dividing (1S7) by (192) 
 
 I 
 3=" 226 
 
 (191) 
 
 (192) 
 
 (193) 
 
 The diameter of a cross-head pin may be determined 
 by (192^ . but as the connecting rod only oscillates on 
 the pin, the heat produced will be less than on an over- 
 hung crank-pin subjected to the same pressure. The 
 length of the cross-head pin need not therefore be so 
 long as the corresponding overhung crank-pin. 
 
 The two connecting rods, Figs. 247-252 and Fi^s. 
 257-260, show that the ratio of the length of cross-head 
 pin to the length of overhung crank-pin may be taken 
 as 0-72 ; taking this ratio for granted, we have for 
 cross-head pins 
 
 3 = 0163 
 a 
 
 \ n 
 
 (194) 
 
 The crank-pins of the cranked shafts. Fi«js. 269 and 
 271. are partly subjected to the pressure, P, transmitted 
 through the connecting rod. and partly to a force, P p 
 which the journal has to transmit' from the one 
 crank-arm to the other. Let the length of the 
 
DETAILS OF STEAM ENGINES. 
 
 225' 
 
 journal be I, then the bending moment due to P will 
 
 PZ 
 
 be Ms = — - and that due to P x will be M&i = PjZ. 
 
 The resultant bending moment will be 
 U, = I J f 
 
 the shaft of the driving engine. The weight of the 
 armature and shaft is W, and C is its centre of gravity. 
 The shaft has two bearings, A and B ; and E x and B 2 
 are the reactions due to these bearings. 
 The equilibrium of W, B 1; and E 2 requires 
 
 144 
 
 + *V 
 
 The diameter of the journal will be 
 d 
 
 -V 
 
 32 
 
 S 
 
 Mr 
 
 (195) 
 
 As P and ~P 1 vary with the relative position of the 
 crank and the connecting rod, the maximum value of 
 M r must be taken in calculating the diameter. The 
 length of the journal may be taken from (187). 
 
 Strength of Shafts and of Shaft Journals. — Shafts 
 are usually subjected to combined torsion and bending. 
 If the shaft at any transverse section be subjected to a 
 bending moment, M &, and to a twisting moment, M t , 
 then the stress in the material at that section will be 
 as if the shaft were subjected to a bending moment, 
 M r> which, according to Eeuleaux, may be taken as 
 
 M r = 0-975 M b + 0-25 M f . . . (196) 
 
 when Mb > M ti and as 
 
 M r = 0-625 M 6 +0'6M| . . . (197) 
 
 when Mt > M 6 . The diameter of the shaft at this 
 particular section should not be less than 
 
 d 
 
 V: 
 
 32_ 
 
 M r 
 
 (198) 
 
 where S is, as before, the maximum stress to be allowed 
 in the material. 
 
 The effect of torsion is to twist the shaft a certain 
 angle, the magnitude of which should not be more 
 than ^ of a degree per foot run. For this reason, 
 the diameter of a steel shaft at such sections which 
 are subjected to torsion should not be less than 
 
 d = 0-3 Vm ( inches . . . (199) 
 
 If the number of horse-power transmitted by the 
 shaft be N, and the speed of rotation be n, then 
 
 12 x 33,000 x N = 63)()00 !N nds and incheg 
 2 7r n n 
 
 M 
 
 Inserting this value for M t in (199) we obtain 
 
 <Z = 4-75 
 
 V 
 
 inches 
 
 (200) 
 
 A 
 
 < a. 
 
 -- * 
 
 -a 
 
 t JB 
 
 
 
 
 
 / 
 
 
 
 
 
 
 . i ' 
 
 
 
 
 
 c 
 
 A 
 
 1 
 
 , 
 
 
 
 
 
 t 
 
 R, 
 
 y 
 
 
 n. 
 
 
 \ 
 
 
 
 ■' 
 
 ,w 
 
 
 
 
 Fig. 276. 
 
 Example. — Fig. 276 represents the armature and 
 shaft of a dynamo machine, which is coupled direct to 
 
 E, = W 
 
 -^— andE 2 = W 
 
 . (201) 
 
 a x + a 2 a-i + a 2 
 
 The maximum bending moment to which the shaft 
 is subjected, will be 
 
 M 6 = E 2 x a 2 = W ?LiL^. 
 <?! + a 2 
 
 The twisting moment will be 
 
 M e = 63,000 — 
 n ' 
 
 Take W = 2,6001b. ; N =» 70 ; n = 500 ; a t = 52in., 
 and a 2 = 40m., then we shall have 
 
 Ei = 1,1301b.; E 2 = 1,4701b. : M» = 58,800; and 
 
 M« = 8,820. As M 6 is greater than M t we have 
 
 M r = 59,535 pounds and inches. 
 
 Formula (198) gives the diameter of the shaft cZ = 
 3'51in. 
 
 If the shaft had only to withstand torsion, its 
 diameter should be determined by (200), and would 
 be d = 3in. 
 
 The greater value, 3-51in., must, of course, be taken. 
 If a key-way is to be cut into the shaft, having a radial 
 depth of, say, fin., then the diameter of the shaft 
 should be 4Jin. 
 
 Journal B. — This journal is subjected to a twisting 
 moment, M t = 8,820, and to a bending moment, M b =■ 
 ^E 2 Z 2 , where l 2 is the length of the journal. If Von 
 Eeiche's rule is to be followed, the length of the 
 journal should be 
 
 l 2 = 0-00102 V1.470 V500 3 = 4&in. 
 The bending moment would then be 
 
 M» = i x 1,470 x 4^ = 769. 
 As M t > M b the diameter of the journal should be 
 
 3 / 32 ~ 
 
 2 = V 7r x 14,000 [0 ' 625 x 769 + °' 6 x 8 >820] =lfin. 
 As the journal must not be twisted more than ^ of 
 a degree per foot run, its diameter must not be less 
 than 
 
 d. 
 
 d, = 4-75 
 
 V- 
 
 70 
 500 
 
 3in. 
 
 We may, however, make the length of this journal 
 greater than that given by Von Eeiche's rule, without 
 producing an excessive stress in the material. Suppose 
 we take Z 2 =2'5 diameters, then in our case 1. 2 would be 
 equal to 7 - 5in. The bending moment, M b would be 
 1x1,470x7-5 = 1,378. 
 
 The greatest stress, S, in the material would be found 
 by the following 
 
 3 
 
 d 2 = 3 
 
 3 /~32 
 
 (0-626x1,378 + 0-6x8,820) 
 
 which gives S = 2,3231b. per square inch only. 
 
"226 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 = 0142 ^500 = 3-175 
 
 Journal A.' — This journal is subjected to bending 
 only. Its diameter, a\, and length, l u may be deter- 
 mined by (189) and (190). 
 
 ^ = 0-00719 Vl,130 4 V500 = say ljin. 
 
 k 
 a\ 
 
 lj = 3|in. 
 
 Should it be considered advisable to make l± longer 
 than the value given above, then we must make 
 
 s /jg _ 
 
 d i -= V —, j x E x l T = 0-0714 8 VEj I, inches. 
 
 In practice l x is often made excessively long, for 
 instance, equal to l 2 , which in our case would require 
 tfj = 0714 8 Vl,l30x7-5 = say, 1 Jin. 
 
 Fig. 277. 
 
 Sh 
 
 U 
 
 Fig. 278. 
 
 by 
 
 The work consumed by friction would thus be increased 
 1-5 
 
 1-25 
 
 1-2. 
 
 The diameter, d ly will also be found to be greater 
 than the value given above — it is usually between 
 d 2 and 0'75 x d 2 ; the work consumed by friction would 
 be increased by 2*4 and T8 respectively. The greatest 
 stress in the material will be 2,3231b. and 5,5031b. per 
 square inch respectively, both of which are very low. 
 
 Eccentrics. — Axl eccentric is a crank, but the radius 
 of the crank-pin is greater than the crank arm, plus the 
 radius of the crankshaft. The crank-pin is made in 
 the form of a sheave, with a hole which fits on the 
 shaft. The distance between the centre of this hole 
 and the centre of the sheave is the length of the crank 
 arm, and is called the eccentricity. The sheave is 
 embraced by two straps, to one of which a rod — the 
 eccentric rod — is fixed. This rod with the straps con- 
 stitute a connecting rod. The other end of the 
 eccentric rod is connected to the piece of mechanism 
 whose motion is to be reciprocating, be it a pump or a 
 slide-valve. As the friction between the straps and the 
 sheave of an eccentric is greater than between an 
 ordinary crank-pin and a connecting rod, eccentrics are 
 only used when the crank arm is required to be small. 
 
 {a) Figs. 277-280 illustrate a common eccentric. The 
 sheave, sh, is made of cast iron ; C" is the centre of the 
 sheave, and C the centre of the hole which fits on the 
 crankshaft, C S. The eccentricity is the distance C C". 
 The circumferential surface of the sheave forms a groove 
 
 into which the two gunmetal straps, Sp, fit. The 
 straps are connected by two bolts, B, but do not meet 
 metal to metal, for the two open spaces, O' and O", are 
 filled with thin brass liners, which are removed one by 
 one when it becomes necessary to compensate for wear. 
 The eccentric rod, E E, is fixed to the one strap by 
 
 Fig 279. 
 
 \ v r 
 
 i 
 
 i _y 
 
 > 
 
 i^j i -I 
 
 j i 
 
 e <■ 
 
 Fig. 280. 
 
 a cotter. With this kind of eccentric, the rod is usually 
 screwed into the sheave 'and held in position by a 
 binding nut. Oh is an oil-hole. 
 
 (i) The cast-iron sheave of the eccentric, Fig. 281, 
 is made in two parts, which are connected by studs and 
 cotters. By this arrangement the sheave can easily be 
 removed from the shaft. The circumferential surface 
 
 Fig. 281. 
 
 of the sheave is cylindrical, and the two gunmetal 
 straps have square cornered grooves, into which the 
 sheave fits. The straps are connected by bolts, B. 
 O is the centre of the crankshaft, and the line C P 
 indicates the position of the crank relative to the 
 eccentric. E E is the flat eccentric rod which is 
 fastened to the one strap by bolts, B 1 . L is the 
 lubricator. 
 
 The Condenser. 
 The exhaust steam of condensing engines is let into a 
 closed cast-iron box, the condenser, where it meets with 
 
DETAILS OF STEAM ENGINES. 
 
 227 
 
 cold water, and thus will be partially condensed. The 
 steam-water, together with the vapour and air, are 
 pumped out of the condenser into the hot-well by the 
 air pump. As the temperature of the steam-water will 
 always be less than 212 deg. F., it follows that the 
 pressure in the condenser will be below that of the 
 atmosphere. 
 
 Injection Condenser. — The condensing water may be 
 introduced into the condenser as a spray or jet, and by 
 coming into direct contact with the steam it will con- 
 dense the latter. The air pump must remove the con- 
 densing water as well as the steam-water and air from 
 the condensing chamber. 
 
 Each valve consists of a flat indiarubber ring, I E, 
 which, when shut, rests on a perforated brass grid, Gr. 
 The opening of a valve is limited by a cup-shaped 
 guard, Gd. The pump piston, P P, is made of brass with 
 cast-iron and steel spring rings ; the piston rod, P R, is 
 of steel, and is the continuation of the tail rod of the 
 low-pressure steam piston. The stroke of the pump is 
 24in., and the volume swept by the piston in one 
 stroke is about one-ninth of the volume of the condensing 
 chamber. 
 
 The connection between a similar condenser and the 
 low-pressure cylinder of a girder compound condensing 
 engine is shown in Figs. 285 and 286. L P C is the low- 
 
 Fig. 282. 
 
 As the pressure in the condenser is very small, the 
 air pump must be placed below the condensing 
 chamber, so as to allow the weight of the water to 
 assist in opening the suction-valves. 
 
 Paxman's injection condenser is shown in Figs. 282 
 and 283, which are sections through A B and CDE F. 
 Fig. 284 is a plan of the condenser with cover removed. 
 The condenser is cast in one with the air pump barrel, 
 B P. The condensation takes place in the condensing 
 chamber, C C, into which the exhaust steam is brought 
 through ESI, the condensing water entering at the 
 opposite end. The injection-pipe is perforated between 
 lines a and b, so as to cause the water to form a spray. 
 The flow of condensing water is regulated by adjusting 
 the lift of valve I V. The mixture of steam- water and 
 condensing water falls to the bottom of C C, and then 
 through the suction-valves, S V, into the air pump, 
 whence it is forced through the delivery- valves into the 
 hot-well, H W. The latter is separated from C C by a 
 horizontal and a vertical partition, which run through 
 the whole length of the condenser. From the hot- 
 well the water runs over a baffle, Bf, into an overflow 
 pipe, OP. 
 
 The air pump is double acting, and there are two sets 
 of indiarubber suction-valves, as well as two sets of 
 delivery-valves, separated by partitions, Pt 1 and Pt 2 . 
 
 ^Y^Y^vk 
 
 Fig. 283. 
 
 pressure cylinder, C the condenser, EP the exhaust-pipe, 
 P R the piston rod of the air pump, I V the injection- 
 valve, F F the foundation frame, and G T the girder 
 head. 
 
 In Fig. 287 is shown an injection condenser made 
 by Messrs. Musgrave and Sons, of Bolton. The con- 
 
 Fig. 284. 
 
 densing water enters at the top through an opening, W I, 
 whereas the exhaust steam enters at the side through 
 ESI. The jet of condensing water rushes through a 
 set of nozzles, N x> N 2 , N 8 , and N 4 , inducing the steam 
 and the air to follow, the steam at the same time being 
 
228 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 condensed. The rotary air pump at the bottom other by tooth-wheels fastened on the end of the 
 discharges the water through the outlet-pipe, P, cylinder spindles. 
 
 Fig. 285. 
 
 IZ I 2 3 
 
 4 5 
 
 Fig. 
 
 Instead of a piston, the air pump has two tooth- wheel- 
 like cylinders, the teeth of which fit accurately into the 
 
 pgg 
 
 Fig. 287 
 
 corresponding gaps. The one cylinder is driven by the 
 engine or by a separate motor, and then drives the 
 
 6 7 5 3 10 II IZFut 
 I I I I I I — ■ 
 
 286. 
 
 The quantity of condensing water required for a 
 given size of engine can be calculated as follows : 
 
 Let the quantity of steam, in pounds, to be con- 
 densed, be denoted by S 
 
 Let the ratio of quantity of condensing water to 
 that of steam to be condensed, be denoted by y 
 
 Let the temperature of condensing water in 
 degrees of Fahrenheit be denoted by t" 
 
 Let the temperature of condenser in degrees of 
 Fahrenheit be denoted by t' 
 
 Let the pressure, in pounds per square inch, in 
 the condenser be denoted by p 
 
 We may take the heat contained in lib. of steam, 
 reckoned from 32 deg. F., as 1,180 B.H.U., and we 
 shall therefore have 
 
 1,180 x S + y S (f - 32) = S (1 + y) {? - 32), 
 and consequently 
 
 1212 - f 
 
 y=~F~r ■ ■ ■ cm 
 
 Example 1.— Take r = 60, p b = 2, therefore H = 126-8, 
 and we shall have 
 
 y = 16-5. 
 Example ±—Foip b =1, f' = l02 and f' = 60, we find 
 
 y=2G'o. 
 ^hich of the two, non-condensing or condensing 
 engines, will be the cheapest, depends upon the ratio 
 of the local prices of fuel and water, and may be deter- 
 mined as follows : 
 
DETAILS OF STEAM ENGINES. 
 
 229 
 
 Let fuel, in pounds, per I.H.P. hour for a non- 
 condensing engine, be denoted by B„ 
 
 Let fuel, in pounds, per I.H.P. hour for a con- 
 densing engine, be denoted by B„ 
 
 Let steam, in pounds, per I.H.P. hour for a 
 non-condensing engine, be denoted by S„ 
 
 Let steam, in pounds, per I.H.P. hour for a 
 condensing engine , be denoted by S e 
 
 Let the price of lib. of fuel be denoted by P„ 
 
 Let the price of lib. of water be denoted by P w 
 
 For the same boiler and steam-pipe arrangement we 
 must have 
 
 |^ = J^ = JL and consequently |^- = ^= S. 
 
 The cost of one I.H.P. hour for a non-condensing 
 engine will therefore be 
 
 O71 X x. iff ~H -Dm X _fcc 
 
 and for a condensing engine 
 
 S„ (1 + y) x P,„ + B e xP ( . 
 
 We must therefore have 
 
 > 
 
 S„ x P w + B„ x P c =S C (1 + y) xPb+B.xP, . (203) 
 < 
 
 according to whether the non-condensing engine is 
 less, equally or more economical than the condensing 
 engine. From (203) we obtain 
 
 Example. — If the fuel be good coal and the boiler be 
 
 g 
 an economical one, we may take „ = 8. For y = 20, 
 
 B 
 4 
 we shall probably have 8 = — , and consequently the 
 
 o 
 
 right-hand side of (204) will be 472. 
 
 In London we may take the price of coal as 
 eighteen shillings per ton, and that of water from 
 water works as ninepence per 1,000 gallons ; in this 
 case 
 
 ^=107. 
 
 It is therefore more economical to use a non-condens- 
 ing engine in London, except on the river side . 
 
 Surface Condenser. — In some cases it may be prefer- 
 able to keep the condensing water and the steam apart ; 
 the condensing chamber is then filled with a great 
 number of thin tubes of small diameter, through which 
 the condensing water is passed by means of a pump — 
 the circulating pump. The condensation of the steam 
 takes place on the cold surfaces of the tubes, and the 
 air pump has only to remove the steam-water and the 
 air which might have leaked into the condensing 
 chamber, 
 
 A surface condenser, designed by Messrs. Musgrave 
 and Sons, is shown in Fig. 288. It is a cast-iron 
 cylinder divided into five compartments, A, B, C, D, 
 and E, through which seven sets of tubes are passed. 
 
 Fig. 288. 
 
 Bach set consists of a lin. gas-pipe, T 3 , screwed into the 
 partition between A and B, and two brass tubes, T 1 and 
 T 3 , of which the first is 3in. in diameter, and is screwed 
 at the bottom into tube-plate, T P 2 , and is provided 
 at the top with packing and gland ; T 2 , whose diameter 
 is 2in., is screwed at the top into the partition between 
 B and C, and is shut at the bottom by a cap which 
 also keeps it from touching T r 
 
 The exhaust steam enters compartment C by inlet 
 S I, and is condensed while passing through the annular 
 space between tubes T x and T 2 , and drops as water 
 into compartment E, from where it is removed by the 
 air pump through outlet SWO. To secure good 
 drainage, the floor of E has a fall of |in. towards the 
 outlet. 
 
 The condensing water enters the condenser through 
 inlet C W I, filling up compartment D, from where it 
 passes through the lateral channel into A, then down 
 through the gas-pipes, up again between T 2 and T 3 into 
 B, and at last leaves the condenser through outlet 
 CWO. By this circulation, the condensing water is 
 much more effective than if forced straight through a 
 number of tubes. 
 
 The condensing chamber is thus composed of the 
 annular spaces between tubes T 1 and T 2 , which offer a 
 total cooling surface of 46 square feet to the steam. 
 
 Figs. 289 and 290 are plans of compartments C and 
 E, and the position of the condensing water inlet jg 
 
230 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 shown by 
 plate T Pi, 
 
 dotted lines. Fig. 291 is a plan of tube- 
 Figs. 292 and 293 are elevation and plan 
 
 Fig. 289, 
 
 Fig. 291. 
 
 Fig. 290. 
 
 Fig. 292. 
 
 The Governor. 
 
 Engines required for driving dynamo machinery must 
 run at a constant speed of rotation, independent of 
 the load. This condition is impossible to fulfil abso- 
 lutely, but by proper design, the variation of speed 
 can be kept within limits which will not effect the 
 purpose for which the engine is designed. There are 
 two distinct apparatus which act as speed regulators. 
 
 (1) To obtain a uniform motion, the effort must be 
 equal to the resistance. This cannot be obtained in a 
 steam engine, but we can arrange that the energy 
 exerted on the crank-pin during one revolution is equal 
 to the energy consumed by the load during the same 
 interval of time. This is the function of the Governor. 
 
 (2) The angular velocity of the crankshaft should 
 remain constant, notwithstanding the variation of the 
 crank effort during a revolution. This is the function 
 of the Flywheel. 
 
 The action of the governor is to diminish the quan- 
 tity of steam let into the cylinder, should the speed of 
 the engine increase, and to increase the quantity of 
 steam should the speed of the engine fall. The prin- 
 ciple of the governor is therefore an imperfect one, 
 as it requires the speed of the engine to rise or fall in 
 order to act. 
 
 of the condenser, and Fig. 294 is a horizontal section, 
 through compartment D, 
 
 Fig. 295. 
 
 The governor commonly used in steam engines is 
 the so-called centrifugal governor, designed on the 
 principle of the conical pendulum, Fig. 295. This con- 
 sists of a pendulum attached to the vertical axle, A B, 
 by a horizontal pin, thus allowing the pendulum to 
 turn in a plane through AB. By rotating the axle 
 round itself, the -pendulum will fly out until it makes 
 an angle, a, with the axle, the magnitude of which 
 depends upon the speed of rotation. Suppose the 
 angular velocity of the axle is w ; then the pendulum 
 will be in equilibrium when the deviating force is 
 
 Tl W 
 
 -r = — x a> 2 x r pounds . . . (205) 
 
 where W is the weight of the ball in pounds, the mass 
 of the arm being neglected. 
 
 But it is evident that P is also equal to W x tan a, 
 and, consequently, 
 
 a) 2 ?- 
 
 tan a ■= 
 
 9 
 
 r 
 
 h' 
 
DETAILS OF STEAM ENGINES. 
 
 231 
 
 by which we obtain 
 
 tB 2 = 
 
 (206) 
 
 Ai^-l'Svr)- ■*»> 
 
 or diminished to 
 
 1 + Kx 
 
 i .Q-f 
 
 Watt's Governor. — A diagram of this governor is 
 shown in Fig. 296. It consists of two pendulums 
 
 suspended in b and b. A sleeve, e e, which can slide * h\ L W 
 
 on the axle, rises and falls with the balls, and thereby The three formill8B (20 7), (208), and (209) give us 
 
 -VfC 
 
 ) . (209) 
 
 Fig. 296. 
 
 actuates the apparatus which controls the quantity of 
 steam let into the cylinder. Let Q be the weight 
 which has to be overcome in lifting the sleeve, and W 
 the weight of each ball, then the moment of the 
 deviating force with regard to b will be 
 
 Q 
 2 cos g * be x sm (a + ft) + ~W x bd x sin a. 
 
 On the other hand, the deviating force, P, at the 
 centre of the ball must be 
 
 W 
 P=— w*xfd, 
 
 g J 
 
 and its arm will be M x cos a. The governor will, 
 therefore, be in equilibrium in the position shown in 
 the diagram if 
 
 — to 2 xfd x bd x cos a 
 9 
 
 Q 
 
 = 2cos/3 x & c x sm (a + /3) -f W x 6 <2 x sin a. 
 
 Let now I, L, h, and K denote bc,bd, fd x cot a and 
 
 tan a + tan ft respectivel then we shall have 
 2 tan a 
 
 A(^14) 
 
 (207) 
 
 This formula shows that the loaded governor must be 
 run at a higher speed than the unloaded one. 
 
 Besides the weight, Q, the governor will have to 
 overcome friction between the parts of the controlling 
 apparatus, be it a throttle- valve or a cut-off valve. 
 Assuming that this friction, F, is the same in both 
 directions, then the load on the sleeve will be Q + F 
 when rising, and Q - F when falling. The governor 
 will therefore not act until the angular velocity has been 
 increased to 
 
 co 2 - or = or - o>i 
 
 (210) 
 
 The governor will evidently be the more sensitive for 
 overcoming friction the smaller we can make 
 
 0>2 
 
 — =1- 
 
 (211) 
 
 The angular velocity, w, will be approximately equal 
 to the mean of w 2 and w lt so we shall have 
 
 <o 2 + ft>i 
 &> = —=— — - 
 
 (212) 
 
 The three formulas (210), (211), and (212) give us 
 
 ft> 2 "" ft> 
 
 F 
 
 I K ^ 
 
 (213) 
 
 The last formula clearly shows that a governor loaded 
 with dead weight will be more sensitive for overcoming 
 friction than the original Watt's governor. 
 
 Porter's Governor. — A governor of this class is shown 
 in Fig. 297. The pedestal, Pd, which carries the 
 governor is bolted on the foundation-frame of the 
 engine. Pulley, PI, is driven by belt, and its rotary 
 motion is transmitted to the governor spindle, Sp, by 
 bevel wheels, B W 2 and B W r The governor is loaded 
 with a dead weight, D W, which is cast in one piece 
 with the sleeve, SI ; and the two pendulums, A x B x and 
 A 2 B 2 , are suspended on one pin, p, at the top of 
 the spindle. The sleeve and the dead weight move 
 up and down with the balls, and thereby turn valve 
 arm, V A, by means of lever L 3 , which has a fulcrum 
 at Fc, and the vertical rod B. D P is a dash-pot to 
 prevent sudden jerks. The governor spindle has a 
 bearing, P B, and rotates within the brass tube, B T. 
 
 The application of formula (213) on this governor is 
 very simple ; we have L = I, and as angle ft is approxi- 
 mately equal to angle a, K will be equal to unity, and 
 therefore 
 
 F 
 
 . . ■ . (214) 
 
 i = 
 
 W + Q 
 
 The sensitiveness of this governor to overcome 
 friction would be increased by adding a weight, w, to 
 
 F 
 W + Q, as we should then have n =^7 ^ • but 
 
 W + (j + w ' 
 
 it is evident that w should be added to the dead weight, 
 and not to the ball, as in the latter case, the total 
 weight added to the governor would be 2w. The 
 sensitiveness of the Porter governor thus depends more 
 on the dead weight than on the balls, which can, there- 
 fore, be made small. 
 
232 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 The governor illustrated in Fig. 297 is made by 
 Messrs. Browett, Lindley, and Co., of Salford, and is 
 applied by them to drive their Lindley-Eider automatic 
 expansion gear. 
 
 of the governor depends, therefore, on the variation of h 
 being small as compared with that of angle a, and the 
 pendulum, Fig. 295 is consequently more sensitive 
 than that in Fig. 298. 
 
 \^\\\\s ^\\\\\\\\\\^^^^ 
 
 4^\ v ^^^^^\V}As^^^^^s^^^^^^^^W 
 
 Fro. 297. 
 
 Stability of Governors. — Formulae (206) and (207) 
 require that h shall diminish when the angular velocity 
 increases ; this condition is satisfied when the governor 
 pendulum is suspended, as in Figs. 295 and 298. The 
 pendulum in these two cases will take up a definite 
 position corresponding to the velocity, w, at which it is 
 turned round the axle, and will not be in equilibrium 
 in any other position unless the velocity be changed. 
 The governor is therefore said to be stable, but it is 
 evident that it cannot keep the speed of the engine 
 constant, as a new position of the pendulum requires 
 the engine to run faster or slower, The sensitiveness 
 
 Fig. 298. 
 
 In Fig. 299, the curve in which the centre of the 
 governor ball moves is a parabola. As the centre of 
 curvature lies in the normal to the curve, h must be 
 the projection on the axle of that part of the normal 
 
 Fig. 299. 
 
 which is situated between the centre of the ball and 
 the axle. This projection is the same for all points in a 
 parabola, and consequently we have a governor which 
 is perfectly sensitive. As the ball is in equilibrium at 
 
 Fig. 300. 
 
 any point of the curve, it follows that the parabola 
 must be specially constructed for the speed at which 
 the governor is to be run. Should, however, the load 
 on the engine be diminished, the engine will begin 
 to race, the consequence of which will be that the 
 governor ball will not be in equilibrium at any point of 
 
DETAILS OF STEAM ENGINES. 
 
 233 
 
 the parabola, but will fly out as far as the mechanism 
 of the governor permits. The steam will thus be shut 
 off, the engine will run slower, but the governor ball 
 will not move until the speed has fallen below that 
 
 corresponding to the parabola. At this moment the 
 ball will suddenly fall as far as it can, and open for the 
 steam. The engine will begin to race, the ball will fly 
 
 Uc3 
 
 Fig. 303. 
 
 there is a certain position of the pendulum where h is a 
 maximum. This position is determined by 
 
 D¥ 
 
 Fio. 302. 
 
 Sin 
 
 «2 
 
 V L • • 
 
 (215) 
 
 where L is the length of the pendulum arm. 
 
 The governor will therefore be stable as long as 
 a > a 2 , and will be unstable when a < a 2 . In the 
 position 2, corresponding to a 2 , the governor will behave 
 
 "V 
 
 Fig. 304. 
 
 out and shut off the steam, and so on. The result will like a parabolic governor. The pendulum must be 
 
 be that the engine will run unsteadily. prevented from falling below position 2 in order to 
 
 The pendulum might be suspended as shown in Fig. avo id the unstable positions. 
 
 300. The peculiar feature of this arrangement is that Parabolic Governor.— Fig. 301 shows a parabolic 
 
234 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 governor which has actually been constructed, but on 
 account of its great sensitiveness has not been adopted. 
 A B is the governor spindle, to which the curved arms, 
 C C, are fixed. The centres of the balls, B x and B 2 , 
 move in a parabola, and thereby control the position of 
 the sleeve, SI. 
 
 A Governor with Grossed Arms is shown in Fig. 302. 
 The pendulums are suspended at C x and C 2 . Arms 
 1\ and F 2 are forked, and prevent the pendulum from 
 falling into a position below that corresponding to 
 angle a 2 in formula (215). Pd is a pedestal carrying the 
 governor, and Sp is the spindle which receives motion 
 from the crankshaft by means of bevel wheels. The 
 up and down motion of the sleeve, SI, is transmitted 
 to the throttle-valve, T V, by a set of levers. As 
 angle /3 is equal to angle a, we shall have 
 
 J- 
 
 h \ l + WL/' 
 
 The sensitiveness of the governor for overcoming 
 friction will be the greater, the smaller we have 
 
 i = 
 
 F 
 
 x W + Q 
 
 I 
 
 Paxman's Governor.— The latest model of this well- 
 known governor is illustrated in Figs. 303 and 304. 
 The two pendulums, A 1 and A 2 , are suspended on pin, 
 p 3 ; the end of A 1 fits into a slot in the spindle, Sp, 
 whereas the end of A 2 being forked embraces the 
 spmdle. Each arm carries two friction rollers on pins 
 p 1 and p 2 ; these rollers bear against a guide-plate, G, 
 whereby the dead weight will be lifted when the 
 pendulum arms fly out. The pressure on each set of 
 friction rollers is \ x Q, and remains constant for all 
 positions of the pendulum. 
 
 The spindle, Sp, is cylindrical through the pedestal, 
 Pd, and the dead weight up to a ; between a and b the 
 spmdle is rectangular and then again cylindrical. A 
 final adjustment of the governor is made by applying 
 spring, Sg, on the top of the spindle. By taking hold 
 of the sleeve, SI, with the hand while the governor is 
 running, and undoing screw Sc, the sleeve can be 
 pushed down, thus compressing the spring. The nut 
 n, is then made to follow the sleeve, and Sc is screwed 
 back. In this way the pressure on the friction rollers 
 can be adjusted, and the sensitiveness of the governor 
 increased. 
 
 The lever, L, which turns on two journals, t, moves 
 up and down with the dead weight and actuates the 
 steam controlling apparatus. 
 
 The peculiar shape of the pendulum arms allows the 
 bahs to be m a position where the pendulums are most 
 sensitive, and still permits the deadweight to be raised 
 considerably. 
 
 Let now the distance between centres of pin p* and 
 the balls be denoted by L\ the distance between centres 
 
 ?? m l Fi* 1 * P \ 01 ' P3 and p2 ^ '• the an S le ^tween 
 L and I by p, and the angle between H, and the axis 
 
 of rotation by a, then the governor will be in 
 equilibrium when 
 
 ^Zxsin(a + /3) + W hx tana = — w 2 h 2 x tana 
 
 and therefore 
 
 A( 
 
 Q I sin(a+/3) 
 
 1+ WL* X 2si 
 
 ina / 
 
 We will also have 
 
 1=TT 
 
 F 
 .^ 2 sina 
 
 + Q 
 
 L 1 
 
 As — is about 3, and /3=30 deg., then in the posi- 
 tion of the pendulums shown in the drawing, we shall 
 have - 1 x , 2 sm a „ = 3, and therefore 
 
 I sin (a+/3) 
 
 F 
 
 '"8-WTQ (216) 
 
 By comparing (216) with (214) for Porter's governor, 
 it will be seen that the balls of the Paxman governor 
 are three times as efficient as those of Porter's. 
 
 The Pickering Governor.— The governor shown in 
 Fig. 305 is made by Messrs. E. Garrett and Sons, 
 of Leiston, and is of the Pickering type. It has 
 two or more balls, B, which are mounted on flat steel 
 springs, Big. These springs are fixed to discs, T\ 
 and D 2 , which rotate with the balls. D 2 is prevented 
 from moving up and down by the collar, Cr, which is 
 screwed to the tubular spindle. T S. The rotary 
 motion of the horizontal spindle, S 2 p, is transmitted to 
 the balls by two bevel wheels, of which the one is fixed 
 to D 2 The spindle, Sp, carries at the one end an 
 equilibrium throttle-valve, TV, and at the other end 
 it is fixed to D x , the lift of the throttle-valve being 
 adjusted by nuts, n. Cv forms the cover of the valve- 
 case^ and Sp works steam-tight up and down in 
 stuffing-box, SB. 
 
 The horizontal spindle, S^, carries an adjustable 
 
 T m f ' ™ g ' '?? ° Qe end of which is fi xed to worm 
 wheel, W while the other end is fastened to the 
 forked end of the spindle. By turning the worm screw, 
 VS. the tension of the spring will be increased or 
 diminished, and thus adjust the pressure of the forked 
 end of the spindle against collar, Cr*. As this pressure 
 has to be overcome by the balls, the speed of the 
 governor will in this way be adjusted. 
 
 The Acme Governor. -Meesrs. Browett, Lindley, and 
 Co. s Acme governor is illustrated in Figs. 306 and 
 
 arJ'hin 77 ° f tW ° half - baI1S > B * and B,, Which 
 
 27 y PmS ' Pl aUd *> t0 a <*oss- P iece which 
 
 cannot move up and down. The balls are held 
 
 oTe Th 7 SPnngS ' Slg ' Which m *& the Aviating 
 force These springs are connected to cross-piece Cr, 
 by links, L, and L, When the speed of the governo 
 
 lin^^^t t0 < r* C °™ th ° tension of ?he 
 spring,, big, the throttle-valve, which is attached to 
 
DETAILS OF STEAM ENGINES. 
 
 235 
 
 10 Inches 
 
 Fig. 305. 
 
 Cr by spindle, Sp, will be pusbed down and tbus regu- Tbe speed of tbe engine can be adjusted by a double 
 late tbe supply of steam. bell-crank lever, CI, wbicb turns on pin, p 8 . One arm 
 
236 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 of this lever is at right angles to screw, Sc, and engages the position of the nut. The spring will thus cause a 
 in a sleeve on spindle, Sp. The one end of spring, Sg, downward pressure on Sp when the nut is screwed 
 
 Fig, 306. 
 
 Fig. 307. 
 
 Fig. 308, 
 
 is attached to nut, n and the tension of this spring will upwards, and an upward pressure when the nut is 
 turn CI in the one direction or the other, according to screwed downwards. 
 
DETAILS OE StEAM ENGINES. 
 
 23f 
 
 The Throttle-Valve. 
 
 The most common way of regulating the quantity of 
 steam to be admitted to the cylinder is by diminishing 
 or increasing the inlet-opening by a valve — the throttle- 
 valve. This valve is actuated by the governor, as shown 
 in several of the preceding illustrations. 
 
 Figs. 309 and 310 illustrate an arrangement which 
 is often used, especially in small engines. S V is the 
 stop-valve, which opens by turning the hand-wheel, 
 
 Fig. 309. 
 
 H.W 
 
 lever, L, by a bolt through one of the holes, h. The 
 valve openings will be larger or smaller, according to 
 the load which the engine has to overcome. 
 
 The Flywheel. 
 
 The flywheel is a wheel of large diameter, having a 
 rim of considerable mass, which moves at great velocity. 
 The moving rim is the seat of a large quantity of kinetic 
 energy, which can be augmented or diminished without 
 causing an appreciable variation of velocity. 
 
 If the effort which produces the rotary motion is vari- 
 able while the resistance to be overcome is constant, or 
 vice versd, then there will be a period during which the 
 energy exerted by the effort is greater than that con- 
 sumed by the resistance. The excess of energy exerted 
 will be accumulated in. the flywheel as kinetic energy, 
 whereby the velocity of the rim will be increased from 
 v 1 to v 2 . This period will be followed by another one, 
 during which the energy consumed by the resistance is 
 greater than that exerted by the effort; the flywheel 
 will during this period give off the energy it received 
 during the first one, and its velocity will .thereby fall 
 from v 2 to v x . 
 
 Let us now by W denote the energy which causes 
 the irregularity of speed, by M the mass of the flywheel, 
 and by E its radius of gyration ; then we must have 
 
 Fig. 310. 
 
 H W ; the throttle-valve, T V, is fixed on spindle, Sp, 
 which is turned in the one or the other direction by 
 the governor, as shown in Fig. 302. 
 
 The valve, T V, will be full open while the engine is 
 stopped, and will be almost shut when the engine is 
 running empty. 
 
 In Fig. 308 is shown a double-beat throttle-valve. 
 It consists of two valves, V! and V 2 , which are fixed 
 on the same spindle, Sp. The governor is attached to 
 
 2 ' 
 
 and therefore 
 
 v -v,= 
 
 2W 
 
 (v 2 +v 1 ) M 
 
 As (« 2 - Uj) must be small, the mean velocity of the 
 rim will be 
 
 v 2 + v 
 ~2~ 
 
 and therefore 
 
 V2-V! 
 V 
 
 W 
 
 X, andM = 
 
 W 
 
 where X is a coefficient which must be the smaller, the 
 steadier the engine is required to run. With engines 
 driving dynamos for electric lighting, X should probably 
 not be less than 0005-. 
 If the wheel makes n revolutions per minute, we 
 
 have ZA? = v , and therefore also M= (JUL)* x J^L 
 
 If W be given in foot-pounds and E and v in feet, 
 then we shall have 
 
 M = 
 
 32-187 x W 
 w 2 X 
 
 Kirn/ 
 
 32187 W 
 E 2 X 
 
 pounds . (217) 
 
 "We may take for E the mean radius of the rim ; 
 M will then be the mass of the rim plus about one- 
 third of the mass of the arms of the wheel. 
 
 Strength of Flywheel— -The flywheel is made of cast 
 iron, and consists of a rim carried by a number of arms 
 the other ends of which are joined to a boss, by which 
 the wheel is keyed to the shaft. 
 
'238 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 (218) 
 
 The rotating rim is subject to a centrifugal force, 
 which tends to tear the rim along a diametrical plane 
 through the axis of rotation, just in the same way as 
 the steam pressure tends to burst a cylindrical boiler 
 shell. The thickness, 8, of the rim can therefore be 
 determined by the same formula which we used on 
 page 108 for calculating the plate thickness of a boiler 
 shell — viz. : 
 
 s= Dx^ A .... 
 2 T 
 
 where D, in this case, is the diameter of the wheel, p 
 the centrifugal force in pounds per square foot of face- 
 surface, and T the tensile strength of cast iron. Let 
 8 be so small that each particle of the rim can be con- 
 sidered to move with the same velocity ; then the mass 
 of the rim per square foot of face- surface will be — 
 
 32-187 
 
 Where y is the mass in pounds of a cubic foot of cast 
 iron, we shall also have 
 
 P 
 
 = 2Mtt 2 ^ 28 yv* 
 
 D 
 
 32187D 
 
 pounds. 
 
 Inserting the latter expression for p in (218), we have 
 
 V 
 
 32187 x T 
 
 v x f 
 
 feet per second, 
 
 which is independent of the dimensions of the wheel. 
 
 The ultimate tensile strength of cast iron may be 
 taken as 2,255,0001b. per square foot, and the maxi- 
 mum working stress as 615,0001b. per square foot. 
 Taking y as 4501b., we shall have 
 
 v = 209ft. per second 
 
 as the maximum velocity for the rim. 
 
 In practice, the circumferential velocity of a flywheel 
 does probably not exceed 100ft. per second. 
 
 If the flywheel be used for driving with belt, then 
 v may be taken on an average as 45ft. per second, and 
 when it is used for driving with ropes, v may be taken 
 as 80ft. per second. 
 
 Determination of W. — In practice, the number of 
 revolutions the engine is to make per minute, as well 
 as the diameter of the flywheel, are given ; the mass, 
 M, of the wheel can therefore be calculated when 
 we have determined the energy, W, which causes the 
 irregularity of speed. 
 
 Fig. 311 represents the theoretical indicator diagram 
 of an engine taken at an early cut-off; OV is the 
 absolute vacuum-line, and O C is the clearance-line. 
 For simplicity's sake, we will assume that the length of 
 the connecting-rod is great compared with that of the 
 crank arm, in which case the diagrams taken from both 
 ends of the cylinder will be identical. The effective steam 
 pressures per square inch of piston area are represented 
 by the ordinates to the diagram, Fig. 312, a b cdefa, 
 measured from the base-line, af. The semicircle, 
 a kf, represents the path of the crank-pin during one 
 
 stroke. When the crank has turned through an angle, 
 <j>, and the piston has moved through a distance, a g, 
 then the crank effort per square inch of piston area 
 will be represented by h g x sin <j>. It will be seen 
 that the crank effort will be positive during that part 
 of the stroke which is represented by a d, and it will 
 be negative for the remainder of the stroke on account 
 of the compression on the other side of the piston. 
 
 Fig. 311. 
 
 In Fig. 313 the length of the base-line, a kf, is equal 
 to the length of the semicircle, a kf, in Fig. 312, and 
 the ordinates to the curve, a c 1 h l d 1 f, are taken propor- 
 tional to the crank effort as determined in Fig. 312 ; 
 thus k h 1 and a k are equal to h g x sin and a A; in 
 Fig. 312. The work done during one stroke by the 
 crank effort per square inch of piston area is, therefore 
 
 Fig. 312. 
 
 represented by area a o 1 d 1 a, minus area d 1 e 1 fd 1 . The 
 ordinate, a a 1 , to the straight line, a 1 / 1 , represents the 
 constant resistance to be overcome by the crank effort. 
 The area of rectangle, a&Pfa, will therefore repre- 
 sent the work consumed by this resistance per square 
 inch of piston area, which must be equal to the 
 work done by the crank effort per square inch of 
 
 Fig. 313. 
 
 piston area. It is now evident that area m c 1 n m, 
 multiplied by the area of the piston in square inches, 
 must represent the energy, W, which causes the 
 irregularity of speed, and which must be subdued by 
 the flywheel. 
 
 In the case just described, we have assumed the 
 piston speed to be so low that the inertia-pressure 
 
DETAILS OF STEAM ENGINES. 
 
 239 
 
 due to the moving masses can be neglected. With 
 high speeds of revolution, however, the inertia-pressure 
 will entirely alter the flywheel diagram. Let, for in- 
 stance, the indicator diagram of a high-speed engine be 
 that in Fig. 311, the stroke of the engine be 1ft., the 
 reciprocating masses reduced on the piston be 21b. per 
 square inch of piston area, and let the engine make 300 
 revolutions per minute, then, according to (179), page 
 213, the inertia-pressure at the beginning of the stroke 
 will be minus 30 - 251b. per square inch of piston area. At 
 the end of the stroke, the inertia-pressure will increase 
 the effective pressure by 30"251b., and at the middle 
 of the stroke, the inertia-pressure will be zero. In 
 Fig. 312 is drawn a straight line, e q, in such a manner 
 that the ordinate, a q, equal to ef, represents 30 • 251b. 
 
 c' 
 
 L 
 
 wm^' 
 
 G^ssS^i^* tt - 
 -% 
 
 I 
 Fig. 314. 
 
 The effective pressure on the piston of the high- 
 speed engine must therefore be estimated from the line 
 e q, instead of from af. When the piston has traversed 
 distance ag, the effective pressure will thus be g 1 h, 
 and the crank effort per square inch of pistoD will be 
 g 1 h x sin <j>. In this manner, flywheel diagram, Fig. 314, 
 has been produced. It will be seen that the energy to 
 be subdued by the flywheel per square inch of piston 
 area has been considerably reduced by the inertia 
 pressure. 
 
 When the engine has two cylinders with cranks at 
 180 deg., then the energy, W, will be twice that due to 
 one cylinder. If a two-cylinder engine has cranks at 
 
 > w^ A 4^ -^V 
 
 i 
 
 Fig. 315. 
 
 right angles, then the two crank efforts will assist each 
 other, and the irregularity will be diminished. Fig. 315 
 represents the flywheel diagram for half a revolution 
 of a two-cylinder engine with cranks at right angles 
 and with indicator diagram, as in Fig. 311. As the 
 resistance to be overcome is twice that of a single- 
 cylinder engine, it follows that a a 1 in Fig. 315 is 
 twice a a 1 in Figs. 313 and 314. It will be seen 
 that the energy to be subdued by the flywheel 
 is much smaller than in the single-cylinder engine, 
 although the flywheel diagram of each cylinder would 
 be that in Fig. 313. 
 
 In single-cylinder engines, as well as in two-cylinder 
 engines with cranks at 180 deg., the maximum energy 
 to be subdued by the flywheel will be produced when 
 
 the ratio of expansion is about 2 ; whereas the greatest 
 irregularity in two-cylinder engines with cranks at 
 right angles will occur at an early cut-off. 
 
 It is worth noticing that in designing the flywheel 
 and the governor for an engine, the governor should be 
 less sensitive than the flywheel, or, in other words, t] in 
 (211), page 231, should be greater than X in (217). 
 
 The Slide- Valve. 
 
 The piece of mechanism generally used for the 
 purpose of regulating the distribution of steam in the 
 cylinder is the slide-valve. Moving to and fro, the 
 slide-valve shuts and opens the steam-ports, thereby 
 allowing the steam to enter the cylinder alternately 
 on the one and the other side of the piston, and again 
 letting the steam already used escape into the exhaust 
 chamber. 
 
 As the reciprocating motions of the steam-piston 
 and the slide-valve are produced by the same form of 
 mechanism — viz., " crank and connecting-rod " — it will 
 be necessary to investigate the laws of this kind of 
 motion, in order to understand properly the distribution 
 of steam by the slide-valve. 
 
 In Fig. 316 is drawn a circle, with O as centre, and 
 with the crank arm, r, as radius. C 2 B 2 and C 4 B 4 are 
 two positions of the connecting-rod corresponding to 
 the two positions OC 2 and OC 4 of the crank arm, and 
 B 2 and B 4 of the cross-head pin. The other two posi- 
 tions, B x and B 3 , of the cross-head pin correspond to 
 the two dead-point positions of the crank arm. 
 
 When now the crank has turned through an angle 
 <j>, so that the crank-pin is at C 2 , then the cross-head 
 pin will have traversed distance S 1( which, by making 
 B 2 N=L, will be equal to G ± N ; we have now 
 
 C!N=r(l-cos0)-L+ VL 2 -r 2 sin 2 ^ = S 1 . (219) 
 
 Formula (219) will thus express the motion of the 
 cross-head pin, and therefore also that of the piston 
 during the stroke from B x to B 3 . 
 
 In the return stroke, the cross-head pin will traverse 
 distance S 2 , while the crank turns through an angle <f>. 
 By making B 4 N 1 equal to L, we have 
 
 N 1 C 3 = S 2 = r(l-cos0) + L- VL 2 =r 2 sin 2 <^ 
 
 We can therefore express the motion of the cross- 
 head pin by 
 
 S = r(l-cos 0) + L ± Vl 2 - r 2 sin 2 <f>, 
 
 where the upper sign must be used for the stroke from 
 B x to B 3 , and the lower sign for the return stroke. As 
 we have approximately 
 
 t l^- n — o ■ 9 r * sin 2 4> 
 L- VL 2 -»- 2 sin 2 = a-r > 
 
 we shall also have approximately, 
 
 S - r a - oos *) T t!™!*. . (220) 
 
 In applying formula (220) to the motion of the slide- 
 valve, we must substitute the eccentricity, f, of the 
 eccentric for r, and angle \},, Fig. 319, for 0. At the 
 
240 
 
 PRACTICAL ELECTRICAL ENCllVEER/JVC. 
 
 same time, as f is very small compared with the length 
 of the eccentric rod, the second term in (220) may be 
 neglected, and the formula for the slide-valve will be 
 
 S = f (1 - cos V0 • • • (221) 
 
 We thus see that the motion of the slide-valve will 
 be the same in both directions. 
 
 in shape and size. The longer, however, the connecting- 
 rod is, compared with the crank arm, the smaller will 
 the last term in (220) be, and the more will the two 
 motions of the piston be alike. 
 
 Zeuner's Slide-Valve Diagram. — The slide-valve, 
 S V, in Figs. 317 and 318 is an ordinary D-valve, which 
 is placed at its central position in Fig. 317 ; b and e 
 
 Fig. 316. 
 
 It may be remarked that the stroke of the slide- 
 valve is termed the " travel " of the slide-valve, and 
 is equal to the " throw " of the eccentric, or twice 
 the eccentricity. 
 
 Fre. 317. 
 
 JCL b £ 
 
 ,- - - -J> - • - « ,, 
 
 ^% 
 
 Fig. 318. 
 
 The motion of the steam-piston, however, will follow 
 formula (220), and will therefore be different in the 
 two strokes. It is for these reasons that the indicator 
 diagrams taken from the two ends of the cylinder differ 
 
 are the lengths of the steam-ports and exhaust-ports 
 respectively. 
 
 a is called the outside lap, and causes the steam-ports 
 to be opened later and shut earlier. 
 
 c is the inside lap, which increases the period of com- 
 pression and delays the release. 
 
 Fig. 318 shows the valve moved through a distance, 
 x, from its central position. 
 
 The line, X, in Fig. 319 represents the direction 
 of the piston-rod, as well as that of the slide-valve rod. 
 is the centre of the crankshaft, and Y is perpen- 
 dicular on OX. f is the eccentricity of the valve 
 eccentric, and with that as radius, a circle, B 1 D 2 , is 
 drawn. 
 
 When the crank is in the dead-point position, C lf 
 then the eccentric centre will be at T) v and the centre 
 line, D x , of the eccentric will form an angle S, the 
 angle of advance, with Y. The eccentric centre will 
 be at D 2 , and the centre line of the slide-valve will be 
 in position Ej, as in Fig. 318, when the crank has 
 turned through an angle 0. The slide-valve will be 
 at its central position, E, when angle xjr is 90 deg. 
 
 The problem now is to find by a simple method the 
 relation between the distance, x, which the valve has 
 moved from its central position, and the angle $ 
 through which the crank has turned. 
 
 This problem can be solved graphically by means of 
 a diagram due to Dr. Zeuner, and which is shown in 
 Fig. 320. Draw first the two lines O C 1 and OY at 
 right angles to each other, the former representing the 
 
DETAILS OF STEAM ENGINES. 
 
 241 
 
 direction of the piston-rod ; then draw line Z, forming describe circle V, V 2 ; then V, R will be equal to x a 
 an angle S, the angle of advance, with Y. With f as which is the opening of the p^r to teaT Witt the 
 diameter, and with centres m line Z, draw two circles inside lap, o, as radius and (As centre J describe cirde 
 
 Fig. 319. 
 
 through O-these circles are called the slide-valve W r W 10 ; as x -c is the opening of the port to exhaust, 
 circles ; also draw C 2 to represent the position of the inside lap circle will serve to find the positions of 
 tiie crank when it has turned through an angle ; release and compression. 
 
 Fig. 320. 
 
 then it can be proved that OP 2 is equal to the dis- With as centre, and a + 5 as radius, draw the 
 
 tance x, which the slide-valve has moved from its circular arc through P 3 , and likewise with b + c a, 
 
 central position. radiuSj draw arc through P 8 
 
 With as centre and the outside lap, a, as radius, We will now proceed to discuss the diagram When 
 
242 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 the crank is in position C — i.e., just before it reaches 
 the dead centre — the port will begin to be opened to 
 steam, showing that the steam will begin to enter the 
 cylinder before the piston has completed its stroke. 
 
 In position C 8 . 
 In position C 4 . 
 
 In position C 5 . 
 
 The port is full open to steam. 
 The port is still full open, but the 
 
 valve begins to close it. 
 The cut-off takes place. 
 
 ft! 
 
 AJ - 
 
 «sal 
 
 
 ,-J 
 
 At the dead-point position, Cj, of the crank, the 
 opening to steam is V x P! ; the latter is called the 
 "lead." 
 
 The various positions of the crank, which are of 
 interest, are marked C 3 , C 4 , etc. ; 
 
 In position C a . 
 In position C 7 . 
 In position C s . 
 In position C 9 . 
 
 The valve is in its central position. 
 The release takes place. 
 The port is full open to exhaust. 
 The port is still full open to exhaust, 
 but the valve begins to close it, 
 
DETAILS OF STEAM ENGINES 
 
 243 
 
 In position C 10 . The compression begins. 
 
 In position C n . The valve is again in its central 
 
 position. 
 In position C . Admission begins. 
 
 We have now followed the crank round one revolu- 
 tion while we have discussed what takes place on the 
 one side of the piston. The position of the piston 
 corresponding to the various positions of the crank 
 must be determined by means of formula (220). 
 
 Meyer's Variable Expansion Gear. — By the simple 
 valve motion just described, the steam will always be 
 cut off at the same point, whatever may be the load 
 which the engine has to overcome. The speed of the 
 engine can therefore only be controlled by throttling the 
 steam before it enters the cylinder, and this is done, 
 as already mentioned, by the governor shutting the 
 throttle-valve more or less, according to the load. 
 Moreover, the slide-valve diagram shows that the ratio 
 of expansion is small. It is true that we could make 
 the valve cut off earlier, by increasing the outside lap 
 and the angle of advance, and also by diminishing the 
 throw of the eccentric, but these can only be varied 
 within narrow limits without causing practical diffi- 
 culties in producing sufficiently large openings to 
 the steam. 
 
 The principle of the Meyer's expansion gear now to 
 be described is the application of a second valve, which 
 slides on the back of the first valve. By the relative 
 motion of the two valves, the second one will cut off 
 steam from the first one, before the latter has reached 
 the position at which it cuts off steam from the 
 cylinder. 
 
 Fig. 321 is a drawing of a steam-cylinder to which 
 the Meyer's expansion gear is applied. SV is the 
 main slide-valve which, if it were alone, would 
 distribute the steam to the cylinder in the same way 
 as the D-valve in Pig. 318. The cut-off valve consists 
 of two blocks, ~B X and B 2 , which slide steamtight on the 
 first valve, and are moved by a separate eccentric. The 
 parts of the valve-rod, C V B, which pass through the 
 two blocks, are screwed, but in such a manner that 
 the thread through B x is left-handed and that through 
 B 2 is right-handed ; if we therefore turn the hand 
 wheel, H W, which is fixed on the tail-rod, C T, one 
 way or the other, the blocks will either be separated or 
 be moved closer together. The action of the cut-off 
 valve is now simply this, that the further the blocks 
 are removed from one another, the earlier will they 
 cover the ports, P 2 , of the slide-valve, and the greater 
 will the ratio of expansion be. By this arrangement, 
 the ratio of expansion can be varied from about 10, down 
 to that due to the slide-valve, S V, alone. It will be seen 
 that C V B can turn freely in the cross-head C x Hj, 
 whereas the slide-valve rod, S V B, is fixed in the 
 cross-head C 2 H 2 . By turning H W, a small pointer is 
 made to slide along a scale fixed on the bracket, thus 
 showing the relative position of the blocks, Bj and B 2 , 
 and therefore also the cut-off. 
 
 Jn order to produce a relative motion of the two 
 
 valves, the angles of advance, of the two eccentrics 
 must be different. The angle of advance for the slide- 
 valve may thus be 15 deg., and that for the cut-off 
 valve 85 deg. The motions of the two valves will 
 therefore be in the same direction in some part of the 
 stroke, and in another part of the stroke, the relative 
 motions of the valves will be in opposite directions. 
 
 As the cut-off, as shown in the drawing, must be 
 varied by hand, it follows that the engine must be con- 
 trolled by a throttle-valve unless the attendant is 
 always at the hand wheel. It is, however, possible 
 to make the governor turn the cut-off valve-rod, and 
 thus vary the ratio of expansion automatically, but the 
 motion is very slow. 
 
 Paxman's Automatic Expansion Gear. — Fig. 322 is a 
 horizontal section through the cylinder and valve-chest 
 of an engine provided with an automatic expansion 
 gear designed by Mr. Paxman. The distribution of 
 the steam is done by a three-ported main slide-valve, 
 S V, of a similar construction to that shown in Fig. 
 321, and by a separate double-ported cut-off valve, C V. 
 Between the two valves is a stationary anchor-plate, 
 A P, which is double-ported on the side next the main 
 valve, and is four-ported on the side next the cut-off 
 valve. Both valves are worked by valve-rods with two 
 nuts at each end. 
 
 The main valve cuts off at 75 per cent, of the stroke, 
 and is worked by a single eccentric, E c v with a 2fin. 
 throw, and an angle of advance of about 27 deg. The 
 valve has a t 9 ^ 11 - outside lap, but no inside one. The 
 stroke, or travel, of the cut-off valve is made to vary by 
 means of a link motion, Pig. 323, which is worked by 
 two eccentrics, Ec 2 and E c 3 , and which is controlled 
 by the governor, as shown. The cut-off valve has no 
 lap, and controls the steam to be let into the cylinder 
 by opening and shutting the four ports on the anchor- 
 plate. 
 
 The throws of the two eccentrics, E c 2 and E<, 3 , are 
 ljin. and T 9 F in. respectively, and their angles of advance 
 are about 60 deg. and minus 90 deg. ; they are therefore 
 also termed the positive and the negative eccentrics. 
 As the eccentric-rods are very long, the slot in the link, 
 L K, will be almost straight, and as the throws of the 
 eccentrics are very short, the obliquity of the link will 
 be small, and consequently the force required for 
 moving the link will also be comparatively small. 
 The travel of the cut-off valve will thus be diminished 
 when the balls of the governor fly out, and the valve 
 will thereby increase the ratio of expansion. The 
 cut-off valve-rod is forked at the end where it carries 
 a slide-block, B, which fits into the slot of the link. 
 
 The length of each of the four ports of the anchor- 
 plate, next the cut-off valve, is ^-in., whereas the 
 length of each of the ports on the main valve side is 
 lin. These ports have a T Vin. lap, and as the main 
 valve has a T 9 ¥ in. lap, it follows that the port of the 
 main valve will open to the port of the anchor-plate, 
 before it opens to the cylinder-port. 
 
 By this expansion gear the cut-off may be varied 
 from to 75 per cent. 
 
244 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 U uj LI 
 
 
DETAILS OF STEAM ENGINES. 
 
 245 
 
 Lubricators. 
 
 The object of lubrication is to diminish the coeffi- 
 cient of friction between the wearing surfaces of the 
 
 engine. If the lubrication fails, the friction will 
 increase and the efficiency of mechanism will diminish ; 
 besides which, the cooling surfaces will not be ample 
 enough for the increased heat to be dissipated, the 
 
 Fig. 325. 
 
 temperature will rise, and the wearing surfaces will 
 seize. 
 
 The lubricator most commonly used for main 
 bearings, connecting-rod ends, etc., consists of a box 
 
 of gunmetal or iron, with a lid on the top to keep 
 dust out. In the middle of the box are one or two 
 tubes, which are continued down to the wearing 
 surface to be lubricated. In each tube is a pin, at one 
 end of which a wick is attached ; the other end of the 
 wick is immersed in the oil with which the lubricator 
 is filled. The lubrication will thus continue by the 
 capillary action of the wick, as long as there is any oil 
 in the box. The rate of lubrication will vary with the 
 height at which the oil stands in the box, and will thus 
 diminish as the oil runs out. Such a lubricator will 
 therefore require a great deal of attention, or else it 
 will be either wasteful or not sufficiently efficient ; and 
 
 Fig. 326. 
 
 Fig. 327, 
 
 Fig. 328. 
 
 when the engine is stopped, the pin must be taken out 
 of the tube, or the lubrication will continue. 
 
 It is for these reasons that great attention has been 
 paid to the designing of lubricators with adjustable 
 feed, whereby the rate of lubrication can be made 
 constant and sufficient for the rubbing surfaces which 
 are to be lubricated. 
 
 A lubricator of this class is the " Crosby Visible 
 Drop -Feed Lubricator," which is illustrated in 
 Pig. 324. 
 
 At the bottom of the glass cup, and inside the tube, 
 is a small cone valve, which is held down on its seat 
 by a spiral spring when the lever, C, is in a horizontal 
 
246 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 position. By turning the lever into the vertical 
 position, as shown in the engraving, the cone valve is 
 lifted and oil can pass out of the cup. The lift of the 
 valve, and thereby also the rate of lubrication, can be 
 adjusted by screwing nut A up or down. B is a lock- 
 nut which prevents A from turning when adjusted. 
 
 Steam Cylinder Lubricators. — A very simple, but 
 rather wasteful, apparatus for lubricating the cylinder 
 and slide-valve of a steam engine is the grease-cup, 
 illustrated in Pig. 325. It is screwed into the top of 
 the cylinder. The oil which is contained in the 
 chamber, V, will, when the cock C 2 is opened, or 
 
 rniR 
 
 Fig. 329. 
 
 Fig. 330. 
 
 The feed can be stopped and started by turning lever, 
 G, without interfering with the adjustment. As the 
 cup is made of glass, and the feed is visible, the driver 
 can see at a glance whether the apparatus is in proper 
 working order. 
 
 partly opened, bo displaced by steam from the 
 cylinder, and thus fall into the latter. The lubri- 
 cator is refilled by shutting C„, and gradually opening 
 cock C v When steam has ceased to issue from the 
 opening, oil is poured into the cup, Cp, until V is 
 
DETAILS OF STEAM ENGINE'S 
 
 24? 
 
 Fig. 335. 
 
 Fig. 331. 
 
 Fig. 332. 
 
248 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 filled. C is then shut. The cylinder receives in this 
 way a charge of oil now and then, and the rate of 
 lubrication is most irregular. 
 
 In Dewrance's " Window Lubricator," which is illus- 
 trated in Figs. 326, 327, and 328, the rate of lubrication 
 can be regulated by the cone valve on the end of the 
 spindle, S. This lubricator is also screwed into the 
 top of the cylinder, and when the cone valve is opened, 
 the steam from the cylinder will pass through hole 
 h 4 and tube, T, into the hpllow space inside spindle, 
 Sp. Some of the steam will condense and fall as 
 water through the hole h into the container, V, 
 which is filled with oil. The water will buoy up the oil, 
 which will pass through hole h, and if now the steam 
 pressure in the cylinder falls, then some of the contents 
 inside spindle, Sp, will be drawn through tube, T, into 
 the cylinder. This lubricator is thus designed to work 
 where there is a variation of pressure such as is found 
 in a steam-cylinder. It has a window of thick annealed 
 glass, G, which is made tight with asbestos packing, 
 and which enables the driver to see the rate of feed 
 and to regulate it by means of the cone valve. 
 
 In order to fill the lubricator, the cone valve must 
 be screwed up tight, and spindle, Sp, unscrewed four 
 turns ; the oil is then poured into cup, C, and runs 
 through hole h into the lubricator. When the oil is 
 gone, the water is discharged into the cylinder through 
 hole h 3 by turning spindle, Sp, half a turn arid opening 
 the cone valve. When used where the steam is very hot 
 and dry, the lubricator may be worked by unscrewing 
 the spindle, Sp, one quarter of a turn, and regulating 
 the feed by the cone valve; the oil will then pass 
 through hole h 3 . Figs. 327 and 328 are horizontal 
 sections through hole h 2 and spindle, S, respectively. 
 
 The most rational, and therefore also the most 
 economical, way of lubricating a slide-valve and a 
 steam-cylinder is by the use of a sight-feed lubricator, 
 in which the quantity of oil supplied to the cylinder is 
 visible to the driver and under his control. 
 
 In Figs. 329 and 330 is shown a sight-feed lubricator, 
 designed and made by Messrs. Chas. Winn and Co., of 
 Birmingham. The lubricator is bolted to a bracket on 
 the engine steam-pipe, with which latter two connec- 
 tions are made — one at A nearest the boiler, and the 
 other at B, both on the boiler side of the stop-valve. 
 Some steam from the steam-pipe will find its way 
 through the connection at A into the condenser, C, 
 where it will be turned into water. By opening the 
 cone valve at D, water will pass from the condenser 
 into the container, F, which is filled with oil. The 
 water will now buoy up the oil, some of which will pass 
 through pipe P, and when the regulating valve at J is 
 opened, the oil will find its way through nozzle, N, and 
 rise as drops through the water with which the sight- 
 glass, K, is filled. The oil will accumulate underneath 
 plug, L, and when the valve at M is opened, will find 
 its way through the connecting pipe at B into the main 
 steam-pipe, where it will be thoroughly mixed with the 
 steam which is intended for the engine. 
 
 As the steam pressure is the same on both openings, 
 
 A and B, it follows that the pressure by which the oil 
 is forced out is due to the head of water in the condenser 
 above the oil-pipe at B. 
 
 Fig. 330 is a section through the container in a plane 
 at right angles to Fig. 329. H is a gauge-glass to indi- 
 cate the contents of oil and water in F, and I is a 
 three-way drain-cock through which the container can 
 be emptied. 
 
 To fill the lubricator, all valves must be closed, the 
 lever of the drain-cock, I, must point upwards as shown 
 in the drawing, the two plugs B and L must be taken 
 out, and condenser and sight-glass filled with clear 
 water. Then take out plug P and fill container with 
 oil ; if there is not sufficient oil at hand, fill up quite 
 full with water. The plugs are then replaced and the 
 lubricator is ready for working. 
 
 The lubricator is started by opening each of the 
 valves D and M half a turn, and the feed is regulated 
 by valve J. 
 
 To refill the container, close first valve J, then valve 
 D ; then run water out of container by placing the 
 lever of the drain-cock, I, pointing downwards, then 
 close drain-cock and fill container with oil. 
 
 Figs. 331-335 are drawings of the latest model of 
 Dewrance's sight-feed lubricator. This lubricator has 
 only one connection with the main steam-pipe. In 
 Fig. 331, the pipe, P 1; connects the lubricator with a 
 horizontal steam-pipe, whereas Fig. 332 is designed for 
 connecting it to a vertical steam-pipe. For the purpose 
 of making the action easily understood, the passage 
 of the oil through the lubricator is shown by dotted 
 arrows, whereas the flow of water is shown by Ml 
 arrows. 
 
 The steam passes through tube T lf opens check 
 valve, CV 1( and enters the coil, CI, where it is con- 
 densed ; it then passes through the air-cock, A C, 
 which is open when its lever points upwards; the 
 water passes then through tube T 2 , and shut-off cock, 
 S C, which is open when its lever is horizontal, as 
 shown in Fig. 332, passes then down tube T 4 to the 
 bottom of the container, V, and buoys up the oil, which 
 will fall through tube T 3 . When the regulating valve, B V, 
 is open, the oil will rise as drops through the water, 
 with which the sight-glass, G, is filled. The oil will now 
 pass between the glass and plug, Pg, will open check- 
 valve C V 2 , Fig. 333, pass out of the lubricator through 
 opening, b, Figs. 333 and 331, and will at last drop into 
 the steam-pipe, whence it will be carried by the steam 
 into the steam-chest and the cylinder. 
 
 The process of refilling the container is as follows : 
 First shut regulating valve and shut-off cock, then 
 turn water out through drain-cock, DC, then shut 
 the latter and open filling-cock, F C, and fill the 
 container with oil. 
 
 It will be noticed in Fig. 332 that the filling-cock 
 cannot be opened before the shut-off cock is shut. 
 The object of the air-cock, A C, is to relieve the appa- 
 ratus from air, which will be blown through passage 
 0, Fig. 335, by turning the lever. 
 
 Fig. 333 is a horizontal section through the shut-off 
 
DETAILS OF STEAM ENGINES 
 
 249 
 
 cock, Fig. 334 is a vertical section through same, and engines a special motor, the barring engine, is required 
 
 Fig. 335 is a vertical section through the air-cock. for this purpose. In Figs. 336 and 337 is shown plan 
 
 The Barring' Engine. and elevation of a barring engine, designed by Messrs. 
 
 In order to start an engine with single cylinder, Musgrave and Sons, of Bolton. It will be seen to 
 
 t^t 
 
 Fig. 336. 
 
 Fig. 337. 
 
 or a double-cylinder engine with cranks at 180 deg., have two vertical cylinders with cranks on a horizontal 
 
 or a compound engine, it is necessary to turn the shaft ; the quick motion of the latter is transformed 
 
 crankshaft, so as to give the cranks a proper position, into a slow one by means of a worm gear, as shown. 
 
 AVith small engines this is done by turning the The flywheel of the main engine which is to be 
 
 flywheel by hand, or by using a bar; but with large turned is provided with teeth into which the teeth 
 
'250 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Fig, 33S. 
 
 Sm 
 
 F - 559 
 
 af the cam-wheel of the barring sngine fit. Fig. 83$ ?reas Fig "'" shows the tw ; ; out 
 
 illustrates the barring engine in gear with tl e 
 
DESCRIPTION OF STEAM ENGINES. 
 
 251 
 
 DESCRIPTION OF STEAM ENGINES. 
 
 High - pressure 
 
 Having given an account in the preceding pages Low- pressure 
 
 of the working parts of a steam engine, it now only cylinder 
 
 remains to describe a few types of engines designed Cylinder drain- 
 
 for driving dynamo machines. cocks 
 
 HPC 
 
 LPC 
 
 d r 
 
 Main bearing ... MB 
 High-pressure 
 
 crank HPCr 
 
 Low-pressure 
 
 crank L P Cr 
 
 Crankshaft ... CKS 
 
 HPO-> 
 
 Fig. 340. 
 
 FF 
 
 12 Q 
 
 lllllllllhll 
 
 2 
 
 I 
 
 \FEET. 
 
 Fio. 341. 
 
 In order to assist the reader in understanding the Eeceiver 
 
 drawings now to be described, the following lettering Eeceiver drain- 
 will be adopted throughout. ^ < t? c ] iS , 
 
 Relief- valve 
 
 Steam-pipe 
 
 SP 
 
 Girder trunk ... 
 
 G T Jacket . 
 
 Exhaust-pipe .,, 
 
 B P 
 
 Girder head .,. 
 
 G H Piston-rod 
 
 E 
 
 Edr 
 
 EV 
 
 J 
 
 PE 
 
 Balance disc ... B D 
 
 Flywheel F W 
 
 Flywheel race ... F E 
 
 Foundation frame F F 
 
 Base-plate B P 
 
 Driving-rod ... DE 
 
252 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 High- pressure 
 
 piston HPP 
 
 Low-pressure 
 
 piston L P P 
 
 Cross-head ... CH 
 
 Connecting-rod . C E 
 
 Guide blades ... G B 
 
 Lubricator 
 
 L 
 
 Governor 
 
 Gr 
 
 Governor driving 
 
 
 wheel 
 
 GW 
 
 Dash-pot 
 
 DP 
 
 Automatic expan- 
 
 
 sion gear 
 
 AEG 
 
 formed by an inner cylinder or bush, which is cast 
 separately, accurately turned and bored, and forced 
 into the outer cylinder as described on pages 209 and 
 210. The cylinders are lagged with hair felt, wood, 
 and covered with sheet iron, and are provided with 
 necessary drain-cocks and sight-feed lubricator. 
 
 Fig. 342. 
 
 Horizontal Compound Steam Engines. The pistons and piston-rods are shown in Figs. 22s 
 
 Messrs. Davey,Paxman, and Co. — The engine shown and 230, on page 212, and described on page 211. 
 
 in Pigs. 340 to 343 is a 40 nominal horse-power hori- The piston-rod glands are of cast iron, with gunmeta' 
 
 zontal compound engine of the receiver type, making bush. 
 
 Pig. 343. 
 
 90 revolutions per minute. The cylinders are placed 
 side by side, and mounted on a strong cast-iron girder 
 or foundation frame. 
 
 They are made of special hard, close-grained cast iron, 
 accurately bored and faced. The diameter of the high- 
 pressure cylinder is 12|in., and that of the low-pressure 
 cylinder is 20in. The stroke of the two pistons is 
 24in. The cylinders are jacketed, the jacket being 
 
 The guide-bars, cross-heads, and slide-blocks are 
 shown in Figs. 234-238, and described on page 215. 
 
 The connecting-rods and straps are of wrought iron, 
 and are shown in Figs. 253-256, and described on page 
 217. 
 
 The crankshaft (see Figs. 269 and 270) is made of 
 steel, and extends at both ends beyond the engine bed, 
 so that the flywheel or a pulley can be put on either 
 
DESCRIPTION OF STEAM ENGINES. 
 
 253 
 
 end. The cranks are set at right angles. The crank- 
 shaft runs in three strong cast-iron phimmer blocks, 
 fitted with adjustable gunmetal bearings. The cranks 
 are balanced by means of a balance-disc. 
 
 The flywheel is 8ft. 6in. diameter and 17in. wide, 
 and is turned on the face to receive a driving belt. 
 
 The governor is of the Paxman adjustable high-speed 
 type, shown in Figs. 303 and 304, and driven by a 
 gearing consisting of tooth- wheels and a driving-rod, 
 as shown in the drawings. 
 
 The high-pressure cylinder is fitted with Paxman's 
 
 The engine may be started in almost any position of 
 the cranks, by means of a small valve — the by-pass 
 valve — which opens communication between the steam- 
 pipe and the low-pressure cylinder. 
 
 Undertype Compound Steam Engine and Boiler. — The 
 engine shown in Figs. 344, 345, and 346 is precisely of 
 the same size and construction as the one just described. 
 The boiler, L B, which is of the locomotive type, is 
 placed over the engine, the smoke-box end being sup- 
 ported by a rest, Et, on the cylinders, and the firebox 
 end resting, as usual, on the ashpit-box, A P. 
 
 UT 
 
 CRS 
 
 Fig. 346. 
 
 automatic expansion gear, consisting of two slide- 
 valves — one main and one cut-off valve — precisely as 
 described on page 243, and shown in Figs. 322 and 323. 
 The eccentric working the main valve is marked S VE, 
 and the two eccentrics for the link motion are lettered 
 AEG. 
 
 The valve-chest of the low-pressure cylinder is 
 placed between the two cylinders, and the steam is 
 admitted to the cylinder by a single slide-valve, cutting 
 off at about 53 per cent., and worked by an eccentric, 
 IjPS. The valve-chest is closed by cover, CC. 
 
 The boiler is made entirely of mild ductile steel 
 plates. All edges of plates are planed. The front 
 plate, firebox, and other plates are flanged from the 
 solid with special machinery. The shell is double 
 riveted in the longitudinal seams, and the boiler is well 
 stayed throughout. The shell is in. thick, the internal 
 firebox ^in. thick, and the tube-plates ^in. thick. 
 There are 100 tubes, F T, of 2Jin. external diameter, 
 made of lap-welded iron. 
 
 The boiler is usually fitted with 12ft. of chimney, 
 Cy, on an uptake, UT, and with one double-lever 
 
254 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 3 
 
 o 
 
 PR 
 
DESCRIPTION OF STEAM ENGINES. 
 
 255 
 
256 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 safety-valve, L S V, and also with the usual and neces- inch, and made for a working pressure of 1201b. per 
 sary fittings. square inch. 
 
 Fig. 347. 
 
 1V 
 
 I'ii i 
 
 IV 
 
 OP 
 
 V°l 
 
 
 X 
 
 EP 
 
 SV 
 
 U=&=HI 
 
 FF 
 
 FF 
 
 EP 
 
 FF 
 
 Fig. 348. 
 
 m 
 
 @ HI'C 
 
 rw 
 
 HI'C 
 
 SP 
 
 TZJ 
 
 4§ 
 
 <n -\ 
 
 LI'C 
 
 rr 
 
 ^ 
 
 4-0 2 
 
 hlil I L 
 
 8 10 12 14 FEET. 
 J I I I I I I 
 
 Fig. 349. 
 
 The boiler is tested by hydraulic pressure to 2201b. The stearn-pipe, being perforated within the boiler, 
 per square inch, and by steam to 1201b. per square acts as an anti-priming pipe and passes out of the boiler 
 
DESCRIPTION OF STEAM ENGINES. 
 
 257 
 
 through the smoke-hox. The steam is admitted to the 
 steam-chest of the high-pressure cylinder through the 
 stop-valve, S V. 
 
 condenser, C, and the piston-rod, P E, of the air pump 
 is coupled with the tail-rod of the low-pressure piston. 
 The condenser is of the construction shown in Figs. 
 
 The exhaust-pipe passes from the low-pressure 
 cylinder into the smoke-hox, thus allowing the exhaust 
 steam to assist in promoting the draught. 
 
 282, 283, and 284, and is described on page 227. The 
 governor is driven by two straps of two sets of 
 pulleys, Gr P. 
 
 bia. 351. 
 
 When the engine is a condensing one, as shown in The feed pump, F P, is driven by an eccentric, P Ec, 
 Figs. 347, 348, then the exhaust-pipe passes into the on the crankshaft. The suction-pipe is fixed on the 
 
258 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 DdH 
 
DESCRIPTION OF STEAM ENGINES. 
 
 259 
 
 flange, FSP, and the feed is regulated by an overflow girder type — i.e., the front cylinder cover, cross-head 
 cock, P C. F W P is the feed- water pipe, and C V guide, main bearing for crankshaft, and foot for 
 is the check valve. bolting to foundation, are cast in one piece, to 
 
 Fig. 356. 
 
 Coupled Compound Girder Steam Engine. which the cylinder with valve chest is bolted. 
 
 Messrs. Davey, Paxman, and Co. — The engine illus- The engines are placed side by side, and in such 
 
 trated in Figs. 349 to 352 consists of two separate a manner that the high-pressure engine is on the 
 
 engines coupled together. Bach engine is of the right-hand side, and the low-pressure engine on the 
 
260 
 
 PRACTICAL ELECTRICAL EtiGINEERlNC. 
 
 left-hand side (standing at the cylinder end and 
 looking towards the crankshaft) . The distance between 
 centre lines of the two engines is lift. 
 
 The size of the total engine is 100 nominal horse- 
 power, and the crankshaft makes 60 revolutions per 
 minute. The two cylinders are connected with a 
 receiver, and are provided with drain-cocks and relief- 
 valves. 
 
 The cylinders are made of hard close-grained cast 
 iron. The diameter of the high-pressure cylinder is 
 
 Cr 
 
 Sd 
 
 '<., 
 
 Sc 
 
 EfflL 
 
 Ar 
 
 w 
 
 f 
 
 I 
 
 I 
 
 w- 
 
 T 
 
 4J£ 
 
 |J 
 
 Y 
 
 Fro. 357. 
 
 iiftr 
 
 22in., that of the low-pressure cylinder is 35in., and 
 the stroke of both engines is 48in. The cylinders are 
 jacketed in the usual manner. 
 
 The cross-heads are made of cast steel, and are 
 shown in Figs. 239, 240, and 211, and are described 
 on page 215. 
 
 The connecting-rods are made of wrought iron, with 
 adjustable gunmetal bearings, and are shown in Fies 
 257, 258, 259, and 260. " 
 
 The crankshaft is made of steel, and is shown in 
 Fig. 268. It runs in adjustable gunmetal bearings. 
 The cranks are set at right angles. 
 
 The engine has two flywheels, 14ft. in diameter by 
 18in. wide, which are turned on the face to receive 
 
 driving belt. They are made in two parts, connected 
 by bolts C B. 
 
 The governor and slide-valves are precisely of the 
 same types as described in the preceding engines 
 made by the same firm. The main valve for the high- 
 pressure cylinder is lettered HSV. A S P is a pipe for 
 passing steam into the LPC from the by-pass valve. 
 
 The low-pressure engine may be fitted with an 
 injection condenser in the same manner as shown in 
 Figs. 347 and 348. 
 
 The Robey Compound Engine. 
 
 This engine, which is illustrated in Figs. 353, 354, 
 and 355, is a coupled compound receiver engine with 
 cranks at right angles, the two sub-engines being of 
 the girder type. The diameter of the high-pressure 
 cylinder is 18Jin., and that of the low-pressure cylinder 
 is 30in., the stroke of both sub-engines being 40in. 
 
 The total engine will develop 200 indicated horse- 
 power at its most economical load, and a maximum 
 indicated horse-power of 320 with 801b. pressure on 
 the engine side of the steam stop-valve, and with the 
 crankshaft making 63 revolutions per minute. 
 
 The cylinders are made of close-grained cold blast 
 cast iron bored true, and the ends enlarged to 
 receive the covers. Each cylinder is steam-jacketed, 
 and is protected as well as the steam-chest with 
 lagging and covered by blue steel plates. The 
 cylinders are also provided with indicator cocks and 
 relief- valves. 
 
 Inside the receiver is a heating coil, which is con- 
 nected with pipes to the jackets of the two cylinders. 
 By this means, steam from the boiler will circulate 
 through the coil as well as through the cylinder jackets. 
 
 The pistons, which are described on page 212 and 
 shown in Fig. 232, have phosphor bronze rings with 
 steel springs in the high-pressure cylinder, and cast- 
 iron rings with steel springs in the low-pressure 
 cylinder, the diameter of the steel piston-rod being 
 3 Jin. The crankshaft is made of steel with a diameter 
 of 12in., and is enlarged in the middle to receive the 
 flywheel. The two journals are each 18in. long by 9in. 
 diameter, their bearings having already been described 
 and shown on page 223. A cast-iron crank disc is 
 forced under heavy pressure upon each end of the 
 crankshaft, and is further secured by a steel key. The 
 crank disc, into which a steel crank-pin is fitted, is 
 a large circular plate of 4ft. 3in. diameter, and counter- 
 weighted at the back to balance the moving parts. 
 
 The cross-heads and connecting-rods have been 
 described on pages 216 and 217. 
 
 The cylinder-end of the tubular girder frame is cir- 
 cular, and being turned and faced true with the axis or 
 centre line of the engine, forms the front cover of the 
 cylinder. A rim upon this circular end enters some 
 distance into the mouth of the cylinder; the joint 
 between the two being an accurate fit, is secured by a 
 number of steel studs and nuts. The girder head or 
 crank end of the girder frame forms a plummer-block 
 for the crankshaft bearing. 
 
DESCRIPTION OF STEAM ENGINES. 
 
 26 L 
 
 The most important feature of this engine is the 
 peculiar manner in which the distribution of the steam 
 is effected. The energy required for working the slide- 
 valves of an engine is no doubt in many cases very 
 considerable, especially where steam of high pressure 
 is used. In this engine the usual slide-valve is 
 dispensed with, and the steam is admitted from the 
 steam-chest into the cylinder, and again cut off by 
 double-beat lifting-valves, one at each end of each 
 cylinder. These valves are situated close to the ends 
 of the cylinders, so that the clearance to be filled with 
 steam at each stroke is considerably reduced. 
 
 Pig. 356 is a section, at right angles to the piston- 
 rod, through one of the valves, the steam-chest, and 
 the cover C G. The end, E, of the eccentric-rod, Ec E, 
 will describe an arc, with centre at Ec ; during part of 
 this motion, the tripper, Tr, which can turn on the 
 pin through E, will depress the outer end of the lever, 
 Lr, which, turning on its fulcrum, F 1 ^ will cause the 
 gunmetal valve, D V, to be lifted from its seat, S V. 
 The lifting of the valve begins just before the com- 
 mencement of the stroke. Owing to the different 
 arcs described by E and Lr, the tripper will at a 
 certain point slip out of contact with Lr, and the 
 valve will drop instantaneously and cut off the supply 
 of steam. To prevent the valve from being injured 
 by coming too heavily on its seat, its spindle, V E, 
 ends in the piston of a dashpot, D P, which in Eig. 356 
 is cut away, in order to show the governor which is 
 situated behind it. 
 
 The eccentric, which works the tripper, is fixed on 
 the driving-rod, D E 1 - 
 
 The lead of the valve, D V, can be adjusted by the 
 nut, A N, on the valve spindle. If the nut be screwed 
 upwards, the lead would be decreased ; screwing it down- 
 wards towards the valve, would increase the lead. 
 More lead can be obtained by shortening the eccentric- 
 rods, E C E, by means of their adjusting nuts. If the 
 rods be lengthened, the reverse will take place. The 
 point at which the tripper slips out of contact with Lr 
 can be adjusted by screw Sc 2 . In this way the ratio of 
 expansion can be made to be the same for each stroke. 
 
 This mechanism is sufficient for admission and 
 cut-off of the steam at fixed points of the stroke, and 
 is all that is required for the low-pressure cylinder. 
 
 In the high-pressure cylinder, however, the cut-off 
 of the steam should be automatic — i.e., should vary 
 according as the load or steam pressure varies, so as 
 to keep the speed of the engine constant. This is 
 effected by the governor, which in rising lifts lever 
 It 1 !, and thereby moves the fulcrum, FC 2 , of Lr to the 
 left, thus causing the tripper to slip out of contact 
 at an earlier point. The opposite will take place when 
 the flyers fall. The governor is loaded with a spring 
 inside its spindle, and upon this spring the governor pen- 
 dulums press. The speed of the engine may be varied 
 a few revolutions by adjusting the compression of the 
 spring by means of the nut, AN r To make the 
 engine run faster, screw the nut down, and unscrew it 
 to make the engine run slower. A cord, Cd, may be 
 
 attached to the lever L T r, and led away to any part of 
 the building. In case of accident to life or machinery, 
 by pulling the cord, the lever L J r draws lever Lr 
 completely clear of the tripper ; no steam can enter 
 the cylinders, and the engine is brought to a standstill. 
 
 The governor is driven by the driving- rod, D E, 
 which again is driven by D E 1 . 
 
 When the engine is used for electric lighting, an 
 electric governor may be applied, which consists of 
 two solenoids, Sd, Fig. 357, each of which has an iron 
 core, Cr, fixed to an armature, Ar. The end of the 
 governor lever, L a r, rests on the screw Sc 1 , which is 
 attached to the armature. By proper adjustment 
 of screw, Sc 1 , and the movable weight, A W, the 
 speed of the engine can be regulated by the strength 
 of the current passing through the solenoid. 
 
 The exhaust-valves, E V, are placed underneath the 
 cylinders, and, like the inlet-valves, close to the pistons, 
 so as to reduce the clearance. The exhaust-valve is a 
 gridiron valve, sliding upon a treble-ported face formed 
 upon the upper surface of the exhaust branch, and 
 worked by an eccentric upon the driving-rod, DE 1 . 
 Although these valves are slide-valves, they require 
 comparatively little power to move, as the pressure 
 upon them is small. On account of the position of the 
 exhaust-valve underneath the cylinder, the latter is 
 kept drained of any water which may accumulate 
 within it. The exhaust-ports are closed when the 
 piston has completed about seven-eighths of its stroke. 
 
 The engine can be turned by means of the bar, Br, 
 Eig. 354. The engine stop-valve is lettered SV, in 
 Figs. 353, 354, and 355, and the high-pressure exhaust- 
 pipe is lettered H E, in Figs. 354 and 355. L S is the 
 low-pressure steam pipe. 
 
 Vertical Side-by-Side Compound Quick-Speed Engine. 
 
 The engine illustrated in Figs. 358 to 362 is a 
 compound receiver engine with vertical cylinders 
 placed side-by-side, and made by Messrs. Browett, 
 Lindley, and Co. 
 
 The engine is intended to make 200 or 250 revolu- 
 tions per minute, as required, and is fitted with the 
 "Acme" governor, which is described and shown 
 on pages 234-236. The I.H.P. is 250, the high- 
 pressure cylinder being 15Jin. and the low-pressure 
 one 23|in. in diameter, both having 16in. stroke. 
 
 Figs. 358, 359, and 360 are sections through the 
 cylinders and the valve-chests. The steam in the high- 
 pressure cylinder is distributed by means of a piston- 
 valve, P V, which is shown in detail in Fig. 363. It 
 will be seen that the high-pressure steam, as well as 
 the exhaust steam, surrounds the piston-valve, and 
 therefore the friction which has to be overcome in 
 moving the valve will only be that due to the pressure 
 of the spring rings, Sg, against the liner, Lr. 
 
 The drawings also show that the steam from the 
 boiler is let through S I into the space S S, which 
 in ordinary engines is occupied by the exhaust steam. 
 When the top piston-rings in the up-stroke of the 
 valve have passed over ports h ; in the liner, steam 
 
262 
 
 PRACTICAL ELECTRICAL ENGINEERING.. 
 
DESCRIPTION OF STEAM ENGINES. 
 
 263 
 
264 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 from the steam-space, S S, will pass through the ports 
 h 2 and h x into the cylinder. At the same time, the 
 lower piston-rings having passed over ports h 3 allow 
 the low-pressure steam at the other end of the cylinder 
 to exhaust through ports h 3 into the high-pressure 
 exhaust space, HBS, The cut-off will take place at 
 the moment the top piston-rings, in the down-stroke 
 of the valve, shut ports h r The distribution of the 
 steam during the up stroke of the high-pressure 
 piston, HPP, is effected in the same manner. 
 
 The piston-valve is fixed to the rod, S P E, 
 which moves steam-tight through a stuffing-box, S B 3 , 
 
 both for steam and exhaust. This valve has two 
 internal steam passages, I P, which pass through the 
 valve and thus communicate at both ends with the 
 steam-chest. The rod, which works this valve, ends 
 also in a cross-head, which slides on a pair of guide- 
 blades. The eccentric, which actuates the low- 
 pressure valve, is fixed on the crankshaft, and the 
 cross-head end of the eccentric-rod is forked, as shown 
 in Fig. 361. 
 
 The piston of the high-pressure cylinder is of cast 
 iron, in one piece, and fitted with steel rings, which 
 are sprung into place and are expanded by a cut steel 
 
 Cn 
 
 is 
 
 J-i-i i i 
 
 2 FEET 
 
 Fig. 360. 
 
 and ends in a cross-head which slides on a pair of 
 guide-blades, G B*. The valve is actuated by an 
 eccentric, which is adjustable to vary the cut-off by a 
 screw mounted on a plate at one side, and a nut 
 attached to the eccentric, which is fitted with slides 
 and locking device. The eccentric-rod, which in 
 Pigs. 361 and 362 is marked H P V, is fitted at' the 
 top end to the cross-head of the valve-rod. 
 
 The high-pressure steam passes through the receiver 
 into the low-pressure steam-chest, L S C. The distri- 
 bution of the steam in the low-pressure cylinder is 
 effected by a flat slide-valve, which is double-ported 
 
 coil. The low-pressure piston is of cast steel, and is 
 fitted with rings and springs in the same manner as the 
 high-pressure piston. This construction of the pistons 
 dispenses with junk rings and screws, which are liable 
 to jar loose and cause breakdowns. The piston-rods 
 pass through stuffing-boxes, S B x and S B 2 . 
 
 A special feature with this engine is the cross-head, 
 which consists of a plain rectangular block of steel 
 forged solid with the piston-rod, and slotted out to 
 receive the cross-head brasses. The cross-head is 
 closed at the top by a forged steel cap, which is slipped 
 over at each end, and embraces the jaws to prevent 
 
DESCRIPTION OF STEAM ENGINES. 
 
 264a 
 
 Fig. 361. 
 
 the latter from spreading. The cap bolts, which pass The cross-head brasses are divided in two parts, and 
 through the block, serve also to receive the slipper- are tightened by a steel wedge of the same width as the 
 slide on the back of the block. brasses. The wedge is adjusted by a screw on the front 
 
264b 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 of the cap-plate. The cross-head pin is of steel, and is 
 3£in. in diameter by 6|in. wide, and has a hole through 
 its centre, which is 2Jin. diameter. 
 The connecting-rods are forked at the cross-head 
 
 lOin. long, and the two outside ones 12Jin. long. The 
 cranks are slotted, and are fitted with cast-iron 
 counter-balances. The diameter and length of the 
 crank-pins are 6|in. 
 
 Fig. 362. 
 
 end, and are of the same type as that shown in Fi^s 
 265 and 266 on page 221. b ' 
 
 The crankshaft is made of steel, and its diameter is 
 7m. at each end. It is carried by three bearings, the 
 diameter of which is G£in., the centre one being 
 
 The cylinders are cast with feet, Ft, which are 
 bolted to the vertical standard, and they are further 
 supported in front by three steel stays or columns, Cn. 
 The two cylinders are held together by bolts, Bt. 
 
 The stop-valve, S V, is opened and shut 'by means 
 
Description of steam engines. 
 
 2640 
 
 3 
 S 
 
 5- 
 
564d 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 of lever, Lr. On the left-hand side of the stop-valve, Willans Central Valve Compound Engine. 
 
 Fig. 361, is a by-pass-valve, which opens and shuts by The engine, Fig. 364, consists of two single-acting 
 
 IW//W 
 
 IS 
 
 ''I'll 
 
 IfiT 
 
 Fm. 364. 
 
 turning hand-wheel, H ; and whereby high-pressure tandem engines with their cranks at 180 deg. The 
 
 steam can be let through the steam-pipe, ASP, into reciprocating parts are therefore well balanced 
 
 the low-pressure steam-chest. At a steam pressure of lmh . per re inch and 
 
DESCRIPTION OF STEAM ENGINES. 
 
 264e 
 
 the crankshaft making 460 revolutions per minute, the 
 engine will develop 80 I.H.P. or 72 A.H.P. 
 
 The stroke is 6in., the diameter of the high-pressure 
 cylinder is 9|in., and that of the low-pressure cylinder 
 is 14in. The effective areas of the two pistons are 67 
 and 141 square inches respectively. 
 
 "Where higher pressure is available, the engine can 
 be arranged for a constant load of 100 I.H.P. 
 
 keeping the eccentric-rod in compression upon both 
 the up and the down stroke, and preventing knocking 
 in the eccentric strap. 
 
 V 2 , V3, V 4 , V 5 , and V 6 are provided with spring rings 
 pressing against the inside surface of the piston-rod. 
 In order to make the piston-rod move steamtight 
 up and down through the covers which separate the 
 cylinders, the glands, G lf G- 2 , an d G3, contain phosphor 
 
 Fig. 365. 
 
 Each line of pistons is connected to its corresponding 
 crank by two exactly similar connecting rods, with 
 a space between, in which works an eccentric, Ec, 
 forged solid on the crank-pin. The piston-rod, which 
 connects the two pistons belonging to the same line, is 
 hollow and provided with steam-ports through which 
 the steam can enter and leave the cylinders. The 
 distribution of the steam through the ports is effected 
 by a set of piston valves, V 3 , V 4 , V 5 , and V 6 , moving 
 inside the piston-rod and worked by the eccentric, Ec. 
 
 Fig. 366. 
 
 Fig. 368. 
 
 The reason why the eccentric is fixed on the crank-pin, 
 and not on the shaft as usual, is that the required valve 
 motion must be a motion relative to the piston-rod, 
 and as the latter moves with the crank-pin, the eccentric 
 must also move with the crank-pin. 
 
 Vj is a guide for taking the side thrust of the 
 eccentric, and V 6 separates the steam-chest from the 
 cylinders, except when the top ports are open; it 
 also serves to transmit the full steam pressure at all 
 times through the line of valves to the eccentric, 
 
 bronze packing rings with cast-iron spring rings, similar 
 to piston rings, but pressing inwards against the piston- 
 rod instead of outwards. 
 
 The lower end of the piston-rod is fixed on a guide 
 piston, Gr P, which works inside a guide cylinder, 
 G- C, and takes the side thrust of the connecting-rods. 
 The connecting-rod ends work on hardened steel pins, 
 which are fixed inside the guide piston. 
 
 The engine being a single-acting one, the connecting- 
 rods must always be in compression, never in tension, 
 
264f 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 For this reason, the upper crank-pin brasses, Br, 
 Fig. 365, of the connecting-rods are wider than the 
 lower ones, which are only intended to act as a stand-by 
 in case of accident. As the working brasses never 
 leave the crank-pins, it follows that no wear which can 
 take place can lead to knocking, as the connecting-rods 
 will follow up the wear automatically. But in order 
 that the lower brasses may be useful as a stand-by, 
 they should not be too far from the crank-pin ; the 
 wear should be taken up when it becomes excessive — 
 say, as soon as it exceeds ^m.. Sufficient slack, how- 
 ever, should always be left to ensure an audible knock 
 if the engine is allowed to race. In that case — i.e., 
 if the speed is allowed to exceed that at which the 
 momentum of the moving parts, upon the up stroke, 
 is balanced by the cushioning, the moving parts will 
 leave the crank-pin at some point before reaching 
 the top of the stroke, and will strike it again with 
 more or less force so soon as steam is admitted to the 
 cylinder. 
 
 The wear of the brasses can be watched in the 
 following manner : 
 
 A small hole is drilled in each guide piston, \m.. in 
 diameter. The hole is just visible below the bottom 
 edge of the guide cylinder when the crank-chamber 
 door is removed, and when the piston is at the bottom 
 of its stroke; when the entire diameter of the hole 
 is in view below the guide cylinder, it is time both to 
 set up the brasses so as to reduce the play, and to pack 
 up the connecting-rods by inserting packing-pieces 
 between the big ends of the connecting-rods and the 
 brasses. The connecting-rods, however, must not be 
 packed up sufficiently to take the hole quite out of 
 sight. Its lower side must still be in sight under the 
 edge of the guide cylinder. If the hole goes out of 
 sight entirely, there will not be enough clearance for 
 safety between the pistons and the top of their 
 cylinders. 
 
 The eccentric-rod is intended to work always in 
 compression, in the same way as the connecting-rods, 
 the holding-down power being furnished by the 
 pressure of the steam in the steam-chest acting con- 
 stantly upon the uppermost piston-valve, V 6 . It may 
 sometimes happen, if the engine is running without load 
 but at full speed, that the pressure in the steam-chest 
 is insufficient to keep the eccentric-rod in contact 
 with the eccentric upon the up stroke. If so, a slight 
 knocking may be heard, as the lower eccentric-strap is 
 purposely left a very easy fit upon the eccentric. Such 
 knocking will cease as soon as the engine is given work 
 to do. 
 
 The cylinders are mounted upon the crank-chamber 
 Cr C, which contains the bearings upon which the 
 crankshaft rests. The crank-chamber is partly filled 
 with a mixture of oil and water. The crank-pins dip 
 bodily into the lubricant at every revolution, and in 
 doing so they splash it to the upper ends of the 
 connecting-rods and eccentric-rods, and into the guide 
 cylinders, as well as into that part of the hollow 
 piston-rod where the guide, V ; , works, 
 
 The main bearings, which have no top brasses, are 
 at all times partly immersed in the lubricant. As the 
 lubricant consists of oil and water, its temperature 
 cannot rise above that of boiling water, and there 
 is, therefore, a greater guarantee against hot bearings 
 than in most other engines, so long as the supply 
 of water is maintained. 
 
 It has, however, been found that when the larger 
 non-condensing engines exhaust into the atmosphere, 
 some of the heat of the exhaust steam in the exhaust- 
 chamber, E Ch, passes by conduction down the sides of 
 the crank-chamber, CrC, and heats the lubricant. In 
 the smaller engines, where the radiating surface is 
 larger in proportion, this effect is not observed. Nor 
 is it present in the larger engines, except after long 
 runs, while in condensing engines the low temperature 
 of the exhaust steam prevents any trouble. In large 
 non-condensing engines, however, it is usual to fit two 
 small cross tubes in the crank-chamber, to which a cold- 
 water service may be connected. This may with 
 advantage be a portion of the feed-water. 
 
 The oil used in the crank-chamber should be best 
 castor oil or good olive oil, and must contain no 
 acid. The quantity of lubricant in the crank- 
 chamber can be ascertained by means of the gauge, 
 Fig. 366. A pipe replacing the plug at F 
 will serve as an overflow. A V acts as an air 
 vessel, to prevent violent oscillations of the surface 
 in the gauge. As the gauge communicates only 
 with the lowest part of the crank-chamber, very 
 little oil will pass into the gauge, and any overflow 
 which may take place consists, therefore, principally 
 of water, and can be caught in buckets and returned to 
 the crank-chamber when cold. The surface of the 
 lubricant in the gauge generally rises about fin. while 
 the engine is running ; this is due to the level being 
 lowered where the revolving cranks scoops out a path 
 in the lubricant. When water is required to be added, 
 it should be poured through the open top of the gauge. 
 Oil should be admitted through the funnel, Fn. A tap 
 is fitted at D for the purpose of occasionally drawing 
 the lubricant from the crank-chamber for the purpose 
 of cleaning the latter. 
 
 One sight-feed lubricator, fixed on the throttle- valve, 
 is sufficient for the cylinders and the valves while the 
 engine is running. Grease-cups are also fitted in the 
 holes, which on the drawing are closed by plugs, Pg 1 ; 
 but they are only used as a stand-by, or for giving 
 the engine a flush of oil at starting and just before 
 stopping. 
 
 The steam distribution is effected, as already 
 mentioned, by the piston-valves shutting and opening 
 the ports in the hollow piston-rod. In Fig. 364, 
 the left-hand line of pistons is shown in section, and 
 the line of piston-valves in elevation. The pistons are 
 upon the down or steam stroke. Fig. 368 is a section 
 through the line of piston-valves. 
 
 The high-pressure steam from the steam-chest, S C, 
 enters the piston-rod through ports 6, passing between 
 the piston-valves V and V B into the high-pressure 
 
DESCRIPTION OF STEAM ENGINES. 
 
 264c, 
 
 cylinder through ports 5. This cylinder will continue to 
 take steam as long as the ports 6 are still in the steam- 
 chest ; but on reaching the gland G s , the ports 6 will 
 pass out of the steam-chest, and cut-off will take 
 place. This will happen early or late, according as 
 the ports are cut lower down or higher up in the 
 piston-rod., or, what is the same thing, according 
 as the gland rings, G 3 , are more or less raised 
 above the cylinder cover by a distance-piece placed 
 below them. Normally, the ports are so placed as 
 
 or pressure, from the upper to the lower side of the 
 piston — i.e., from the high-pressure cylinder to the 
 receiver, E. 
 
 During the down stroke, the low-pressure cylinder 
 takes steam from the receiver through ports 3, the 
 piston-rod, and ports 2. This will go on as long as 
 ports 3 are in the receiver, but the cut-off will take 
 place at the moment ports 3 reach the gland G 2 . 
 
 The cut-off in the low-pressure cylinder will also 
 take place late or early according to the position of 
 
 u w 
 
 ~u U" 
 
 ■"•-. EO 
 
 AEC HPS HPCr fCW 
 
 BP 
 
 LPCr 
 
 12 
 
 llllll I'lllll 
 
 4 FEET 
 
 Fig. 369. 
 
 to give cut-off at either 0*4 or 0"6 of the stroke, 
 when the gland rings are as low as they can be fitted. 
 The height to which the rings are raised determines 
 the actual cut-off. 
 
 At about three-quarter stroke the ports 5 are 
 closed by the (relative) upward movement of V 5 . 
 
 Shortly before the piston reaches the bottom of its 
 stroke, V 5 , still moving upwards, begins to pass above 
 the ports 5, thus putting the ports 5 and 4 into com- 
 munication, and allowing the steam to be transferred 
 upon the up stroke, with no practical change of volume 
 
 ports 3 in the piston-rod and to the height of the gland 
 rings, G 2 . Also here, shortly before the end of the 
 stroke, the valve V 3 will commence to uncover ports 2, 
 and to allow the low-pressure exhaust steam to pass 
 through the piston-rod, and out again through ports 1 
 into the exhaust-chamber, E Ch. 
 
 The right-hand line of pistons, shown in elevation, 
 Fig. 364, are upon the up or exhaust stroke. The 
 valves are necessarily invisible, but the course of the 
 exhaust steam is indicated by arrows. 
 
 The water above the high-pressure piston drains 
 
264h 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 downwards through the ports into the receiver, and 
 that above the low-pressure piston drains through the 
 ports into the exhaust-chamber during the exhaust 
 stroke. 
 
 The pistons, under a recent patent of Mr. Willans, 
 are dished on the upper side, and as the exhaust- 
 ports travel with the pistons, and are, so to say, at 
 the bottom of the hollow, both the rush of the steam 
 through the ports and the upward movement of the 
 pistons combine to drive out the water during the 
 exhaust stroke. 
 
 The valves, V 2 and V 4 , by rising at each stroke 
 above the bottom of the ports 1 and 4 respectively 
 prevent the accumulation of water in any part of the 
 valve motion. 
 
 parts can only be kept in compression upon the up 
 stroke by very powerful cushioning, which in most 
 engines is obtained by an excessive compression of the 
 steam. 
 
 Sometimes when the steam is exhausted into a 
 vacuum, sufficient cushioning cannot be obtained by the 
 usual means. In this engine, the cushioning is provided 
 by the guide pistons, which, on the up stroke, com- 
 press the air contained in the guide cylinders. Thus 
 any desired amount of cushioning can be obtained, 
 according to the clearance allowed in the cylinder. 
 
 The energy expended in compressing the air is given 
 out again by its expansion on the succeeding down stroke. 
 There are holes, h, in the guide cylinders, which are 
 uncovered by the guides at the bottom of the stroke, and 
 
 Cr n 
 
 I 
 
 o © 
 
 & o 
 
 FW 
 
 Fig. 370. 
 
 Besides the relief-valves shown in the drawing, the 
 partition between the low-pressure cylinder and the 
 receiver is fitted with an unloaded valve, through 
 which water in the cylinder can be discharged quickly 
 into the receiver. 
 
 Both cylinders and receiver are provided with drain- 
 cocks, which discharge into the crank-chamber. 
 
 The plugged holes, Pg 2 and Pg 3 , are intended for 
 fixing indicators. 
 
 Beference has already been made to the fact that the 
 connecting-rods and eccentric-rods are constantly in 
 compression — a condition rendered possible only by 
 the fact that the engine is single-acting, giving no 
 pull to the crank upon the up stroke, but only a 
 push upon the down stroke. In any engine running 
 a.t a high speed of revolution, however, the moving 
 
 as the casing which surrounds the guide cylinders and 
 forms part of the framing of the engine is open to the 
 atmosphere, it is evident that the air compression 
 always commences at atmospheric pressure, and is 
 constant in its results whatever alteration may be 
 made in the pressure of the exhaust steam. 
 
 Air or relief-cocks, A C, are fitted upon the guide 
 cylinders, in order to avoid compressing the air in them 
 when the engine is turned by hand, and to facilitate 
 starting. If the cocks are opened at starting, they 
 must be closed again immediately, and they must never 
 be open when the engine is running at speed, or the 
 necessary cushioning will be wanting. A drain-cock is 
 fitted to the exhaust-chamber, which should be con- 
 nected, by a copper pipe, with some convenient place 
 for discharging hot water ; but if the exhaust-pipe is led 
 
Description of steam engines. 
 
 264 1 
 
 downwards, and is itself properly drained, the exhaust- 
 chamber drain may be dispensed with. 
 
 The high-pressure pistons can be inspected by first 
 disconnecting the steam-pipe at S I, and the governor 
 spindle by unscrewing nut N 2 . The joints Jt 2 should 
 then be broken, when the steam-chest and the high- 
 pressure cylinders can be lifted off in one ; the hollow 
 piston-rods will draw through glands G- 3 . 
 
 If it be required to inspect one of the low-pressure 
 pistons, then the bolts in the high-pressure piston 
 should be unscrewed, and the catchers, which enter 
 some of ports 4, should be removed. The high-pressure 
 
 spindle should be disconnected, and the connecting-rod 
 ends, as well as eccentric-strap, should be undone 
 through the crank-chamber door. The joint at the 
 level of the crankshaft should then be broken, and the 
 upper part of the engine should be lifted ; the crank- 
 shaft will then be left lying on the main bearings. 
 
 The governor, which is fixed on the crankshaft, con- 
 sists of two flyers, Fl, attached to bell-crank levers 
 with fulcrums at Fc. The springs Sg 2 , which supplies 
 the deviating force, cause the bell-crank levers to 
 press against the sleeve, SI, which otherwise would 
 be moved by the force of spring Sg 1 in the direction of 
 
 © © € 
 
 © 
 
 © 
 
 
 © 
 
 cc 
 
 © 
 
 © 
 
 
 © 
 
 © © < 
 
 Fig. 371. 
 
 piston can then be lifted up together with the upper 
 part of the piston-rod, and the valves V 6 and V 5 will 
 be exposed. Then by breaking the joint Jt 1 , the low- 
 pressure cylinder can be lifted. 
 
 The guide piston can be examined by removing the 
 low-pressure piston, and unscrewing the guide cylinder 
 cover, which will draw over the piston-rod. Then 
 through the crank-chamber door undo the connecting- 
 rod ends and the eccentric-strap, after which the guide 
 piston can be lifted out of its cylinder. 
 
 If it be required to expose the crankshaft and the 
 main bearings, then the steam-pipe and the governor 
 
 the arrow. The action of the governor is, therefore, to 
 release the sleeve, more or less, according to the rotary 
 speed of the crankshaft, and thereby cause the throttle- 
 valve to shut or open for the steam. 
 
 The tension of spring Sg 1 can be adjusted by 
 screwing nut N x up or down, with the virtual effect 
 of adding to, or decreasing the centrifugal action of the 
 flyers, thus enabling a given opening of the flyers, and, 
 therefore, a given opening of the valve, to be obtained 
 at a somewhat increased or oininished speed as may 
 be desired. The length of the governor spindle, 
 which is in two parts, E x and E 2 , can be adjusted 
 
264j 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 by means of the long nut, Fig. 367. The top part 
 of this nut is split, and can be opened sufficiently to 
 receive the part R 2 of the spindle ; and by screwing 
 the nut N 2 down, R 2 will be squeezed in the hole 
 and held in position. 
 
 The "Windsor" Quick-Speed Engine. 
 
 This engine, which is made by Messrs. Davey , Paxman, 
 and Co., and illustrated in Figs. 369, 370, and 371, 
 is a compound receiver engine with inverted cylinders 
 placed side-by-side, and with cranks at 180 deg. The 
 cylinders are mounted on two cast-iron standards, 
 Std, and are supported in front by two steel stays, Sy. 
 The whole engine is securely fixed on a strong cast- 
 iron bed-plate. 
 
 The size of the engine is 25 nominal horse-power, 
 and is designed to work with a steam pressure of 
 1201b. per square inch, and at a speed of 160 to 220 
 revolutions per minute. The diameter of the high- 
 pressure cylinder is 10in., and that of the low-pressure 
 cylinder is 16in., the stroke of both pistons being 12in. 
 
 The details of this engine have already been described 
 and shown : The pistons on page 211 and Figs. 229, 
 230, aDd 231 ; the cross-heads on page 216 and Figs. 
 
 242 and 243; the connecting-rods on page 221 and 
 Figs. 265-267 ; the crankshaft on page 222 and Figs. 
 271 and 272 ; the governor on page 234 and Figs. 303 
 and 304 ; and the automatic expansion gear on page 
 
 243 and Figs. 322 and 323. 
 
 The governor is driven by a series of three wheels, 
 GW, the first one of which, is fixed on the crank- 
 shaft, the second one on a spindle, supported by 
 bracket Br 4 , and the third one is mounted on a spindle 
 which drives the governor-spindle by means of a pair 
 of bevel-wheels. The governor is mounted on a 
 bracket, Br 3 , and its spindle is supported by a pivot 
 bearing on bracket, Br 2 . The governor works the link 
 motion by means of the three levers L x , L 2 , L 3 , of 
 which the two first ones are fixed to the horizontal 
 spindle, Sp, supported by Br. 
 
 The cross-head pins, as well as the crank-pins, are 
 lubricated from the oil-box, O B, the oil being carried 
 through the four tubes, O T. 
 
 The high-pressure and the low-pressure distributing 
 sliding-valve gears are lettered HPS and LPS respec- 
 tively, and the slipper guides, which are bolted to the 
 front of the standards are lettered S G. 
 
SWITCHES AND SWITCHBOARDS. 
 
 265 
 
 CHAPTER XIII. 
 
 SWITCHES AND SWITCHBOAEDS. 
 
 SSUMING that the engines and dynamos of 
 the installation are all in order, it is seldom 
 convenient to send the current direct from 
 the terminals of the dynamo into the con- 
 ducting wires. The intermediate apparatus 
 in large central stations is often a rather 
 complex affair, but in early days, and with series 
 circuits, it often consisted of a single make-and-break 
 switch. It will be better, in this brief description, to 
 commence with the simplest case, and gradually con- 
 tinue more and more complicated arrangements till we 
 come to those of the present time. For the most part, 
 the study of the illustrations given will be sufficient 
 information after the underlying principles have been 
 discussed. 
 
 -E3 i«— 
 
 A B 
 
 <*> 
 
 %t c I 
 
 
 A ^* B 
 
 
 4 
 
 s ci 
 
 — WIILMM 3 
 
 c 
 
 A B 
 
 Figs. 1, 2, and 3. 
 
 are the fixed contact-pieces in the circuit, EPF the 
 moving piece pivoted at P carrying conducting pieces 
 at E and E separated by non-conducting material. 
 When E P F is in the position shown by the dotted 
 line the circuit is closed. When it is in the positive, 
 shown by the thick lines, there are two gaps in the 
 circuit, one at each pole of the machine. 
 
 In turning the switch, and especially if it is worked 
 by inexperienced hands, it is quite possible to hesitate, 
 or to turn the switch slowly. We shall see shoitly 
 that it is necessary to avoid such slow action. 
 
 In the first place, however, a good switch, when used 
 as a bridge to make contact, so completes the circuit 
 that there is no increase of resistance other than there 
 would be if the gap was completed by an unbroken 
 continuity of the main conductor. Then it is of the 
 
 A B 
 
 — EH5> 
 
 i — > 
 C D 
 A I 
 
 P.: 
 
 C D 
 
 Figs. 4 asb 6. 
 
 The switchboard, or rather the switch in its simplest 
 form, merely consists of an apparatus to complete or to 
 break the circuit. In Figs. 1 and 2 let D C represent a 
 conductive circuit, with a break at A B. Suppose A B to 
 be terminals which can be bridged or connected by the 
 conductor, S. One end of S may be fixed to one 
 terminal as a pivot, or fixed as shown. When the con- 
 ductor, S, is in the position shown in Figs. 1 and 2, there 
 can be no current, but when it is brought into the 
 position shown in Fig. 3 the circuit is completed, and 
 current is possible. 
 
 Switches that simply make and break contact at one 
 point of the conductor, or rather at one point leading 
 from one pole of the machine, are called single-pole 
 switches ; those that break and make contact at two 
 points in the conductor, or rather at two points leading 
 from the two poles, are called double-pole switches. 
 
 Figs. 1, 2, and 3 represent single-pole switches, Figs. 
 4 and 5 double-pole switches. The arrangement of 
 these switches is shown in the figures, where A B, C D 
 
 greatest importance that when the circuit is broken the 
 switch should act quickly, or an arc will be formed, 
 which soon does a considerable amount of damage. 
 Just examine the conditions existing when breaking 
 contact. Suppose the pressure used is considerable, 
 such as would send a current through a very high 
 resistance. At one moment in the action of breaking 
 the circuit the bridge-piece of the switch must neces- 
 sarily be just leaving the contact-piece, and the air 
 space, or gap between them, is infinitesimally small. 
 The pressure is sufficient to send a current through this 
 small increase of resistance, and it does, but in so doing 
 tears off minute particles of the conductor, raises these 
 particles to incandescence, and bridges over the gap 
 with a bridge of these incandescent particles. This 
 bridge is termed an arc, undoubtedly a corruption of 
 arch. The particles are mostly torn from one side of 
 the conductor, which, if the action occurs often, is soon 
 burnt up. We require the most absolute contact 
 between two conductors to avoid arcing. As the 
 
266 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 distance between the conductors increases, the 
 resistance increases and the current decreases, till at 
 length the resistance becomes so great that the current 
 ceases altogether. If then a break is made slowly so that 
 the arc continues over an appreciable length of time, and 
 the space is filled with incandescent particles, the con- 
 ductor is soon burnt away and must be renewed. Again, 
 the resistance decreases more slowly when the break is 
 
 handle merely releases 
 without interference. 
 
 the spring, the latter acts 
 
 Fig. 6. 
 
 Fig. 9. 
 
 slow than when it is quick, because the bridge of In other cases the switch, when on, is held in 
 particles can be more complete, and the length of the position by a strong spring, and the hand in turning 
 gap must be longer than would be necessary in their 
 absence. 
 It would be a mistake under these circumstances to 
 
 depend upon the hand alone for turning off the switch, 
 hence it is usually supplemented by a strong spring, or, 
 better still, the action of the spring alone is utilised. 
 The switch piece may either be rigidly connected to its 
 handle, when both hand and spring may aid in break- 
 ing contact, or the manipulating handle may not 
 be rigidly connected, but be quite independent, and 
 turn loosely upon the pivot. Then when moved, the 
 handle moves a stud or projection of some kind releases 
 the spring, which alone acts in moving the contact- 
 
 Fig. 10. 
 
 off the switch has to exert a considerable force, which, 
 as in the case of the switches at the Post Office, 
 
 -l~ 
 
 Fig. 8. 
 
 piece. In the case 
 
 where hand and sprint 
 together the hand may really, and often in practice contact rapidly, 
 does, retard the action of the spring, but where the The metal of 
 
 work renders it almost impossible to do other than break 
 ntact rapidly. 
 The metal of which the switch is made should be a 
 
SWITCHES AND SWITCHBOARDS. 
 
 267 
 
 good conductor — hard copper or gunmetal for pre- 
 ference — or in some cases, for small switches, of brass. 
 The pieces carrying the current must be of sufficient 
 section to carry the largest possible total current 
 without heating. The danger of heating in a switch is, 
 
 these reasons it is advisable to have the rubbing 
 contacts large and the springs which press them 
 together strong, and in large switches adjustable to 
 take up the wear. For the second reason it is advisable 
 never to allow the mechanical pivot of the moving 
 contact-piece to act also as an electrical contact, but to 
 
 make the two functions independent, and so arrange 
 the contact-piece that it shall be a pure electrical bridge- 
 over from terminal to terminal. This is happily done in 
 the Post Office switches referred to above and illustrated 
 below. With the very best arrangement of contacts 
 
 Fig. 12. 
 
 however, more to be feared from accidental increase of 
 resistance arising between the surfaces which make the 
 contact. This may occur from separation of the 
 contacts due to either wear and tear, or to loosening of 
 the screws of the switch, or it may be due to dust, dirt, 
 or oil getting between the surfaces. For the first of 
 
 Fig. K. 
 
 there is still some danger of the parts heating, and it is 
 therefore advisable to mount the whole upon an in- 
 combustible base, as of slate or porcelain. In addition, 
 it is necessary to see that screws which fasten contact- 
 
 pieces do not pass through the base, but are counter- 
 sunk and firmly bedded, where not open to inspection, 
 in plaster or some similar incombustible material. The 
 aim of all designs of switches, then, is to have : 
 1. Sufficient thickness of conductor in the contact- 
 pieces to prevent the slightest heating. 
 
268 
 
 practical electrical engineering. 
 
 2. A good broad rubbing contact for making the circuit : 
 
 if possible double and adjustable for large switches, 
 the contact-pieces being, of course, so insulated 
 that no current leaks through moving parts. 
 
 3. A very quick and sufficiently long break . when 
 
 breaking the circuit — a loose handle moving 
 independently of the contact-piece being prefer- 
 able in order to ensure sudden spring off. 
 
 4. No circuit through the pivot, but a pure bridge- 
 
 over. 
 
 5. Incombustible base. 
 
 6. Absence of metal screws, etc., at the back. 
 
 7. Simplicity in its parts, and ease in fixing. Besides 
 
 which, for high-tension switches there should be : 
 
 8. Absence of all possibility of accidental contact with 
 
 the conductors by any person standing near, or 
 manipulating the switch. 
 
 Various devices are used to ensure good contact. 
 There are two pieces of conductor required to obtain 
 this contact. Let us call one the fixed piece, the other 
 the moving piece. In the first place, both fixed and 
 moving pieces are generally slightly bevelled, so that 
 there shall be no jar or trouble of getting them one on 
 the other. The fixed piece is frequently in the shape 
 of C the moving piece coming between the two arms, 
 and so, from the springiness of the arms, making 
 good rubbing contact. The spring of the arms also 
 compensates for the wear and tear of contact rubbing. 
 Other makers use a spring to force the moving piece 
 upon the contact-piece, or the fixed piece is a spring 
 itself. Instead of a single moving piece to make con- 
 tact, a moving piece made of wires or thin strips of 
 conductor is used. In other cases the moving piece is 
 split so as to form a spring. 
 
 Contact can be made and broken without the aid of 
 a switch. In many cases a conical plug conductor is 
 used, Fig. 6. In this case we have two fixed pieces so 
 bored that the plug fits accurately. Contact is made 
 by pushing the plug home, at the same time giving it a 
 slight turn to ensure good contact. 
 
 Keferring again to .Figs. 1 to 5, it will be as 
 well to show how the single and double switches are 
 arranged in the circuit in practice. It is customary to 
 call the wire going from the + terminal of the dynamo, 
 the + lead ; and the wire going to the - terminal, the 
 - lead. Fig. 7 shows the use of tbe single switch. 
 The - wire is broken and the ends fixed to the binding 
 screws, A and B, of the switch. So long as A and B 
 are not conductively connected, the circuit is broken 
 and the lamp, L, is not lighted. Connecting A and B 
 by means of the switch, the lamp is put into the circuit 
 and receives current. The double-pole switch, Fig. 8 
 has both + and - wires, broken and connected to the 
 terminals AB, CD, as shown. The switch is shown 
 making contact, so that the lamp, L, is in circuit The 
 wire connections are usually made through perforations 
 in the base of the switch, so that there is no possibility 
 of accidental contact with the moving parts. 
 
 Switches are mounted on a variety of materials, the 
 larger ones sometimes on wood faced with asbestos, or 
 more usually upon slate, the smaller ones on porcelain. 
 The smaller ones for internal work must, of course, be 
 useful, but the principal aim of design is to make them 
 both ornamental and useful. These will be dealt with 
 when considering the fittings for interior work. Mean- 
 while the practice with larger switches will best be seen 
 by illustrations. Then, in addition to switches for the 
 purposes hitherto mentioned, switches are required for 
 a number of purposes — for regulating the charging and 
 discharging secondary batteries, automatic switches, 
 multiple switches, etc., all of which entail some differ- 
 ence in design, but none in principle. 
 
 Fig. 16. 
 
 Fig. 9 shows a wedge contact switch, with loose 
 handle. The spring of the contact-pieces, which clutch 
 the moving piece as in the figure, is strong enough to 
 cause a considerable power to be applied to the handle 
 when a break is necessary. This results in contact 
 being very sharp and abrupt. 
 
 Fig. 10 shows a double contact switch with brush 
 contact-pieces, and Fig. 11 shows a multiple contact- 
 piece similarly constructed. Figs. 12 and 13 show 
 double-pole switches. The latter figure shows the 
 switch in a case. In circuits where high pressure is 
 used the switches should always be protected. Fig. 14 
 shows a ring contact switch ; but instead of illustrating 
 some scores of switches, for their name is legion, the 
 engineer will prefer details of construction of one or 
 two selected specimens. Such details are shown in the 
 
270 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 7^ 
 
 m 
 
 m 
 
 m 
 
 o o o o 
 
 o o o o 
 
 °o° °o° 
 
 o o 
 
 n \ r 
 
 o 
 
 
 o 1 
 
 
 o 
 o" 
 
 o 
 
 
 
 6 
 
 
 
 o 
 
 
 o 
 
 
 o 
 
 o 
 
 
 o 
 
 
 b" 
 
 
 
 
 
 
 o 
 
 
 o 
 
 
 o 
 
 
 
 
 
 
 MH 
 
 Ema 
 
 in 
 
 03=0 
 
 m n 
 
 °o° °o° 
 
 m_m 
 
 o o o o 
 
 III 
 
 
 
 III 
 
 o o 
 
 O 
 
 
 
 
 
 
 o a o 
 
 III 
 
 III- 
 
 O 
 
 o 
 
 o 
 
 
 "o 
 
 
 r o 
 
 
 r o 
 
 
 
 
 
 
 
 o_ 
 
 [o. 
 
 
 o 
 
 
 o 
 
 
 o 
 
 
 o 
 
 
 o 
 
 
 
 
 
 
 o 
 
 
 o 
 
 
 o 
 
 
 
 
 
 
 o 
 
 ^ 
 
 o 
 
 -■ - 
 
 o 
 
 mm 
 
 [P^GHHa Bma 
 
 r 
 
 Fh:. -".-CKOMl'TOV- "' 
 
SWITCHES AND SWITCHBOARDS. 
 
 •271 
 
 7I\ 
 
 7K 
 
 7K 
 
 =**§ 
 
 oi 
 
 O! 
 
 ■n 
 
 M 
 
 ^ 
 
 & 
 
 V 
 
 ') i 
 
 ©I ©I ©l ©I 
 
 ^* 
 
 
 u,4 
 
 III 
 
 III 
 
 o o 
 
 
 o o 
 
 — M 
 
 nr_m 
 
 o o o o 
 
 °0° 
 
 III 
 
 o oT 
 
 TIT 
 
 O O 
 
 o o 
 
 °o° 
 
 III 
 
 o o 
 
 ITT 
 
 o p 
 
 o o 
 
 o 
 
 
 o 
 
 
 o 
 
 
 o 
 
 o 
 
 
 
 
 
 o 
 
 
 o 
 
 o 
 
 
 ~o 
 
 
 o 
 
 
 
 
 
 
 o 
 
 
 o 
 
 
 o 
 
 
 
 
 
 
 o 
 
 
 o 
 
 
 o 
 
 
 
 
 
 
 AZ 
 
 (€M 
 
 O O O O 
 
 III 
 
 III 
 
 O o 
 
 o o 
 
 Ml 
 
 °0° 
 
 III 
 
 c o 
 
 o • 
 
 o 
 
 
 o 
 
 
 o 
 
 
 o 
 
 
 
 
 
 
 
 
 o 
 
 
 o 
 
 
 o 
 
 
 o. 
 
 "o 
 
 
 "o 
 
 - 
 
 o 
 
 o 
 
 
 o 
 
 
 o 
 
 o 
 
 
 
 o 
 
 <■ 
 
 "o 
 
 
 
 
 
 o o 
 o o 
 o o 
 
 ^ 
 
 o o 
 o o 
 -O o 
 
 T 
 
 SWITCHBOARD. 
 
!74 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 NLUAllMt. h' EEDERS 
 
 1V-. £». -CI, >M1'T0^ 
 
 S' 
 
 J 
 
SWITCHES AND SWITCHBOARDS. 
 
 275 
 
 \ 
 
 oHAHGme cvRfiF-nr inpicatoks 
 
 ■O 
 
 -r 
 
 CHARGE 
 
 METER 
 
 0<~ 
 C6> 
 
 CHARGE 
 METER- 
 
 
 "IS 
 
 -r-a 
 
 \ UHOySWAY 
 
 \ K m itch 
 
 *J. 
 
 A5. 
 
 4 5. 
 
 ^% 
 
 & 
 $ 
 
 WiS^ 
 
 SM*1£ REQULATIfitO 
 
 k 
 -k 
 
 V \ &WITOH 
 
 <i!!/ flEGUt.ATWG 
 
 & 
 
 SHITVH 
 
 S0_ 
 fl- 
 
 sz 
 
 3- 
 
 et\ switch 
 
 W 
 
 * « 
 
 Pi 
 
 O 
 
 1 
 
 J)r*V+nno iZlrvvi-K 
 
 VOLTS 
 
 POSI Tl VE FEEDERS 
 
 6 6 6 6 6 6 6 6 
 DQ DO DD DD DD DD DD CD 
 
 PO^>lT I V £ FEEDER PLUG BOARD 
 
 CIRCUIT V O LT M £T T g /? "=; 
 
 ill 
 Z>/. 
 
 o 
 
 [<?: 
 
 B-t. 
 
 3a. 
 
 O 
 
 "6 
 
 o 
 
 o 
 
 2?3 
 
 o 
 
 ] — [ 
 
 o 
 
 POSIT/ y^ D Y/VSIMO PLUGBOARD 
 
 ; :h~i 
 
 1 [ 
 
 o 
 
 D4-. 
 
 O 
 
 o 
 
 Bi. 
 
 B* 
 
 O 
 
 o 
 
 o 
 
 £H UfJT 
 
 ft£SI^T4AC£ SWTC»a 
 
 *°iX- //VO/CATrtg. 
 
 
 o 
 
 °1 
 
 >\" SWITCH!* 
 
SWITCHES AND SWITCHBOARDS. 
 
 277 
 
 accompanying illustrations, which, being drawn to scale, used for interchanging, dynamos. All the drawings 
 will require little or no explanation other than the are one-quarter full size, the material used in this and 
 figures themselves give. These details are of switches in the other switch being slate for the base, generally 
 
 O : - 
 
 i rrirp - 
 
 Fig. 17. 
 
 used in the Post Office installation, under the control 
 of Mr. W. H. Preece, F.B.S., and under the immediate 
 supervision of Mr. Probert, to whose courtesy we are 
 indebted for the drawings. Fig. 15 shows a switch 
 
 bronze for the contact-pieces, and ebonite for the 
 insulating parts. Pig. 16 shows a single-pole dynamo 
 switch, also one-quarter full size, while Fig. 17 shows 
 one switch section on the engine-room switchboard. 
 
278 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Here the plan is one-eighth full size, the details being 
 one-quarter full size. A glance at either plan or details 
 will show the simple and effective way in which contact 
 through moving parts is avoided. A spring U piece is 
 placed just in front of handle and pivot, the moving 
 contact-piece falls into this U piece and into a second 
 U piece at the centre, which forms the other contact 
 terminal. 
 
 It is seldom, however, even in the simplest installa- 
 tion that a switch can be placed at every point where it 
 might be required. For convenience sake the circuits 
 are so arranged, and the wires so run, that the required 
 switching apparatus is collected together and placed in 
 a position that is easy of access to those in charge. 
 Generally the switches are arranged upon a common 
 back, behind and through which the necessary circuit 
 
 Fig. 18. 
 
 connections are made, so that the whole or a large part 
 of the installation may be under the control of one man. 
 Switchboards are said not to repay the time and trouble 
 spent upon them, yet they are necessary to a proper 
 maintenance of the installation, and are often very com- 
 plex and, with the instruments, very expensive. Con- 
 sider the simplest case — one dynamo supplying current 
 to one series circuit. The man in charge requires a 
 means of putting on or cutting off the current. A 
 switch enables him to do this. He also requires to 
 know the current passing into the circuit — hence must 
 have a current measurer (ammeter) provided ; he also 
 wants to know the pressure at which he is working, and 
 therefore a voltmeter must be provided. Besides these 
 instruments, provision has to be made in case of 
 accidentally getting too high a pressure and too much 
 
 current, which might destroy much valuable apparatus. 
 To guard against this, a fusible plug or cut-out of some 
 kind is provided, which acts automatically either by 
 fusing or in some way breaking the circuit when the 
 current reaches a dangerous quantity. Usually then the 
 switch or switches, the necessary measuring instruments 
 and the cut-outs are arranged upon the switchboard. In 
 large installations arrangements have to be made for the 
 exciter circuits, for the charging and discharging of 
 batteries, for the shifting of the load from one dynamo, 
 or several dynamos in parallel, to another or other 
 dynamos. Each separate installation has to be studied, 
 and its switchboard arranged according to the special 
 requirements. It will not be necessary to describe in 
 detail each switchboard illustrated, or to reiterate over 
 and over again that each circuit needs its own am- 
 meter, and so on. Fig. 18 then shows a simple type 
 of switchboard as used for accumulator installa- 
 tions, and designed by Messrs. Drake and Gorhain. 
 The dynamos and lamp switches are of the ring 
 contact pattern, the ammeter a steelyard ammeter, also 
 introduced by this firm. The board is fitted with a 
 simple form of combined charge and discharge regu- 
 lator, so that lamps can be run and cells charged at the 
 same time. The pressure at the battery terminals 
 while being charged is greater than that required in the 
 lamp circuit, and moving down the right-hand handle 
 connects the proper number of cells to give the lamp 
 circuit its required pressure. A four-way switch, as in 
 the figure, is that usually employed for a 50- volt circuit, 
 a six-way switch being used for 100 volts. Figs. 19 
 and 20 show a type of switchboard constructed by the 
 General Electric Company. Fig. 21 also shows a 
 switchboard by the same makers, in use at the installa- 
 tion at the Hotel Metropole, Brighton. 
 
 The accompanying illustrations, Figs. 22 and 23, 
 represent one of Messrs. Crompton and Co.'s large 
 central station switchboards of the most recent design. 
 
 The board, as shown, is composed of seven large 
 slates, each 7ft. high, 3ft. 6in. wide, and l|in. thick, 
 all of which are mounted on a strong wrought-iron 
 frame in panel form. 
 
 The whole framework is supported on heavy cast- 
 iron brackets, bolted to the wall about 10ft. from the 
 ground to economise space, and to enable the attendant 
 on the platform to watch all that is going on in the 
 engine-room below. 
 
 Ample space is left between the back of the slates 
 and the wall for access to the connections, as will be 
 seen from Fig. 24. 
 
 The switchboard is designed for a station similar to 
 those of the Kensington and Knightsbridge Electric 
 Lighting Company, working on the three-wire system, 
 with accumulators. In this case, however, arrange- 
 ments have been made for six large dynamos giving 500 
 amperes at 250 volts, for two smaller dynamos giving 
 the same current at half the pressure, for two double 
 batteries, and for eight pairs of feeders, each capable of 
 carrying 500 amperes. 
 
 To minimise the chances of short circuits, and for 
 
SWITCHES AND SWITCHBOARDS. 
 
 279 
 
 Fig. 21. 
 
SWITCHES AND SWITCHBOARDS. 
 
 281 
 
 convenience in the connections, the switchboard is dynamo current indicators, automatic cut-outs, volt 
 divided into three distinct parts. The three slates on indicators, shunt resistance switches, and dynamo plug 
 
 Fig. 19. 
 
 the right-hand side are entirely positive, the three on boards, as well as the cut-outs and current indicators 
 the left-hand are negative, while that in the centre may and plugboards for the feeders. 
 
 Fig. 20. 
 
 be termed the intermediate slate. On the board the On slates 3 and 5 are placed all the battery switches, 
 slates Nos. 1, 2, 6, and 7, Fig. 22, contain the large while slate Xq, 4 contains the switches and instru- 
 
282 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 ments for the two smaller dynamos, which can be 
 thrown, by means of the double-pole switch above and 
 the two small plugboards, on to either side of the 
 system for equalising purposes. 
 
 By means of the plugboards one pair of batteries 
 and any of the feeders and dynamos may be worked 
 entirely apart from the others, and, if necessary, at a 
 different pressure. 
 
 One of the advantages of this system is that any of 
 the switch contact-pieces may be quickly and safely 
 taken down for examination or repairs without 
 interfering in any way with the running of the 
 station. 
 
 The connections in the battery-room are made in a 
 similar manner, and the copper is in these places care- 
 fully painted to protect it from the acid spray. 
 
 Fis. 24. 
 
 The connections at the back of the board are made 
 with drawn copper rods of high conductivity, varying 
 in size from Jin. to 2in. in diameter. The joints in 
 these rods, and the connections with the fittings, are 
 made by means of copper flanges which are screwed 
 on to the rods. These flanges are carefully faced 
 tinned, and drawn together by means of two bolts' 
 
 We will now follow through the connections of the 
 diagram, assuming in the first place that the station is 
 working under ordinary conditions. 
 
 Starting with dynamo No. 1, Fig. 23. The current 
 passes from the right-hand brush direct to the vertical 
 bar on the positive dynamo plugboard, D x . 
 
 We will only take into consideration double battery 
 
SWITCHES AND SWITCHBOARDS. 
 
 283 
 
 No. 1, which is composed of 112 cells, and is practically switch. The 12-way switch will be in such a position 
 two 56-cell batteries coupled together in series. on the battery that it remains practically inactive. 
 
 £ }dK'.;<iK_j 
 
 Fig. 25. 
 
 By plugging B 1 with the top horizontal bar B lf we The position of the large two-way switch will be as 
 couple the dynamo with the 12-way battery switch, shown, and the current will consequently pass direct to 
 Ej + , and the bottom contact of the main two-way the top horizontal bar of the positive feeder plugboard, 
 
284 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 and thence by means of plugs through the current 
 indicators and duplex cut-outs to the lamps, returning 
 again through the negative duplex cut-outs and current 
 indicators to the top horizontal bar of the negative 
 feeder plugboard. 
 
 The current now passes through the two-way switch, 
 as shown, connecting it with the 12-way battery 
 switch, ~R 1 - , which is also placed practically in a 
 
 It will be readily seen that battery No. 2 may be 
 employed in the same way as battery No. 1, by substi- 
 tuting in each case the second horizontal bar of the 
 plugboards for the first. 
 
 By altering the position of the 12-way regulating 
 switches in regard to the cells at either end of the 
 batteries, these may assist the dynamos, or receive a 
 charge themselves ; and in cases where the dynamos are 
 
 ifj^i, 
 
 o 
 
 n 
 
 o 
 
 Fig. 26. 
 
 position of inactivity, and so on to the top horizontal 
 bar of the negative dynamo plugboard, Bl. By 
 plugging on to this bar the dynamo bar Bl negative 
 the circuit is completed through the automatic cut-out 
 and current indicator to the negative brush of the 
 dynamo. 
 
 Any of the large dynamos may be similarly con- 
 nected either separately, or at the same time, as desired. 
 
 not running, the lights may be maintained for a time 
 at the proper pressure by switching into circuit the 
 available end cells. 
 
 Under the ordinary working conditions, it is, of 
 course, impossible to charge these end cells, as the 
 dynamos would have to run at too high a pressure for 
 the lamps. This difficulty is overcome in the following 
 manner. 
 
SWITCHES AND SWITCHBOARDS. 
 
 285 
 
 The dynamos are coupled as before, and the end cells 
 which require charging are put in circuit by the 12-way 
 switches. Before this is done, however, the two-way 
 switches are turned in the position in which they are 
 shown with battery No. 2, which throws the feeders on 
 
 All the batteries have a common connection at the 
 third wire, at which point the battery meters and 
 current indicators are coupled; these are provided with 
 two-way switches, as shown, to cut them out of circuit 
 when the batteries are discharging. 
 
 Fig. 27. 
 
 to the small eight-way battery switches, by means of 
 which proper pressure is kept on the circuits. 
 
 Before moving the two-way switch, care must be 
 taken that the large and small regulating switches are 
 on the same cell, otherwise a short-circuit would take 
 place across the cells which the two switches overlap. 
 
 In order to provide for the irregularities of load on 
 either side of the system, the two 125-volt machines 
 are employed. 
 
 It will be seen that A dynamo, by means of the 
 double-pole switch, is coupled on to the third wire and 
 the vertical bar of the small positive plugboard, while 
 
286 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 B dynamo is coupled to the third wire and vertical bar 
 of the small negative plugboard. By plugging the small 
 horizontal bars with the vertical bars, either or both of 
 these small dynamos can be made to assist either or 
 both sides of the system. 
 
 The automatic cut-outs are a new combination of a 
 magnetic and fusing cut-out, and are specially adapted 
 for heavy currents. 
 
 The battery switches are so arranged that not the 
 slightest flicker is noticed when a cell of the battery is 
 passed into or out of circuit. 
 
 lamps are needed — then the others can be switched out 
 without trouble. The examination of the circuits is 
 thereby facilitated. A lamp or lamps going wrong 
 causes a glance at the measuring instruments, and it is 
 at once known whether that particular circuit is taking 
 too much or too little current, or the circuit where the 
 difficulty arises can be at once determined. Of course, 
 any movement of the switches upon any of the distri- 
 buting boards is known in the dynamo-room because of 
 the load taken off or put on the machines, and of the 
 indications on the instruments of the main switchboard. 
 
 NEGATIVE 
 
 Fro. 28. 
 
 Fig. 25 may be given here, although it really illus- 
 trates another part of our subject. Hitherto we have 
 been speaking of main switchboards, but other switch- 
 boards are used where, as in hotels, factories, or ware- 
 houses large numbers of lamps are used. It is necessary 
 to control such lamps from the room or the immediate 
 vicinity in which they are used. This figure then 
 illustrates a distributing switchboard in use on the 
 ground floor of the Post Office, and is one-eighth full 
 size, while Fig. 26 shows the details of this board also 
 one-eighth full size. That is, in this case, we hav fall 
 the lamps upon this floor controlled from a switchboard 
 m a convenient position. It may be that only a few 
 
 A different arrangement has been adopted at the 
 new Opera House, and a little consideration will show 
 that the ordinary switching on and off of lamps would 
 be unsuited for theatrical purposes ; and what is said 
 here of the new Opera House applies equally well to 
 almost all theatres. Hence special apparatus has to 
 be designed and constructed to fulfil the varying con- 
 ditions that arise in such installations. This apparatus, 
 again, must be simple, and so unlikely to get out of 
 order, for a panic in a theatre is one of the things to be 
 avoided at all hazard. Itiswell known that the exigencies 
 of theatre lighting necessitate great variations of light. 
 In some brilliant scenes all the light possible is required, 
 
SWITCHES AXD SWITCHBOARDS. 
 
 287 
 
 _E 
 
 L 
 
 t 
 
 Fig. 29. 
 
 £ 
 
 u 
 
 L 
 
 _Ei 
 
 tr 
 
 13 
 
 Fio. 31. 
 
 and from these well-lighted scenes there must be every opera of " Ivanhoe " the load varies some eight or ten 
 gradation of light to almost total darkness. In the times in the evening— from 1,400 to 300 amperes. 
 
288 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 The manipulation of light is managed by a somewhat 
 different kind of switch to any we have hitherto 
 mentioned. It is done by means of a series of small 
 wheels, supplied with handles fitted independently on 
 an axle running through them all. These hand wheels 
 communicate, by means of endless wire ropes, with 
 switches which control the resistances by short-circuit- 
 ing some of the coils in circuit. This method does 
 away with all sparking, and if the switch should get out 
 of order it does not put the light out, but merely in- 
 troduces all the resistance into the circuit. A large 
 lever, Fig. 27, at the end of the axle enables all the 
 lamps in the building that are supplied with resistances 
 to be raised or lowered, while the independent hand 
 
 $ $$$ 
 
 Fig. 30. 
 
 wheels can be utilised to vary any desired combination, 
 ine limits of variation in this particular installation 
 are from 110 to 50 volts. 
 
 For quick changes' of scene for which the curtain is 
 not let down, but which have to be done in the dark a 
 large mam switch is used to break the main circuits 
 carrying 1,400 amperes. This switch is fitted with copper 
 blocks to take the arc, and is short-circuited, after the 
 circuit has been completed, by two large mercury cups 
 and a connection. The object of the blocks of copper 
 gauze in the switch is to facilitate renewal of the 
 contacts, and they also have the effect of making the 
 switch much more durable than if the contacts were 
 made and broken on springs, as is usual, because the 
 small globules of copper produced by breaking the 
 circuit are flattened into the gauzy material when con- 
 
 tact is next made. Thus, while breaking contact may 
 be said to deteriorate the gauze, making contact assists 
 to maintain it in an efficient state. 
 
 Fio. 32. 
 
 We have now given the details of switchboard con- 
 struction, but it may not be out of place to refer to 
 others ; for, as we say, each installation needs its special 
 
SWITCHES AND SWITCHBOARDS. 
 
 280 
 
 design, although the general type of design is similar 
 throughout most of the switchboards constructed by 
 the same maker. A type of a high-pressure central 
 station switchboard has been designed and adopted by 
 
 Fig. 33. 
 
 the Brush Company, as has also a type of low-pressure 
 switchboard. The same idea forms the basis of both 
 types of switches, being that of mechanical arrange- 
 ment to allow of a certain number of dynamos to be 
 
 connected with all or any of a given number of circuits 
 in such a manner that these can be changed over with 
 ease and certainty. In the low-pressure switch a solid 
 heavy bar slides along, by means of an ebonite handle, 
 into jaws consisting of split pieces of stout copper. 
 To these are connected one end of the cable, and 
 the other end is connected to stout springs, which 
 press up underneath the sliding piece, thus making 
 good contact. In the switchboard a series of these 
 sliding bars are arranged in the centre of the board, and 
 each one connected to its own dynamo. At the top 
 and at the bottom are corresponding receiving jaws, 
 to either of which they can connect. Each of these 
 jaws has a handle, to which a movable contact is fitted; 
 these contacts can be lifted and fitted down over either 
 of the transverse bars. The transverse bars are joined 
 in couples, top and bottom, each couple connected to 
 a separate circuit. To throw over a dynamo to another 
 circuit, therefore, the jaw not in use corresponding to 
 that dynamo is lifted and pressed down to the required 
 circuit bar. The switch handle is then pushed hard 
 across, and the current is, of course, shunted to the 
 new circuit. 
 
 In the high-tension switchboard the same arrange- 
 ment of dynamos and circuits is made, but the switch 
 is of a different pattern, giving a long break to obviate 
 the spark. In this a handle on a loose pivot catches 
 against a small trigger, and releases an arm carrying a 
 contact-piece. This arm is actuated by a powerful 
 spring, and the act of moving the handle causes it to 
 fly round instantaneously, the break being, of course, 
 twice the radius of the arm. The connection of 
 dynamos to the various circuits is made in the same 
 manner as the former to the transverse bar connection. 
 A locking arrangement prevents the arm flying back 
 accidentally, and is only unlocked by again altering 
 the circuit contact. 
 
 Fig. 28 shows the connections of the switchboard 
 designed by Mr. G. Kapp for use at the installation at 
 Madame Tussaud's. Here there are three circuits — two 
 main circuits and a panic circuit. As will be seen, 
 each of the two main circuits has 10 switches, the 
 panic circuit being controlled by its own switch. The 
 switchboard is arranged to allow connection of either 
 dynamo to either battery, of the supply of current 
 direct to either circuit, or from the batteries, or for the 
 charging of the batteries while partially lighting, or the 
 discharge of the dynamos and batteries together. Fig. 
 29 shows a type of board used by Messrs. Poole and 
 White for private and small installations. Fig. 30 
 shows the connections of this board, which are suffi- 
 ciently distinct to explain themselves. It will be seen 
 that the board is somewhat similar in type to that 
 illustrated in Fig. 18. 
 
 Besides arranging for connecting circuits and 
 measurements, it is necessary to provide also an auto- 
 matic arrangement to break contact should the cur- 
 rent by any means become too large. This, as has 
 been said, is usually obtained by the insertion of a 
 fusible piece of metal, called a cut-out or fuse. 
 
290 
 
 PRACTICAL ELECTRICAL ENC-lNEEklNG. 
 
 
 i/> 
 
 ^r 
 
 
 J" 
 
 *J 
 
 J~ 
 
 r^ -1 © 
 
 r 
 
 t— e 
 
 9-1 
 
 r 
 
 -•© 
 
 '->« 
 
 r 
 
 eJ"i 
 
 f 
 
 ; 
 
 1 — o 
 
 ' — 
 
 •r^e 
 
 J 
 r 
 
 
 X 
 
 'p© 
 
 ?, 
 
 r 
 
 9 1 
 
 <r 
 
 L 
 
 
 .r 1 
 
 « i 
 
MEASUREMENT. 
 
 291 
 
 Frequently a magnetic apparatus is adopted. We 
 refer to the subject of cut-outs here only to give 
 Figs. 31, 32, and 33, which show the elevation one- 
 eighth full size and the details half full size of the 
 fuseboard at the Post Office. 
 
 The Post Office installation, to which we have so 
 
 frequently referred, has been so carefully designed and 
 the details thought out, that it may well form a pattern 
 for work of a similar character. The main switchboard 
 is, of course, a collection of switches similar to that 
 illustrated in Fig. 17, while Fig. 34 shows the connec- 
 tions of the dynamo switches. 
 
 CHAPTER XIV. 
 
 MEASUEEMENT. 
 
 Introductory. 
 
 HE most important knowledge for an elec- 
 trical engineer is that relating to measure- 
 ment. It may be said, without fear of 
 contradiction, that the instruments in the 
 hands of electricians enable them to make 
 more delicate and more exact measure- 
 ments than is possible in any other branch 
 of scienee, not even excepting chemistry, and far more 
 delicate than in any other industry. Most of the 
 instruments employed are the subject of patents, and 
 thus at present cannot be generally manufactured. 
 Many of the instruments, however, used in ordinary 
 work wherein measurements to "1 per cent, are suffi- 
 ciently accurate, are not controlled by patentees. All, 
 however, are based upon similar principles, so the plan 
 adopted here will be to deal, in the first place, with the 
 scientific principles underlying the measurements, and 
 subsequently with the application of these principles to 
 the instruments. Frequently it will be best to give the 
 principles and the application together, so a hard and 
 fast plan cannot be followed. The history of the 
 development of electrical measurement is interesting, 
 but will form no part of these papers. The work of the 
 British Association Committee is writ in the annals of 
 that association, and to the reports of that committee 
 the reader desirous of knowing this history must be 
 referred. Under that committee chaos has given place 
 to order, and we have to use what is, not what was. 
 The aim of the committee was to devise a system of 
 measurement founded upon a firm scientific basis, and 
 hence capable of general application. The system 
 devised is known as the B.A., or the absolute, 
 or the C.G.S. system. The meaning of B.A. 
 is simply the initials of British Association. The 
 term " absolute " indicates not absolute accuracy, by 
 independence of the properties of any particular instru- 
 ment, or of other physical quantities, except those of 
 length, mass, and time. It is termed the C.G.S. 
 system because the units involved are derived directly 
 
 from the centimetre as the unit of length, the gramme 
 as the unit of mass, and the second as the unit of time. 
 
 From the units adopted in the C.G.S. system are 
 derived the units adopted in practice, so that some- 
 thing must be known about what may be termed the 
 fundamental system before that used in practice can be 
 thoroughly understood. 
 
 There are two systems of electrical units based upon 
 the C.G.S. units, known respectively as the electro- 
 static and the electromagnetic. In this latter system all 
 the quantities are expressed in units derived from a 
 magnetic pole of unit strength, and unit pole is defined 
 as " a pole which, placed at unit distance from an equal 
 and similar pole, repels it with unit force." 
 
 Fundamental Units in the C.G.S. System. 
 Length, 1 centimetre = -393708in. = about fin. 
 Mass, 1 gramme = 15'43235 grains = about 15; grains. 
 
 = "035274 ounce avd. = about tjJj ounce avd. 
 
 = ■032151 ounce troy = about ^fu ounce troy. 
 Time, 1 mean solar second, or asioo part of the mean solar day. 
 
 A caution must be given here to the effeet that 
 weight is not the same as mass, although it is 
 frequently assumed to be identical. Mass is invari- 
 able, weight is variable. The mass of a body is the 
 same at the poles as at the equator, but the weight is 
 not the same owing to the varying pull of gravity. To 
 put this plainly, we may suppose the pull of gravitation 
 to be towards the centre of the earth, and that the 
 earth acts, so far as gravity is concerned, as if it were 
 concentrated at the centre point. The further a body 
 is from the centre of gravity the less is the pull ; so if 
 we take a piece of iron and determine the pull of 
 gravity upon it in a London laboratory, that pull will 
 be found to differ very considerably from the pull 
 determined upon the same piece of iron at the equator 
 or the poles. The number of molecules which con- 
 stitutes the mass of the iron is unchanged, but the 
 pull of gravity upon each molecule, and therefore upon 
 the whole iroD, is changed. 
 
 In the C.G.S. system the unit of mass is taken as the 
 
292 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 quantity of distilled water at the temperature of 4deg. 
 C. contained in 1 cubic centimetre, which under 
 standard pull of gravity (usually represented by g) 
 corresponds to 1 gramme weight. 
 
 In England — outside of pure science — the standard 
 of mass is the Imperial standard pound avoirdupois, 
 or rather a piece of platinum marked " P. S. 1844, 
 lib.," preserved in the Exchequer Office. The choice 
 of these fundamental units is entirely arbitrary, and 
 many would prefer a foot-pound-second system to the 
 G.G-.S. system. In engineering matters other units 
 are required, especially dynamical units. The most 
 important of these are the units of force and work. If 
 we allow a body, such as a gramme weight or a pound 
 weight, to fall fully for one second under the influence 
 of gravity alone, it will have acquired at the end of that 
 period a velocity of 981 cm. per second, or about 32'2ft. 
 per second. This does not mean that the body falls 
 981 cm., or 32'2ft., but that at the end of a second, if 
 allowed to travel uniformly onwards at the velocity thus 
 acquired, it would move through 981 cm., or 32"2ft. 
 each succeeding second. 
 
 The actual space fallen through in the first second is 
 I- of 981 cm., or \ of 32'2ft. — that is, 490'5 cm., or 
 161ft. It is easy to show that under the influence of 
 gravity the actual fall of a body in the second second 
 will be 1471-5 cm., or 48'3ft. Now, if a body is con- 
 stantly acted upon by a force less than gravity it will 
 acquire a less velocity per second ; if the force is greater, 
 then a greater velocity is acquired. 
 
 The C.G.S. unit offeree has been defined as "that 
 force which, acting upon a mass of 1 gramme for 1 
 second, will impart to it a velocity of 1 cm." This unit 
 is called the " dyne," and evidently is equal to ^ T g. 
 
 When the foot-pound-second system is used, the unit 
 of force is called the " poundal," and is that force which 
 acting for 1 second on a mass of lib., generates a velocity 
 of 1ft. per second, and is evidently ^ g. 
 
 " Work " represents force applied through distance 
 or length, and the unit of work in the C.&S system 
 is called an "erg," and equals the unit of force x unit 
 of length ; or equals the work done in lifting 1 gramme 
 vertically against gravity through the distance of 1 cm 
 In practical work, this unit being so small is of little 
 use, and another, 10,000,000 times as large, called the 
 joule, is sometimes used, though as yet it has not 
 commended itself to engineers. 
 
 In the foot-pound-second system the unit of work 
 is termed a foot-pound-that is, an amount of work 
 equal to that done in lifting a pound vertically against 
 gravity through the distance of a foot. 
 ^ Sir W. Thomson has suggested the use of the term 
 activity, as equivalent to the time rate of doing work 
 The unit rate of working adopted by engineers in this 
 country is 1 horse-power, which is equivalent to 33 000 
 foot-pounds per nnnute, or 550 foot-pounds per second 
 This is a far more practical unit than the ero- ))er 
 second. b f ex 
 
 Some idea of the amount of work represented by an 
 erg may be obtained thus: 1 gra mme raised vertically 
 
 against gravity a distance of 1 cm. = dynes x cm. = 
 981 x 1 = 981 ergs. In English measures it takes about 
 160 ergs of work in raising a grain vertically against 
 gravity through an inch, or 1,920 ergs to raise 1 grain 
 similarly through the distance of a foot. More nearly, 
 a foot-pound = 13,562,600 ergs ; which shows what 
 large numbers would have to be dealt with if engineers 
 adopt this unit. 
 
 So far as the amount of work is concerned it does not 
 matter whether the pound is lifted in a second or in a 
 year, but it does matter where " activity" is concerned. 
 
 The English unit horse-power is generally taken as 
 equivalent to 7,460,000,000 ergs per second, for the 
 horse-power is 33,000 foot-pounds per minute or 550 
 foot-pounds per second, and 550 foot-pounds = 76 kilo- 
 gram-metres, i.e., 76 kilogrammes raised 1 metre in 
 1 second— hence 
 
 1 horse-power = 76 kilogram-metres. 
 
 = 76 x 1,000 gramme-metres. 
 
 = 76 x 1,000 x 100 gramme-centimetres. 
 
 = 7,600,000 gramme-centimetres, 
 but we have seen that to raise 1 gramme vertically 
 against gravity 1 cm. high requires an expenditure of 
 work equivalent to 981 ergs. 
 
 Therefore 1 horse-power = 7,600,000 x 981. 
 
 = 7,455,600,000 ergs per second. 
 
 As was said above, the round number 7,460,000,000 
 is generally taken as the equivalent. 
 
 Energy expended or gained is measured by the work 
 done by or upon the body, and is measured by the same 
 units as work. 
 
 The units referred to may again be tabulated : 
 Fundamental Units. 
 
 Length. 
 
 Mass. 
 
 Time. 
 
 Force. 
 
 Work. 
 
 Energy. 
 
 Activity. 
 
 Velocity. 
 
 Acceleration. 
 
 C.G.S. (centimetre- 
 gramme-second). 
 Centimetre. 
 Gramme. 
 Mean solar second. 
 
 F.P.S. (foot-pound- 
 second. 
 Foot. 
 Pound. 
 Mean solar second. 
 
 Derived Units. 
 
 D y ne - Poundal. 
 
 Er g- Foot-pounds. 
 
 Er S- Foot-pounds. 
 
 Erg per second. Foot-pounds per sec. 
 
 Centimetre-second.- Foot-second. 
 
 Centimetre-second. Foot-second. 
 
 It is necessary to reiterate that the measurements 
 required by electrical engineers of to-day are very 
 different from those required in telegraphy, either as 
 regards land lines or submarine cables. The work done 
 by currents used in either branch of telegraphy is very 
 small compared with that done in electric lighting or 
 transmission of power ; hence the ordinary engineering 
 units are more applicable than those used in telegraphic 
 work. Foot-pounds, horse-power, etc., are terms in 
 common use, appealing to the uneducated as well as to 
 the educated, and not requiring a special training to 
 make comprehensible. It may be that horse-power 
 J™.™? means nothing, except the conventional 
 33,0001b. raised 1ft. in one minute; that no horse will 
 
ME A S UREMENT. 
 
 293 
 
 continue to work, if it is able to work at all, at this rate, 
 but for all practical purposes the horse-power is a very 
 good unit, and needs not to be superseded by one that 
 may be more in accord with the notions of pure 
 science. 
 
 The Earth's Magnetism. 
 
 In the vast majority of instruments used in practice, 
 a magnet forms an important feature. It must be 
 remembered that such a magnet is always subject to 
 the influence of the earth's magnetism, and, therefore, 
 before the effect of any other influence can be measured, 
 registered, or tabulated, it is necessary to know exactly 
 the effect of the earth's magnetism upon the magnet. 
 
 Any magnet so suspended as to be free to move in 
 any plane will finally come to rest in a position point- 
 ing almost due north and south with one pole a little 
 higher than the other. The geographical poles occa- 
 sionally, and only after long intervals of time, coincide in 
 direction with the magnetic poles of the earth. Thus, 
 when we say a magnetic needlepoints north and south, 
 the earth's magnetic north and south is meant, and not 
 the geographical north and south. The angle between 
 the magnetic meridian and the geographical meridian of 
 a place is called the " magnetic declination " of that 
 place. 
 
 If a magnet, as above mentioned, be suspended, 
 this magnet will set itself in a direction to the west of 
 the geographical N and S, there being at the present 
 time an angle of about 17"25 deg. between lines showing 
 the two directions. That is, the magnetic north pole 
 is at the present time 17 25 deg. west of the geographical 
 north pole. The variation from the geographical 
 meridian is termed the declination of the magnet, and 
 the angle the angle of declination. The declination 
 at any place, therefore, is the value of the angle in 
 degrees at that place between the magnetic and 
 geographical meridian of that place. 
 
 The nearer the magnet needle is carried to the mag- 
 netic pole, the more and more, if free to move vertically, 
 does one end or pole dip downwards, till if carried 
 quite to the pole the needle would point straight down. 
 At the earth's north magnetic pole the true south pole 
 of the needle (commonly, but unscientifically, called 
 the north pole of the needle because it points 
 towards the north) is downward, and at the earth's 
 south magnetic pole the true north pole points 
 downwards. The angle formed by the magnet pointing 
 downwards at any place and a horizontal axis, is called 
 the dip of the needle. 
 
 Both "dip" and " declination," it must be remem- 
 bered, change with position on the earth's surface, and 
 the declination, if not the dip, differs at the same position 
 at different periods of time. 
 
 We assume, then, that the earth acts as a magnet, 
 and, if so, we must assume that the lines of force, con- 
 stituting the earth's magnetic system, possess similar 
 properties to the lines of force of any other system. 
 Inasmuch as the lines of force of any two magnetic 
 systems brought within each other's influence always, 
 when free to act, place themselves, or tend to place 
 
 themselves, in the same or in parallel directions, it will 
 be seen that other magnets under the influence of the 
 earth's magnetism merely obey this invariable rule. 
 
 If, then, we have a magnet, it naturally, if free to do 
 so, as has been said, takes up a position pointing to the 
 magnetic north and south. It is this directive influence 
 of the earth's magnetism upon small magnets free to 
 move in a horizontal plane that gives us a most valu- 
 able aid to navigation in the compass. If a magnet is 
 moved from its natural position of rest, it requires 
 force to move it against the action of the earth's mag- 
 netism, just as a lump of metal requires force to lift 
 it against the action of gravity. We can measure the 
 force employed, or the work done, in lifting a pound 
 against gravity, and similarly we can measure the force 
 employed, or the work done, in forcing a magnet out of 
 its natural position against the force of the earth's 
 magnetism. 
 
 In considering the earth's influence upon a magnet 
 we shall notice : 
 
 1. How to find the magnetic meridian. 
 
 2. How to find the horizontal component of the 
 earth's magnetism. 
 
 3. How to find the vertical component of the earth's 
 magnetism. 
 
 Determination of the Magnetie Meridian. 
 
 The reader is not supposed to have at his command 
 the apparatus usually found in physical laboratories, 
 but to require a fairly correct method of determination 
 sufficiently accurate, in fact, for all practical purposes. 
 
 Take an ordinary bar magnet. If the magnetic axis 
 of the bar corresponds with the geometric axis, the 
 meridian is easily found. At each end of the bar 
 magnet, midway between its vertical sides, and therefore 
 in the plane of the geometric axis, attach a small, short 
 pointer — a bristle will do. Having determined the 
 table or work bench upon which the meridian is to be 
 marked, fix a hook in the ceiling above this place. 
 
 Fig. 35. 
 
 Fasten to the hook a silk thread or a small steel wire, 
 according to the weight of the magnet, and by the silk 
 thread or the wire suspend the bar magnet, which for 
 convenience may be carried in a stirrup of light card- 
 board, Fig. 35. When suspended, the magnet should 
 just swing freely upon the table, or rather over a piece 
 of looking-glass, which should be placed upon the table 
 face upwards, The glass may be kept in position by a 
 
294 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 few pins, not by nails or screws of iron. In fact, the 
 less iron in the vicinity of the magnet the better. 
 
 The suspending thread must be free from torsion. 
 To obtain this, put a piece of lead, of equal weight, to 
 the magnet in the stirrups, twist the hook in one direc- 
 tion or the other, till the lead hangs approximately in the 
 magnetic meridian, a position which may be determined, 
 as shown, by an ordinary compass needle. When the 
 lead is perfectly stationary in this direction, hold the 
 stirrup and replace the lead by the magnet, taking care 
 that the N pole of the magnet is directed towards the 
 north, otherwise, on freeing the stirrup, the magnet will 
 swing halfway round, and give a twist on the thread of 
 180 deg. of torsion. The magnet being properly placed, 
 set it vibrating through an arc of 5 deg. to 10 deg., and 
 then allow it to come to rest. From above the magnet 
 look down so that the image of the projecting bristle or 
 wire, as reflected by the mirror, coincides with the object 
 itself. By this means any error due to parallax is 
 avoided. Mark the position of the images, at both ends 
 of the magnet, on the surface of the mirror with dots, 
 and call these dots A and K x . Now reverse the magnet 
 in the stirrups so that what was the top face is put 
 underneath — the north pole, of course, still pointing to 
 the north. Eepeat the observations at each end of the 
 magnet, marking the images as before, calling these new 
 marks B and B x . The object of this reversal of the 
 magnet is to correct any errors arising from the mag- 
 netic axis of the needle not corresponding with its 
 geometric axis. 
 
 Eemove the magnet. 
 
 Join A Aj and B Bj with straight lines. 
 
 Draw a line bisecting the angle formed by these two 
 lines, and the straight line thus bisecting the angle 
 will be the plane of the magnetic meridian required, 
 Fie;. 36. 
 
 Fig. 36. 
 
 The table or bench, if fixed, can be permanently 
 marked, or permanent marks can be placed along the 
 floor and up the opposite walls of the room. 
 
 Having determined the line of the meridian, we are 
 in a position to determine the horizontal component of 
 the earth's magnetism, and from the value so found to 
 calculate the total magnetic force. 
 
 To And H, the Horizontal Component of the Earth's 
 Magnetism. 
 
 A freely-suspended magnet needle, if free to move in 
 any plane, sets in the direction of the line of dip — that 
 is, in the direction of the earth's lines of force. In other 
 words, whatever force is acting upon it, is acting along 
 the line of dip. The total force thus acting does not 
 admit of being readily measured by any direct method, 
 but by resolving it into components acting at right 
 angles to each other, one vertically and one hori- 
 zontally, according to the well-known principle of the 
 
 parallelogram of forces, we can measure the force of 
 either component, and so calculate the total force. 
 
 Let n represent the magnet needle inclined to the 
 horizontal by the angle of dip d ; also, let H represent 
 the horizontal component, V the vertical component, 
 and T the total force, Fig. 37. 
 
 Then 
 
 or, 
 
 T = 
 
 T 
 
 H 
 
 cos d 
 
 V 
 
 sin d sin d 
 
 Of the two, H and V, the determination of H is less 
 liable to error, and is therefore the one usually selected 
 to measure. 
 
 Fig. 37. 
 
 Employing Gauss's method, which consists of deter- 
 mining two values : 
 
 First. — The product of the moment (M) of the magnet 
 into the horizontal component of the earth's mag- 
 netism, H = (M H), as shown by the period of 
 vibration of a magnet freely swinging in a horizontal 
 plane under the earth's magnetism alone ; and 
 
 Secondly, the ratio — , as measured by the angle through 
 
 which a short compass needle is deflected by the same 
 magnet at a known distance. 
 
 The magnetic moment of a magnet is the product of 
 the strength of either pole into the distance between 
 the poles. 
 
 A convenient and inexpensive set of apparatus for 
 determining H has been designed by Mr. J. T. 
 Bottomley, of Glasgow University. It consists essen- 
 tially of two distinct parts : 
 
 1. A wooden chamber or box, the base of which is 
 supported by levelling screws — one at each corner. 
 This is used for determining the period of vibration. 
 
 2. A magnetometer, consisting of a small closed 
 chamber, in which is suspended a short magnetised 
 needle with mirror attached— similar to the arrange- 
 ment in Sir W. Thomson's reflecting galvanometer. 
 This is employed in measuring the angle of deflection 
 produced by the magnet, whose period of vibration 
 has already been ascertained. 
 
 Make a box, Fig. 38, about 12in. high, Gin. wide, 
 and 6in. deep, with the top, bottom, and two sides' 
 
MEASUREMENT. 
 
 295 
 
 made of wood, the two remaining and opposite sides 
 being glass plates, capable of sliding in vertical grooves 
 cut in the wooden sides. These glass plates when 
 raised give access to the interior of the box, and when 
 lowered into their ordinary position close the box and 
 effectually prevent the vibrating magnet needles being 
 disturbed by air currents. To a wooden peg, P, so 
 
 Fig. 38 
 
 fixed as to be able to be turned in a hole in the centre 
 of the top of the box, is fastened a single unspun silk 
 fibre of convenient length, which carries at its lower 
 end a light paper stirrup, in which the magnet used 
 in the experiment" is placed. A fine pencil line, or 
 preferably a fine scratch made with a diamond, must be 
 drawn vertically down the centre of each glass side, so 
 that when the box is carefully levelled in a true vertical 
 position, the two scratches arid the silk fibre are all in 
 the same place. 
 
 To Determine M H. 
 
 Place the box on the bench or table in such a position 
 that the two scratches on the glass doors are in the 
 magnetic meridian, as previously determined on page 
 294, and level the box until the silk fibre hangs in the 
 same plane. Now raise one of the doors, and place in 
 the stirrup a brass or copper rod of equal weight to the 
 magnet which will subsequently replace it. If on the 
 rod coming to rest its axis does not coincide with 
 the magnetic meridian, it can be made to do so 
 by a little manipulation of the peg at the top. The 
 peg must be gently turned till the copper rod lies ' in 
 the plane of the magnetic meridian. 
 
 Now raise one window, take hold gently of the stirrup, 
 and replace the copper rod by a well-magnetised piece 
 of hard steel wire of the same weight, and from 3in. to 
 4in. in length. Some care is required in this operation, 
 and especially as to the poles of the magnet. If the 
 student is standing on the south side of the apparatus 
 the needle must be held by the south-seeking pole, and 
 the north-seeking pole pushed through the stirrup, 
 otherwise, on releasing the stirrup the needle will swing 
 round and impart a torsion of nearly 180 deg. on to the 
 silk fibre. If he stands on the north side of the appa- 
 ratus the action must be the reverse of the above — the 
 
 north-seeking pole will be held, and the south-seeking 
 pole pushed through the stirrup. If the replacement of 
 the copper by the magnet needle has been carefully 
 done, the needle should hang in the magnetic meridian, 
 but if it does not — that is, if it is slightly out — it may be 
 brought into position by turning the peg. It is difficult 
 to fix up a fibre absolutely without torsion, and the 
 copper is first placed in the, stirrup to get rid of any 
 torsion that happens to exist. Assuming, then, we have 
 the magnet in position, we want to make the needle 
 swing slightly on both sides of the scratches on the 
 glass. This we can do if we approach a small magnet 
 with one pole pointing towards the centre of the sus- 
 pended needle, giving a slight movement to the magnet, 
 which will set the needle vibrating. When it is vibrat- 
 ing through an arc of about 8 deg. or 10 deg., let an 
 assistant observe the times of successive transits, using 
 a watch with a seconds' dial. To make this observation, 
 place the eye in a line with the two scratches, and 
 observe the transit made by the nearest end of the 
 needle across the plane of the scratches. What we 
 require to know is the time of a " complete vibration," 
 that is, the time between one crossing the meridian to 
 the next crossing of the meridian in the same direction — 
 thus it is only necessary to notice the transits in one 
 direction. The recurrence of the tenth, twentieth, 
 thirtieth, fortieth transit (if the needle continues in 
 vibration so long) should be recorded. The mean time 
 of these readings divided by the number of transits, 
 gives approximately the mean time of a complete vibra- 
 tion. A series of three or four such sets of readings 
 should be made, and the mean of the series taken for 
 the true value. The time, T, required by a magnet 
 whose moment (vide next paragraph) of inertia = K, to 
 make one complete vibration through a small arc, is 
 found by the following formula : 
 
 T = 2tt 
 
 Vmb 
 
 MH 
 
 or 
 
 therefore 
 
 T 2 = 4 7T 2 
 
 K 
 MH 
 
 MH = 
 
 4jrK 
 
 (A) 
 
 Moment of Inertia, K. 
 
 The moment of inertia of a material point, referred 
 to an axis round which it revolves, is its weight 
 multiplied by the square of its distance from the 
 axis. The moment of inertia of a body round any 
 axis is the sum of the moments of all its particles round 
 that axis. The value of K depends only on the form 
 and mass of the body, and is not altered by magnetisa- 
 tion. The following table gives the values of K for a 
 few bodies of symmetrical form. 
 K = moment of inertia, W = weight of body, r = radius. 
 
 1. A long, thin, uniform bar, whose 
 thickness and width are small com- 
 pared with its length, I, suspended 
 at its centre of gravity. 
 
 Value of K. 
 
 K=W. 
 
 12 
 
296 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 2 A rod of length I, with square cross 
 section, each side of breadth b, the 
 axis of suspension passing through 
 the centre of gravity. 
 
 3. A cylinder of length I, and radius 
 r, suspended by a perpendicular to 
 the middle point of the axis. 
 
 4. A hollow cylinder, inner radius r, 
 outer radius r lt referred to its cen- 
 tral axis. 
 
 5. A cylinder referred to its axis. 
 
 Value of K. 
 
 K = W 
 
 12 
 
 K = W ll2 + 
 
 
 K 
 
 r 2 + ?y 
 
 K = W. 
 
 K=Wr 
 
 6. A ring whose thickness is small 
 compared with its radius, r, sus- 
 pended at its centre of gravity and 
 at right angles to plane of ring. 
 
 In the case of the magnet needle, we deal with a body 
 either of rectangular or circular section. If the former, 
 carefully measure length, I, and breadth, b. If the 
 latter, measure the radius, r, and in both cases W, the 
 weight. 
 
 Then, for a rectangular-shaped needle or bar 
 
 For a cylindrical-shaped needle or bar 
 
 K = W(W 
 
 4/ 
 
 M2 ' 4:J 
 
 Having thus ascertained the time of vibration, T, and 
 the moment of inertia, K, we can proceed to the 
 
 Determination of the Ratio - . 
 
 H 
 
 Another apparatus, termed a magnetometer, Fig. 39, 
 is employed in obtaining this ratio. It consists of a 
 
 Fig. 39. 
 
 circular wooden base, B, supported on three levelling 
 screws lo this base a vertical pillar is fixed, 6in high 
 2m. .wide, and lin. thick; having cut nearly along its 
 face a narrow groove opening near the bottom into a 
 flat circular chamber, and at the top continued by a 
 small aperture to the outside. At the top of this small 
 aperture is a peg A single fibre of silk goes through 
 the aperture, and the silk is fixed by means of the ptg 
 
 and a little wax. At the lower end of the silk fibre is 
 fastened a small mirror, to which is attached a small 
 magnet. The length of the silk thread is regulated 
 before fastening at the top, so that the magnet- hangs 
 freely in the flat circular chamber near the bottom of 
 the vertical piece. A glass plate covers the face of the 
 vertical piece in front of the mirror. This plate can be 
 kept in position by a couple of elastic bands, and it 
 prevents any disturbance of the mirror from air 
 currents. 
 
 Besides the magnetometer, we require a long flat 
 rule at least a metre long, divided into centimetres and 
 millimetres. Place this rule flat on the table so that 
 its centre point, fiftieth centimetre, comes upon the 
 meridian line marked on the table, Fig. 36. Set the 
 magnetometer with its mirror facing E or W, accord- 
 ing to position of magnet needle on the back of the 
 mirror — that is, so that the north-seeking pole points to 
 the north. Take care that the suspended fibre is free 
 from torsion, and hangs approximately over the point 
 of intersection of the meridian and the flat rule. 
 
 Fig. 40. 
 
 Opposite the glass face of the magnetometer is placed 
 a vertical screen, having upon it a scale in millimetres 
 as shown in Fig. 40. This scale may be of paper 
 pasted on the screen. Just below the central division 
 of the scale is a small circular aperture having a thin 
 wire across its vertical diameter. A paraffin lamp 
 placed behind the screen, with its flame set edgeways, 
 will send a circular beam of light through the aperture, 
 and this falling on the mirror of the magnetometer, is 
 reflected back on the scale of the screen, the shadow of 
 the wire showing most distinctly the movement of the 
 mirror. By interposing a convex lens between the hole 
 and the mirror, and varying the distance of the lamp, 
 a sharply-defined image of the cross wire is projected 
 on the scale. It will simplify subsequent calculation if 
 the distance between the mirror and scale be made 
 exactly 1 metre. 
 
 Care should be taken that the stand carrying the 
 vertical scale is placed parallel to the meridian line on 
 the table, and that the spot of light and image of cross 
 wire stands exactly at the zero point of the scale, if the 
 divisions are numbered outwards from the centre, or at 
 the 500 division if they are numbered consecutively 
 from left to right (assuming the scale to contain 1,000 
 divisions). 
 
 Having satisfactorily arranged the magnetometer and 
 
MEASUREMENT. 
 
 297 
 
 scale, take the magnetised needle, whose time of 
 vibration has been determined, and place it " end on " — 
 i.e., with one of its poles directed towards the magneto- 
 meter, and in such a position that its centre shall be 
 directly over, say, the 200th* division of the horizontal 
 ruler or scale, and supported by a convenient block of 
 wood, at such a distance above the table that the axis 
 of the needle produced would pass through the centre 
 of the small magnet and mirror of magnetometer, 
 Fig. 41. 
 
 M . 
 
 Then the ratio of — is given by the formula 
 H 
 
 M 
 H 
 
 = ^r 3 tan d 
 
 (B) 
 
 Instead of the deflecting magnet being placed "end 
 on," it may be placed " broadside on " — i.e., the hori- 
 zontal metre scale may be placed so that its 500th 
 division shall be immediately underneath the magneto- 
 
 Fig. 41. 
 
 Suppose the magnetised knitting needle to be placed 
 on the W side of, and with its N pole turned towards, the 
 magnetometer. Bead. off deflection of spot of light. Then 
 reverse position of magnet so that the S pole is directed 
 towards magnetometer, the centre of needle remaining 
 at same distance as before, the spot of light will now 
 be directed towards the other side of zero of scale. 
 Again read off and take mean of the two readings. Now 
 remove the needle and its supporting block to the other 
 (E) side of the magnetometer, and place it so that its 
 centre shall be at a distance from the centre of the 
 magnetometer corresponding to that at which the two 
 previous readings were taken (300 mm). To do this, 
 evidently we must place the centre of the needle over 
 the 800th division of the flat scale. Proceed to take the 
 mean of two morereadings, using alternate poles of the 
 needle. The mean of the two means so obtained may 
 be taken as the value of the deflection in divisions of 
 the scale. Let it be called the " observed deflection." 
 But since the angular velocity of the spot of light is 
 twice that of the mirror attached to the small magnets of 
 magnetometer, half the "observed deflection" must be 
 taken as representing the true value of the deflection at 
 the distance of 300 mm. Assuming, as we have done, 
 that both the horizontal and vertical scales are divided 
 equally— i.e., in millimetres— then half the observed de- 
 flection divided by distance of scale from mirror will give 
 us the tangent of the angle through which the magnet 
 has been deflected by the magnetised needle at a dis- 
 tance of 300 mm. ; call this tan d. By taking readings 
 with magnetised needle alternately on each side of 
 magnetometer, any error in distance arising from non- 
 coincidence of the suspended magnet with the 500th 
 division of the horizontal scale is corrected ; and half 
 the distance between the position occupied by the centre 
 of the magnet during the two deflections (in the above 
 case 200 mm. and 800 mm.) = ?°°z20° = 300 may be 
 
 safely taken as the distance, r, of centre of magnet from 
 centre of deflecti ng needle. 
 
 * Any distance may be taken which will give a convenient deflec- 
 tion ; the stronger the magnetism of the needle the greatei the 
 distance at which it can be placed from the magnetometer. 
 
 Fig. 42. 
 
 meter as before, but this time parallel to the meridian 
 
 instead of at right angles to it, and the needle, with its 
 
 supporting block, placed alternately at equal distances 
 
 on the X and S sides of the magnetometer, and with 
 
 its poles pointing E and W, Eig. 42. 
 
 Two readings with poles reversed are taken at each 
 
 position, and their mean value calculated as described 
 
 in the "end on" method. Half this value in scale 
 
 divisions divided by distance of magnet from scale, 
 
 gives as tan d. In this arrangement, with tan d and 
 
 distance r of centre of deflecting needle from magneto- 
 
 , M 
 meter, 
 
 H 
 
 is found as follows 
 
 — = r 3 tan d. 
 H 
 
 From these data we can calculate in absolute 
 measure either the value (H) of the horizontal compo- 
 nent of the earth's magnetism ; or (M) the magnetic 
 moment of the magnet employed during the investi- 
 gation. 
 
 To Find H. 
 
 Divide first equation (A) by second equation (B) and 
 extract the square root. 
 
 4 7T 2 K 4 7T 2 K 
 
 MH 
 
 rp2 
 
 or 
 
 IJI2 
 
 M \ r 3 tan d (if end on) r 8 tan d (if broadside on) 
 
 MH 2 
 
 M 
 
 =m= 
 
 4tt 2 K 
 £ T 2 r't&nd 
 
 or 
 
 H 
 
 _2tt / 
 - "TV : 
 
 K 
 
 l^tand 
 
 To Find M. 
 
 Eliminate H by multiplying first arid second equation 
 together, and extract the square root. 
 
 \[H x — = M2H = M 2 = 47r2K . h r> tan d (if end on) ; 
 H H T- 
 
 or, 
 
 4tt 2 K 
 
 T 2 
 
 r 3 tan d (if broadside on) ; 
 
 .'. M = ^J i K r> tau d. 
 
298 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Example 1. — The following value of H was obtained, 
 employing the method and apparatus described in the 
 text. 
 
 Time of one complete vibration. 
 
 First series. Second series. 
 
 3h. 5' 40" 1 3h. 22' 00"! 
 
 U61"'5 f=62" 
 
 10th vibration 3h. 6' 41"-5J 3h. 23' 02" J 
 
 } = 62"-0 ) = 62" 
 
 20th „ 3h. 7' 43"-5J 3h. 24' 04" j 
 
 30th „ 3h. 8' 46" J 3h. 25' 06"J 
 
 186-0 186-0 
 
 Mean 62" 
 
 62 
 Mean time of one vibration — = 6 2" = T 
 
 Length of magnetised needle 8 cm. = I 
 
 Radius of magnetised needle - l cm. = r 
 
 Weight of magnetised needle l - 848 gramme = W 
 
 / 72 r -2\ 
 
 Therefore, moment of inertia K =■ W i — 
 
 \12 4/ 
 
 Q2 -12 . 
 
 = 1-848 if- + 4- = 9'99 
 12 4 
 
 Distance of scale from 
 
 magnetometer = 600 mm. 
 
 Observed deflection of spot 
 
 of light = D = = 70 ,, (scale divisions) 
 
 True deflection of magnet 
 
 = g = d = 35 " 
 
 Tangent of d = |^ = "0581 
 600 
 
 Distance of centre of deflecting magnet from centre of 
 mirror 22 cm. 
 
 r s = 10648 
 
 2 7T = 6-2832 
 
 Hence H = ~J K 
 
 T v i r 3 tan d 
 
 _ 6-2832 /~^99 
 6'2 V 309-3 
 
 = 1-013 ^0^32 
 = 1-013 x -18 
 = -182 dyne 
 
 Comparison of the Horizontal Intensities of Two Magnetic 
 
 Fields. 
 
 The following method of measuring the strength of 
 the field surrounding the pole of any given magnet at 
 varying distances, or of comparing the forces or 
 " fields " of two or more magnets at the same distance, 
 depends upon the law of pendular vibrations, " that the 
 force is proportional to the square of the number of 
 vibrations executed in a given time." 
 
 Suspend in the paper stirrup of apparatus, Fig. 35, a 
 short magnetised needle, about lin. long, following the 
 instructions given on p. 295 as to the position of the 
 
 chamber with respect to the meridian, absence of 
 torsion, etc. Since the pole of a magnet vibrating in 
 the field produced by fixed magnet is also under the 
 influence of the horizontal component of the earth's 
 magnetism, means have to be taken to ascertain the 
 later quantity, in order to eliminate it. To do this, 
 set the needle vibrating through a small arc by bringing 
 a weak magnet near the outside of the case, taking care 
 to prevent any swinging motion being given to the 
 suspending fibre, and proceed to count the number of 
 transits of the nearest pole of the needle across the 
 plane of the scratches on the glass doors of the instru- 
 ment during any convenient period of time, say, 1 
 minute; call this number n lt then the force of the 
 earth's magnetism (or strength or field) acting on the 
 needle is measured byw 1 2 = "E." 
 
 Now place a long bar magnet with its axis in the 
 magnetic meridian, and raised to such a distance 
 above the table that its axis produced would pass 
 through the centre of the needle and paper stirrup, 
 and remembering to place the bar magnet so that 
 the pole nearest the vibrating needle shall be of 
 opposite polarity to that of the needle. Again set the 
 needle in vibration, and count the number of transits 
 made in the same interval of time as before (1 minute) 
 = w 2 , calling the force due to the pole of the bar 
 magnet at distance d 1 "Mj"; then the joint effect of 
 the earth and magnet, or E + M, is measured by w 2 2 ; 
 therefore M = E + M-E=n 2 2 
 
 Mi 
 
 — g-r-T— r - t?P\ 
 
 ■" •ttnr ^ p rrT^i rn.iMi niiiii.ri ^* IJ J 
 
 Q _£<unp' 
 
 Flu. 43. 
 
 The distance d v or the space separating the poles 
 (not the ends)* of the needle and bar magnets, must 
 now be carefully measured, and afterwards the pole of 
 the magnet removed to a distance d 2 and another obser- 
 vation made of the number of transits in 1 minute. 
 Calling this number n s , and the force at d 2 "M 2 ," we 
 have the combined force of the earth and magnet at d? 
 measured by n 3 2 ; and, therefore, M 2 = E + M 2 - E 
 = % 2 - V ; or, Mj : M 2 : : n 2 2 - n* : w 8 2 - nf. 
 
 Since the same magnet has been employed during 
 the whole time of the experiment and only the distance 
 separating its pole from that of the needle has varied, 
 we shall find, on comparing the distances d 1 and d 2 
 with the forces M. 1 and M 2 , that the latter vary inversely 
 as the square of the former. 
 
 Example 2. — A short magnet suspended by the silk 
 fibre, Fig. 38, when under the influence of the hori- 
 zontal component of the earth's magnetism alone, 
 made 17 complete vibrations per minute = n v The 
 N pole of a long bar magnet was then placed at a 
 distance of 25 cm. = d\ from the centre of the vibrat- 
 ing magnet, when the number of vibrations executed 
 in 1 minute was 24 = n.,. On removing the pole of the 
 
 * In long thin steel magnets the pole is situated at a distance of 
 from ^ to T 'j the length of the bar from its ends. 
 
MEASUREMENT. 
 
 299 
 
 bar magnet to a distance of 50 cm. = d. 2 , the number of 
 vibrations in 1 minute fell to 19 = n z . 
 Then as 
 
 1 . 1 
 
 n.y - 11, . Wo- ?!, 
 
 24 2 - n-: 19 2 - 17 s : 
 
 dj 2 ' d.f 
 
 ■ L-_l 
 
 ' 25 2 "50 2 
 
 576 -289 : 361 -289 ::-| ■ 1^ 
 
 287 
 
 r2 ::4 : 1 
 
 measurements of the Magnetic Moments of the same Magnet 
 after different increments of Magnetism have been im- 
 parted to it ; or, Comparison of the Magnetic Moments of 
 two or more Magnets with each other. 
 
 We saw on page 297 how the magnetic moments, M, 
 of a magnet might be determined in absolute units, the 
 value of H being obtained at the same time. The fol- 
 lowing method, known as the " deflection method," 
 enables us readily to compare the moments of a series 
 of magnets if the value of H at the place of observation 
 is known. 
 
 The magnetometer, scales, lamps, etc., are arranged, 
 as in Fig. 41, for the " end on " method, only as we shall 
 probably find the magnets we are using more powerful 
 than the magnetised knitting needle used in the deter- 
 mination of H, they will necessarily have to be placed at 
 a greater distance from the magnetometer ; therefore, 
 instead of placing the 500tb division of the metre rule 
 underneath the centre of the magnetometer, it will be 
 advantageous to have the whole length of the metre 
 rule on one side of the magnetometer, the suspending 
 fibre hanging just over the zero end of the rule, as 
 shown in outline in Fig. 43. 
 
 Having everything arranged with spot of light exactly 
 over centre of vertical scale, place the deflecting magnet 
 " end on " at such a distance from the magnetometer as 
 will produce a suitable deflection. Let r = this distance 
 between the centres of the magnet and the needle, and 
 d\ the deflection obtained. Eeverse the magnet so that 
 its opposite pole is now directed towards the magneto- 
 meter, the centre of the magnet remaining at the same 
 distance as before, and again note deflection, <L 2 (this 
 time on the opposite side of zero). Take the mean of 
 
 these two deflections ^~ 2 , and call it the " observed 
 
 A 
 
 deflection " d a , then the magnetic moment, M, of the 
 
 magnet is 
 
 M^Hr 3 tan d a . 
 
 Eemoving the magnet, and varying its magnetism by 
 any of the known methods, and afterwards replacing it 
 in its former position (or, in the case of comparing two 
 separate magnets, replacing the first magnet by the 
 second, and taking care that the distance between the 
 centres of the magnet and needle is accurately main- 
 tained), take the mean of two other observations, with 
 reversed poles, let this observed deflection bed*, then, as 
 
 before „ 
 
 M^JHr 3 tand 6 , 
 
 and since H and r 3 are common to both equations it 
 follows that 
 
 M : Mj : : tan d a : tan d*. 
 
 It has been mentioned that the definition of a 
 single magnetic pole was purely ideal, but that by taking 
 a long thin bar and magnetising it, the distance between 
 its poles becomes so great in comparison to the distance 
 separating the poles of the neighbouring magnet that 
 the effects due to the more remote pole may be neglected. 
 It is only under such conditions that the law of "inverse 
 squares " holds good. When the length of the magnet, 
 and therefore the distance between its poles, is small, 
 the needle of the magnetometer is influenced by both 
 poles of the deflecting magnet, in which case it may be 
 proved that the total action of the magnet must dimini sh 
 in an inverse ratio to the third power of the distance, 
 provided the action of a single pole really stands in an 
 inverse ratio to the square of the distance. 
 
 Weber has by this means indirectly proved the law 
 of inverse squares. 
 
 Arrange the apparatus and proceed as follows. 
 Either the magnetometer, Fig. 38, or a delicately- 
 suspended compass needle may be employed. 
 
 Let a 6 be a short magnet 10 cm. long, whose centre 
 is placed at 100 cm. from C, Fig. 43, the distance of 
 one pole, say b, will be 95 cm., and the other pole, a, 
 105 cm. from C. 
 
 If we call the attraction b.c = 1, at a distance of 1 cm., 
 
 the attraction at 95 cm. will be -— = „ nr . if the action 
 
 do- 9,025 
 
 of the pole stands in an inverse ratio to the square of 
 
 the distance. 
 
 From the same data, it follows that the repulsion of 
 
 the pole a.c is — - = — - : therefore the total action 
 
 exerted by the short magnet a. b on the magnetometer, 
 
 C, is 
 
 1 1 20 
 
 9,025 11,025" 9,950 
 
 If now the magnet be removed to double its former 
 distance, so that its middle be 200 cm. from C, and the 
 distance b.c being = 195 cm., and a.c = 205 cm., the 
 
 - , andtherepulsiona.c 
 
 attraction b.c will be 
 
 195 2 38,025 ' 
 and the total action = 
 
 38,025 42,025 
 
 J^ 1 
 
 205 ~ 42,025' 
 
 40 
 159,800' 
 
 Therefore, by removing the bar magnet from 100 cm. 
 
 20 
 
 to 200 cm., its action diminishes in the ratio 
 
 9,950 
 
 to 
 
 40 
 
 provided the action of each separate pole stands 
 
 159,800 
 
 in the inverse relation to the square of the distance. 
 
 20 . 40 _ 1 2 1,990 _ 1 
 
 995 
 
 But 
 
 9,950 159,800 995 15,980 15,980 8 
 
 or at double the distance the action is only £ as intense ; 
 and 8 is the third power of 2. 
 
300 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 'Example. — Compare the magnetic moments of two 
 magnets, A and B, from the following data. In each 
 case the centre of the magnets were 50 cm. from the 
 magnetometer. A gave an " observed deflection " of 
 40 scale divisions, while B produced a deflection of 36. 
 Distance of scale from magnetometer 600 mm. 
 
 Value of d= Scale reading 
 2 
 
 .'. A : B : : tan da, : tan db 
 
 .. 20 JJ3 
 
 ' 600 ' 600 
 
 : : -033 
 
 •030 
 
 11 : 10 
 
 Determination of the Angle of Inclination or Dip. 
 
 The determinatiou of the angle of dip is of great 
 importance, as coupled with the previously obtained 
 value of H it enables us to calculate the total intensity 
 of the earth's magnetism at the place of observation. 
 
 Place the instrument so that the graduated vertical 
 circle is parallel to the magnetic meridian, and since 
 the meridian line is already marked upon the table 
 this is easily effected; but supposing the meridian 
 were not known, its direction could be determined with 
 sufficient accuracy by simply turning the inclination 
 instrument through such an horizontal angle that the 
 needle hangs vertically. When this is the case the 
 plane of vibration of the needle is approximately east 
 and west, and 90 deg. from this plane will be the 
 meridian. 
 
 Set the needle in vibration by bringing pole of small 
 magnet near it, remove magnet, and allow the needle 
 to settle, and take readings at both ends of needle. 
 Repeat the operation, and take the mean of the four 
 readings. 
 
 By reading at both ends of needle, any error due to 
 eccentricity of axis of suspension to centre of circle is 
 corrected, and by repeating the observation and making 
 the mean the probable error is still further reduced. 
 
 Take out the needle from its bearings and rotate it 
 through an angle of 180 deg. round its own axis— i.e., 
 so that the side formerly facing the observer is now 
 turned away from him— and again take the mean of four 
 similar readings. The object of reversing pivot on 
 bearings is to eliminate any possible lateral displace- 
 ment of centre of gravity from axis of pivot. 
 
 The needle must again be removed, and its magnetism 
 reversed by carefully stroking it with a powerful bar 
 magnet. This is effected by laying the needle on the 
 table and bringing over, say, its N-seeking end— the S 
 seeking pole of the bar magnet (the magnet being held 
 m a vertical position) and gradually drawing the 
 magnet along the needle to its other end. Now lift off 
 the magnet and carry it back at a distance from the 
 needle to its former position, and repeat the action • 
 then turn the needle so that the side in contact with 
 the table is now upwards, and repeat the stroking 
 The polarity of the needle will be completely reversed 
 
 The needle is again mounted on its bearings, and 
 another series of observations made as before — i.e., two 
 with one face of the needle towards you and two with 
 its face reversed, taking care to read at both ends of 
 the needle each time. By reversing the magnetism of 
 the needle any error arising from longitudinal displace- 
 ment of its centre of gravity from the axis of the pivot 
 is corrected. 
 
 The mean of these 16 observations will be the angle 
 of dip = d. 
 
 Fig. 44. 
 
 Example. — With a simple form of dip needle, con- 
 structed as in Fig. 44, the following results were 
 obtained : 
 
 S-seeking Mean of the 
 N-seeking pole. Mean of pole. four readings. 
 
 253-5 - 180 = 735 74"0 — 
 
 253-7 - 180 = 73-7 7 74'2 73"8 
 
 After reversing pivots on bearing. 
 246-5 - 180 = 66-5 — 66-2 — 
 
 246-6 - 180 = 66-6 66'3 64"4 
 
 After reversing magnetism, but on same bearings as in H. 
 
 253-5 - 180 = 73-5 — 73"5 — 
 
 253-3 - 180 = 73-3 73'5 • 73"4 
 
 After reversing pivots on bearings to A again. 
 
 244-0 - 180 = 66-0 — 63"5 — 
 
 244-5 - 180 = 66-5 64"0 65"0 
 
 Angle of dip = 
 
 69--15 
 
 If the plane of the dip circle be turned out of the 
 magnetic meridian, the angle of dip increases until at 
 90 deg. from the meridian the needle stands vertically. 
 This arises from the fact that while the vertical com- 
 ponent, V, of the earth's magnetism remains constant, 
 the horizontal, H, has become nil, its effect being 
 merely to put a strain on the pivot. At any inter- 
 mediate angle, a, say 60 deg., while V remains 
 unchanged, H becomes diminished. 
 
 Resolving the horizontal force acting on the end of the 
 needle into two components, E^ in the present plane of 
 the needle and H 2 at right angles thereto, the force H 2 
 merely acts as a strain on the pivot, while B^ represents 
 the new value of H. But H x in this case equals H cos a, 
 
 , . V 
 and since = = tan d, the tangent of the new angle, a, 
 
 will be 
 
 V = _V__ _ V 
 H x H cos a H 
 angle of dip = d, and a the 
 
 x , and since the true 
 
 cos a 
 
 angle through which the 
 
ME A S UREMENT. 
 
 301 
 
 plane of the needle has been turned, the tangent of the 
 
 new angle of dip, D, will be . 
 
 cos a 
 
 Example. — The angle of dip d being 69 deg. (see 
 preceding example), the plane of the circle was turned 
 through an angle of 60 deg. = a, when the angle of dip 
 increased to 79'8 deg. = D. 
 
 By above formula, tan D = ^ = tan 69' = 2j625 
 
 cos a cos 69 deg. "5 
 
 = 5'25. Referring to table of tangents, we shall find 
 that 5-25 = 79-5 deg. 
 
 Having now measured the horizontal force, H, and 
 the angle of dip, d, the total force and its vertical 
 component can be calculated. 
 
 For let any inclination needle free to vibrate in 
 plane of the magnetic meridian take up its position, 
 then if we let d = angle of dip, and T the total force, 
 resolving this total force into two components, H and 
 
 V, at right angles to each other, we have — = cos d ; 
 
 :. T = 
 
 H 
 
 cos d' 
 
 and, again, since 
 
 V 
 H 
 
 tan d, 
 
 ;. V = H tan d. 
 
 Hence, H and d being determined by experiment, we 
 have only to divide the value of the horizontal com- 
 ponent, H, by the cosine of the angle of dip in order to 
 obtain the total force ; or, multiply the horizontal com- 
 ponent by the tangent of the angle of dip to get the 
 value of the vertical component, V. 
 
 Example. — The horizontal component, H, having 
 been determined and found equal to - 182 dyne, and the 
 angle of dip to be 69 deg. 15 min., what is the value of 
 the vertical component, and also the total force of the 
 earth's magnetism ? 
 
 The vertical component, V = H tan d. 
 
 = H tan 69 deg. 15 min. 
 
 = ■182x2-6396. 
 = -480 dyne. 
 The total force, T, 
 
 H "182 _ -182 
 
 cos d cos 69 deg. 15 min. "3527 
 
 The Tangent Galvanometer. 
 
 When a current of electricity flows through a con- 
 ductor we have seen that a magnetic field is produced 
 in the space surrounding the conductor, the effect being 
 to cause any magnet capable of movement in its imme- 
 diate neighbourhood to set with its magnetic axis at 
 right angles to the conductor. 
 
 If the wire be bent into an arc of a circle and a 
 current passed, the centre of the arc behaves exactly 
 like the pole of an ordinary magnet, and any pole of 
 a magnetic needle placed there will be attracted or 
 repelled according to the direction of the current, If 
 
 = • ")1 dyne. 
 
 the conductor, hereafter called the wire, be bent into a 
 circular shape, with radius of circle = r, then length of 
 wire in circle I = 2-wr, and with current = C, the force 
 acting at centre of circle, 
 
 2ttt „ 2tt 
 
 F = C 
 
 = C 
 
 If a magnet pole of strength M be placed at the. 
 centre of the circle, the force will be represented by the 
 product of the strength of pole and the intensity of the 
 field, or 
 
 p _ MC2tt 
 
 r 
 
 It is impossible to have a magnet with only one pole, 
 but we can approximate to the ideal by having a 
 magnet relatively small to the radius of the coil in the 
 centre of which it is suspended, the force exerted upon 
 either pole by the current being then practically the 
 same. 
 
 This is the principle adopted in the tangent galvano- 
 meter. 
 
 This instrument consists of a vertical ring of wire ; 
 in some instruments several coils of thick and thin wire. 
 The thick coils are used for large currents ; the thin 
 coils are for use with smaller currents, and can be used 
 in series, in parallel, or separately. The coils are 
 carried on a base supported by levelling screws, so that 
 they may be kept in a vertical plane irrespective of 
 inequality of surface upon which the instrument stands. 
 
 E tevculijort 
 
 Thorny. 
 
 Fig. 45. 
 
 Fig. 45 shows one form of the instrument. At the 
 centre of the coil, suspended on a delicate pivot, or by 
 a silk thread, is a small magnet, whose length should 
 not be more than ^ oi i of the diameter of the 
 ring. The magnet carries a light pointer at right 
 angles to its axis, so as to enable the graduated card 
 just underneath the magnet to be easily read. 
 
 In setting up the instrument the plane of the ring 
 must be placed in the magnetic meridian, when the 
 pointer should point to zero on the scale. 
 
 Since the magnetic field produced at the centre of the 
 ring is very uniform, and the length of the magnet 
 small in comparison to the radius of the coil, we may 
 without serious error assume the poles to be at equal 
 distances from the ring, whence a current through the 
 wire will act upon each pole of the magnet with equal 
 and opposite forces. The effect of this couple is to 
 rotate the magnet round its point of suspension. When 
 
302 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 a current passes, the magnet is rotated through a 
 deflection, d. The extent of the rotation depends upon 
 the ratio of the force due to the current in the coils, and 
 the horizontal component of the earth's magnetism. In 
 all questions relating to a magnet we must consider the 
 effect of the earth's magnetism, or in some way 
 neutralise the effect. Eoughly speaking, the earth acts 
 as a magnet with its poles near to but not quite in the 
 same direction as the geographical poles. The position 
 of the earth's magnetic poles is constantly, though 
 through small periods of time, slightly changing. Just 
 as we term the imaginary lines passing through any 
 place, and through the geographical poles, the geo- 
 graphical meridian of that place, so an imaginary line 
 drawn through a place and through the earth's 
 magnetic poles is called the magnetic meridian of that 
 place. The action of the earth's magnetism upon the 
 suspended needle is to direct the axis of the needle 
 into a position parallel to the magnetic meridian — in 
 other words, to make the needle point north and south 
 towards the north and south magnetic poles. If the 
 needle be moved out of the meridian, the horizontal 
 intensity, H, of the earth's magnetism tends to pull it 
 back into the plane of the meridian. 
 
 Now we are in a position to determine the law which 
 governs the needle under the action of a current in the 
 coil. It must be remembered that the action on one 
 pole is equal and opposite to that on the other. 
 
 Let N S, Fig. 46, represent the magnet in the plane 
 of the coil, both magnet and coil being in the plane of 
 the magnetic meridian. 
 
 Let Nj Sj be the new position of the magnet when a 
 current is sent through the coil. 
 
 Let d be the angle through which the needle is 
 deflected. 
 
 H, as before, is the horizontal intensity of the earth's 
 magnetism. 
 
 Let F be the force due to the current. 
 
 Then the moment of couple due to earth's magnetism 
 = H x AB. 
 
 And the moment of couple due to current = F x 
 D E, but when the magnet is at rest these couples are 
 equal. 
 
 That is, FxDB = HxAB 
 
 whence F = H x 
 
 = H x 
 
 AB 
 DE 
 
 AO 
 DO 
 
 = H tan d 
 which gives the law : " The magnetic force which, 
 acting at right angles to the magnetic meridian! 
 produces on a magnet the deflection d is equal to 
 the horizontal force of the earth's magnetism at 
 that point multiplied by the tangent of the angle of 
 deflection." 
 
 At the same point H may be taken as constant 
 hence the current varies as the tangent of the angle 
 of deflection. It must be noted that the strengths of 
 
 current are relative only. Currents can thus be com- 
 pared with one another. 
 
 In using a tangent galvanometer the aim should be 
 to get deflections as near to 45 deg. as possible, because 
 this is the most sensitive deflection of the instrument. 
 If two currents are to be compared, obtain deflections 
 as near as possible at equal distances on the two sides 
 of 45 deg. Thus if it is desired to get a current double 
 of another, let the apparatus be so arranged as to give 
 with the smaller current, say, a deflection of 30 deg. 
 
 Tan of 30 deg. = -5773503. 
 
 A current of double strength will give a deflection in 
 degrees of which 
 
 the tan = 2x -5773503 = 1-1547006 or 46 deg. 6' nearly. 
 
 The tangent galvanometer enables us readily to 
 determine strength of current in absolute measure, and 
 then a division by 10 gives us strength of current in 
 amperes, because the practical unit of electrical current 
 = 10 _1 or T V of the C.G.S. unit of current. To obtain 
 the value of the current in absolute measure, or in 
 
 Fro. 46. 
 
 amperes, it is necessary to know the reduction factor or 
 the constant factor of the galvanometer for converting 
 indications of the galvanometer into absolute measure, 
 
 which is - — with one turn of wire in the coil, and 
 
 r H 
 
 with n turns of wire in the coil. 
 
 2 7r n 
 
 The formula for current then becomes 
 
 C = _!L_ H tan d, 
 
 2, tt n 
 
 and for current in amperes C = 
 
 ~ - H tan d -L. 
 
 2t»i 10 
 
 With the foregoing theoretical considerations we are 
 in a position to define "unit current"; it is "that 
 current which, flowing through a wire formed into 
 a circle of unit (1 cm.) radius, acts on a unit magnetic 
 pole placed at the centre with unit force (1 dyne) per 
 unit (1 cm.) length of the circumference." 
 
 Hence, if the length of the arc be equal to 1 radian, 
 or 57 deg. 17' then unit current will repel a unit magnetic 
 pole with a force of 1 dyne, whereas, if the wire forms 
 
MEASUREMENT. 
 
 303 
 
 a complete circle, the force at the centre will be 2 -k 
 dynes = 6-28318 dynes. 
 
 The relation of the forces acting on the poles of a 
 short magnet placed at their centre by currents flowing 
 through coils of different radii is very clearly illustrated 
 by the following arrangement, due to Professor Poyn ting, 
 of Mason College, Birmingham. 
 
 On a wooden base, supported on three levelling 
 screws, is fixed a vertical board, about 18in. square, 
 having a circular hole bored through its' centre. The 
 diameter of the hole must be sufficiently large to allow 
 the small mirror with attached magnet belonging to 
 the magnetometer to swing freely, when suspended by 
 a silk thread attached by a brass pin to the face of the 
 board. An iron nail must not be used, as it would pro- 
 bably affect the magnet, and the silk thread must be 
 kept a little distance from the face of the board, to 
 enable the magnet and mirror to vibrate freely. 
 
 "With a radius of 4in. from the centre of the hole, a 
 single layer of silk or cotton covered wire is attached to 
 the face of the board either by some kind of cement, or 
 by a few brass wire staples, the two ends of the coil 
 being led down to the base, and soldered to two binding 
 screws. A second coil of wire is also attached to the 
 board in a similar manner, but with a radius of 8in., 
 and making two complete turns, its two ends being led 
 down and secured to two other binding screws. At a 
 suitable distance in front of the board and its rings 
 (which, as in the case of a tangent galvanometer, must 
 be set in the plane of the magnetic meridian) is placed 
 the scale and lamp. When all is arranged satisfactorily, 
 and the spot of light is focussed on the zero division of 
 the scale, join up a single Daniell cell (and sufficient 
 resistance to keep the spot of light well on the scale) 
 with the terminals of the inner ring, and note deflection 
 produced. On removing the connecting wires from the 
 terminals of the inner to those of the outer circle of 
 wire, the same deflection will be produced, for though 
 the current now goes twice round the magnet, it does 
 so at twice the distance. Therefore, the force at centre 
 of first coil will be 
 
 = C 
 
 2-7TT 
 
 = C 
 
 2 x 3-1456 x 10 
 
 10 2 
 
 = Cx -62832, 
 
 and at centre of second coil 
 
 = G 2jrnZ = C 2 x 3 ' 1416 x 2 X 20 - C x -62832, 
 
 r 2 20 2 
 
 as before, and since we are using equal currents in each 
 case, C cancels out, and we see at once that the two 
 forces are equal to each other. This fact can be still 
 more strikingly shown by joining up the two coils "in 
 series," but so that the current will flow round the two 
 coils in opposite directions, when no deflection of the 
 magnet will occur, the effect of one coil being exactly 
 neutralised by that of the other. 
 
 If the student has grasped the fundamental principles 
 on which the tangent galvanometer is based, he will be 
 in a position to follow the directions given in the 
 following pages as to the method of using the instrU' 
 
 ment ; and as it is important that every student should 
 possess one of these useful pieces of apparatus, we 
 insert the following description taken from the South 
 Kensington syllabus of the course of practical instruc- 
 tion in physics, whereby those students who are not 
 prepared to incur the expense of purchasing an instru- 
 ment from a professional maker can make one for 
 themselves, which will be quite good enough for 
 ordinary purposes. 
 
 " Cut three strips of thin card lin. wide, and long 
 enough to go round a wood block 6Jin. diameter. Lay 
 on the first piece dry and fix the other two with weak 
 glue; the joints should just meet, but not overlap, 
 wind round with string and leave during the night to 
 dry. On the middle Jin. of the card wind 20 turns of 
 thin insulated wire in two layers of 10 turns each. 
 Tie the ends together, leaving them about 6in. long. 
 On either side of the thin coil, and touching it, wind 
 one turn of thick wire and leave the ends about 8in. 
 long. Varnish, and leave to dry. Divide the card ring 
 into four equal quadrants, taking care that one of the 
 divisions is opposite the ends of the wire. This, when 
 the card is mounted, will be the lowest point in the 
 ring. 
 
 To the middle of the opposite division, which will 
 be the highest point for a minute ring of fine wire, 
 on either side of the remaining two divisions, fix a 
 small block of wood so as to hold a 6£in. disc of card 
 between them exactly at right angles to the plane of 
 the ring. Mark each block with a central line, so that 
 when a diameter of the card meets these lines the card 
 is fixed centrally also. Divide the card into degrees 
 and tangents. The degrees may be marked from a 
 protractor, but the tangents are best found thus : Fix 
 the card on a table and put a pin through its centre. 
 At exactly lOin. from the centre of the zero line fix a 
 scale of inches divided into tenths, exactly at right 
 angles to the zero line — say a straight edge against the 
 pin and over each division of the scale, and where the 
 straight edge cuts the circle of the card make a mark. 
 These marks will form a scale of tangents. Make a 
 hole in the centre of the card fin. in diameter. Fix 
 the card in the ring, so that the diameter at right 
 angles to the zero line is in the plane of the ring. Fix 
 the ring vertically in the centre of an 8Jin. square of 
 wood, using blocks of wood to hold it in position ; pass 
 the ends of the wires through holes in the board, bring- 
 ing the ends of the thin wires to binding screws at one 
 corner of the board, and the thick wires to the opposite 
 corner. Support on three levelling screws. Cut a 
 piece of steel -i x i x £ ; harden and magnetise. Fix 
 with thin wire at right angles to steel a capillary glass 
 pointer about 5in. long. Support magnet and pointer 
 by single silk fibre to small ring at top of card ring. 
 Adjust length of silk so that magnet hangs in central 
 aperture of card disc with pointer free to move. Turn 
 ring into plane of magnetic meridian, and twist levelling 
 screws till magnet swings freely in central hole. 
 
 With weak currents use thin wire coil, and with 
 strong the thick wire." 
 
304 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Electromotive Force, or Electrical Pressure, and Resistance. 
 
 In the preceding pages we have denned the electro- 
 magnetic unit of current, and briefly explained how 
 currents can be determined in absolute measure ; it 
 now remains to define the absolute units of electro- 
 motive force (pressure) and resistance, and from these 
 and the unit of current to derive the practical units — 
 i.e., volt, ohm, ampere, coulomb, watt, joule, and farad. 
 
 Ohm's law teaches us that the strength of a current 
 flowing through a wire whose resistance is E, the ends 
 of such wire being maintained at a difference of potential 
 Vi - V 2 , is expressed by the formula 
 Vr-V 2 _ E 
 G It "E 
 
 where E represents the electromotive force due to the 
 difference of potential Vj-V 2 . Evidently in above 
 equation the unit representing any one of the three 
 quantities is dependent upon the other two. This 
 connection between the three quantities will be made 
 apparent by considering the action of an imaginary 
 dynamo. 
 
 Suppose that in an uniform magnetic field — -such 
 as that due to the earth's magnetism — we place two 
 parallel rails so that their plane is at right angles to 
 the line of dip, Fig. 47. Let one end of each rail be 
 
 W 
 
 connected to the terminals of a galvanometer, and the 
 circuit completed by means of a metal slider, free to 
 move along the rails. Assume for a moment that the 
 system has no resistance. The slider being stationary, 
 insert the N-seeking pole of a bar magnet between the 
 rails, and therefore in the line of dip, when, supposing 
 the rails to extend in a N and S direction, there will 
 be a reverse current — that is, in the opposite direction 
 to that in which the hands of a clock move — induced 
 in the circuit, flowing from the east rail through the 
 slider to the west rail, and thence through the galvano- 
 meter back to the east rail again. 
 
 On withdrawing the N-seeking pole of the magnet, a 
 " direct " purrent— that is, in the same direction as the 
 
 hands of a watch move in — will be produced. Now on 
 inserting the N pole of the magnet, the effect is simply 
 to increase the number of lines of force enclosed within 
 the system comprising the rails, slider, and galvano- 
 meter, and withdrawal is accompanied by a corre- 
 sponding decrease. 
 
 Hence we see that "increasing the number of loops 
 or lines of force in a given area gives us a ' reverse ' 
 current," and decreasing the number furnishes a 
 " direct " current. 
 
 Dispensing with the magnet, and simply using the 
 magnetic field due to the earth, it is obvious that we 
 can vary the number of lines of force enclosed by 
 moving the slider. Suppose the slider to move with a 
 velocity of 1 cm. per second, and call the electromotive 
 force developed E, then a velocity of 2 cm. per second 
 will yield an electromotive force of 2 E — in other 
 words, the electromotive force increases with the 
 velocity of the slider. 
 
 If the field be of unit strength, the distance between 
 the rails, and therefore the length of the slider, 1 cm., 
 and the distance travelled in unit time (1 second) be 
 1 cm., then the difference of potentials, or electromo- 
 tive force, will equal one absolute limit. 
 
 Having our units of electromotive force and current 
 fixed, it becomes an easy matter to define the remain- 
 ing unit of resistance. 
 
 Eeferring again to our rails and slider, we see (since 
 we have assumed that they are without resistance) that, 
 in accordance with Ohm's law, as we increase the 
 electromotive force by increasing the velocity at which 
 the slider moves, we increase the current in the same 
 ratio. Now letting the slider move 1 cm. per second 
 as before, let us put so much resistance into the circuit 
 that unit current flows ; the resistance so added will 
 equal one absolute unit of resistance, or we may define 
 unit resistance as that resistance which under unit 
 electromotive force conveys one unit of current per 
 E_ 
 E' 
 
 If, therefore, we increase the velocity of the slider, 
 we must, in order to keep the current equal to unity, 
 increase the resistance in the same proportion, or 
 E n 
 E n 
 
 These absolute units of electromotive forqe and 
 resistance are far too small for actual work, therefore 
 multiples of these absolute units have been taken 
 to form a practical series — i.e.: 
 
 The Volt, or unit of electromotive force, is equal to 
 100,000,000, or 10 8 absolute units, so that our ideal 
 slider, in order to produce a difference of potentials 
 equal to 1 volt, would have to traverse one hundred 
 million centimetres per second. 
 
 The Ohm, or unit of resistance, would, as we have 
 
 seen above, in order to maintain unit current, have to 
 
 If 1 T^ 1 9 
 
 equal 10 9 absolute units, for C = — = But, as 
 
 1 E E10" 
 
 before mentioned, the practical unit of current {the 
 
 second, C = - 
 
 C = 
 
ME A S UREMENT. 
 
 305 
 
 ampere) is only of an absolute unit, therefore we 
 must, in order to reduce the current ^, increase the 
 resistance 10 times, or our slider must move with a 
 velocity of 10 9 centimetres per second. 
 
 It would be impossible to construct our ideal rails 
 and slider, but a closed circuit consisting of a coil of 
 wire, in the centre of which a small magnet is sus- 
 pended, may be easily constructed and caused to rotate 
 in the magnetic field due to the earth's magnetism. 
 
 Then, knowing the velocity of rotation, the strength 
 of the field, length and resistance of the wire, and the 
 deflection of the needle, it becomes possible to calculate 
 the electromotive force, and therefore the resistance, 
 necessary to permit unit current to flow. This was the 
 method employed by the committee of the British 
 Association in the determination of the "B.A. " unit 
 of resistance, or " ohm." 
 
 The Coulomb, or unit of quantity, represents the quan- 
 tity of electricity conveyed by unit current in unit time — 
 i.e., by a current of 1 ampere in 1 second. It is equal 
 to 1CH or "1 of an absolute unite of quantity. 
 
 The Farad, or unit of capacity. A condenser has 
 unit capacity if it contains unit quantity at unit differ- 
 ence of potential, or 1 coulomb under an electromotive 
 force of 1 volt. It is equivalent to 10 -9 or one thousand 
 millionth of an absolute unit of capacity. Since the 
 farad is too large for general purposes, the microfarad = 
 10~ 15 absolute unit (or one millionth of a farad) is 
 usually employed. The Atlantic cables have, on the 
 average, a capacity of about "35 microfarad per knot. 
 
 The Watt, or unit of power, equals 10 7 ergs, or 
 10,000,000 absolute units of power. It represents the 
 power conveyed by an ampere of current under a differ- 
 ence of potentials of 1 volt, or the power done by a 
 current of 1 ampere through a resistance of 1 ohm. 
 It is equal to ttw horse-power. 
 
 The Joule, or unit of work or heat. Like the watt, 
 or unit of power, it is equal to 10 7 ergs, but the factor 
 time is introduced. It represents the work done, or 
 heat generated by an ampere flowing through a resist- 
 ance of 1 ohm or 1 coulomb falling through a difference 
 of potential of 1 volt in 1 second. It will heat "2405 
 gramme of water 1 deg. C. 
 
 The Wheatstone Bridge. 
 
 For the measurement of resistances some form of 
 Wheatstone bridge is essential, it being far more con- 
 venient and expeditious than any of the substitution or 
 differential methods. 
 
 We shall first describe the method of using the 
 ordinary Post Office form of bridge, and subsequently 
 the kind known as Kirchoff's "slide wire or metre 
 bridge," a form of bridge that can be made by anyone 
 possessing ordinary mechanical skill, and which, in 
 combination with a set of resistance coils, will enable 
 anyone to make the greater portion of the measure- 
 ments described in the following pages. 
 
 For a detailed explanation of the theory of the 
 Wheatstone bridge, the student must consult some 
 standard text-book, such as Kenyon. We shall confine 
 
 ourselves to practical details, merely remarking that 
 the Wheatstone bridge is an arrangement of a series of 
 resistances, by varying which we can, as in an ordinary 
 proportional sum (three of the values being known), 
 determine the fourth. 
 
 Theoretically, the resistances are placed as in 
 Fig. 48. The three known resistances being placed 
 in the arms AB, B C, and AE, while the unknown 
 resistance, whose value is to be determined, occupies 
 
 Fig. 48. 
 
 the fourth arm, C B . The poles of one or more cells 
 of a battery (Daniell's are very suitable) are connected 
 to the points B E, having a key, K l7 in circuit. A 
 delicate galvanometer, G, and key, K 2 , are joined up 
 with the points AC, and a certain ratio being given to 
 A and B, K x is depressed. The resistance, D, is varied 
 until, on closing K 2 , no deflection of the galvanometer 
 
 needle occurs, 
 that as 
 
 or, 
 
 Fig. 49. 
 
 When this is the case it can be proved 
 B : A : : D : X, 
 
 AD 
 B ' 
 
 X = : 
 
 Assuming that the ordinary Post Office form of 
 bridge is available, the connections are made as indi- 
 cated in Fig. 49, where the lettering is similar to the 
 theoretical diagram, Fig. 48. 
 
306 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 The wires leading from the battery are joined to 
 terminals B 1 E. These are usually marked " copper " 
 and " zinc." This arrangement is useful in testing 
 telegraph lines, but for measuring ordinary resistances 
 it is immaterial to which of the two terminals, ~B 1 E, 
 the copper is attached. The galvanometer is joined up 
 to the two terminals, A C and the resistance to be 
 measured is connected to the two terminals C E. 
 Having everything arranged, take out the two 1,000- 
 ohm plugs in the arms AB, BC, and also one of the 
 infinity plugs. 
 
 Close the right-hand key, ~K lt so as to complete the 
 battery circuit, and immediately afterwards the left- 
 hand key, K 2 (to bring in the galvanometer). The 
 object of closing the battery key before the galvano- 
 meter key is to prevent the violent throw of the needle 
 due to self-induction in the resistance coils. 
 
 The keys being held down, note the direction in 
 which the needle of the galvanometer is deflected. This 
 enables us to judge, when making subsequent adjust- 
 ments, whether the resistance unplugged in AE is too 
 high or too low. Suppose that, with the infinity plug 
 removed, the deflection was to the right-hand side of the 
 observer. Then a similar deflection obtained during 
 the actual measurement shows that the resistance, A E, 
 is greater than the unknown, C E, whereas a deflection 
 to the left informs us that AE is less than CE. AE 
 can then be varied till balance is obtained, and the 
 needle remains stationary. 
 
 Having ascertained this necessary information, 
 reinsert the plugs, and proceed with the actual 
 measurement. 
 
 If the resistance to be measured is small, the two 
 10-ohm plugs in the arms A B, B C should be removed 
 and A E adjusted till balance is secured, while with 
 higher resistances the two 100, or 1,000, ohms should 
 be employed instead of the two ] ohms, as the nearer 
 the values of the resistances in the four arms approxi- 
 mate to each other, the greater is the accuracy obtained. 
 If with the resistances in the two arms, A B, B C 
 equal— i.e., 10:10,100:100; or 1,000 : 1,000, as the 
 nature of the case may require — we find, on introducing 
 a certain resistance in A E, that balance is obtained ; 
 then it follows that the unknown resistance equals that 
 in A E, which is known, for as 10 : 10 : : A E : C E. 
 
 But it will frequently happen that balance cannot be 
 secured by altering the value of A E, the arms A B 
 and B C remaining equal. Suppose that with A B and 
 B C = 10 ohms each, that 25 ohms in A E is too low, 
 and 26 ohms is too high, obviously the true value of C E 
 lies between these two values. Keplace the 10-ohm 
 plug in AB, at the same time removing the 100-ohm. 
 On making A E 250 ohms, we shall find it is too low 
 and 260 too high, but increasing the value of the 
 250, unit by unit, we shall probably find that balance 
 is obtained, say, when A E = 256 ohms ; then the value of 
 the unknown resistance will be as 
 
 100 : 10 : : 256 : 256 ohms. 
 Supposing we still fail to obtain a balance, say, 256 
 
 is too low and 257 too high, change the ratio of the 
 arms AB, B C to 1,000 : 10, and make A E 2,560, and 
 increase step by step, as before, until balance is obtained, 
 suppose when A E = 2,569 ohms then the unknown 
 resistance, CE, is 
 
 1,000 : 10 : : 2,569 : 25'69 ohms. 
 Should the resistance to be measured be greater than 
 the whole resistance in the box, usually about 11,000 
 ohms, its value may be ascertained by varying the ratio 
 of AB : BC, but in the opposite direction to that 
 described above. 
 
 Suppose that with A E = 10,000, the deflection shows 
 that it still falls short of that necessary to balance by 
 making A B = 100 and BC = 1,000, we could get any 
 value up to 100,000 ohms, for as 100 : 1,000 : : 10,000 : 
 100,000, and by making A B 10, and B C 1,000, we 
 have a range of 1,000,000 ohms, but with these high 
 ratios the results can only be considered approximately 
 true. 
 
 With the bridge properly adjusted, proceed as follows 
 to measure the resistance of three lengths of wire. 
 
 (1) Separately. 
 
 (2) In series. 
 
 (3) In parallel. 
 
 Calling the three wires respectively A, B, and C, 
 fasten the two ends of A to terminals C E, and measure 
 its resistance, and the same for wires B and C. 
 
 Next join the three wires end to end "in series," and 
 measure the combined resistance ; it should be equal to 
 the sum of the separate resistance A, B, and C. 
 
 Care must be taken that the ends of the wires do 
 not overlap more than is necessary to form a junction, 
 otherwise the total length of the wires will be reduced 
 and the resistance too low. 
 
 Next proceed to measure the resistance of the wires 
 in "parallel" — by pairs — and prove that the resistance 
 of two wires in parallel equals the product of their 
 separate resistances divided by their sum, or E = 
 A x B 
 A + B' 
 
 Lastly, join the three wires in parallel, and show that 
 the total resistance corresponds to that given by the 
 
 formula B = , ^ ^ — -. 
 
 (A B) C + A B 
 
 Example. — Three wires measured separately gave 
 the following resistances : 
 A = 2431 
 
 B = 4-09 } 11-17 ohms. 
 C = 4-65J 
 
 Measured in series the resistance was 
 
 A + B + C = 11-16 ohms. 
 With two of the wires in " parallel " the resistances 
 were : 
 
 By calculation from 
 
 By actual measurement. 
 
 A + B = 1-520 ohms 
 B + C = l-596 „ 
 B + C = 2-180 „ 
 
 formula K = 
 
 AB 
 
 A + B 
 1-520 ohms 
 1595 „ 
 2-181 „ 
 
MEASUREMENT. 
 
 307 
 
 Lastly, the three wires in parallel gave 
 
 = 1-149 ohms 1-148 ohms. 
 
 If the student has not a properly constructed bridge 
 of the kind already described, he can readily make for 
 himself one of the "slide wire" or metre bridge 
 pattern. This is a very valuable and efficient form of 
 bridge, and if carefully made will yield very satisfactory 
 results. 
 
 To a well-seasoned board, 110 cm. long, 10 cm. wide, 
 and 2| cm. thick, are attached, by screws, three thick 
 
 injuring the surface of the slide wire by sparking on 
 opening battery circuit. If the resistance in a : x be 
 low, Kj should be closed for as short a time as possible 
 to prevent the coils in a becoming heated. 
 
 Assuming, for the sake of comparison with the dia- 
 gram of the Post Office form, that the arrangement in 
 Fig. 51 is adopted. A key, K, is placed in the battery 
 circuit, and the free end of the galvanometer wire is laid 
 on different portions of the German silver wire until 
 no deflection of the galvanometer needle occurs. It will 
 thus be seen that merely touching the slide wire with 
 
 ' mm 
 
 
 
 - n 
 
 c> 
 
 n 
 
 <~> 
 
 
 
 i i i i i i i i i i i i i i i I I l l I i i i i r i i i i i i i i i i i i 
 
 «— ' 
 
 
 
 Fig 50. 
 
 strips of copper. One, 90 cm. long, having three binding 
 screws soldered to it, one at each end, the other at the 
 centre, is fastened parallel to the length of the board ; 
 the other two strips are each 6 cm. long, and have one 
 binding screw soldered to one end of each strip. These 
 two short strips are fixed transversely to the length 
 of the board, and at a distance of exactly 1 metre 
 between their inner faces. To the free ends of these 
 
 the galvanometer wire replaces K 2 in the Post Office 
 bridge. In the more elaborate kinds of the slide wire 
 bridge the galvanometer wire is attached to a movable 
 block and spring key, A 2 . This block slides along the 
 German silver wire, and by depressing the spring contact 
 is made. 
 
 To make a measurement, place the set of resistance 
 coils in the gap A, and the unknown resistance in the 
 
 E 
 
 3 
 
 
 f° C 
 
 
 
 E 
 
 
 3 tO 
 
 
 
 
 Ol C 
 
 
 A 
 
 
 l"¥ 
 
 
 X 
 
 
 B — 
 
 ^ 
 
 At"{ n ( 
 
 . u^ D — 
 
 
 ^UL_J> 
 
 Jig. 51. 
 
 short strips are soldered the ends of a well-drawn 
 German silver wire, and underneath the wire a paper 
 scale (1 metre long) divided into 1,000 mm. is pasted 
 to the board, Fig. 50. The other ends of the short 
 transverse strips are provided with binding screws. 
 
 The ways of connecting up this kind of bridge is 
 shown in Fig. 51. 
 
 Sometimes the positions of the battery and galvano- 
 meter are reversed, but the above arrangement is to be 
 preferred, as there is less likelihood of the galvanometer 
 needle receiving a violent throw than when placed in 
 the other position ; also, there is not so great a risk of 
 
 other gap, X, and slide the contact wire along till 
 balance is obtained ; then, as before, 
 
 B : D : : A : X 
 
 or, 
 
 X = 
 
 AD 
 B " 
 
 As with the other kind of bridge, it is necessary to 
 ascertain whether a deflection of the galvanometer, 
 say, to the right, indicates the resistance in A as being 
 too high or too low. Of course, this is done making A 
 or X infinitely great at first. 
 
308 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Having made the slide wire bridge, a set of resistance 
 coils is necessary. It is better to obtain them from 
 makers who are accustomed to the work, but there 
 is nothing difficult in constructing such resistances. 
 With the slide wire bridge the number of coils need 
 not be so large as with the Post Office bridge for this 
 reason : in the latter form we only have three ratios 
 between the arms, a b — -viz., 10 : 10, 10 : 100, and 
 10 : 1,000 ; whereas with the former we can get any 
 ratio ranging between 1 : 999 and 999 : 1, although the 
 values obtained when the slider approaches the ends of 
 the slide wire are not comparable to those obtained 
 when it is nearer the centre. A set of six coils of the 
 following values will be found useful : 1, 5, 10, 25, 50, 
 and 100 ohms. The wires for the three smaller values 
 may consist of silk-covered copper, but for the remain- 
 ing coils silk-covered German silver wire must be used, 
 as the length of copper wire necessary to give the 
 required resistance would be considerable. Should the 
 worker have access to a set of coils, he may measure 
 off his wires directly ; but if not, he must procure from 
 some dealer or manufacturer of resistance coils a length 
 of wire of known resistance. 
 
 " Lead one end of the first coil of wire under the 
 board, and solder to the rivet of the first mercury cup. 
 Lead the second end of the first coil and also the first 
 end of the second coil to the rivet of the second cup — 
 the second end of the second coil and the first end of 
 the third coil to the rivet of the third cup, and so on. 
 Make thick copper wire staples l|in. long and turned 
 down £in. at each end for connecting the cups. Make 
 also some longer staples, 3in., 4fin., and 6in. long. 
 Mark the value of the coil connected with each pair of 
 mercury cups on the board between them ; amalgamate 
 the loose copper connections." 
 
 We may supplement the above concise description 
 by saying that the reason why the wires are bent in the 
 middle and then wound on the reels, so that the two 
 halves lie side by side, is to prevent self-induction, and 
 also any action on the needles of galvanometers placed 
 in the neighbourhood of the set of coils. 
 
 The amalgamation of the copper connections is to 
 ensure good contact with the mercury filling the cups, 
 and for this reason the flat-headed rivets mentioned 
 in above quotation ought to be amalgamated likewise. 
 
 In order to amalgamate the loose contacts, the copper 
 
 Q Q O O Q O 
 
 ^ ^ W ( 4iiy?ii 
 
 & 
 
 s y ,o y 25 y 50 y l00 ^ 
 
 Fig. 52. 
 
 " Measure off (by bridge) lengths of wire having resist- 
 ances in ohms or fractions of ohms, as above men- 
 tioned. Double the wire in middle, and laying this 
 middle part on a cotton reel, commence winding until 
 all the wire is laid on except a few inches at each end ; 
 soak in melted paraffin. In order to ensure the 
 greatest accuracy, some makers, instead of measuring 
 in the first instance the full length of wire, measure a 
 length of wire which will be about (very slightly over) 
 half the required resistance. Two lengths of such 
 wire are laid side by side, and at the ends which 
 represent the point where the bend would come in a 
 single wire, a drop of solder connects the two wires in 
 series. The resistance of the whole is now measured, 
 and if too great the wires are soldered a little further 
 up, and so on until great accuracy is obtained. It is 
 said that this plan is more practical than the doubling 
 plan, and answers the same purpose. Make mercury 
 cups l£in. apart in a row along one edge of the board. 
 Fasten the bobbins of wire along the centre of the board 
 in a row opposite interspaces between the mercury cups. 
 Through a hole in each mercury cup push a flat-headed 
 copper rivet, so that the head forms the bottom of the 
 cup, Fig. 52. 
 
 must first be cleaned by dipping the ends into a little 
 nitric acid, and afterwards into some nitrate of mercury, 
 washed, and burnished up. 
 
 Another indispensable piece of apparatus (indeed 
 without it the Wheatstone bridge will be valueless) is 
 a sensitive galvanometer. For ordinary bridge work a 
 carefully made astatic galvanometer answers exceed- 
 ingly well ; but for determinations of capacities a mirror 
 galvanometer is necessary. The construction of the 
 astatic form of the instrument is so simple and so 
 thoroughly explained in most text-books that it need 
 not be referred to here ; but a few words respecting the 
 mirror galvanometer may not be out of place. To 
 construct a highly-sensitive mirror galvanometer, and 
 accompanying box of shunts, require considerable tech- 
 nical skill; but an instrument possessing sufficient 
 delicacy for ordinary purposes may be made without 
 much difficulty. 
 
 Get a boxwood reel turned to shape and size shown 
 in section on Fig. 53 ; place this reel in a lathe, or ou 
 an horizontal spindle capable of being rotated by a 
 handle, and carefully wind on No. 36 silk-covered 
 copper wire until the reel is filled, leaving about 
 2ft. of each end of the wire projecting to be attached 
 
ME A S UREMENT. 
 
 309 
 
 to two terminals. It will be found an improvement 
 to have these projecting ends of thicker wire carefully 
 soldered to the thin wire, as by doing so there is less 
 risk of their being broken by any strain they may 
 be accidentally subjected to owing to their exposed 
 situation. 
 
 The reel being filled with the wire, it may be mounted 
 on a suitable pillar and fixed to the top of a strong flat 
 box, 8in. x 8in. x 2in., the two wires being led down 
 into tbe box. 
 
 The object in using a box instead of a solid board for 
 the base is that it forms a convenient receptacle for the 
 " shunt coils," which must be prepared as follows : 
 
 Carefully measure the resistance of galvanometer 
 coil. Let us suppose it to be 1,000 ohms. It will fre- 
 quently happen that if the whole of the current we may 
 be dealing with at the time is sent through tbe galvano- 
 meter, the deflection will be so great that the spot of light 
 will be thrown off the scale. To prevent this, part of 
 the current is " shunted "—i.e., allowed to traverse our 
 derived path lying between the two terminals of the 
 galvanometer. 
 
 Fig. 53. 
 
 The proportions of the total current which will 
 under these conditions pass through the galvanometer 
 and shunt, will depend upon the ratio of their respective 
 resistances. 
 
 A short reference to the theory of shunts will not be 
 inappropriate at this point. 
 
 The joint resistance of two or more wires in "parallel" 
 is equal to the reciprocal of the sum of the reciprocals 
 of the resistances of the several wires measured, sepa- 
 rately; or, 
 
 1 
 
 E =- 
 
 1 1 1 
 - + — + — 
 r r-L r 2 
 
 If the galvanometer terminals, G, be connected by 
 a shunt, S, and if a current, C, be made up of G + S 
 G 
 
 parts, then C 
 
 G + S 
 
 will be the value of that portion 
 
 S 
 
 making S = 
 meter will be 
 
 G 
 
 n 
 
 C 
 
 G 
 
 G- 
 
 G 
 
 C 
 
 n 
 
 n-1 
 
 Hence, to reduce the galvanometer current to -th of its 
 
 n 
 
 of the current going through the shunt, and C 
 
 G + b 
 
 the value of the remaining portion going through the 
 
 galvanometer. 
 
 Obviously, if we make S = G, then the current 
 
 f O 
 
 through the galvanometer will be C — = - ; and 
 
 G + G 2 
 
 -, the current through the galvano- 
 
 former value, the shunt employed must have a re- 
 1 
 
 sistance of 
 
 n-1 
 
 th that of the galvanometer, or, 
 
 S = 
 
 G 
 
 n 
 
 Usually a set of shunts are supplied with the 
 galvanometers, whereby the current may be reduced 
 to its xjf and x^ or j^^nr part, the resistance of these 
 shunts being, according to above formula, \, i, and 
 a^-9 that of the galvanometer. 
 
 The combined resistance of a galvanometer and 
 
 any given shunt may be calculated by the method of 
 
 reciprocals, or by the simpler method depending upon 
 pa -I 
 
 the fact that B = -, or if the value of the -th 
 
 G + S n 
 
 power of the shunt used be known, by dividing the 
 
 i r 1 
 
 galvanometer resistance bv the value - or B = — . 
 8 n n 
 
 If with an unshunted galvanometer the deflection is 
 found to be too large, and a shunt has to be employed, 
 the relative value of the current C, which flowed 
 through the unshunted galvanometer to the current C x , 
 flowing after the shunt is inserted, is found by the 
 equation 
 
 C = C, x &+J. 
 
 This value — - — is called the multiplying power ol 
 
 the shunt. 
 
 Assuming the resistance of our galvanometer, G, to 
 be 1,000 ohms, as before mentioned, we find the re- 
 spective resistances of tV, nj, and ia 1 oa shunts must 
 
 G 
 
 be as follows, since S 
 
 n-1' 
 
 -o 4.1, 1 x, 4- a L 000 
 For the — shunt S = — — — 
 
 1,000 
 
 = 111 ohms. 
 
 For the — - shunt 
 
 1,000 
 100-1 
 
 1,000 in> , , 
 = ' =10 , lohms 
 99 
 
 -r, xi. 1 v, 4. a 1.000 1.000 , , 
 F or the shunt S = — -j— — - = ' =1 ohm 
 
 1,000 
 
 1,000-9 999 
 
 The wire for the shunts should be of the same metal 
 as that used for the galvanometer coil — i.e., copper — 
 but the error will not be great if German silver be used, 
 thereby reducing the length of wire required ; and since 
 space is not of so great importance here as it was 
 with the galvanometer coil, cotton-covered wire may 
 be employed. Having cut off the necessary lengths of 
 wire, and measured the resistance of the wires for the 
 shunt, bend each wire in its middle, and laying the 
 middle point of the wire on a bobbin, proceed to wind 
 
310 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 the two halves side by side, allowing a few inches of 
 the ends to hang free from the bobbin. 
 
 Screw the three bobbins by brass screws (on no 
 acoount must iron screws be employed, as they may 
 become magnetised by the currents circulating through 
 the coil and affect the needle of galvanometer) to under 
 surface of top of box, and solder the ends of the shunt 
 and galvanometer wires to a set of brass strips fixed to 
 top of box, Fig. 54. 
 
 The shape of these strips and the method of con- 
 necting up the wires will be easily understood from 
 the diagram. 
 
 A and B are two L-shaped pieces of brass, their longer 
 sides being about 4in. and the shorter lin. in length, and 
 having a semi-circular notch cut out of the ends of the 
 shorter sides, so that when they are placed in position 
 inserting a conical plug will make contact between the 
 
 Fig. 54. 
 
 two strips. On the inside of B three similar semi- 
 circular notches are cut, and opposite each notch a 
 short brass strip is placed having corresponding notches 
 cut in their ends facing B. The ends of the wires 
 belonging to the coil and shunts are soldered to the 
 bottom of the strips, as indicated in the diagram. 
 1, 2 are two terminals to attach the wires leading from 
 the other parts of the circuit. 
 
 In its normal condition (with the plug out) the whole 
 of the current goes through the galvanometer. Insert- 
 ing the plug at a, b, and c successively, J^, 5 ^, and 
 xinnr P art of the total current passes through the gal- 
 vanometer, while putting the plug in d the galvanometer 
 is " short-circuited," the whole current flowing direct 
 through the plug from 1 to 2 or 2 to 1 as the case 
 may be. 
 
 The galvanometer and shunts being satisfactorily 
 arranged, we require to mount the mirror and its 
 accompanying magnet. For the mirror, a circular 
 microscopic glass cover, about £in. in diameter, may be 
 taken and silvered by following process : 
 
 A. Dissolve 154 grains of silver nitrate in 17oz. of 
 
 distilled water. Add ammonia till precipitate first 
 formed is nearly redissolved. Filter and dilute to 
 34oz. 
 
 B. Dissolve 31 grains of nitrate of silver in 34oz. of 
 distilled boiling water. Dissolve 23 grains of Eochelle 
 salt in a little water. Add to boiling nitrate till preci- 
 pitate formed becomes grey. Filter and allow to cool. 
 
 K'4 
 
 Fio. 55. 
 
 Fig 56. 
 
 Clean glass object with nitric acid and caustic potash, 
 well washing with water before and after the potash, 
 wash in alcohol, and, lastly, well-distilled water ; place 
 in clean dish, and while still wet pour over equal quan- 
 tities of A and B. In about two hours silvering is 
 complete ; take out of dish, dry, and varnish back. 
 
 To the back of the mirror is attached, by a little 
 cement, a piece £in. long of well-magnetised watch 
 spring, the whole being suspended by a single unspun 
 silk fibre, in such a manner that the magnet hangs in a 
 horizontal position, Fig. 55. 
 
 A cardboard (or, better still, a thin brass) tube, lin. 
 long and fin. diameter, has a small hole drilled midway 
 between its ends, and the silk fibre carrying the mirror 
 and magnet is threaded through the hole and secured 
 by a bit of cement, so as to allow the mirror to vibrate 
 freely in the centre of the tube when lying in a hori- 
 zontal position, Fig. 56. 
 
 The tube carrying the mirror slides freely into the 
 circular aperture in the centre of the galvanometer. 
 
 The only thing remaining to complete our galvano- 
 meter is the controlling magnet. The object of this 
 magnet is to diminish or neutralise the effects of the 
 earth's magnetism. 
 
 A brass rod placed behind the galvanometer with 
 a ring (similar to a retort stand), capable of being 
 clamped at different positions along the rod, will 
 answer exceedingly well. The ring arm should pro- 
 ject over centre of galvanometer, and the controlling 
 magnet being placed upon it, Fig. 57, its height can be 
 varied at pleasure, and a slight movement in a hori- 
 zontal plane will serve to bring the spot of light to the 
 zero of the scale whenever necessary. 
 
MEASUREMENT. 
 
 311 
 
 The scale and lamp used with the magnetometer, 
 Fig. 41, can be employed with the galvanometer, a lens 
 being used to focus the spot of light on the scale. 
 
 Figrure of Merit of Galvanometer. 
 
 By the " figure of merit " of a galvanometer is under- 
 stood the reciprocal of the current necessary to produce 
 a deflection of 1 deg. under a given electromotive force. 
 
 Join up a Daniell cell, resistance-box, and tangent 
 galvanometer in simple circuit, Fig. 58, and unplug 
 resistance till a suitable deflection is obtained, d ; the 
 figure of merit can then be easily calculated as under. 
 
 <n 
 
 Fig. 58. 
 
 Example. — A tangent galvanometer, G, = 2 - 5 ohms, 
 gave a deflection, d, = 36 deg., when joined to a Daniell 
 cell (whose electromotive force = 1 volt, and internal 
 resistance, r, = 3'5 ohms) by means of a wire of 10 
 ohms resistance = B. What is its figure of merit ? 
 
 Since the current necessary to produce a deflection 
 of 36 deg. 
 
 C = = t: — ~ — =-r^ = '0625 ampere, 
 
 r + B+C 35 + 10 + 2-5 16 * ' 
 
 .'. the current required to give a deflection of 1 deg. 
 
 •0625 tan 1 deg. = ;0175 x . Q625 = . 0Q15 
 
 tan 36 deg. -7265 ^ 
 
 and its figure of merit is — = 
 
 ■0015 
 
 -666. 
 
 Another way of regarding the figure of merit of a gal- 
 vanometer is to ascertain through what resistance an 
 electromotive force of one Daniel cell (practically 1 volt) 
 will produce a deflection of 1 deg. In the above example 
 this would be 666 ohms. 
 
 With a sensitive mirror galvanometer the deflection 
 due to the current from one Daniell, with all the avail- 
 able resistance in circuit, will generally be sufficient to 
 throw the spot of light off the scale. In this case the 
 galvanometer must be shunted, and the resistance, E 1( 
 in current calculated according to following formula. 
 Let r = integral resistance of cell. 
 B = interposed resistance. 
 G = galvanometer resistance. 
 S = resistance of shunt. 
 d = observed deflection in scale division. 
 Then to get the amount of resistance necessary to 
 
 obtain a deflection of 1 deg. we must have, if the gal- 
 vanometer be unshunted, 
 
 E 1 = d(r+B + G) 
 
 Since it is obvious that if a given resistance, E, we 
 get a deflection of d divisions, the resistance necessary 
 to reduce this to one division must be d times as great. 
 
 Should a shunt, S, have to be employed the formula 
 becomes 
 
 G + S 
 
 E = d 
 GS 
 
 G + S / , ^ ^ GS x 
 
 and ^ ° being the multiplying power and 
 S G + S 
 
 the resistance of the shunted galvanometer respectively. 
 
 With a mirror galvanometer, r will be so small in 
 comparison to the high resistance in circuit that it 
 may be neglected. 
 
 Example. — A Thomson galvanometer, G, of 6,000 
 ohms resistance, and an interposed resistance of 9,000 
 ohms (r being neglected) gave a deflection of 330 
 scale divisions when joined up with one Daniell cell = 
 1 volt. What was its figure of merit ? 
 
 E 1 = d(E + G) 
 
 = 330 (9,000 + 6,000) 
 = 330 x 15,000 
 = 4,950,000 ohms. 
 
 Example. — Using the same galvanometer and cell, 
 but inserting the — shunt S = 666 ohms, and an in- 
 terposed resistance of 14,400 ohms, d = 33 scale 
 divisions, therefore 
 
 1 S \ G + S/ 
 
 _ 33 6,000 + 666 / 14 4Q0 6,000 x 666 \ 
 666 V ' 6,000 + 666/ 
 
 = 33 x 10 (14,400 + 600) 
 = 330 x 15,000 
 = 4,950,000 ohms. 
 
 N.B. — The value of — - — is simply the reciprocal 
 S 
 
 of the denomination of the shunt — = 10 ; and the 
 
 value of x = — (seepage 309). 
 G + S n 
 
 Determination of the Reduction Factor of a Tangent 
 Galvanometer. — By the reduction factor of the galvano- 
 meter is understood that constant by which the value 
 of the angle of deflection obtained must be multiphed 
 in order to obtain the value of the current in absolute 
 measure, or in amperes. 
 
 Supposing the value of H at the place of experiment 
 to be known, and also the mean radius, r, of the coil, 
 and the number of complete turns, n, made by the 
 wire, then the reduction factor, K, can be calculated 
 from the following formula, 
 
 K = 
 
 2 wn 
 
312 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Example. — A certain tangent galvanometer, whose 
 coil consisted of 40 = n complete turns of wire, the 
 mean radius, r, being 1175 cm. Assuming the value 
 of H to be '176 dyne. Eequired, the reduction factor in 
 absolute units. 
 
 11-75 x 176 2-068 
 
 Answer. — K 
 
 2 x 3-14159 x 40 251-3272 
 
 •008228 in absolute units, or "008228 x 10 = "08228 in 
 amperes. 
 
 Hence the value of any current (in amperes) producing 
 a given deflection, d, with this particular instrument 
 would be 
 
 C = K tan d = "08228 tan d 
 
 If the dimensions of the galvanometer are not known, 
 its reduction factor maybe ascertained by the following 
 method, which depends upon a knowledge of the 
 quantity of a metal liberated from a solution of one of 
 its salts by a known quantity of electricity — i.e., the 
 electro-chemical equivalent of the metal : 
 
 Take two pieces of clean sheet copper about 15 cm. 
 x 3 cm., and place them between three dry strips 
 of wood, with their copper ends projecting sufficiently 
 above the wood strips to allow of a wire being attached 
 to each, by means of a clamp, the whole being held 
 firmly together by two strong indiarubber bands. 
 
 Eest the wooden support on the top of a beaker con- 
 taining a saturated solution of copper sulphate (CuSOJ 
 so that from 6 cm. to 8 cm. of the copper plates are 
 immersed in the liquid ; join up with a good Daniell 
 cell, a key, and the tangent galvanometer whose reduc- 
 tion factor is to be determined. 
 
 Before connecting up, the copper plate which is to 
 serve as the cathode — i.e., the one connected with the 
 sine pole of the battery, must be carefully weighed on a 
 good balance, and its weight in grammes = W X recorded. 
 After ensuring that the galvanometer is properly 
 adjusted, close the circuit and read the deflection of 
 the pointer at intervals of 10 minutes for one hour, and 
 take the mean of the six readings as the true value of 
 the deflection, S. 
 
 By taking these readings at short intervals, any 
 variation in the strength of the current will be detected 
 and the average current estimated. At the expiration 
 of the hour disconnect the copper plates, and after 
 carefully washing the cathode, first with distilled water 
 and then with alcohol, and lastly with a little ether (by 
 this treatment the plate dries quicker and there is less 
 danger of the copper becoming oxidised) it must be 
 placed at once into an air bath or oven, and exposed to 
 a temperature of about 100 deg. C, until it is thoroughly 
 dried. After cooling under a dessicater, or if this is not 
 available, under an inverted beaker, the cathode is to 
 be again carefully weighed ; call this W 2 . 
 
 Let W = weight of metal liberated, or W 2 - Wj. 
 G = current in amperes. 
 
 e = electro-chemical equivalent of metal (copper) 
 t = time. 
 6 = deflection, 
 
 If W x be weight of cathode at commencement, and 
 W 2 be weight of cathode at end of experiment, we have 
 
 W r W 1 = W= e CT 
 
 and C = W 
 
 e t 
 
 But C = Ktan(5, 
 
 whence the reduction factor, K, is known, lot 
 
 tan S 
 
 Example. — With the galvanometer referred to in last 
 example the following data were obtained : 
 
 W 2 = 10-383 e = "00033 
 
 Wj = 10-285 t = 60 x 60 = 3,600 seconds 
 
 W = -098 8 = 45 deg. 30' (mean of six readings) 
 
 Then W = C e t 
 
 ■098 = C x -00033 x 3600 
 
 = C x 1-188 
 
 . r _ W -098 - QO _ 
 
 . . O = — = — — = 0H25 ampere. 
 
 But we saw above that C = K tan S, and substitut- 
 ing for C the value -0825, we get 
 
 C = -0825 = K tan 45-30 
 
 = K x 1-0176 
 
 ■ K= ' 0825 
 1-0176 
 
 = -081, 
 
 instead of 08228 as calculated from the known dimen- 
 sions of the instrument. 
 
 Arrange battery power and resistance so that the 
 deflection is as near 45 deg. as possible— as that is the 
 angle of maximum sensitiveness. 
 
 Eleetro-Chemieal Equivalent. 
 
 One of the phenomena connected with electrical 
 currents is the decomposition of chemical solutions, and 
 the weight of a substance in solution decomposed by 
 the passing of one ampere for one second is called the 
 electro-chemical equivalent of the substance. The 
 following table gives the electro-chemical equivalent of 
 the more important substances usually dealt with : 
 
 Electro-Chemical Equivalents. 
 
 Substance. Electro-chemical equivalent in 
 
 grammes per ampere per second. 
 
 Hydrogen 0"00001035 
 
 Oxygen 0-00008282 
 
 Chlorine 0-0003675 
 
 Sodium 0-0002381 
 
 Gold 0-0006780 
 
 Silver 0-0011180 
 
 Copper 0-0003261 
 
 Zinc 0-0003364 
 
 Lead 0-0010684 
 
 This table is given not because we believe in its 
 accuracy to even a fifth decimal place, but because it 
 
MEASUREMENT. 
 
 313 
 
 shows a most curious example of the desire of modern 
 science to get exact numerical values in its investiga- 
 tions, whenever possible. The figures can only be 
 taken as approximately accurate. 
 
 Experimental Illustrations of Ohm's Law. 
 
 The relation to each other of the three quantities 
 which constitutes Ohm's law may be verified by the 
 following simple experiments : 
 
 LetC = ^, 
 
 where C = current, E electromotive force, E = resistance 
 of cell and external resistance. 
 
 First Case. — E remaining constant, varies directly 
 as the electromotive force. 
 
 Join up in circuit, Fig. 58, a good Daniell cell, a 
 Thomson galvanometer of known resistance, and a 
 box of resistance coils, and note deflection, d, obtained. 
 
 A tangent galvanometer might be employed, but in 
 this case the internal resistance of the cell must be 
 ascertained by any of the methods to be subsequently 
 described, as, owing to the resistance of an ordinary 
 tangent galvanometer being generally low, the internal 
 resistance of a Daniell cell becomes too great a factor 
 of the total resistance to be neglected. Whereas, with 
 a Thomson's high-resistance galvanometer, the internal 
 resistance of the cell is negligible. 
 
 Increase the battery power from 1 to 2 and 3 cells 
 when the values of the deflections will become 2d, 3d, 
 etc., or, with a tangent galvanometer, tan d becomes 
 2 tan d, 3 tan d, etc., and since these values are propor- 
 tional to the currents producing the deflection, we have 
 E„ 
 
 C« — 
 
 E 
 
 Example. — With a Thomson galvanometer of 6,000 
 ohms shunted down ^u, the total resistance being 5,000 
 ohms, one Daniell cell gave a deflection of 36 scale divi- 
 sions, while two cells gave 71"5 divisions. 
 
 Second Case. — E remaining constant, C a inversely 
 as E. With one cell and a total resistance, E, in circuit, 
 note deflection, d. Increase the resistance to 2 E, and 
 subsequently to 3 E, when the deflections will be reduced 
 
 to — and — respectively. 
 
 Example. — With the same shunted galvanometer and 
 one Daniell cell, with 5,000 ohms in circuit the deflection 
 was as before — 36 ; on doubling E to 10,000 the deflec- 
 tion fell to 17 8. 
 
 Third Case. — The current remaining constant, the 
 electromotive force varies directly as the resistance. 
 
 Note deflection, d\, obtained with one cell and re- 
 sistance, E, in current. On introducing two and three 
 cells, we must, in order to keep C constant, increase E 
 to 2 R - 3 E respectively, or 
 
 C = 
 
 «E 
 
 Example. — Again employing the same galvanometer 
 and cells, one Daniell yielding through 5,000 a deflection 
 of 36 ; when two cells were used in series, the value of E 
 
 had to be increased to 10,000 in order to maintain the 
 same deflection. 
 
 Should the deflection produced in either of the pre- 
 ceding measurements be so great that the spot of light 
 moves off the scale, the galvanometer may be shunted 
 with one of the sets of shunts supplied with it, as in 
 the examples given ; in this case the resistance of 
 galvanometer and shunt combined will be equal to 
 
 their product divided by their sum E = -^ ^, which is 
 
 Gr + D 
 
 equivalent to the 
 
 Eesistance of the galvanometer 
 
 Eesistance of the multiplying power of shunt n 
 Measurement of Galvanometer Resistance. 
 
 We shall describe four methods of measuring the 
 resistance of a galvanometer, from which the student 
 
 Fig. 59. 
 
 can select the method most suitable to the kind of 
 instrument he has to deal with. We shall assume that 
 he has provided himself with an ordinary astatic 
 galvanometer to use with the bridge. This need only 
 have a coil composed of a few turns of comparatively 
 thick wire, the resistance of which will be so small that 
 it need not be known, but the resistance of the tangent 
 
 GX 
 
 Fie. 60. 
 
 and mirror galvanometer ought to be accurately deter- 
 mined. 
 
 A. Direct Bridge Method.— Place the galvanometer 
 in the fourth arm of the AYheatstone bridge, and 
 measure it as an ordinary resistance, Eigs. 59 and 60. 
 This is the only trustworthy method for low-resistance 
 galvanometers. 
 
 Example. — A tangent galvanometer, whose resistance 
 being required, was placed in the fourth arm of the 
 Post Office bridge, and on making B = 100 and A = 10, 
 
314 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 it was necessary to make D = 280 before balance was 
 obtained. Wbat was the resistance ? 
 
 x-d| = 
 
 280 x 10 
 100 
 
 = 28 ohms. 
 
 A mirror galvanometer was substituted for the 
 tangent galvanometer, and making B and A 1,000 ohms 
 each, D had to be made 6,000 ohms. Eequired its 
 resistance. 
 
 6,000 x 1,000 
 
 X 
 
 1,000 
 
 = 6,000 ohms. 
 
 B. Half Deflection Method. — Join up the galvano- 
 meter with a cell of low internal resistance, and a 
 box of resistance coils, Fig. 58. If the resistance 
 of the galvanometer be low, the resistance of the cell 
 should be known, as it may form an appreciable part 
 of the total resistance in circuit ; in the case of a high 
 resistance galvanometer it may be neglected. Note 
 deflection, d v produced with a resistance, E 1( unplugged 
 in box. Increase E : to B 2 , so that the new deflection, 
 d it is equal to half d^ Then, 
 
 G = B 2 - 2 E x . 
 Example. — The tangent galvanometer before men- 
 tioned, using a cell of very low resistance, gave a 
 deflection of 47 deg. (tan 47 deg. = 1-07) when Ej = l 
 ohm. In order to reduce this deflection to one-half, or 
 28 deg. (tan 28= -53) E 2 had to be made 30 ohms. 
 .-. G = B 2 -2 Ej. 
 = 30-2x1. 
 = 28 ohms. 
 
 Fig. 61. 
 
 The mirror galvanometer, in order to keep the spot 
 
 of light on the scale, had to be shunted down to - 
 
 10' 
 Its actual resistance wduld therefore be i* = 6 ' 000 - 
 600 ohms (see page 309). 
 The value obtained by this method was 
 B 1 = 5,000 ohms. d\ = 64 scale divisions. 
 
 E 2 = 10,600 ohms. 
 
 d,-32 
 
 .". G = 10,600-2 x 5,000 
 = 600 ohms. 
 Note.— Make Bj and E 2 as low as possible but 
 remembering that with the tangent galvanometer the 
 best results are obtained when d, and d, have equal 
 
 values on either side of 45 deg. — i.e., tan 55 deg. and 
 tan 35 - 5 deg. 
 
 The connections are shown theoretically in Fig. 58 
 and practically in Fig. 61, using the Post Office bridge. 
 The arms, A and B, play no part in the arrangement 
 only the set of resistance coils, D, being utilised ; in fact 
 the bridge becomes a simple resistance-box. The in- 
 finity plug, I P, being removed. 
 
 C. Equal Deflection Method. — Join up galvanometer 
 cell of low resistance, and resistance, E lf as in last 
 method, only interposing a shunt, S, between the 
 terminals of the galvanometer, G. Note deflection, d. 
 
 Fig. 62. 
 
 The shunt, S, is now removed, causing an increased 
 deflection, d 2 . Increase B L to E 2 , so that d\ is again 
 produced, then 
 
 ~ET~ - 
 
 With the Post Office resistance-box, to insert the 
 shunt, S, put in the infinity plug, I P, and remove one 
 of the plugs between B and A. To remove S we have 
 simply to unplug I P, Figs. 62 and 63. 
 
 Make S as small and E : as large as possible. 
 
 Example.— By this method the tangent galvanometer 
 gave the following data : d = 16 deg., E t = 20 ohms, 
 E 2 = 77 ohms, and S = 10 ohms. 
 
 .-. G = S Ba ~ B i, 
 
 77-20 
 20 ' 
 57 
 20' 
 
 = 28-5 ohms, 
 and the mirror galvanometer, shunted down to — as 
 
 before, d = 20 divisions. E x = 1,000 ohms. E 2 = 7,100 
 ohms, and S = 100 ohins. 
 
 = 10 
 = 10- 
 
MEASUREMENT. 
 
 315 
 
 . G = 1(?0 7,100 - 1,000 
 1,000 
 
 = 100 
 
 6,100 
 
 1,000 
 
 = 100 x 61 
 
 = 610 ohms. 
 
 Thomson's Method. — Connect up galvanometer in 
 fourth arm of bridge, and place battery and key, K 2 , 
 between the points AC. Join up B E through ~K lt 
 unplug resistance in A and B, and adjust resistance d 
 till no alteration in the deflection of the mirror is pro- 
 duced either by opening or closing K x , Figs. 64 and 65. 
 
 Fig. 64. 
 
 With the Post Office bridge, hold the left-hand key 
 firmly down to bring in the battery, and having given 
 a suitable ratio to b d, adjust d, closing Kj at intervals. 
 "When the proper value has been given to d, closing E^ 
 will produce no effect on the galvanometer. Attention 
 must be directed to one very important point. 
 
 Make B greater and A less than G. 
 With a slide wire bridge the connections would be 
 as follows : Make A small and D large, Fig. 66. 
 
 Example. — In measuring the resistance of Thomson 
 galvanometer shunted down ^. With b = 1,000 and 
 a = 100 ohms, it was necessary to make d = 6,090 
 ohms. 
 
 . n 6,090 x 100 anQ , 
 
 • • G = 1,000 = 609ohms - 
 
 Fig. 65. 
 
 Example. — With a slide wire bridge, and tangen 
 galvanometer, no effect was produced on needle when 
 A was 25 ohms, and contact made at the 470 division 
 = B. 
 
 . AD = 25 x 53 
 B 470 
 
 G 
 
 = 28-2 ohms. 
 
 Fig. 66. 
 
 Since on closing K^ and bringing in the current the 
 mirror is permanently deflected, the spot of light 
 moving off the scale, some means must be employed to 
 counteract this effect of the current. This is best secured 
 by placing a bar magnet on one side of the galvanometer, 
 and at such a height and distance from it that the attrac- 
 tion due to the magnet exactly balances the repulsion 
 arising from the current ; when this is done the spot 
 will return to zero, the magnet of the galvanometer being 
 at the same time exceedingly sensitive to any alteration 
 in the strength of the current. 
 
 Then, as in ordinary bridge measurements of resist- 
 ances, 
 
 AD 
 B- 
 
 G = : 
 
 Summary. 
 
 Of the above four methods, the first with the Wheat- 
 stone bridge is to be preferred when a second galvano- 
 meter is at hand ; when such is not the case, and 
 especially when measuring a high-resistance galvano- 
 meter, Thomson's method is very satisfactory. Since, 
 as in ordinary bridge determinations, it is a null 
 method, the measurement being taken when the 
 needle is undeflected ; it is therefore independent of 
 the values of the graduations of the galvanometer 
 scale, and also of the internal resistance of the battery. 
 
 Of the remaining two methods (B and C), that of 
 '* equal deflection " has the advantage of being applic- 
 able to any kind of galvanometer, since we have only 
 to reproduce a given deflection, therefore its value 
 
316 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 need not be known, whereas with " half deflection " 
 method is restricted to those instruments in which 
 the values of the different deflections are comparable. 
 In either case the internal resistance of the battery 
 must be exceedingly small in comparison to that of the 
 galvanometer. 
 
 For this reason they are not adapted for measuring 
 low-resistance galvanometers ; in fact, the only reliable 
 method for this class of instruments is that first 
 described with the Wheatstone bridge. 
 
 We shall see that the tangent galvanometer as 
 described is very little used in the ordinary work of 
 electric lighting, but the principle of the tangent galva- 
 nometer underlies many, if not most, of the instruments 
 used. In many operations it is not of very great im- 
 portance to read resistance, say, to the hundredth of an 
 
 Standard Direct Reading 1 Electric Balances. 
 
 These instruments are founded on the mutual forces, 
 discovered by Ampere, between movable and fixed 
 portions of an electric circuit. The shape chosen for 
 the mutually-influencing portions is circular, and each 
 such part will be called for brevity an ampere ring ; or 
 sometimes simply a ring, whether it consists of only 
 one turn or of any number of turns of the conductor ; 
 or an arc when it consists of less than a whole turn. 
 
 In each of the balance instruments, except the kilo- 
 ampere balance, each movable ring is actuated by two 
 fixed rings, all three approximately horizontal. There 
 are two such groups of three rings — two movable rings 
 attached to the two ends of a horizontal balance arm 
 pulled, one of them up and the other down, by a pair 
 of fixed rings in its neighbourhood. The current is in 
 
 Fig. 67. — Centi-ampere Balance. 
 
 ohm — it is sufficiently practical to read to tenths ; the 
 same may also be said with regard to amperes and 
 volts. Hence it will be found that the practical instru- 
 ments are often of such a character that they most 
 admirably fall in with this laxity of accuracy. They are 
 only approximate. A well-known electrical engineer 
 goes even further than this, and maintains that the 
 best-known makers are not too particular in measuring 
 the accuracy of the resistances in the resistance-boxes 
 they send out. Be that as it may, the user, when 
 necessary, ought to be able with a Kirchoff ' bridge 
 and a good galvanometer to ascertain for himself 
 whether his apparatus is accurately described or 
 not. 
 
 We have already said that to Sir W. Thomson more 
 than to anyone we are indebted for the great advance 
 that has taken place since 1850 in electrical measuring 
 instruments, and we cannot do better than to briefly 
 and rapidly describe the most recent forms of his 
 standard instruments. 
 
 opposite directions through the two movable rings to 
 practically annul disturbance due to horizontal com- 
 ponent of terrestrial or local magnetic forces. In the 
 kilo-ampere balance the whole current passes through a 
 single fixed ring and divides through two halves of a 
 movable ring, which are urged one up and the other 
 down by the resulting amperian force. 
 
 In all the instruments the balance arm is supported 
 by two trunnions, each hung by an elastic ligament of 
 fine wire, through which the current passes into and 
 out of the circuit of the movable rings or ring. 
 
 In all the balance instruments, in which the movable 
 ring is between two fixed rings, the mid-range position 
 of each movable ring is in the horizontal plane nearly 
 midway between the two fixed rings which act on it. 
 The current goes in opposite directions through the 
 two fixed rings, so that the movable ring is attracted 
 by one of the fixed rings and repelled by the other. 
 The position of the movable ring equidistant from the 
 two fixed rings is a position of minimum force, and the 
 
MEASUREMENT. 
 
 317 
 
 sighted position, for the sake of stability, is above it at 
 one end of the beam and below it at the other, in each 
 case being nearer to the repelling than to the attracting 
 ring by such an amount as to give about -^ per cent, 
 more than the minimum force. 
 
 In the balance instruments to measure alternate 
 currents (which may be also used for direct currents) 
 of from 1 ampere to 600 amperes the main current 
 through each circle, whether of one turn or of more 
 than one turn, is carried by a wire rope of which each 
 component wire is insulated by silk covering, or other- 
 wise, from its neighbour, in order to prevent the induc- 
 tive action from altering the distribution of the current 
 across the transverse section of the conductor. 
 
 The balancing is performed by means of a weight 
 which slides on an approximately horizontal graduated 
 arm attached to the balance ; and there is a trough 
 fixed on the right-hand end of the balance into which a 
 proper counterpoise weight is placed, according to the 
 particular one of the sliding weights in use at any time. 
 
 from a hook carried by a sliding platform, which is 
 pulled in the two directions by two silk threads passing 
 through holes to the outside of the glass case. 
 
 Four pairs of weights (sliding and counterpoise), of 
 which the sledge and its counterpoise constitute the 
 first pair, are supplied with each instrument. These 
 weights are adjusted in the ratios of 1 : 4 : 16 : 64, so 
 that each pair gives a round number of amperes, or half- 
 amperes or quarter-amperes, or of decimal sub-divisions 
 or multiples of these magnitudes of current, on the 
 inspectional scale. 
 
 The useful range of each instrument is from 1 to 100 
 of the smallest current to which its sensibility suffices. 
 The ranges of the different types of this instrument 
 regularly made are 
 
 I. Centi-ampere balance : from 1 to 100 centi-am- 
 peres (Fig. 67). 
 
 II. Deci-ampere ,, ,, 1 to 100 deci-am- 
 
 peres. 
 
 Fig. 68. — Deka-ampere Balance. 
 
 For the fine adjustment of the zero a small metal flag 
 is provided, as in an ordinary chemical balance. This 
 flag is actuated by a fork, having a handle below the 
 case outside, as shown in the drawing, Fig. 67. To 
 set the zero the left-hand weight is placed with its 
 pointer at the zero of the scale, and the flag is turned 
 to one side or the other until it is found that, with no 
 current going through the rings, the balance rests in 
 its sighted position. 
 
 To measure a current, the weight is slipped along the 
 scale until the balance rests in its sighted position. 
 The strength of the current is then read off approxi- 
 mately on the fixed scale (called the inspectional scale) , 
 with aid of the finely-divided scale for more minute 
 accuracy, according to the explanations given below. 
 Each number on the inspectional scale is twice the 
 square root of the corresponding number on the fine 
 scale of equal divisions. 
 
 The slipping of the weight into its proper position is 
 performed by means of a self-releasing pendant, hanging 
 
 ILL Deka-ampere (Fig. 68) from 1 to 100 amperes. 
 IV. Hekto-ampere balance ,, 6 to 600 ,, 
 V. Kilo-ampere ,, ,, 25 to 2,500 ,, 
 
 VI. Composite „ „ "02 to 500 „ 
 
 and from 100 to 50,000 watts (at 
 100 volts). 
 The following table shows for each type of instru- 
 ment the value per division of the inspectional scale 
 corresponding to each of the four pairs of weights : 
 
 I. II. in. IV. V. 
 
 Centi- Deci- Am- Am- Am- 
 
 am- am- peres peres peres 
 
 peres peres per per per 
 
 
 per divi- 
 
 per divi- divi- 
 
 divi- 
 
 divi- 
 
 
 sion. 
 
 sion, sion. 
 
 sion. 
 
 sion. 
 
 1st pair of weights.. 
 
 . 25 
 
 •25 '25 
 
 15 
 
 5 
 
 2nd 
 
 ■50 
 
 ■5 -5 
 
 3-0 
 
 10 
 
 3rd 
 
 10 
 
 1-0 i-o 
 
 60 
 
 20 
 
 4th 
 
 2-0 
 
 2-0 2-0 
 
 12-0 
 
 50 
 
 The fixed inspectional scale shows, approximately 
 enough for most purposes, the strength of the current ; 
 
318 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 the notches in the top of the aluminium scale show 
 the precise position of the weight corresponding to 
 each of the numbered divisions on the fixed scale, 
 which practically annuls error of parallax due to the 
 position of the eye. When the pointer is not exactly 
 below one of the notches corresponding to integral 
 divisions of the inspectional scale, the proportion of 
 the space on each side, to the space between two 
 divisions, may be estimated inspection ally with 
 accuracy enough for almost all practical purposes. 
 Thus we may readily read off 34 - 2 or 34'7 by esti- 
 mation with little chance of being wrong by 1 in 
 the decimal place. But when the utmost accuracy 
 is required, the reading on the fine scale of equal 
 divisions must be taken, and the strength of current 
 calculated by aid of the table of double square roots 
 appended to these instructions. Thus, for example, if 
 the reading is 292 we find 34 - 18, or, say, 34'2 as the 
 true scale reading for strength of current ; or, again, if 
 the balancing position of the pointer be 301 on the fine 
 scale, we find 3470 as the true reading of the inspec- 
 tional scale. 
 
 The centi-ampere balance, with a thermometer to 
 test the temperature of its ampere rings, and with 
 platinoid resistances up to 1,600 ohms, serves to 
 measure potentials of from 10 volts to 400 volts. 
 
 Constant op the Centi-Ampeee Balance when 
 used as a voltmetee. 
 
 Weight Resistance in Volts per division 
 
 used, circuit.* of fixed scale. 
 
 1st pair of weights 400 1-0 
 
 800 2-0 
 
 1,200 3-0 
 
 1,600 4-0 
 
 * Including resistance of the instrument, which is about 50 ohms. 
 
 If the second pair of weights is used the constants will 
 be double of those noted above. 
 
 When using, the instruments should be levelled in 
 accordance with the indications of the spirit-level 
 attached, by means of the levelling screws on which 
 the sole-plate of the instrument stands. 
 
 In the centi and deci-ampere balances, the beam can 
 
 be lifted off its supporting ligaments by turning a handle 
 
 attached to a shaft passing under the sole-plate of the 
 
 instrument. This shaft carries an eccentric, on the 
 
 edge of which rests the lower end of a vertical rod 
 
 which is fixed at its upper end to a tripod lifter. When 
 
 the instrument is to be removed from place to place. 
 
 the lifter should be raised ; but when it is fixed up for 
 
 regular use it is advisable to keep the beam always 
 
 hanging on the ligaments. In the deka, hekto, and 
 
 kilo-ampere balances there is no lifter, but the beam is 
 
 packed by placing distance-pieces between the two 
 
 halves of the suspension trunnions and screwing them 
 
 together by means of milled-headed screws provided for 
 
 the purpose. When the instrument is unpacked and 
 
 prepared for use, these distance-pieces must be taken 
 
 out and placed in receptacles provided for them in the 
 
 weight box. 
 
 A set of four sliding weights, of which the carriage 
 constitutes one, is supplied with each instrument. The 
 carriage is fitted with an index to point to the movable 
 scale, and is intended to remain always on the rail. 
 One or other of the weights is to be placed on the 
 carriage in it in such a way that the small hole and slot 
 in the weight pass over the conical pins. The weights 
 are moved by means of a slider, which slides on a rail 
 fixed to the sole-plate of the instrument, and carries a 
 pendant with a vertical arm intended to pass up 
 through the rectangular recess in the front of the 
 weight and carriage. The slider and weight are shown 
 in position in the figures. The slider is moved by silk 
 cords which pass out at the ends of the glass case. 
 When the cords are not being pulled for shifting the 
 weight their ends should be left free, so that the 
 pendant may hang clear of the weight. When a 
 weight is to be placed on or removed from the carriage 
 the slider should be drawn forward at the top until it 
 is clear of the weight, and then pushed to one side until 
 the weight is adjusted, when it may be replaced in 
 position in a similar manner. 
 
 Cylindrical counterpoise weights with a cross-bar 
 passed through them are supplied for the purpose of 
 balancing the sliding weights when they are placed at 
 the zero of the scale. The sliding weight should be 
 placed so that the index of the carriage points to the 
 zero of the scale, and the proper counterpoise weight 
 should be placed in the trough, fixed to the right-hand 
 end of the beam, with its cross-bar passing through the 
 hole in the bottom of the trough. The flag which is 
 attached to the cross trunnion of the beam should then 
 be turned by means of the handle projecting from 
 under the sole-plate, until the index on the end of the 
 movable scale points to the middle one of the five 
 black lines on the fixed scale opposite to it. Care must 
 be taken when making this adjustment that the fork 
 which moves the flag is not left in contact with it, as 
 this would impede the free swing of the beam. The 
 fork should be turned back a little after each adjust- 
 ment of the flag, and, when the flag is being adjusted, 
 it is better to watch the flag itself, and make successive 
 small adjustments until the beam stands at zero, than 
 to make successive trials by pushing round the handle 
 while watching the position of the index. 
 
 If the ligament has stretched since the instrument 
 was standardised, the index at one end of the movable 
 scale will be found to be below the middle line on its 
 vertical scale, when the index at the other end is 
 correctly pointing to the zero position. The error so 
 introduced would be a small one, but it may be easily 
 put right by slightly loosening the screws fixing the 
 pillared frame, which supports the movable beam, to 
 the base-plate, and raising it by slipping one or two 
 thicknesses of paper below it until the indices simul- 
 taneously point to their zero position. 
 
 A lens is supplied with each instrument for facili- 
 tating accurate observation either when reading the 
 position of the weight or adjusting the zero. 
 
 The vibrations of the beam may be checked so as 
 
ME A S UREMENT. 
 
 319 
 
 to facilitate reading by bringing the pendant, whicb 
 moves the weight, Ughtly into contact with it in such 
 a way as to give a little friction without moving the 
 weights. 
 
 In using the centi-ampere balance as a voltmeter, 
 when great accuracy is required, care must be taken 
 that the effect of change of temperature in changing 
 the resistance of the coils of the instrument, and of the 
 external resistance coils, is allowed for ; and in this 
 use of the instrument it is advisable to employ currents 
 such as can be measured by the lightest weight on the 
 beam. When the instrument is to be used as a volt- 
 meter, four resistances are provided, three of which 
 are each 400 ohms, and the fourth is less than 400 
 ohms by the resistance of the coils of the instrument 
 at a certain specified temperature. The smallest 
 resistance is intended to be included by itself in the 
 circuit when the lowest potentials are being measured, 
 and in series with one or more of the others when the 
 potential is so high as to give a stronger current than 
 can be measured with the lightest weight on the beam. 
 The correction for temperature is, for the copper coils 
 of the balance about - 38 per cent, per degree centi- 
 grade, and for the platinoid resistances, about 0"024 
 per cent, per degree centigrade. 
 
 Composite Balance. — The instrument is similar in 
 form to the centi and deci-ampere balances, but the 
 pair of fixed coils at one end of the beam are made of 
 a rope of insulated wires similar to that used for the 
 coils of the hekto-ampere balance. Separate electrodes 
 are provided for the rope coils, and for the fine wire 
 coils. A switch which allows the movable coils either 
 to be included in the circuit by themselves or in series 
 with the fixed fine wire coils, is attached to the under 
 side of the sole-plate of the instrument. When the 
 handle of the switch is turned to "Watt," the movable 
 coils alone are in the circuit ; but when the handle is 
 turned to "Volt," both the movable and the fixed fine 
 wire coils are in the circuit. 
 
 The composite balance can be used as hekto-ampere 
 balance, or as a wattmeter, or as a voltmeter, by 
 following the instructions given below. 
 
 In using this balance it should, like the standard 
 balances, be levelled and the stop screws turned back 
 out of contact with the cross trunnion and the front 
 plate of the beam so as to leave it free to oscillate. 
 
 To use the instrument as a centi-ampere meter or as 
 a voltmeter the switch is turned to " Volt," and one or 
 other of the weights marked V.W l; V.W 2 , V.W 3 , used. 
 The current flowing through the instrument is then to 
 be calculated from the constant given in the certificate 
 sent with the instrument. The volts on the terminals 
 are calculated from the current in amperes and the 
 resistance in ohms (including the anti-inductive resist- 
 ance, if any) in circuit. If V be volts, C current, and K 
 resistance, 
 
 V = CE. 
 
 The anti-inductive resistance is arranged so that the 
 instrument reads a round number of volts per division. 
 
 To use the instrument as a hekto-ampere meter the 
 switch is turned to " Watt " and the thick wire coils 
 inserted in the current circuit in such a way that the 
 right-hand end of the beam is repelled up. Either the 
 sledge alone or the weight marked W W is to be used in 
 this case. A measured current is then passed through 
 the suspended coils and the constants given in the 
 certificate for the balance used in this way are calculated 
 on the assumption that this current is, as there stated, 
 0'25 amperes. Any other current which is convenient 
 in the circumstances may be used, and if this current 
 be C amperes the corresponding constant is obtained 
 by multiplying the constant given in the certificate by 
 C/'25 or 4 C. The current through the suspended coils 
 may be measured by means of the instrument itself 
 arranged for the measurement of volts. This may be 
 done by first measuring the current which the differ- 
 ence of potential between the supply conductors of an 
 electrical installation, or between the poles of a battery, 
 causes to flow through the coils of the instrument and 
 its external resistance, and then turning the switch to 
 " Watt," and at the same time introducing a resist- 
 ance into the circuit equal to the resistance of the fixed 
 coils. 
 
 To use the instrument as a wattmeter one terminal 
 of the fine wire coils is joined to one end of the anti- 
 inductive resistance and the other terminal to one of 
 the leads, the other end of the resistance being joined 
 to the other lead. The thick wire coils are inserted in 
 the main circuit as described in the previous paragraph. 
 With the instrument thus joined up, the current through 
 the suspended coils and the electromotive force between 
 the leads may be obtained by the operations described 
 two paragraphs above, since the presence of the thick 
 wire coil in the circuit causes no appreciable error ; or 
 the electromotive force may be taken from the electro- 
 static voltmeter used on the circuit, and from its indica- 
 tions the current in the suspended coil circuit calculated. 
 The watts are then to be calculated from the electro- 
 motive force on the leads and the current through the 
 thick wire coils by the formula 
 
 W = VC = cCE, 
 
 where c is the current in the suspended coil circuit, C 
 the current in the thick wire coils, and E the resistance 
 in the circuit. 
 
 The weights sent out with the instrument are 
 arranged to give a round number of watts per division 
 of the scale with a known anti-inductive resistance in 
 series with the fine wire coils. 
 
 Portable or Marine Voltmeter. 
 For the measurement of potential in connection with 
 electric lighting or power installation on board ship, 
 the mass of the moving part of the balance voltmeter 
 and engine-room voltmeter is too great to be convenient 
 for accurate use. The marine voltmeter now to be 
 described is specially suitable for such a purpose, but 
 it is also equally useful as a portable voltmeter for 
 general use, Figs. 69 and 70. 
 
3i>0 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 The instrument consists of a small oblate of soft 
 iron supported on a stretched wire in the centre of a 
 solenoid of fine copper wire connected in series with 
 platinoid resistances, variable according to the poten- 
 tial to be measured; and is founded on the principle 
 that an oblate spheroid of soft iron, movable round a 
 diameter, tends to turn its equatorial plane parallel to 
 the lines of force in a uniform magnetic field. The 
 pointer is fixed relatively to the oblate in such a 
 manner that, when the pointer is at the zero position 
 of the scale, the equatorial plane of the oblate is 
 inclined about 45 deg. to the lines of force of the 
 solenoid. 
 
 The suspending wire is stretched between the two 
 ends of a brass tube, being fixed at the bottom end 
 and carried at the upper end by a torsion head, which 
 
 The resistances to enable the instrument to be used 
 as a voltmeter are wound anti-inductively on two brass 
 cylinders (vide above) , and the lower one of these may 
 be arranged to serve as a convenient means of support- 
 ing the instrument on a table or shelf. When, as is 
 most commonly the case, the mean potential to be 
 measured is 100 volts, the platinoid resistance is 
 adjusted to make up, along with the fine copper wire 
 solenoid (of which the resistance is about 60 ohms), a 
 total resistance of 1,000 ohms. Thus, the direct read- 
 ing of potential on the scale is in volts. 
 
 In order to save time in taking readings a checker 
 is provided. A brass arc, capable of moving in a 
 vertical direction, is placed parallel to and slightly 
 below the plane in which the pointer moves, and b^ 
 means of a handle this arc may be brought gently and 
 
 <3>^ 
 
 Fig. 69. 
 
 is secured by screwing down upon it the movable cap of 
 the top resistance coil (vide second paragraph below). 
 Portions of the tube are cut away to permit of easy 
 access to all parts of the instrument for adjustment or 
 inspection. In order to prevent damage to the sus- 
 pending wire or accidental disturbance of the torsion 
 head, two brass cylinders, which also serve to carry the 
 resistance coils (vide second paragraph below), are 
 placed covering the two ends of the supporting tube, 
 and are fixed by screws to the sheath. 
 
 The scale is graduated from zero to 140. but for 
 convenience of observation the first marked division 
 is 50. It is placed in a horizontal box with a glass 
 cover fixed to the sheath, and the pointer shows, by 
 inspection, direct reading of currents of from 50 to 140 
 milliamperes. 
 
 Fig. 70. 
 
 momentarily into contact with the pointer so as to 
 quickly stop its oscillations. 
 
 A\hen the instrument is to be used for very accu- 
 rate work, a means of observing and annulling any 
 error due to residual magnetism in the oblate may be 
 provided by a reversing key placed below the scale box, 
 and two magnets screwing into the sheath. The current 
 through the instrument is made in one direction when 
 the handle of the reversing key is in the top position, 
 and made in the opposite direction when the handle is 
 in the bottom position. The current is broken when 
 the handle is on either side. The residual effect in the 
 instrument is very small, and it is found to be sufficiently 
 accurate for all practical purposes without this adjust- 
 ment. 
 
 When in use this instrument should be supported 
 
MEASUREMENT. 
 
 321 
 
 with its scale approximately horizontal. For use as a 
 marine voltmeter it is found convenient to place the 
 instrument in a bracket, and secure it by passing a 
 collar over the upper end of the tube. 
 
 The adjustment for annulling any effect due to 
 residual magnetism in the oblate may be tested by 
 taking two readings with a constant current passing 
 through the coil, first in one direction and then in the 
 other. If the readings so obtained do not agree, the 
 current should be reversed through the coil several 
 times by turning the handle of the reversing key — in 
 most cases this will be found sufficient to restore the 
 adjustment. Should it not do so recourse must be had 
 to the two compensating magnets in the case, which 
 should be simultaneously screwed out or in as required 
 until the desired equality of readings is obtained. "\A~hen 
 this is done, the magnets should be again clamped by 
 means of the nuts provided for that purpose. AYhen 
 it is found necessary to screw out the magnets to make 
 the above compensation, and on screwing them out as 
 far as possible a perfect adjustment cannot be obtained, 
 the two magnets should be interchanged, and the 
 desired compensation will then be speedily secured by 
 screwing them in a little. 
 
 The adjustment of the zero is made by the maker, 
 and it will rarely, if ever, require revision. Should 
 such be found necessary at any time it may be veiy 
 easily effected as follows : 
 
 (a) I nscrew the movable cap of the top resistance 
 coil. 
 
 (b) Turn the torsion head until the needle points to 
 zero. 
 
 (c) Eeplace the cap. screwing it down firmly. 
 
 Adjustable Magnetostatie Current Meter. 
 
 The magneto-static current meter, Fig. 71, consists 
 essentially of a small steel magnet or system of magnets 
 suspended in the centre of a uniform field of force due 
 to two coils, each having one or more turns of copper 
 ribbon or wire, and also under the directive influence of 
 two systems of powerful steel magnets. The suspended 
 system of magnets is attached toone endof a vertical shaft 
 passing down centrally through an opening in the sole- 
 plate of the instrument from an indicating needle, which 
 is supported by a jewelled cap resting upon an iridium 
 point. The two systems of directive magnets are circular 
 in form, and each ring is composed of two semi-circular 
 magnets placed in a brass cylindrical frame with their 
 similar poles together. Each system is securely fixed 
 to a circular brass frame, which fits on to the cylindrical 
 case of the instrument in such a manner that the 
 systems are capable of being turned round, together or 
 separately, as explained below. The instrument has a 
 "tangent scale," which is adjusted in its position 
 before the instrument is sent out, so that the needle 
 indicates equal differences of readings for equal differ- 
 ences of current. The scale consists of a hundred 
 divisions, and for most purposes it is convenient to set 
 the field magnets in such a position that the needle 
 points to 0, and to use the scale from that point 
 
 upwards towards 100. Sometimes, however, it may be 
 found convenient to measure currents, whose direction 
 is being occasionally reversed, without being at the 
 trouble of reversing the electrodes in the contact clip ; 
 in that case the zero should be set to the division 50 at 
 the middle of the scale, and readings taken on each 
 side of it. It must be remembered that when the point 
 taken as zero is changed, the constant, by which the 
 indications of the instrument have to be multiplied to 
 give the current in amperes, is changed in proportion 
 to the cosine of the angle between the zero point and 
 the middle of the scale : and as this angle is 60 deg. the 
 constant with the zero at 50 on the scale is exactly 
 double the constant with the zero at on the scale. 
 
 The instrument is provided with a " lifter," which 
 serves to raise the needle off the iridium point when it 
 is being moved about from place to place. This lifter 
 is in the form of a ring placed below the needle, and 
 
 Fig. 71. 
 
 may be raised or lowered by turning the handle 
 attached to an eccentric passing through the side of 
 the instrument on a level with the scale. It also serves 
 as a checker, bj T bringing it lightly into contact with 
 the pointer, so as to stop its vibration. 
 
 The two grades of this instrument which are found 
 most convenient are : 
 
 The milliampere-meter, which has an effective range 
 of from '3 to 300 milhamperes, and is usually adjusted 
 to read 2 milliamperes per division. 
 
 The ampere-meter, which has an effective range of 
 from '3 to 300 amperes, and is usually adjusted to read 
 ] ampere per division ; in both grades with the zero 
 at on the scale. If desired, instruments can be 
 supplied having the constants adjusted to any value. 
 The very wide range of accurate measurements given 
 by these instruments make them invaluable for labora- 
 tory use. 
 
322 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 The instrument has an advantage, important for some 
 practical purposes, of being available as an accurate 
 direct-reading current meter, through a continuous 
 range of from 1 to 100 times its smallest current, 
 which may be anything from half a milliampere to 
 4 amperes, according to the number of turns in the 
 coils supplied with the instrument. It is not, however, 
 available as an alternate-current instrument, and it 
 must be remembered that the magnetism of the steel 
 directing magnet does not remain absolutely constant. 
 With good quality of steel, a proper preliminary ageing 
 of the magnet (by heating it several times in boiling 
 water and cooling again, and subjecting it to somewhat 
 varied rough usage) brings it to a condition in which 
 its magnetism is found to remain exceedingly nearly 
 constant month after month and year after year. Still 
 it should never be relied upon as absolutely constant, 
 and for accurate laboratory work it is therefore neces- 
 sary to have some means of retesting the instrument 
 at any time. This is always easily done with the 
 utmost accuracy if one of the balance instruments is 
 available as a standard. 
 
 Another advantage which the instrument has is that, 
 when a standard instrument is available, its constant is 
 capable of being varied to any desired value down to 
 one-tenth of that which it has with its directive magnets 
 in their strongest position. Thus, if the constant should 
 be 3 amperes per division of the scale, with the similar 
 poles of the magnets coinciding, it may be adjusted to 
 any value down to 0'3 ampere per division. 
 
 One very convenient use of the instrument is to act 
 as a lamp-counter for indicating the number of incan- 
 descent lamps in use in an installation. For this 
 purpose it is best to standardise it by putting on a 
 known number of lamps and adjusting, as described 
 below, until the desired reading is obtained on the 
 scale. Of course this numbering of lamps is not 
 possible to any great accuracy, because the lamps 
 themselves are not all rigorously equal in the amount 
 of current which each takes, but the lamp-counter 
 serves the important practical purpose of showing at 
 any time the number of lamps in use nearly enough for 
 practical purposes. In private houses this is very 
 useful as a check against some lamp or lamps being 
 left accidentally alight in a cellar, or safe-room, or box- 
 room, or other place where the fact of its being alight 
 might escape observation for days or weeks together. 
 
 To count larger numbers of incandescent lamps up 
 to 1,000 or more, the instrument is made with smaller 
 rings of more massive conductor, and the same pro- 
 portionate accuracy is attained as with the 100 lamp- 
 counter. 
 
 The milliampere-meter, on account of the low resist- 
 ance of its copper coil— about 40 ohms— may conve- 
 niently be used as a voltmeter. To adapt it for this 
 purpose a copper cylinder, wound anti-inductively with 
 two platinoid resistances, is supplied. The first of 
 these, together with the resistance of the instrument, 
 makes up 100 ohms, and the second alone is 900 ohms! 
 Thus, taking the constant of the instrument at 2 milli- 
 
 amperes per division, by joining the smaller in series 
 with the instrument, the reading on the scale will be 
 1/5 of a volt per division. With both resistances in 
 series with the instrument the reading will be 2 volts 
 per division. 
 
 The magnetostatic current meters when used should 
 be levelled, and the pointer adjusted to zero. The 
 adjustment is made by 
 
 (a) Loosening the two lower milled-headed screws 
 clamping the magnet frame, and turning the frame 
 round till the pointer stands at zero ; 
 
 (b) Then reclamp the frame by tightening the two 
 screws. 
 
 The scale is firmly clamped in its place before send- 
 ing the instrument out, and this position is marked by 
 two lines on the outside of the case, one horizontal and 
 the other vertical, just below the of the scale. The 
 horizontal line is engraved below the movable top of 
 the instrument, and the vertical one on the side of the 
 case. Should the top of the instrument have been 
 inadvertently moved, and the scale thus put out of 
 adjustment, it may be set right by slightly loosening 
 the two slotted screws and turning the top round till 
 the extremities of the two lines coincide. 
 
 If the needle should by accident be bent, it may 
 firstly be made as straight as possible by the hand, and 
 finally adjusted as follows : Set the zero, by the field 
 magnets, to the division 50 at the middle of the scale, 
 then join the instrument in series with another current 
 instrument of convenient form, and pass a current 
 through both sufficient to give a deflection of about 40 
 divisions on the magnetostatic instrument. Eeverse 
 the current on the magnetostatic instrument only, and 
 set the scale so that equal deflections, read in divisions, 
 are given on each side of the zero for equal currents, as 
 indicated on the auxiliary instrument. The zero must, 
 of course, be reset by the magnets every time the scale 
 is moved. When the scale has been adjusted to this 
 position, firmly clamp the top of the instrument by the 
 two slotted screws, and again mark the position of the 
 horizontal line on the outside of the case. 
 
 The constant may be quickly varied as follows : Join 
 the instrument in series with any reliable current 
 instruments of known accuracy, such as the deci- 
 ampere balance, and pass a convenient current through 
 both instruments, observing the readings. Break the 
 current, loosen the two upper pair of milled-headed 
 screws, and turn the top system of magnets relatively 
 to the lower, so that the similar poles of the two 
 systems are brought closer together or moved further 
 apart, according as it is desired to make the instrument 
 respectively less or more sensitive. Eeclamp the 
 screws and adjust the zero as previously described. 
 Again, make the current and note the reading on the 
 two instruments. The desired reading on the magneto- 
 static may be obtained quickly after one or two approxi- 
 mations, care being always taken to readjust the zero 
 after each movement of the top magnets. 
 
 When convenient it is always best to standardise the 
 instrument in the place where it is to be used, but when 
 
MEASUREMENT. 
 
 323 
 
 it is intended to move it from place to place it should 
 be standardised in such a position that when the needle 
 is pointing to zero it is in a direction approximately 
 east and west. 
 
 The Ampere Gauge. 
 
 The ranges of the different types of this instrument 
 usually made are : 
 
 amperes. 
 
 I. 
 
 From 
 
 •25 to 5 
 
 II. 
 
 J J 
 
 1 „ 20 
 
 III. 
 
 J J 
 
 5 „ 100 
 
 IV. 
 
 if 
 
 10 „ 200 
 
 V. 
 
 
 25 „ 500 
 
 The instrument, Fig. 72, is of simple construction, 
 having a vertical slate hase-plate, to which are attached : 
 
 (a) A solenoid of special form. 
 
 (6) Brass bearing-plates supporting a balance which 
 carries a soft iron plunger on its one arm, and a 
 brass counterpoise weight on the other. 
 
 (c) A brass arc having a scale graduated to give direct 
 readings in amperes. 
 
 (d!) A hinged arm which bears a light checker. 
 
 Fig. 72. 
 
 The solenoid is built up of copper plates with mica 
 insulation between them, and is fixed to the base-plate, 
 so that its core is vertical. 
 
 The balance is supported on knife-edges at such a 
 distance below the solenoid that the top end of the 
 plunger is slightly entered into the core. The plunger 
 is made from a thin soft iron wire about 20 centimetres 
 long, and is supported by a cross-bar with two hooks 
 on it, which pass over two knife- edge stirrups on the 
 arm of the balance. It has a brass weight hung on its 
 lower end in order to keep it in a vertical position and 
 prevent its being attracted against the side of the 
 solenoid. 
 
 An indicating needle, or pointer, formed from a strip 
 of platinoid, passes down from the trunnion of the 
 balance to the brass arc bearing the graduated scale. 
 As the plunger is attracted upwards, this pointer passes 
 round the scale and indicates the strength of current 
 passing through the solenoid. 
 
 When the instrument is packed for carriage, the 
 brass counterpoise on the arm of the balance and the 
 weight hung on the end of the plunger — which is the 
 larger of the two — should be removed and placed in 
 the receptacles provided for them. The pointer should 
 also be placed in the slot at the left-hand side of the 
 scale and secured by the button. 
 
 The instrument should be secured to a wall by 
 means of its brass support provided for the purpose, so 
 that the pointer is in the same plane with the scale and 
 stands at when no current is passing through the 
 solenoid. "When the instrument is thus supported, 
 the plunger should be found hanging parallel to and 
 between the two white lines engraved on the slate base- 
 plate. 
 
 The Electrostatic Voltmeters. 
 
 These voltmeters have the great advantage of being 
 available as accurate measures of potential on direct and 
 alternating systems, arid, being electrostatic, they use 
 no current, and consequently require no temperature 
 correction. They are therefore free from the causes of 
 error so prevalent in instruments of the electromagnetic 
 type, whose accuracy is impaired by variations of tem- 
 perature, and which when used on alternating systems 
 are affected by errors due to self-induction varying with 
 the period of alternation. The chain of electrostatic 
 voltmeters measures from 40 to 100,000 volts, and is 
 composed of three distinct types — viz., the multicellular 
 electrostatic voltmeters, the vertical electrostatic volt- 
 meters, and the electrostatic balance. 
 
 The ranges of the separate instruments as usually 
 
 made are : 
 
 160 
 
 100 
 
 240 
 
 130 
 
 400 
 
 240 
 
 800 
 
 600 
 
 1,600 
 
 1,300 
 
 4,000 
 
 8,000 
 
 12,000 
 
 50,000 
 
 100,000 
 
 The instruments are made on the principle of an air 
 condenser, having one of its parts movable about an 
 axis, so as to increase or diminish the capacity. The 
 condenser is enclosed in a metal case, for the double 
 purpose of protecting the movable part from air currents, 
 and from the disturbing influence of any electrified 
 body, other than the fixed portion, differing from it in 
 potential. In all the instruments, except the electro- 
 static balance, the fixed portions consist of two sets of 
 quadrant-shaped cells in metallic connection with each 
 
 Multicellular Electrostatic Voltmeter-! 
 
 Vertical 
 
 Electrostatic Balance 
 
 range 
 best of range 
 / range 
 ^best of range 
 
 (range 
 best of range 
 / range 
 \best of range 
 / range 
 \best of range 
 range 
 
 40 to 
 
 50 „ 
 
 60 „ 
 
 70 „ 
 
 80 „ 
 
 100 „ 
 
 200 „ 
 
 300 „ 
 
 500 „ 
 
 700 „ 
 
 200 „ 
 
 400 „ 
 
 soo „ 
 
 2,500 „ 
 5,000 „ 
 
324 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 other, and formed by a number of parallel brass plates. 
 These cells are fixed by an insulating support to the 
 case of the instrument, and a terminal passes from 
 them to an insulated binding screw on the outside of 
 the case. 
 
 The movable portion in all the instruments is in 
 metallic connection with the surrounding case. In the 
 multicellular voltmeters this connection is made through 
 the suspending wire, and in the vertical scale voltmeter 
 and electrostatic balance through the knife-edges which 
 support the movable part. The movable portion 
 carries the pointer which indicates by direct readings 
 the difference of potential between the two parts of the 
 condenser. 
 
 The action of the instrument, shortly stated, is as 
 follows : "When the fixed and movable plates are con- 
 
 Fig. 73. 
 
 nected respectively to two points of an electric circuit, 
 between which there exists a difference of potential, 
 the movable plate tends to move so as to augment the 
 electrostatic capacity of the instrument, and the mag- 
 nitude of the force concerned in any case is proportional 
 to the square of the difference of potential by which it 
 is produced. In the use of the vertical and electro- 
 static balance instruments this force of attraction is 
 balanced by the horizontal component of a weight of 
 any convenient amount hung on the knife-edge in con- 
 nection with the movable part, while in the case of the 
 multicellular it is balanced by the torsion of the sus- 
 pending wire. 
 
 The arrangement of the parts of the multicellular 
 electrostatic voltmeter is shown in Figs. 73 and 74. 
 These figures apply to an early form of the instrument] 
 
 and differ in two matters of detail from the voltmeter 
 as now made. For simplicity in manufacture the cells 
 are now made with straight backs, and the plates 
 looked at in plan are, therefore, triangular instead of 
 square, as shown in Fig. 74. A coach spring has now 
 been interposed between the suspending wire and the 
 spindle carrying the vanes, as explained below. 
 
 The insulated cells are formed of triangular brass 
 plates fixed into sav/ cuts in a brass back piece so as to 
 be equal distances apart and accurately parallel to each 
 other. Two sets of these cells, C, are fixed relatively 
 to each other, as shown in Fig. 74, by a vulcanite 
 support to the sole-plate, so that their plates are hori- 
 zontal, and are completely enclosed within the brass 
 cylindrical case of the instrument. 
 
 On the top of this cylinder is a shallow horizontal 
 circular scale-box containing the scale of the instrument 
 and having a glass cover, which serves to protect from 
 currents of air the movable indicator, I, and the scale 
 and interior parts from dust. 
 
 Fig. 74. 
 
 For the movable part a number of vanes, V, similar 
 in form to those of the quadrant electrometer are used. 
 These vanes are placed parallel to each other on a 
 spindle with distance pieces between them. The top 
 end of this spindle passes through a small hole in the 
 sole-plate of the instrument, which forms the bottom 
 of the scale-box, and is attached to a small coach 
 spring, which in turn is secured to one end of a fine 
 iridio-platinum wire suspended from a torsion head at 
 the top of. a vertical brass tube. The torsion head may 
 be turned by means of a forked key provided for the 
 purpose, and is clamped, to protect it from accidental 
 displacement, by a cap which screws on to the end of 
 the tube. The coach spring has sufficient rescilience to 
 allow the spindle to touch a guard stop, and so saves 
 the suspension from injury in event of the instrument 
 being roughly set down. 
 
 Two vertical brass repelling plates, which also act as 
 guard-plates to prevent the movable part from turning 
 
MEASUREMENT. 
 
 325 
 
 beyond its prescribed limits, are fixed to the bottom of 
 the sole-plate. These two plates carry a guide-plate, G, 
 with a circular opening in it, through which the lower 
 end of the spindle passes. A little brass disc, or head, 
 D, is attached to the end of the spindle, sufficiently 
 large to prevent its passing back through the hole in 
 the guide-plate. Thus the movable part is effectually 
 secured from swinging about so as to be injiu-ed, and 
 by no possibility can it come into contact with the 
 insulated quadrants. When the instrument is level 
 the spindle hangs free by the suspending wire, so that 
 the vanes are horizontal, and each is in a plane exactly 
 midway between those of two contiguous condenser- 
 plates. 
 
 An aluminium needle attached to the top of the 
 spindle indicates, on the horizontal circular scale fixed 
 to the upper side of the sole-plate, the difference of 
 potential between the movable and fixed portions of the 
 condenser by direct readings in volts. 
 
 Fig. 75. 
 
 To enable the multicellular to be used as an inspec- 
 tional instrument capable of being read from a distance, 
 as across an engine-room, a mirror, supported in a 
 frame which passes over the vertical brass tube, and 
 rests upon the glass cover of the instrument, is supplied. 
 When this mirror is in position, it is at an angle of 
 45 deg. with the plane of the sole-plate, and by reflecting 
 the scale and pointer gives the instrument all the 
 advantages of a vertical scale. The instrument is 
 shown in Fig. 75 with its mirror in position. A small 
 thumb-screw is placed in the centre of the base-plate 
 below the instrument, which can be screwed in so as 
 to lift the weight of the spindle and vanes from the 
 suspending wire and clamp the disc on the end of the 
 spindle against the guide-plate. A lifter or checker is 
 also provided similar to that used in the magnetostatic 
 instruments. A switch is attached to the upper 
 terminal of the instrument by which the voltmeter 
 can be taken out of circuit when desired. The switch, 
 
 after breaking circuit, puts the case and the insulated 
 cells in metallic connection. When received from the 
 maker the indicator needle with attached vanes will be 
 found supported by means of the thumb-screw below 
 the instrument, and also by the circular lifter, or 
 checker, turned up so that the weight of the needle 
 and vanes is taken off the suspending wire. The scale 
 is graduated to read directly in volts. 
 
 To set the instrument up for use (a) unscrew the 
 thumb-screw, and turn down the checker, so that the 
 needle swings clear ; (b) level the instrument so that 
 the spindle of the vanes passes down centrally through 
 the intersection of the two black cross-lines on the sole- 
 plate. To adjust the zero, if necessary, unscrew the 
 
 Fig. 76. 
 
 cap on the top of the tube, remove the washer, turn 
 the torsion head by means of the forked key until the 
 pointer stand at on the scale. Replace the washer 
 and screw on the cap again. Before adjusting the zero 
 turn the switch so that the insulated cells are in 
 metallic connection with the case. When the instru- 
 ment is to be removed from place to place, see that the 
 needle is lifted by turning up the checker, and when 
 it is packed for use as a portable instrument, always 
 screw up the thumb-screw as mentioned above. 
 
 As aluminium is electro-positive to brass, the instru- 
 ment reads about A of a volt too low when the positive 
 pole of a battery or dynamo is attached to the upper or 
 insulated terminal of the instrument ; and about A volt 
 too high if connected in the opposite direction. "W ith 
 alternating currents it is correct. 
 
 The vertical electrostatic voltmeter is shown in 
 Fig. 76, and, as will be seen, the insulated quadrants 
 are supported with their plates vertical, and only one 
 
326 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 large vane is used. This movable plate is supported in 
 a vertical position on knife-edges, so that the plane of 
 its motion is parallel to the two fixed plates which form 
 the insulated quadrants. Its upper end has a fine 
 prolongation which serves as a pointer for indicating 
 the deflections on the scale of the instrument, and at 
 its lower end is fixed the knife-edge for the weights, 
 having its length perpendicular to the plane in which 
 the plate moves. 
 
 In order to save time in taking readings, an arrange- 
 ment is provided for checking the oscillations of the 
 movable plate, and stops are placed to limit its range 
 and prevent damage to the pointer. One of these stops, 
 the left-hand one, is made to act as a support for the 
 vane in the arrangement for portability described below. 
 
 The scale is graduated from to 60, and the divisions 
 represent equal differences of potential — the actual 
 magnitude of the difference per division being de- 
 pendent upon the weight in use at the time. A set of 
 three weights is sent with each instrument, providing 
 for three grades of measurement in the proportion of 
 1:2:4. Thus the instrument shows one division per 
 50 volts with the link (the lightest weight) alone on ; 
 one division per 100 volts with the medium weight 
 hanging on the link, and one division per 200 volts with 
 all three weights on. 
 
 To set up the electrostatic voltmeter in working order, 
 remove the glass door of the case, place the movable 
 plate or vane on its knife-edge support, handling it very 
 carefully unless it be bent or twisted in the operation. 
 A line, drawn lengthwise on the surface of the movable 
 plate, and passing through its intersection with the 
 knife-edge, divides the portions above and below the 
 knife-edge into unequal parts. When the movable plate 
 is properly placed, this line is just seen behind the ver- 
 tical edge of the fixed plate when the pointer indicates 
 zero, and the smaller segments of the movable plate are 
 then hidden from a front view by being between the 
 fixed plates. 
 
 To detect, and if necessary correct, any accidental 
 bending of the pointer, with reference to the attracted 
 portion of the movable plate, hang one of the weights 
 on the lower knife-edge ; take the round pin sent inside 
 the case and with it press the movable plate in between 
 the fixed plates, until it rests in the two V-notches near 
 the upper end of the vertical edges of the fixed plates ; 
 holding the pin so, rotate it about its axis, and observe 
 that the pointer indicates a small red line seen on the 
 scale in the neighbourhood of division number 35. 
 
 Eemove the weight and see whether the movable 
 plate is in neutral equilibrium. If it is so, the index 
 will move very slowly along the scale, and will come to 
 rest somewhere within its range. If the index rest 
 against one of the stop-pins, screw out, or in, the nut 
 on the horizontal screw attached to the lower end of 
 the vane until the pointer comes to rest on the scale. 
 If the index rests very definitely at one point of the 
 scale and vibrates about it, the movable plate has too 
 much stability ; if it is found that the index will rest 
 against both of the stop-pins, but will not rest at any 
 
 other point on the scale, the movable plate has too 
 little stability. The stability can be adjusted by 
 screwing up or down the nut on the vertical screw 
 attached to the lower end of the vane. These adjust- 
 ments are made by the maker, and will generally be 
 found to be nearly enough correct. 
 
 After hanging on the weights, adjust the pointer to 
 zero by means of the screw levelling feet on the case of 
 the instrument. 
 
 The Eleetrostatie Balance. 
 
 The arrangement of the parts of this instrument is 
 shown in Fig. 77. The fixed portion of the condenser 
 in this instrument is a brass disc, B, which is supported 
 from a slate base, S, on three glass pillars, P. The 
 disc is provided with the well-known Thomson " hole, 
 slot, and plane " arrangement, so that it always rests in 
 exactly the same position on its supports. 
 
 Fig. 77. 
 
 A wire thickly covered with indiarubber passes from 
 a terminal, T, through a glass tube, C, C, C, and makes 
 connection with the disc by a spring contact ; the glass 
 tube being filled with paraffin to prevent the lodgment 
 of moisture and give great resistance to disruptive 
 discharge. A sheath formed by a short piece of glass 
 tube pulls up over the terminal, T, and protects it from 
 being touched by accident. The slate base-plate is 
 provided with three screw levelling feet. A brass case 
 fits upon the slate base-plate, and fixed to its top is a 
 metal scale-box with a glass front, which contains the 
 indicator and scale. The movable part, Y, is a round 
 aluminium plate, supported by two long links, which 
 pass through a slit in the top plate of the case to two 
 knife-edge stirrups on one end of the counterpoised 
 indicator, I. The whole movable portion is supported 
 by knife-edges on two brass pillars and has a short arm, 
 A, with a knife-edge stirrup at its extremity attached to 
 its axis. The weights which fix the constant of the 
 instrument hang on this stirrup. 
 
MEASUREMENT. 
 
 327 
 
 The instrument has a scale with divisions corre- 
 sponding to equal differences of potential. The scale 
 is graduated from to 50, and three weights are pro- 
 vided such that, with the first alone hung on, the con- 
 stant is 250 volts per division, with the first and second 
 weights on, it is 500 volts per division, and with all 
 three weights on, 1,000 volts per division. 
 
 The following precautions ought always to be taken 
 for safety in the use of Sir W. Thomson's electrostatic 
 voltmeters in connection with dynamos, whether for 
 direct or alternating currents. 
 
 In all applications in which one of the two con- 
 ductors connected with the voltmeter is kept perma- 
 nently connected with the earth, this conductor should 
 be connected with the outer case of the voltmeter. The 
 other is to be connected with the insulated terminal, and 
 must be carefully guarded against accidental contacts. 
 
 To provide for use in any application not fulfilling 
 this condition, all the electrostatic voltmeters are 
 supplied with thoroughly insulating feet ; and the 
 precautions stated below must be observed. 
 
 The vertical scale voltmeter for from 400 to 12,000 
 volts, when set up for permanent use, should be 
 enclosed in a case (which may be of wood with a glass 
 front) preventing any person from accidentally touch- 
 ing the metal case or the terminals of the instrument. 
 The vibration checker is worked with perfect safety by 
 a silk cord passing through the wood or glass of the 
 protecting case to the front outside. For temporary or 
 experimental applications the user must take his own 
 precautions ; an outer enclosing glass case might be 
 found too cumbrous. For ordinary domestic electric 
 lighting or other applications to less than 200 volts, the 
 multicellular voltmeter may be left unprotected so far 
 as personal danger is concerned ; but, to avoid chances 
 of damage to instruments or wires, or of melting a fuse, 
 its outer case, as well as its terminal insulated from 
 the outer case, ought to be perfectly guarded against 
 accidental contacts when the instrument is set up for 
 permanent use. Glass and vulcanite sheaths are pro- 
 vided for this purpose by the instrument maker when 
 desired. 
 
 Never open the case of the vertical scale voltmeter, 
 to change its weights, nor touch its terminal to con- 
 nect or disconnect (or to secure either connection if 
 imperfectly made), without being sure either that 
 the dynamo is not running, or that both the con- 
 ductors leading to the voltmeter are safely disconnected 
 from its circuit. 
 
 It may be asked, with reference to the vertical scale 
 voltmeter, why is the inner case made of metal ? The 
 answer is, that the electric conditions for definiteness 
 of measurement require the vane to be protected all 
 round from sensibly disturbing influence of any sub- 
 stance, other than the air around it, differing in poten- 
 tial from itself unless at the same potential as the 
 quadrants. Why, then, not coat the metal inner case 
 with wood or vulcanite, or other non-conductmg 
 material ? Answer : The protection thus imagined 
 might be delusive when 10,000 volts is dealt with. 
 
 Safety is more surely secured by an outer case an inch 
 or so from the inner metal, unless, which is always best 
 when it can be arranged for, one of the conductors is 
 kept connected with the earth, and with the metal 
 case of the electrometer also connected with the earth. 
 
 New Engine-Room Voltmeter. 
 
 This instrument is intended for installation work on 
 either direct or alternating circuits, where it is con- 
 venient to have a direct-reading voltmeter with large 
 scale divisions. 
 
 Fig. 78. 
 
 The instrument, as shown in Fig. 78, depends for its 
 action upon the repulsion of a movable coil, M, by a 
 fixed coil, F. The fixed coil, F, bears all the other 
 portions of the instrument attached to it, and is in its 
 turn supported from a vulcanite block fixed to the 
 instrument case. This vulcanite block also bears the 
 terminals of the instrument. The movable coil is sup- 
 ported on knife-edges, and the circuit through it from 
 the fixed coil of the voltmeter is made by two spirals of 
 fine copper wire. A pointer attached to the movable 
 coil indicates by direct readings in volts the difference 
 of potential between the terminals of the instrument. 
 Attached to the pointer on one side, and perpendicular 
 to it, is a short arm, with a screw nut, N, which, 
 together with the sliding weight, S, on the pointer, 
 serves to adjust the balance of the movable parts. To 
 
Missing Page 
 
MEASUREMENT. 
 
 329 
 
 instruments, costing from £20 to £30 ; considerable 
 trouble and expense are incurred in tbe first place to 
 adjust tbe coils with perfect accuracy ; and an elaborate 
 but unnecessary finisb is indulged in to do justice to it. 
 A bridge costing £2 or £3, suitable for ordinary instal- 
 lation work, cannot be obtained. It is by no means 
 intended to depreciate accuracy. An accurate instru- 
 ment is better tban an inaccurate one, otber tbings 
 being equal ; but it would be absurd if a grocer could 
 buy notbing cheaper tban a chemical balance to weigh 
 with, because scale makers found they were capable of 
 very great accuracy and had plenty of room for 
 unnecessary finish. Accurate bridges are required for 
 such work as localising faults in telegraph lines, but 
 rough resistance-boxes right within, say, | per cent., 
 would be good enough for ordinary work. As it is, 
 an installing engineer, who would not scruple to 
 dispense with a voltmeter and judge the electromotive 
 force by the look of the lamps, searches for ground 
 contacts with a 40-guinea bridge in conjunction with 
 one defective Leclanche cell and a pivoted detector 
 which sticks." 
 
 A very useful and at the same time inexpensive 
 form of galvanometer is 
 
 The Holden-d'Arsonval Galvanometer. 
 
 This galvanometer, as shown complete in Fig. 79, and 
 with its cover removed in Fig. 80, is a development of 
 the well-known d'Arsonval dead-beat zero instrument. 
 In the improved galvanometer we have a powerful 
 laminated permanent magnet of circular form, and 
 placed horizontally. The poles of this magnet are 
 brought round to face one another, and are turned out 
 so as to encompass the moving coil. In the centre 
 of the magnetic field, and midway between the poles, 
 a cylindrical rod of very soft iron is placed, serving as a 
 medium for concentrating the lines of force, and making 
 the magnetic field quite uniform. The moving coil 
 is built upon a light silver frame, and is wound 
 with No. 40 silk-covered copper wire, and when 
 mounted with its suspending wires it has a total 
 resistance of about 16 ohms. The current is led into 
 the coil by means of the suspended platinoid or 
 phosphor bronze wire, starting from a spring at the 
 top of the instrument and passing out through another 
 wire fixed at the bottom and attached to a mill-headed 
 screw. The spring holding the top wire serves to give 
 the necessary tension. The bottom screw-nut is used 
 to produce the desired torsion to bring the mirror to 
 its zero mark. A special point about this instrument 
 is an arrangement for lifting out the coil and its 
 enclosed iron core, and thereby affording a ready 
 means of renewing the suspension wires or mirror, 
 and making other adjustments. By simply unscrewing 
 a clamping screw, and removing one connecting wire 
 from a binding clamp, the whole coil system can be 
 removed. The advantage of this arrangement is 
 apparent, as a number of different resistance coils may 
 be supplied with each instrument, and at the same 
 time it reduces the chances of damage in transit. With 
 
 a coil having a resistance of 16 ohms and an added 
 resistance of 100,000 ohms, a deflection of the coil 
 equal to an angle of 1 deg. may be obtained with a 
 pressure of 1 volt. As the scale is found to be of 
 practically equal divisions, the instrument may be used 
 as an ampere or voltmeter by the addition of suitable 
 shunts or resistances. 
 
 Measurement of the Internal Resistance of Batteries. 
 
 Having ascertained the resistance of our tangent and 
 mirror galvanometers, the data so obtained will assist 
 us in measuring the internal resistance of the battery 
 we are using. The student will notice the points 
 of similarity between the methods here given for 
 measuring battery resistance with those for measuring 
 galvanometer resistance, page 313. 
 
 Galvakometer. 
 
 Direct bridge method. 
 Half deflection method. 
 Equal deflection shunt 
 
 method. 
 Thomson's method. 
 
 Battery. 
 
 Direct bridge method. 
 Half deflection method. 
 Equal deflection shunt 
 
 method. 
 Mance's method. 
 
 A. Direct Bridge Method. — This is a very inferior 
 method to either of the others to be described. It 
 necessitates the enployment of two similar cells having 
 exactly equal electromotive force. The two cells are 
 joined up so as to oppose each other, and are then 
 placed in the fourth arm of the bridge, and the 
 resistance of the pair measured in the ordinary way. 
 This value divided by 2 is taken as the resistance of 
 the cell — a pure assumption, as we have no proof that 
 the resistance of one cell equals that of the other. 
 
 The arrangement is shown in Fig. 81. 
 
 Fig. 81. 
 
 B. Half Deflection Method. — Join up the battery by 
 means of short thick wires with resistance-box and 
 tangent galvanometer, and introduce such resistance, 
 By as will allow of a deflection of about 55 deg. = d. 
 Increase Bj to B 2 until the deflection is reduced one- 
 half = cL = -J. then 
 "-2 2 . 
 
 r = E 2 - (2 Bj + G.) 
 
 The connections are the same as for the determina- 
 tion of galvanometer resistance by " half deflection 
 method," Figs. 82 and 83. 
 
 " To avoid mistake in calculation, first double the 
 smaller resistance, to the result add the resistance of 
 
830 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 the galvanometer, and then deduct the total from the 
 greater resistance." — Kempe. 
 
 If with the lowest resistance we can give to B x the 
 deflection should exceed 55 deg., the sensitiveness of 
 the galvanometer may be reduced by employing a 
 shunt, in which case the resistance of galvanometer 
 G x S 
 
 will be 
 
 G+ S 
 
 Example. — The Daniell cell referred to in last 
 example was found to give a deflection of 10 deg. when 
 E x = 4 ohms, using a shunt, s, of 1 ohm. On removal 
 of shunt it required a resistance, E 2 , of 186 ohms to 
 reproduce the first deflection. Eesistance of galvano- 
 meter being 28 ohms. Bequired, r. 
 
 r==s E 2 -E, 
 
 Ej + G 
 
 186-4 182 , „ , 
 4 + 28-82" 5 ' 7d,m8 " 
 
 w 
 
 Fig. 82. 
 
 Fig. 85. 
 
 Example. — "With a tangent galvanometer, G, whose 
 resistance was known to be 28 ohms, the internal 
 resistance of a Daniell cell was found by the half 
 deflection method to be as follows : 
 
 When tfj = 50 deg. (tan 50 deg. = 112) Ej = 20 ohms. 
 „ d 2 = ^ = 30 deg. (tan 30 deg. = -56) E 2 = 74 „ 
 
 r=E 2 -2Ej + G 
 
 = 74-2x20 + 23 
 = 74-68 
 = 6 ohms. 
 
 C. Equal Deflection Shunt Method. — Join up galva- 
 nometer, resistance-box, and cell whose resistance, r, is 
 required with a shunt, s, between its poles, and note 
 deflection, d u produced with resistance, Bj, in circuit. 
 
 <h 
 
 o 
 
 Fio. 83. 
 
 <& 
 
 Fig. 84. 
 
 On removing shunt the deflection will be increased to 
 dj. Increase ^ to B, till tf, is again produced, then 
 
 Ej + G 
 
 A comparison of diagrams 84 and 85 with those given, 
 Figs. 82 and 83, in the corresponding method for 
 measuring galvanometer resistance shows that the 
 relative position of G and r are merely changed — in all 
 other respects the test is made in the same way. 
 
 D. Mance's Method. — Place cell whose resistance is to 
 be measured in the fourth arm of the bridge (Pig. 86) 
 connecting the galvanometer (which should be a 
 Thomson's mirror) in usual manner to terminals, C A, 
 interposing a key between B and E. Then adjust 
 resistance a, b, and d. 
 
 a d 
 
 With a Post Office bridge the connections would be 
 made as in Pig. 87. Having arranged suitable resist- 
 ances in a b, bring in the galvanometer by firmly 
 pressing down K, and then vary the resistance in d 
 till no alteration in the deflection is produced by 
 closing or opening K. 
 
 Fig. 86. 
 
 The great advantage of this method is that the 
 electromotive force need only be constant during the 
 short time necessary to close and open key^ 
 
 Make b as high and a as low as the resistances at 
 command will permit. 
 
 The student will not fail to notice the similarity of 
 this method with that of Thomson for measuring the 
 resistance of galvanometers — the galvanometer and 
 battery simply change places. The directions to be 
 followed are the same in both cases. 
 
MEASUREMENT. 
 
 331 
 
 Example. — The Daniell cell previously referred to 
 required 630 ohms in d, b and a being 1,000 ohms and 
 10 ohms respectively, therefore 
 
 r = 
 
 ad 10 x 630 
 1,000 
 
 6'3ohms. 
 
 A " slide wire " bridge may be employed for this test 
 instead of the Post Office form ; in which case the 
 connection will be as in Fig. 88. 
 
 Fig. 87. 
 
 Placing the cell in r and resistance coils in d, move 
 slider along the wire until on closing the key no change 
 in the deflection of galvanometer is produced, then 
 
 ad = j / d \ 
 T Vl.OOO - a) % 
 
 Make d greater than r, so that b may exceed a. 
 
 In Mance's as in Thomson's method, a controlling 
 magnet must be employed to bring the permanent 
 deflection to the zero of scale. 
 
 Example. — In measuring r of the Daniell cell, d 
 being 10 ohms, balance was obtained when a = 380 
 divisions of the scale. 
 
 — rl ( ^ ^ = 1 380 _ = 1 o 
 YL.000 - a) 1,000-380 
 
 380 
 620 
 
 - §£W - 6-1 ohms. 
 620 
 
 Summary.— -Of the four methods of determining the 
 internal resistance of batteries described above, the 
 
 equal deflection method has the advantage that any 
 sensitive galvanometer may be used, since we have 
 only to reproduce a given deflection — also, in common 
 with Mance's, the resistance of the galvanometer need 
 not be known. 
 
 Where a Thomson reflecting galvanometer is at 
 hand, Mance's method leaves nothing to be desired. 
 
 "With a good tangent galvanometer of moderate 
 resistance, very good results can be obtained with 
 the half deflection method. In every case the higher 
 the figure of merit of the galvanometer, the greater is 
 the degree of accuracy attainable. 
 
 Determination of Electromotive Force. 
 
 The measurement of electromotive force has> the 
 disadvantage that, unlike the measurement of resist- 
 ance, there exists no procurable standard of the 
 practical unit, or volt. Still, the electromotive force 
 of a well-made Daniell cell so nearly approximates to 
 the theoretical value of the volt that we may safely 
 take one of these cells for our standard. 
 
 The Daniell cell is made in a variety of forms, but 
 the one known as " Wheatstone's standard cell " will 
 be found very suitable for our purpose. 
 
 Into an outer glass or glazed earthenware pot (Fig. 
 89) is placed a cylindrical sheet of copper with a wire 
 attached (for positive pole), and inside this a small 
 porous pot, the space between the porous and outer 
 pot being filled with a saturated solution of copper 
 sulphate. 
 
 A few small pieces of zinc are put into the porous 
 pot, and sufficient mercury added to nearly cover them. 
 
 
 Fib. 89. 
 
 The porous pot is then filled up with water to the same 
 height as the copper sulphate in the outer jar, and a 
 short length of guttapercha-covered wire, with the end 
 uncovered, so as to make contact with the mercury and 
 zinc, serves to form a negative pole for the cell. 
 
 "When not in use the cell should be dismantled, and 
 after placing the porous pot in nitric acid for a short 
 time it should be well washed, and then allowed to 
 stand in water until again required. The mercury and 
 zinc may be repeatedly used. The electromotive force 
 of this cell is T079 volt. 
 
 All. the following methods of determining electro- 
 motive forces are based upon the principle of comparison 
 against this standard cell : 
 
 A. Equal deflection method. 
 
 B. Equal resistance method. 
 
 C. Wiedeman's method. 
 
 D. Wheatstone's method. 
 
 E. Poggendorff's method. 
 
 F. Latimer Clarke's method. 
 
332 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Equal Deflection Method. — As in the case of 
 measuring the resistances of galvanometers and 
 batteries by the equal deflection method, any kind 
 of sensitive galvanometer may be used, as we have 
 merely to reproduce a given deflection. 
 
 Join up standard cell, Ej, resistance-box, and galvano- 
 meter in simple circuit, Figs. 83 to 85, and unplug 
 resistance, E 1; till a suitable deflection, d, is obtained. 
 
 Eeplace standard cell, Bj, by the one whose electro- 
 motive force has to be measured, E 2 , altering resistance 
 in box to B 2 till the same deflection is secured as with 
 the standard cell. 
 
 Fig. 90. 
 
 Then, if G = resistance of galvanometer, and r v r 2 
 the internal resistances of the two cells, B x and E 2 , it 
 follows, since the deflections are equal, the currents 
 must be equal also, and by Ohm's law we get 
 
 E: _ B 2 
 
 C=- 
 
 r 2 + E 2 + G' 
 
 »-, + E x + G 
 
 or, calling the total resistances in each measurement 
 Ej and B 2 , 
 
 Ej E 2 
 Ej : E 2 : : Ei : E 2 , 
 
 .-• e 2 =e3. 
 
 from which we see (as proved on page 313, " Experi- 
 mental Proofs of Ohm's Law ") that the electromotive 
 forces of the two cells are directly proportionate to the 
 total resistances in circuit. 
 
 With a high resistance in circuit, ^ and r 2 may be 
 ignored. 
 
 The greatest accuracy is obtained when E x and E 2 
 are made as high as possible. 
 
 Example. — In the following examples the electro- 
 motive force of a Leclanche cell, E 2 , was compared with 
 that of the standard Wheatstone-Daniell, E t . 
 
 Using a tangent galvanometer, we obtained 
 
 d = 25 deg., E x 30 ohms, E 2 40. 
 
 E : : E 2 : : E, : E, 
 
 1 : E 2 : : 30 : 40 
 
 E=^° = l-3volt, 
 
 Example. — Employing the mirror galvanometer, the 
 values of E x and E 2 were 7,000 ohms and 9,000 ohms 
 respectively. 
 
 E, 
 
 1 x 9,000 
 7,000 ' 
 
 = 1'28 volt. 
 
 Equal Resistance Method. — Join up standard cell, 
 Ex, resistance-box, and galvanometer, as in preceding 
 method, Figs. 83 to 85. Adjust resistance till a 
 suitable deflection, d\, is obtained. Note this deflection, 
 and also the total resistance in circuit. Substitute E 2 
 for Ej, and if it has a different internal resistance the 
 resistance in the box must be readjusted so as to 
 
 Fig. 91. 
 
 maintain the total resistance the same as when Ej 
 was in circuit. Note new deflection, d 2 . 
 
 Then, if G x be the current producing d\, and C 2 the 
 current producing d 2 , we have, by Ohm's law, 
 
 Cl =§ = andC 2 =|. 
 
 And since the deflections are proportional to the 
 currents producing them, we have 
 
 Ej ; E 2 '.". Gi : G 2 '. '. d 1 : a\, 
 E 2 = E 1 ^. 
 
 or 
 
 If the tangent galvanometer scale is graduated into 
 degrees of arc, then the tangents of d^d^ must betaken. 
 
 /<^\ 
 
 \ * 
 
 Fig. 92. 
 
 f^h. 
 
 VjulV 
 
 Make E as high as possible, and with a tangent 
 galvanometer let d lt d 3 have equal values on each side 
 of 45 deg. 
 
 Example.— Using a tangent galvanometer and 
 making the total resistance E = 30 ohms, the standard 
 
MEASUREMENT. 
 
 333 
 
 cell gave a deflection, d l7 of 41 deg., and the Leclanche 
 a deflection, d 2 , of 48 deg. 
 
 tan 48 deg. 
 
 _ 111 
 
 ■87" 
 
 tan 41 deg. 
 = 1-27 volt. 
 
 Example. — The shunted mirror galvanometer with 
 E = 10,000, gave d x = 105 and a\ = 11 scale divisions. 
 
 .\ E =1 105 + U 
 
 E 2 = E x 
 
 The mirror galvanometer (shunted down to T V m 
 order to obtain a convenient deflection) gave with the 
 standard cell, E lf and 10,000 ohms in circuit, a deflec- 
 tion, d\, of 40 divisions. On substituting the Leclanche 
 cell, E 2 , a deflection, d 2 = 50"5 divisions was obtained 
 through the same resistance ; therefore 
 
 50-5 
 40 
 
 _ 1 x 505 
 40 
 
 = 1-26 volt. 
 
 Wiedemann's Method. — This is a very good method, 
 since it is entirely independent of the internal resistance 
 of the battery, but a sensitive galvanometer and high 
 resistance should be employed. 
 
 Join up both cells in series, Fig. 92, and adjust the 
 resistance so as to allow a high deflection, d\, to be 
 obtained. 
 
 The current, C^, producing this deflection with a total 
 resistance, E x , in circuit is 
 
 Ej + E 2 
 
 Cx- 
 
 E 
 
 Now reverse the weaker cell, E 2 , so that the two 
 oppose each other, and note the smaller deflection, d. 2 , 
 due to C 2 , E remaining same as before. 
 
 therefore 
 hence 
 
 r _ E t -E 2 
 ° 2 E~~ 
 
 E 1 C 1 -E 2 C = E 1 C 2 + E 2 C 2 , 
 Ei : E 2 : : C l + C 2 : G 1 — C 2 
 
 : : d t + d 2 : d l - d 2 . 
 
 N.B. — Since the weaker cell has to be reversed in 
 above method, E a represents the Leclanche and E 2 
 the standard Daniell, whereas in each of the other 
 methods described the reverse is the case. 
 
 The above formula is applicable to mirror galva- 
 nometers, etc., with tangent galvanometer tan d^ and 
 d. 2 must be taken. 
 
 Example. — With the tangent galvanometer, the 
 Leclanche cell, E 1; and the Daniell, E 2 , in series, gave 
 a deflection, d\ = 54' 5 deg., and with the Daniell 
 reversed a deflection d 3 = 10 deg. 
 
 Ej : E 2 : : tan d^ + tan d 2 : tan d x - tan d„ 
 
 E! : 1 : : tan 54"5 + tan 10 : tan 54"5 - tan 10 deg. 
 
 : : 1-40 deg. + "176 : 1-19 - "17 
 
 : : 1576 : 1224 
 
 ]/576 
 Ei - 1 1-224 
 
 - 1-28 volt. 
 
 105 - 11 
 116 
 94 
 = 1-24 volt. 
 
 = 1. 
 
 Wheatstone's Method. — This method, like the pre- 
 ceding one, is independent of the battery resistances, 
 and any sensitive galvanometer may be used. 
 
 Join up in simple circuit, as in Fig. 93, the standard 
 cell, Bj, galvanometer, and resistance in box until a 
 moderately high deflection, d\, is obtained. Increase 
 the resistance by an amount = E 1; so that a diminished 
 deflection, d 2 , is produced. 
 
 <1> 
 
 Fig. 93. 
 
 Eeplace Ej by the cell, E 2> whose electromotive force 
 is to be measured = E , and readjust initial resistance 
 in circuit till deflection dj is obtained, as with E x . 
 The resistance is now increased by an amount = K2> 
 sufficient to again reduce the deflection to d 2 , as in first 
 case. Then 
 
 Ej : E 2 : : E x : E. 2 ; 
 
 or the electromotive forces of the two cells are directly 
 proportionate to the added resistances. Make the initial 
 resistances as high as the galvanometer will admit of, 
 and the added resistance approximately double this. 
 
 Example. — With an astatic galvanometer and an 
 initial resistance of 10 ohms the standard cell, E 1} gave 
 a deflection of 35 deg. = d\, which was reduced to 25 deg. 
 by the addition of 20 ohms = E x . 
 
 Substituting the Leclanche cell, E 2 , it required 19 
 ohms to reproduce the deflection d^, and to reduce this 
 to d 2 an addition of 25 ohms = E 2 had to be made. 
 
 E, 
 
 25 
 20 
 
 = 1-25 volt. 
 Example. — Employing the mirror galvanometer, it 
 required 2,000 ohms = Ej to reduce d\ to d 2 with 
 the Daniell cell in circuit, whereas the Leclanche 
 required 2,550 ohms. 
 
 E 2 = Eji 
 
 E 9 = lx 
 
 E, = 1 
 
 2,550 
 
 2,000 
 = 1-27 volt. 
 
334 
 
 PRACTICAL ELECTRICAL ENGINEERl^. 
 
 Poggendorff's Method. — With this method, the 
 reading is taken when no current is flowing through 
 the galvanometer, the electromotive force of the cell 
 to be determined being balanced against that of the 
 standard, therefore the relative values of the divisions 
 on the galvanometer need not be known. Hence any 
 form of instrument can be used, a low-resistance 
 astatic combination being very suitable. 
 
 Again, since no current is flowing, all possibility of 
 errors arising from polarisation is prevented. Also, by 
 varying the resistances and taking a second reading 
 
 -VXA/WVWVVVXAA- g 
 
 Fig. 94. 
 
 and combining the two values, we eliminate the 
 internal resistance of the cell under investigation. 
 
 Join up the two cells, galvanometer, and resistance- 
 box, as shown in Figs. 94 and 95. 
 
 The standard cell,E 1( should be stronger than the one, 
 E 2 , to be measured, hence two or more Daniell cells in 
 series must be used as occasion may require. 
 
 Unplug the two 100-ohm = r x in A C and adjust the 
 resistance, 1^, in A E until no deflection is produced 
 on closing the key, K 2 . Eeduce resistance r x to r 2 by 
 
 Fig. 95. 
 
 inserting one of the 100-ohm plugs in A C, and again 
 vary resistance to E 2 in A E, so that balance is again 
 obtained, then 
 
 E x : E 2 : : (E : - E 2 ) + (r x -r 2 ) : (R x - E 2 ) 
 
 or 
 
 E 1 _(R L -E ? ) + (r 1 -r 3 ) 
 
 E, 
 
 (R-E 2 ) 
 
 Example. — On placing two Daniell cells, Ej, and 
 whose electromotive force = 2 volts, between C E and 
 the Leclanche cell, E 2 , to be measured with galvano- 
 meter between A E, and making r 1 = 200 ohms, it 
 required a resistance of 500 ohms in A E to prevent 
 deflection of the needle of an astatic galvanometer. 
 
 On r x being reduced to r 2 = 100 ohms, Ex had to be 
 lowered to E? = 300 ohms. 
 
 E x 
 2:E 2 : 
 
 E 2 (R l -XJ + (r l -rJ:(R 1 --Rj>. 
 (500 - 300) + (200 - 100) : (500 - 300) 
 200 + 100:200 
 3:2. 
 
 E 2 = ^J? = 1-3 volt. 
 
 If the electromotive forces of the two cells com- 
 pared be equal, the test is impossible, and the greater 
 the difference in their electromotive forces the more 
 accurate becomes the measurement. 
 
 Latimer Clarke's Method. — In the development of 
 Poggendorff's method both the standard and the trial 
 cell are compared when no current is flowing in either. 
 The initial difference of potential, to the fall of which 
 the electromotive forces of the two cells are to be com- 
 pared, is set up and maintained by a third battery, E, 
 
 Fig 96. 
 
 of greater electromotive force than either of the two 
 under comparison. 
 
 Instead of an ordinary box of resistance coils, which 
 might be damaged by the heat produced by a strong 
 continuous current, a slide wire resistance may be 
 employed as follows : Take about 20ft. of thin German 
 silver wire and stretch it in a series of zig-zags between 
 two rows of pins stuck into a dry board, the end of the 
 wires being fastened to two binding screws, A and B, 
 Fig. 96. If the length of each diagonal be 1ft., it will 
 facilitate subsequent measurement. 
 
 To terminal A connect the wires leading from the 
 zinc poles of three batteries — E, the one maintaining 
 the difference of potential in the wire ; E^ the stan- 
 dard cell ; and E 2 the cell to be measured. 
 
 Connect wire from positive pole of E through K to 
 terminal, B, of slide wire, and from the corresponding 
 poles of Ej, E 2 lead wires through two galvanometers, 
 G v Go. 
 
 Close K and move end of wire from E lf G x along the 
 slide wire from A towards B. At a certain distance 
 = Pj, the needle of Gj will come to rest. Measure the 
 length of wire from A to P r 
 
 Next touch slide wire with the end of wire from E 2 , 
 G 2 . Note position, P 2 , at which G 2 comes to rest, and 
 
MEASUREMENT. 
 
 measure length of wire, A P 2 . Then, assuming that 
 
 the resistance of the slide wire to be uniform, and to 
 
 vary as its length, the distances A Pj and A P 2 may be 
 
 taken as giving the relative resistances, B x E 2 , of the 
 
 slide wire included in the two circuits of E x and E 2 , 
 
 and if C = current produced by E flowing through B A, 
 
 then 
 
 E 1 = CE 1 , andE 2 = CE 2 ; 
 
 < .E 1= CE lj 
 
 E 2 C E 2 
 
 and since C is common to both measurements, we get 
 
 E 2 E 2 
 or the electromotive forces of the two cells compared, 
 are proportionate to the lengths of the slide wires, 
 AP^AP,. 
 
 Messrs. Glazebrook and Shaw have suggested a very 
 useful modification of above arrangement, by which 
 one galvanometer suffices, a consideration of import- 
 ance to amateur experimentalists. 
 
 The arrangement is very similar to the preceding 
 one, except that the two positive wires from Ej and E 2 
 
 Fig. 97. 
 
 are connected to the two terminals, a, b, of a three- 
 poiDt switch, the remaining terminal, c, being joined 
 to the galvanometer and wire for making contact with 
 the slide wire, Pig. 97. 
 
 Close K and switch E x on to galvanometer by 
 inserting the peg between a c, slide the free end of the 
 wire from galvanometer along A B till needle comes to 
 rest— note this point P r Switch E 2 on to galvano- 
 meter and ascertain position P 2 . 
 
 Measure A Pj and A P 2 , and proceed as previously 
 described. 
 
 Example.— Using a Grove cell, E, to maintain the 
 difference of potential on the wire, a standard Daniell 
 cell required contact to be made at a point, P 1( lift. 6in. 
 from A, whereas with the Leclanche cell P 2 was at a 
 distance of 14ft. 9in. : therefore, as 
 
 lift. 6in. : 14ft. 9in. : : 1 : E 2 , 
 
 E, = 
 
 M^ = 1-28 volt. 
 
 116 
 
 Summary.— Of the above methods of comparing the 
 electromotive force of cells, "the equal deflection 
 method " has the advantage that any delicate galvano- 
 
 meter may be used, but either the internal resistance 
 of the cells must be known, or the interpolar resistance 
 must be so great in comparison, that it may be 
 neglected ; when this is the case it is a very ready 
 and useful method. With the " equal resistance 
 method " the same remarks apply as to internal 
 resistance of cells, in addition to which the values of 
 the deflections must be comparable ; with a good 
 mirror or tangent galvanometer it works well. 
 
 Fig. 98. 
 
 Wiedeman's method is independent of the internal 
 resistance, but to prevent errors arising from polarisa- 
 tion a high resistance, and therefore a sensitive 
 galvanometer, should be put in circuit ; also E and 
 E 2 should not differ materially in value. 
 
 With a good astatic galvanometer, Wheatstone's 
 method leaves little to be desired. It is rather 
 complex, but is independent of the internal resistance 
 of battery. 
 
 Poggendorffs method, especially as developed by 
 Latimer Clarke, has the great advantage that it is a 
 "null" method; the values are compared when no 
 current flows ; therefore there is no risk of polarisation, 
 and it is also entirely independent of the internal 
 resistance of the cells compared. 
 
 Measurement of Capacity. 
 
 For the measurement of capacities some standard 
 condenser is necessary. A £ or £ microfarad will be 
 very suitable ; and, supposing the student has access 
 to a standard condenser, he can easily make a copy for 
 his own future use. 
 
 Fig. 99. 
 
 Take two well-seasoned mahogany boards, 15in. by 
 12in., and bore with a gimlet a series of holes round 
 the four sides and at a distance of half an inch from 
 the edges. Next take some paper — that known as 
 "bank post" is very suitable — and after cutting it into 
 sheets, 14in. by llin., and drying them well, dip them, 
 one by one, into melted paraffin wax. Allow the sheets 
 to drain, and hang them up in a dry room to cool. 
 Now cut a series of sheets of tinfoil, 12in. by 10in., 
 and commence to build up the condenser (Figs. 98 
 and 99). 
 
§36 
 
 PRACTICAL ELkCfklCAL ENG/NEEX/lvG. 
 
 Place one of the boards on the table, and lay upon 
 it two sheets of the paraffined paper ; then a sheet of 
 tinfoil, laying on the sheet a strip of foil sufficiently 
 long to project over one of the edges of the paraffined 
 paper ; next, two more sheets of paraffined paper, then 
 a second sheet of foil with a projecting strip, but with 
 the strip this time turned towards the opposite edge of 
 the paraffined paper. Continue the process until some 40 
 sheets of foil have been laid, each sheet separated by two 
 layers of paraffined paper, taking care that the ends of the 
 strips belonging to the odd numbers lie over each other 
 on one side of the pile, and the corresponding strips 
 belonging to the even numbers forming another group 
 on the other side of the pile. Lastly, place two sheets 
 of paraffined paper on top of last foil, and bend the 
 whole of the " odd " strips up on to the top at one side, 
 and the whole of the " even " strips up on the top at the 
 opposite side ; folding the ends of each bundle of strips 
 round the ends of two short wires, pass these wires 
 through two holes in the top board, which must be 
 carefully placed on the top of the pile, and by means 
 of suitable screw clamps compress the whole firmly 
 together. This being accomplished proceed to compare 
 
 Fig. 100, 
 
 the capacity with that of a standard condenser by one 
 of the methods subsequently described ; should it be 
 found too small, unclamp and insert some more sheets 
 of foil and paper until on further trial the capacity of 
 the condenser is found to equal that of the standard. 
 
 When this is attained, screw the top and bottom 
 boards securely together by brass screws (in the holes 
 previously alluded to), and fill up the space between 
 the edges of the boards with melted paraffin. 
 
 The two wires coming through top board must now 
 be soldered to two brass terminals screwed on to the 
 board, and when not in use these terminals should be 
 connected by a loose brass strap. 
 
 Direct Discharge Method. — The kind of galvanometer 
 most suitable for the measurement of the capacity of 
 a condenser is Thomson's reflector, and before com- 
 mencing the determination the controlling magnet 
 should be turned so that its field tends to neutralise 
 that due to the earth. If this neutral point is arrived 
 at, the spot of light will not always return to zero- 
 hence, although the galvanometer is in its most sensitive 
 state when these conditions are fulfilled, it is necessary 
 
 to regulate the position of the magnet so that the 
 magnetism of the earth sligbtly predominates. 
 
 Having adjusted the galvanometer, connect it up 
 with a standard condenser (£ or J microfarad will be 
 sufficient), a good Daniell cell, and a Morse key, as in 
 Fig. 100. 
 
 On depressing K a current rushes into the condenser 
 charging it to the potential of the battery employed 
 accompanied by a momentary deflection of the spot of 
 light on the galvanometer scale. Note value of deflec- 
 tion, d\, produced, and continue holding the key down 
 until the spot of light has returned to zero, then release 
 key and the condenser will be discharged through the 
 galvanometer, the deflection, d 2 , this time being in the 
 opposite direction. If the galvanometer has been 
 correctly adjusted, d\ will equal d 2 ; should they slightly 
 differ the mean of d\ and a\ may be taken as the true 
 value, but if the discrepancy be considerable, proceed to 
 readjust the galvanometer and scale. Take the mean 
 of five or six sets of readings obtained in this manner 
 and call it D v 
 
 Now substitute for the standard condenser the one 
 whose capacity is to be measured, and take the mean 
 
 0=F - 
 
 Fig. 101. 
 
 of a similar number of readings obtained in the same 
 way as with the standard condenser. Calling this value 
 IX, and the capacities of the standard and trial con- 
 densers Cj and C 2 respectively, we have 
 
 C, : C 2 : : D : : D 2 
 
 or 
 
 Co = C ^ 
 
 Usually the galvanometer is placed as in Fig. 101, 
 in which case we only get a deflection when the con- 
 denser is discharged, but in all other respects the 
 manipulation is the same as in the previous arrange- 
 ment. 
 
 Since, in this method, the galvanometer is not 
 affected on charging, there is no necessity for holding 
 the key down while the spot of light returns to zero, 
 therefore a short, firm contact suffices to charge the 
 condenser, and the discharge may follow immediately— 
 a point of great importance, as we thereby prevent the 
 phenomena known as electrification. 
 
 Example. — Required, the capacity of a newly-made 
 condenser, C 2 , from following data, obtained by the 
 
MEASUREMENT. 
 
 337 
 
 " direct discharge method." Using two Daniell cells in 
 series, C a gave a deflection of 195 scale divisions, 
 whereas a standard "5 mf. condenser, C v gave under 
 similar conditions a deflection of 18 scale divisions. 
 
 •5 : C„ 
 
 18 : 195 
 
 C 2 = 
 
 •5x19-5 
 18 
 
 = -542 mf. 
 
 Thomson's Method. — This method, which is far more 
 accurate than the direct deflection method, is now very 
 generally employed. The theory of the process is very 
 clearly shown in the following diagram, which is but 
 slightly altered from that given in Messrs. Munro and 
 Jamieson's "Electrical Tables." 
 
 On closing Kj the poles of a well-insulated battery of 
 one or more Daniell cells are joined through the two 
 resistances, B, and B 2 , a * the points D, B. 
 
 Calling Vj and V 2 the potentials at junctions, D, B, of 
 the battery wires with the two resistances, B x and B 2 , 
 we have 
 
 Now depress simultaneously the two keys, ~K 3 and K 4 , 
 thereby charging the two condensers, C x and C 2 , respec- 
 tively to the two potentials, V a and V 2 . 
 
 If Cj and C 2 represent the capacities in microfarads 
 of these two condensers, and Qx and Q 2 the quantities 
 given to them, then 
 
 Q i: Q xVaO^VaC,. 
 
 Next release the two keys, K 3 and K 4 , and allow the 
 two opposite charges of + and - electricity to mix 
 through the wire, W, then if Qj = Q 2 on bringing the 
 galvanometer, G, into circuit by closing Kj, there will 
 be no deflection. Should, however, a deflection occur, 
 it shows the capacities are unequal, and the ratio of 
 Bj to B 2 must be changed until on trial no deflection is 
 produced. 
 
 Then V T C 1 =V 2 C„ 
 
 or 
 But 
 
 or 
 
 V 2 : V x : : C x : C 2 . 
 
 V 2 : V a : : E 2 : B lt 
 
 ' . B 2 : Ex : '. Cj : O . 
 
 C 2 =|iC 1 . 
 
 Where accuracy and dispatch are essential, some 
 form of key is necessary (such as Lambert's) which 
 combines the movements required for charging, mixing, 
 and discharging the condensers; but for ordinary 
 experimental work a Post Office Wheatstone bridge, 
 supplemented by two Morse keys, will be found suffi- 
 cient, as seen in Fig. 103, in which the connections are 
 lettered and numbered with those in Eg. 102 for the 
 sake of comparison. 
 
 To make a measurement : 
 
 (1) Arrange apparatus, as in Fig. 103, and commence 
 by taking out the 1,000-ohm plug between D and E. 
 Call this E x . Then unplug 2,000 ohms between E 
 and B, and proceed as follows =£*. 
 
 (2) Firmly close K x . This brings the battery in 
 circuit. 
 
 (3) Firmly close K 3 and K 4 simultaneously, so 
 charging both condensers, but with electricities of 
 opposite sign. 
 
 (4) Eelease K 3 and K 4 , and afterwards K v This 
 allows the two charges to mix, and cuts out the 
 battery at the same time. 
 
 (5) Depress K 2 and allow the surplus charge (if any) 
 to pass through galvanometer, Gr. 
 
 Fio. 102. 
 
 Note direction in which the deflection occurs, and 
 keeping Bj constant, make Bo, considerably less than 
 Ej, say = 100 ohms, and repeat the experiment. If the 
 deflection is now in the opposite direction, it shows 
 that the true value to be given to E 2 lies between 100 
 and 2,000. The resistance, B 2 , in E B must then be 
 varied until no deflection is produced. 
 
 Example. — In comparing the capacities of the two 
 condensers mentioned in last example, putting the 
 standard "5 mf. at C l3 and the unknown at C2rwith 
 B : = 1,000 ohms, it was necessary to make B 2 = 
 
 Fig. 103. 
 
 910 ohms, in order that no deflection should occur, 
 hence 
 
 Cj : C 2 : : E 2 : Bj. 
 
 C.-C3, 
 
 " 5- 910~' 
 = -549 mf. 
 
 A modification of the above method, given in Messrs. 
 G/lazebrook and Shaw's "Practical Physics,"p. 476, has 
 the great merit of considerably simplifying the manipu- 
 lation, and in which the resemblance to the Wheatstone 
 bridge method of measuring resistances becomes at 
 
338 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 once apparent. Fig. 104 shows the theoretical, and 
 Fig. 105 the practical arrangement, using the Post 
 Office bridge as in last method. 
 
 Join up as in Fig. 105 (K may be an ordinary Morse 
 key), the wire from B being joined to central terminal 
 of key, while the back terminal against which the key 
 rests in its normal condition is in connection with the 
 
 Fig. 104. 
 
 point a, thereby keeping the plates of the two con- 
 densers in metallic circuit, and therefore discharged, 
 while on depressing the key the battery is brought into 
 circuit and the condensers charged. Unplug resist- 
 ance E 1( and adjust E 2 , until no deflection occurs 
 when K is either closed or opened, then, as before, 
 G 1 : C 2 : : B 2 : Bj. 
 
 Fig. 105. 
 
 Example.— Employing the same two condensers as 
 in previous examples, putting the -5 mf. at C x and the 
 unknown at C 2 , and giving E x a value of 2,000 ohms, no 
 deflection occurred on closing or opening K when E 2 
 = 1,800 ohms; therefore 
 
 2 'B, X l,800~ 5 x -9— 55 ad- 
 measurement of the Combined Capacities of Two Condensers. 
 
 _ When n condensers are arranged in parallel or mul- 
 tiple arc, Fig. 106, similarly to the so-called quantity 
 
 Fig. 106. 
 
 arrangement with voltaic cells, the capacity of the 
 combination is the sum of the several individual 
 capacities, or 
 
 C-Cj + C 2 + ,etc. 
 
 but if they are joined up in series, Fig. 107, analogous 
 to the " cascade " arrangement with Leyden jars, then 
 the joint capacity is the reciprocal of the sum of the 
 reciprocals of the separate capacities, or 
 
 = 
 
 1 1 + 
 \- 1- < etc. 
 
 C, C, 
 
 Example. — When the two condensers previously 
 used, and which gave separately by the direct discharge 
 method deflections of 8 and 19 • 5 scale divisions respec- 
 
 Fig. 107. 
 
 tively, were joined "in parallel," the discharge reading 
 was 37 divisions ; but when joined up " in series " the 
 deflection was 9'5 divisions, the same battery being 
 used in each determination. 
 
 Calculating from above formula, the deflections 
 should have been : 
 
 In parallel, D = 18 + 19'5 = 37"5. 
 
 1 1 
 
 In series D = 
 
 ■0555 + -0512 -1067 
 
 = 9-4. 
 
 Measurement of Resistances by Fall of Potential. 
 
 If the two poles of a cell whose electromotive force = 
 E be joined by a wire of resistance E, then the fall of 
 potential may be shown graphically, as in Fig. 108. 
 
 If internal resistance of the cell, r, be small in com- 
 parison to E, it may be neglected, and only the differ- 
 ence of potential between the two battery terminals 
 taken into consideration. The fall will vary directly 
 with this external resistance, therefore, taking any two 
 points on this connecting wire, the resistance enclosed 
 between them will bear the same relation to the differ- 
 ences of potential existing between such points as the 
 total resistance of the wire bears to the total difference 
 of potential between the two poles of the battery. 
 
 Fig. 108. 
 
 Or, again, consider 
 A B, B C, then the 
 resistance, E 2 , of B 
 V-Vn between AB 
 orVj-V,. 
 
 E x : E 2 
 
 Hence if we have 
 potential differences, 
 can readily ascertain 
 
 the wire divided into two lengths, 
 
 resistance, B p of A B is to the 
 
 C as the difference of potential, 
 
 is to that existing between B C 
 
 the means of comparing these 
 and a known resistance, Bj, we 
 the value of B 2 . We can do this 
 
MEASUREMENT. 
 
 339 
 
 either by employing a Thomson quadrant electrometer 
 or a mirror galvanometer, having a high resistance in 
 circuit. Employing the galvanometer, and touching 
 the points A B on the connecting wire (so as to 
 put the galvanometer in parallel or derived circuit), 
 the current passing through the galvanometer, and 
 therefore the deflection, a\, will depend upon the 
 difference of the potentials at A and B. Similarly 
 touching B and C the deflection, d. 2 , will represent the 
 difference of potential between B C or 
 
 Ivj : B 2 ■ • "i • **i 
 
 To make a determination, join up a good Daniell 
 cell of very low resistance, the standard resistance, 
 Bj — which may conveniently consist of a moderately 
 thick iron wire of 1 ohm resistance stretched on a board 
 as in Fig. 96 — and the object, B 2 , whose resistance 
 has to he determined in simple circuit. This ensures 
 the same current passing through Bj and B 2 , Fig. 109. 
 
 Connect galvanometer wires to A B and note deflec- 
 tion, dj, with Bj = 1-ohm circuit ; remove to B C and 
 observe deflection, d? ; lastly, replace wires on A B and 
 take another reading, d^. 
 
 Bx : B 2 : : 
 
 Fig. 109. 
 
 The object of taking two readings at A B is to correct 
 any error arising from any possible alteration in the 
 electromotive force of the battery during the test. H 
 ds = d 1 it may be rejected, but should a discrepancy 
 appear, then the mean of the two values must be taken, 
 or 
 
 di + d & . 
 <T~ ' 2 
 
 This method is very applicable to the measurement 
 of low resistances, such as the armatures of dynamos, 
 etc. 
 
 Example. — A Daniell cell, a 1-ohm resistance, B 1( 
 and short length of German silver wire were joined in 
 circuit. On bringing the terminal wires from the 
 galvanometer, with a total of 18,000 ohms in circuit, 
 in contact with the ends of B 1; a deflection, d^, of 52 
 scale divisions was obtained. On touching the ends 
 of the German silver wire the deflection 0% = 100 
 divisions. 
 
 Bj '. E 2 '• ' d-± : d% 
 
 1 : x : : 52 : 100 
 1 x 100 
 52 
 
 = 1-92 ohms. 
 The ordinary bridge method gave for the same piece 
 of wire a resistance of T95 ohms, 
 
 x = 
 
 Joule's Law. 
 When the circuit of a voltaic battery is closed, 
 chemical action ensues, the fall in the potential energy 
 of the zinc and acid being accompanied by the evolu- 
 tion of an equivalent amount of energy in some other 
 form, such as mechanical work, chemical decomposition, 
 magnetism, light, or heat. Joule's law shows that the 
 heat developed in a conductor is directly proportional 
 to 
 
 (1) Its resistance ; 
 
 (2) To the square of the current strength ; and 
 
 (3) To the times during which the current flows ; 
 
 or, B.=G 2 B,t. 
 
 ■p ■ 
 Now since Ohm's law teaches us that C =— , if we 
 
 B 
 
 substitute this new equivalent for one of the C's in 
 
 the above equation, we get 
 
 H = ECU =EC ^ 
 
 or the amount of heat equals the product of the electro- 
 motive force, the current, and the time. 
 
 •p 
 Beplacing the remaining C by — , we obtain another 
 
 B 
 
 expression of the law — i.e.: 
 
 H JE Bf _E j f 
 BE B 
 
 from which we see that the heat developed equals the 
 product of the electromotive force and the time divided 
 by the resistance. 
 
 Finally, Q, the quantity of electricity flowing, equals 
 the current (or rate of flow) into the time = C t = 
 
 ^ t, and substituting in above equation for the value 
 B 
 
 "Cl 
 
 — t its equivalent Q, we see that H = E Q, or the heat 
 E 
 
 equals the electromotive force into the quantity. 
 
 The mechanical equivalent of heat, as demonstrated 
 by Joule, proves that the heat necessary to raise the 
 temperature of 1 gramme of water from deg. to 
 1 deg. centigrade, would suffice to lift the same weight 
 of water through a vertical height of 42,400 centimetres. 
 
 But to lift 1 gramme through 1 centimetre requires 
 an expenditure of 981 ergs, hence the mechanical 
 equivalent of heat expressed in ergs = 42,400 x 981 = 
 41,594,400. 
 
 The electrical unit of heat, or "joule," equals 10' 
 ergs. It is defined as the heat generated in one 
 second by a current of 1 ampere passing through a 
 
 resistance of 1 ohm = ^ =ij = 10 7 , or 10,000,000 ergs. 
 B 10 a 
 
 Assuming that the mechanical equivalent of heat 
 
 equals in round numbers 42,000,000 ergs, we see that 
 
 42 7 
 1 "ioule" would raise -— ='238 gramme of water 
 J 10 7 
 
 1 deg. C. 
 
 If, therefore, the current be expressed in amperes, 
 
 resistance in ohms, and the time in seconds, multiplying 
 
340 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 the right-hand side of equation by "238 gives the heat 
 generated in any conductor in ordinary thermal units, or 
 
 H = C 2 K£x-238. 
 
 The proper quantitative demonstration of Joule's 
 law requires an amount of apparatus, such as delicate 
 thermometers, calorimeters, and care in manipulation, 
 as places it beyond the scope of these " outlines," and 
 the student desirous of prosecuting the subject further 
 must consult some standard works on the subject, such 
 as those by Tyndall, Balfour Stewart, or Clerk-Maxwell, 
 but the facts on which the law is based may be experi- 
 mentally proved with sufficient accuracy by the 
 following simple piece of apparatus : 
 
 Take two ordinary test tubes, one about 6in. and 
 ljin. wide, the other 6in. long by fin., and fix the 
 smaller tube inside the larger by means of an annular 
 bung of indiarubber, Fig. 110, closing the mouth of the 
 
 Fig. 110. 
 
 smaller tube with a good bung of the same material, but 
 having two holes bored through it, one for the insertion 
 of the end of a piece of glass quill tubing— bent once at 
 right angles— forming an index tube, the other for a 
 short piece of similar tube closed by a short india- 
 rubber tube and pinchcock. Through the annular 
 bung are bored four holes, one wide enough to admit 
 the shank of a small funnel or pipet, the other three 
 sufficient to allow short lengths of thick copper wires 
 being passed tightly through a, b, and c. 
 
 A coil of thin platinum wire of exactly 2 ohms 
 resistance has its ends soldered to a, c, and its middle 
 to b — it is thus divided equally into two lengths of 
 1 ohm each. 
 
 The test tube should be supported on a small wooden 
 stand, and a cardboard scale placed upon the horizontal 
 tube. 
 
 To use the apparatus, pour into the space between 
 the two test tubes, by means of a small funnel or pipet, 
 a measured volume of water at a known temperature— 
 i.e., that of the room in which the experiment is made. 
 
 The quantity of water should be sufficient to cover the 
 whole of the platinum wire. Next, open pinchcock 
 and placing the mouth to the end of the indiarubber 
 tube, at the same time dipping the end of the index 
 tube into a little coloured water, gently exhaust the air 
 in the inner tube. This will draw a globule of the 
 coloured water along the tube forming an index. When 
 the globule has reached the zero of scale, close pinch- 
 cock, and any noncoincidence between the index and 
 zero can be rectified by slightly moving the scale along 
 the tube. 
 
 Connect up one or two constant cells of low internal 
 resistance, such as Grove's or Bunsen's, the resistance- 
 board, AB, Pig. 96, a tangent galvanometer and key, 
 to the thick copper wires, a and b. On closing K, a 
 current, C, equal tan d, will flow through the first half of 
 the platinum wire = 1 ohm, heating the water sur- 
 rounding the wire. This in its turn will heat the air 
 inside the inner tube, causing it to expand, and forcing 
 the index along the tube carrying the scale. 
 
 Note deflection, d, of tangent needle, time, t, during 
 which the current flows, and I the number of divisions 
 on the scale over which the index has moved. Open 
 K and proceed to make another determination, as 
 follows : 
 
 Eemove the warm water by means of a small syphon 
 made out of a piece of narrow quill tubing, through the 
 hole in the annular bung, and again pour in an equal 
 quantity at the same temperature as the previous 
 charge ; open pinchcock and bring index to zero as 
 before, close K again for a period = 2 t, when the 
 index will have moved through a distance approxi- 
 mately = 2 I. 
 
 It will not be exactly 2 I, for the outer glass vessel is 
 radiating heat all the time, and a correction would have 
 to be made accordingly, but the approximate value 
 will be quite near enough to show that " the heat 
 varies directly as the time." 
 
 Recharge the apparatus and connect the leading 
 wires to a and c, and send the same current as 
 indicated by the tangent galvanometer through the 
 two ohms of platinum wire — for an equal time, t, 
 the index will move through 2 I, showing that the 
 " heat varies directly as the resistance." 
 
 Since we have an additional resistance of 1 ohm 
 introduced between a c, we must, in order to keep 
 the current the same value, reduce the length of the 
 German silver wire (stretched on the board A B) in the 
 circuit, until the same deflection is produced = tan d. 
 
 Lastly, recharge the instrument, and arrange the 
 connection to include a and b, but vary the resistance 
 on the board, E, or, if need be, electromotive force, by 
 addition of other cells, until a deflection, d^, is produced 
 showing that the new current, C 2 , is equal to 2 C r If 
 this current, C 2 , flows for the original time, t, the index 
 will be moved through 4 time I, proving that the heat 
 varies as the square of the current. 
 
 The relative amounts of electromotive force and 
 resistances necessary for the production of Cj and C 2 
 when a and. b or a and c are in circuit may easily be 
 
MEASUREMENT. 
 
 341 
 
 ascertained by a few preliminary trials ; or an ordinary 
 wire of one or two ohms resistance may be inserted in 
 circuit instead of a, b, c, and when the exact conditions 
 for producing Ci, C 2 are attained, replacing the trial 
 wire by the platinum coils, a, b, c. 
 
 Students who desire to understand the whole theory 
 and practice of testing must, as we have said before, 
 read and study Mr. Kempe's excellent treatise, which 
 is far and away the best book on the subject ever 
 
 Fig. ill. 
 
 issued. In the ordinary work of a central station, or 
 an isolated electric lighting plant, less elaborate instru- 
 ments are used. Generally the switchboard contains 
 ammeters, fairly accurate and easily readable, always 
 in circuit, also voltmeters. The latter are not always 
 permanently in circuit, but can be switched on and off 
 as desired. As full details will be given later on it will 
 be unnecessary to do more here than describe some 
 types of instruments tc be found on the switchboards. 
 
 Fig. 112. 
 
 Ayrton and Perry Permanent Magnet Ammeters. — 
 In these instruments (Figs. Ill, 112, and 113) the shapes 
 of the magnet pole-pieces, needle, and coil are such as 
 to render the deflections proportional to the current 
 flowing. They are calibrated direct in amperes, and 
 the readings given consequently require no multiplica- 
 tion by a constant. The needle being placed in a 
 powerful magnetic field, the instruments are rendered 
 extremely dead-beat, and unlikely to be influenced by 
 the presence of dynamo machines. By the lightness 
 of the needle and the excellence of the jewelling and 
 
 pivoting great sensibility is obtained. These instru- 
 ments are made of two types, which are known 
 respectively as simple and commutator instruments. 
 In the commutator instrument the galvanometer coil 
 consists of 10 strands of wire, which may be placed 
 either in parallel or series by the simple turning of the 
 commutator, Jbig. 112. This gives 2 deg. of sensibility, 
 and enables the instrument, by the aid- of a resistance 
 coil inserted in it, to be calibrated with a current having 
 one-tenth of that which the instrument can measure. 
 An ampere-meter designed to measure 50 amperes 
 when the commutator is placed so as to put all of 
 the 10 strands of the coil in parallel, will give, when 
 it is moved to place them all in series, the same 
 deflection with a current of 5 amperes. In the 
 simple instruments there is no commutator. They 
 must, therefore, be calibrated from time to time by 
 comparison with a standard instrument, and adjusted 
 to read correctly. The adjustment is easily effected, 
 and is performed by moving the needle from a strong 
 part of the field to a weaker, or vice versa, Fig. 113. 
 
 Fig. 113. 
 
 The coil is supported by two screws, and by means 
 of nuts it can be moved as above described. On un- 
 screwing the base board the magnet and coil of the 
 instrument are exposed, and the adjustment can then 
 be made. 
 
 To calibrate the commutator ammeter turn the com- 
 mutator to series, and send a current from a standard 
 cell of known electromotive force through the ammeter, 
 obtaining a deflection D. Pull out the 1 ohm plug, 
 the deflection will now be reduced to D. If E is the 
 electromotive force of the cell in volts, then current 
 
 =E for deflection D, and 1 deg. gives E ~ 
 
 ampere sin series, or 10 E 
 
 D-D 
 
 Dd 
 
 amperes in parallel. 
 
 The adjustment of the coil must be made until the 
 desired value per degree is obtained. 
 
 To calibrate the simple ammeter place it in series 
 with a standard instrument having about the same 
 range, and let the current flow through both, adjusting 
 the coil of the former until the desired value per division 
 is obtained. 
 
342 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Simple Ammeters with Permanent Magnets. — This, 
 as the name implies, is a simpler and less expensive 
 form of instrument, suitable for indicating the current 
 under circumstances in which the great accuracy 
 obtained by the use of the more expensive types is 
 not required (Fig. 114). The deflection of the pointer 
 is proportional to the current flowing; the dial is 
 marked direct in amperes, and the magnet is fitted 
 with adjusting screws for recalibration. 
 
 Electromagnet Instruments. — Instruments in which 
 the permanent magnets are replaced by electro- 
 
 Fig. 114. 
 
 magnets have come greatly into favour in recent 
 years, for the very obvious reason that their con- 
 struction does not necessitate the employment of a 
 magnet which, though called permanent, is liable 
 to a change of strength with time. In the electro- 
 magnet instrument as usually made, the same circuit 
 which produces the deflection by means of a galvano- 
 meter coil circulates round a soft iron bar, and so 
 produces an electromagnet to act as the controlling 
 force. But in instruments so designed the controlling 
 
 Fig. 115. 
 
 force is necessarily small, or, in other words, the field 
 is weak, and besides the disadvantage which the 
 instruments have in being very far from dead-beat, they 
 are extremely likely to be influenced by the presence 
 of dynamo machines near them. In Paterson and 
 Cooper's instruments these defects have been over- 
 come, and this has been done by substituting for 
 the weak force exerted by an electromagnet the 
 powerful force of a spiral spring. The deflecting 
 soft iron needle attached to the spring forms 
 part of a magnetic circuit, and in its normal position 
 
 lies nearly at right angles to a soft iron ring, which 
 comprises the other part. On this ring the wire 
 through which the current is to flow is wound, and 
 when the ring is magnetised by the current the needle 
 tends to set itself in such a way as to close the magnetic 
 circuit. ' Since its movement to effect this is against 
 the spiral spring, and the deflection actually observed 
 is a measure of the force exerted by the current on the 
 spring, and therefore the dial being properly graduated 
 
 Tig. 116. 
 
 of the current flowing. Fig. 115 shows the ordinary 
 form of the instrument for measuring currents up to 
 400 amperes, the magnetic system with the controlling 
 spring being plainly visible. Fig. 116 is a modification 
 of the instrument suited for the measurement of large 
 currents, the wire carrying the current in the smaller 
 instrument being here replaced by a single bar of 
 copper ljin. in diameter, the soft iron ring nearly 
 embracing this bar and leaving only sufficient space 
 
 between its open ends for the soft iron deflecting needle 
 to lie in. This type of the instrument is made to 
 measure currents from 500 to 2,000 amperes. 
 
 Ayrton and Perry Permanent Magnet Voltmeters 
 (Figs. 117 andl_ :_- ~_ '. ■ r '■■ 3 
 
 of the magnet p-«»,e-u»«^H. m -«./;.-. *,,,.: ._.„j 
 
 as to render the cz~zv;:..c:zz "crrr.:" .;.',;..:.; ; r - ~.-- r: ..„t 
 
 flowing. They a_, vii...i,ia',.'..'.. v.-v. .,» ,r, .i . .3 
 
 readings given consequently require no multiplication 
 by a constant. The needle being placed in a powerful 
 magnetic field, the instruments are rendered extremely 
 
MEASUREMENT. 
 
 343 
 
 dead-beat aud unlikely to be influenced by the presence 
 of d3*namo machines. By the lightness of the needle 
 and the excellence of the jewelling and pivoting great 
 sensibility is obtained. These instruments are made 
 of two tvpes, which are known respectively as simple 
 and commutator instruments. In the commutator 
 instrument the galvanometer coil consists of 10 strands 
 of wire, which may be placed either in parallel or 
 series by the simple turning of the commutator. This 
 gives 2 deg. of sensibility, and enables the instrument, 
 by the aid of a resistance coil inserted in it, to be 
 calibrated with a voltage equal to one-tenth of that 
 which the instrument can measure. A voltmeter 
 designed to measure 100 volts when the commutator is 
 placed so as to put all of the 10 strands of the coil in 
 series, will give, when it is moved to place them all 
 in parallel, the same deflection with a difference of 
 potential of 10 volts. In the simple instruments there 
 is no commutator. They must, therefore, be calibrated 
 from time to time by comparison with a standard 
 instrument and adjusted to read correctly. The adjust- 
 
 Gardew Voltmeter. — This instrument, Fig. 119, has 
 been greatly improved in points of detail from time to 
 time, and may now be considered as perfect as this 
 diss of apparatus can be made. Essentially the 
 Cardew voltmeter consists of a tine platinum silver 
 wire '0025in. in diameter, which is heated by 
 the current passing through it. The wire is kept 
 stretched by a spring, but expands on being 
 heated, the expansion given by different tempera- 
 tures due to different potentials being indicated 
 b;v a pointer in collection with suitable gearing. 
 Since there are no magnets or coils in the instru- 
 ment, and no retardation due to self-induction, 
 it is equally applicable to the measurement of alter- 
 nating current; in fact, it is the only instrument that 
 lor the latter kind of work gives a reliable reading. 
 The dial is graduated direct in volts. When there is 
 
 Fig. 318. 
 
 Fig. 119, 
 
 raent is easily effected, and is performed by moving the 
 needle from a strong part of the field to a weaker, or 
 vice versa. 
 
 To calibrate the commutator voltmeter turn the 
 commutator to parallel, and send a current from a 
 standard cell of known electromotive force through the 
 instrument, a deflection D will be obtained. Pull out 
 the plug of the resistance coil, and a new deflection D 
 will be given. If E is the electromotive force of the 
 standard cell, then the difference of potential at the 
 
 terminals of the instrument = E - "—volts for deflection 
 
 D 
 
 D, and 1 deg. gives E — — in parallel, or 10 E 
 
 volts in series. The adjustment of the coil must be 
 made until the desired value per division is obtained. 
 
 To calibrate the simple voltmeter place it in parallel 
 with a standard instrument having about the same 
 range, and let the current flow through both, adjusting 
 the coil of the former until the desired value per degree 
 is obtained. 
 
 an alternating difference of potentials, the indication 
 shows the square root o:' the mean square of the volts. 
 In the instrument there are four lengths of wire, 
 making about 12ft. in all, these forming two continuous 
 lengths of 6ft., which pass over two small pulleys at 
 the top of the tube, whose pivots turn in jewelled 
 centres. An instrument having 12ft. of wire will 
 measure a difference of potentials of 120 volts, this 
 being 10 volts per foot. To save the working wire from 
 fusion, a still finer wire, 'OOloin. diameter, is inserted 
 in the path of the current. 
 
 Pocket Voltmeter.— These instruments, Fig. 120, no 
 larger than a good-sized watch, are very convenient for 
 the measurement of small differences of potentials. 
 Though sometimes wound for 120 volts, they are really 
 not recommended by the makers for more than horn 
 SO to 100. They are constructed with permanent steel 
 magnets, and connection is made to the points between 
 which it is desired to illustrate the difference of 
 potentials by flexible silk-braided wires supplied with 
 the instrument. The graduation on the dial is in volts 
 
344 
 
 PRACTICAL kLBCtklCA'L EttGlNkEklNd. 
 
 and the instrument is particularly applicable for testing 
 the electromotive force of accumulators. 
 
 Ordinary Lineman's Detector. — This handy instru- 
 ment, Fig. 121, consists of an astatic combination with 
 
 Fig. 120. 
 
 the axle lying horizontally, one needle serving as a 
 pointer, the other moving between two coils inside 
 the mahogany case. The system is slightly weighted 
 
 Fig. 121. 
 
 to give the needles a vertical bias. In the detector 
 illustrated there are only shown two terminals for 
 a low-resistance winding, but the instrument is 
 
 Fro. 122. 
 
 frequently supplied with three terminals being wound 
 with two coils* one of high and the other of low 
 resistance. 
 
 Portable Galvanometer— This instrument, Pig. 122, 
 is very sensitive in its action. The needle has an 
 
 agate cup inserted at its centre and the moving 
 system is balanced on a polished steel point, the 
 deflection being read off by means of an aluminium 
 pointer. The instrument has a brass case, and is 
 fitted with a bevelled plate-glass top. It is very 
 suitable for portable use with a Wheatstone's bridge, 
 where the ordinary Thomson galvanometer would be 
 useless on account of the difficulty in setting it up. 
 
 Some instrument makers have devised portable 
 testing sets. The best known are those of Mr. 
 Blackburne, made- by Messrs. Clark, Muirhead, and 
 Co., and the Silvertown instrument. The latter may be 
 described, and the instructions issued with it given. 
 
 The Silvertown Portable Testing Set. 
 
 This set consists of a small collection of the necessary 
 instruments for testing electrically such insulated con- 
 ductors as are used in telegraph, telephone, or electric 
 light work. 
 
 The whole set is contained in two small wooden 
 boxes, of which one holds the batteries and the other 
 the galvanometer, resistance coils, key, and commu- 
 tators required for making the two most important 
 measurements on such circuits. These are measure- 
 ments of the resistance of the conductor, and the 
 efficiency of the insulation. 
 
 For connecting the battery to the testing instruments 
 convenient leads are provided terminated in brass plugs 
 with insulated handles for inserting in the proper plug 
 holes. The instruments are connected up together in 
 their own box in such a way as to secure the greatest 
 economy of space, and to enable the two tests to be 
 taken with the greatest readiness. 
 
 The battery consists of two parts : one— commonly 
 called the bridge battery— is a set of Leclanche cells of 
 low resistance intended to be used in testing conductor 
 resistances only, a purpose for which currents of 
 electricity of sensible magnitude are required. The 
 other part is a set of 30 small Leclanche cells having 
 a total electromotive force of 45"50 volts intended 
 exclusively for measuring insulation resistances, or 
 other resistances of considerable magnitude. These 
 cells are designed to give only very small currents of 
 electricity, and care should be taken not to connect 
 them inadvertently to the Wheatstone's bridge or 
 otherwise put them on a circuit of low resistance. 
 'Ibis battery, called the insulation battery, is subdivided 
 into three sections of 10 cells each, so that electro- 
 motive forces of about 15 volts, 30 volts, or 45 volts 
 can be employed as may be found convenient. 
 
 Of the box containing the instruments, the diagram, 
 Fig. 123, shows all the connections and resistance 
 coils. 
 
 The only part of the instrument which requires 
 detailed description is the galvanometer. This consists 
 of a coil of fine wire on a brass bobbin, in the centre of 
 which a small magnetic needle with an aluminium 
 pointer is hung in the same way as is usual in 
 compasses. The pointer projects through the opening 
 in the end of the coil, and the excursions of the needle 
 
MEASUREMENT. 
 
 m 
 
 are limited by the size of the opening to about 45 deg. 
 on each side of the centre. On removing the glass 
 cover, the needle on its point may be taken out by 
 withdrawing the slide on which it is pivoted from 
 inside the coil. The scale, which is a scale of equal 
 currents, is approximately a scale of tangents, and is 
 obtained empirically by calibrating the instrument. 
 The north end of the magnetic needle points to the 
 left-hand side of the box when it is swinging freely in 
 its zero position. 
 
 On the left-hand side of the box is placed the con- 
 trolling magnet, and the position of this affects the 
 sensitiveness of the galvanometer. When the north 
 pole of the controlling magnet is uppermost, the 
 galvanometer will be most sensitive ; on turning the 
 magnet round so that the south pole is uppermost, the 
 deflection of the needle due to any given current will 
 
 The pointer of the galvanometer will then be found to 
 be swinging near its zero, and may be brought exactly 
 to it by slightly turning the controlling magnet. 
 
 If at any time the galvanometer needle should 
 become insensitive and sluggish it may be due to one 
 of several causes. 
 
 It may be that the needle has become demagnetised. 
 This can be remedied by withdrawing and remagne- 
 tising it with an ordinary horseshoe magnet, care being 
 taken that this is done in the same direction as before. 
 
 It may be that some dirt has found its way into the 
 jewel. This may be removed with a piece of soft wood 
 cut to a fine point. 
 
 It may be that the jewel or needle point is injured, 
 and must be repaired by competent workmen. This 
 will probably have occurred either through the whole 
 instrument having received a blow when the lid is open 
 
 IrloOOO: 
 
 £3 0. 
 
 ( \ INSULT! i 
 
 IN5UU-- 1 § 
 
 '• r " , -™WL ^^^V-^-A 1 -"' INSULA 
 
 Fig. 123. 
 
 be reduced by about 40 per cent. Generally in testing 
 the insulation of well-insulated wires, the galvanometer 
 is required to be as sensitive as possible, and the north 
 pole of the controlling magnet should be at the top ; 
 but for measuring conductor resistances, for which the 
 galvanometer is generally amply sensitive, it will be 
 found more convenient to bring the south pole upper- 
 most, thereby causing the galvanometer needle to 
 oscillate more rapidly. 
 
 Besides thus affecting the sensitiveness of the 
 galvanometer, the magnet is also used to adjust the 
 needle to the zero in its position of rest by turning it 
 slightly in one direction or the other. 
 
 In preparing the test the box should be placed on a 
 table, or some other approximately level surface in 
 front of the electrician, he facing the magnetic east, 
 and the controlling magnet being in a vertical position. 
 
 and the jewel resting on the needle point, or through 
 the brass spring in the lid of the box being bent so that 
 it no longer presses on the lifter when the lid is closed, 
 and the needle has consequently been resting on the 
 point while the box has been carried about. 
 
 Testing Conductor Resistance. 
 
 The method used of measuring the resistance of the 
 conductor of the circuit under examination is that of 
 Wheatstone's bridge. 
 
 The diagram, Fig. 124, shows only those parts of 
 the instrument which are employed in this test, and 
 omits the parts, and their connections, which relate 
 only to insulation testing. The parts employed are 
 the following : 
 
 The adjustable resistance. This, it will be seen, 
 consists of two sets of nine coils each connected to 
 
346 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 circular plug commutators or dials. One set of coils 
 has nine resistances of 10 ohms each, making 90 ohms 
 in all; the other has nine resistances of 1 ohm each, 
 making 9 ohms in all. If the hole marked with 
 any number— say, 5— is plugged in the 10-ohm dial, 
 a resistance of 50 ohms is inserted between the con- 
 necting leads entering and leading away from the 
 dial; and a similar rule applies to the 1-ohm dial. 
 Hence, if the hole 6 be plugged in the 10's dial and 
 the hole 8 be plugged in the units dial, a total 
 resistance is inserted in the two in series of 68 ohms. 
 The lowest resistance that can be obtained is given 
 when both the holes are plugged, when the coil 
 resistance inserted is zero. The highest resistance 
 is obtained by plugging the two 9 holes, when the 
 total resistance is 99 ohms. If no plug is inserted in 
 
 the two ends of the Wheatstone's bridge by depressing 
 the contact key. It will be noticed that the shunt 
 coils, with their plug commutator, are omitted from 
 the diagram. This is done because they are not 
 essential to the test, though they may be conveniently 
 used when the balance of the bridge is not yet approxi- 
 mately correct, and very large deflections are being 
 obtained. 
 
 The battery, as has been already described, consists 
 of four Leclanche cells, having an electromotive force 
 of about six volts. One pole of the battery is connected 
 in the usual way to the middle of the Wheatstone's 
 bridge, and the other to the point where the end of the 
 adjustable dial coils is connected to one of the 
 terminals, to which the conductor under test is 
 attached. The connections are made by inserting the 
 
 9^^m^^&^- 
 
 r/8-.', 
 
 TENS Jr=j\ '.j=l UNITS JhH'- 1 
 
 
 to bridge battery. 
 
 Co tht ends of tht conductor 
 
 Fig. 124. 
 
 one or both dials, the circuit is broken, and the resist- 
 ance is infinity. 
 
 The second part of the apparatus is the double set 
 of proportional resistances, consisting of two coils of 
 10 ohms each, two of 100 ohms, and two of 1,000 
 ohms, these constituting what is known as Wheat- 
 stone's bridge. Of these only one on each side of the 
 centre is to be unplugged for any given test, and a rule 
 is given later on for selecting the resistances to be 
 employed to obtain the greatest possible sensitiveness ; 
 that is to say, for selecting those coils which will give 
 the largest deflection on the galvanometer, when the 
 resistance plugged in the dials varies by a given error 
 from that of the circuit under test. 
 
 The third part is the galvanometer, which has been 
 already described. Its two terminals are connected to 
 
 plugs at the ends of the battery leads, in the two holes 
 marked " bridge," and immediately this is done the 
 current is established in the coils ; the galvanometer 
 circuit is of course not completed till the key is 
 depressed. 
 
 The ends of the conductor to be tested are to be 
 secured under the two terminals marked " bridge 
 terminals," and in measuring low resistances care must 
 be taken that they are very securely attached. 
 
 The test is begun by selecting the coils to be 
 unplugged in the Wheatstone's bridge ; to do this it is 
 necessary to know approximately the value of the 
 resistance to be measured. Generally some idea of its 
 value can be formed, but if it is quite unknown, two 
 coils on the bridge may be chosen at random and a 
 preliminary measurement made. This measurement 
 
MEASUREMENT. 
 
 34? 
 
 will enable the correct coils to be chosen for a further 
 test. The following pairs of coils are to be selected : 
 
 For resistances between 1 ohm and 10 ohms, left- 
 hand coil, 100 ohms ; right-hand coil, 10 ohms. 
 
 For resistances between 10 ohms and 100 ohms, left- 
 hand coil, 100 ohms; right-hand coil, 100 ohms. 
 
 For resistances between 100 ohms and 1,000 ohms, 
 left-hand coil, 100 ohms ; right-hand coil, 1,000 ohms. 
 
 In all these cases coils will be employed in both 
 dials, and a result giving two significant figures will 
 be obtained ; a third figure can always be found 
 in measuring resistances between these limits — viz., 
 between 1 ohm and 1,000 ohms — by observing the 
 deflections of the galvanometer needle on both sides of 
 the zero for different adjustments of the dial resistances 
 near the balancing point. 
 
 For example, we will suppose that the 10-ohm coil in 
 the right-hand side of the bridge and the 100-ohm coil 
 on the left-hand side are unplugged, and that when 
 
 45 ohms are plugged in the dials, and the key 
 depressed, a throw of three divisions of the galvano- 
 meter needle is observed to the right ; and when 
 
 46 ohms are plugged we get a throw of two divisions to 
 the left on the galvanometer scale. It is clear that the 
 resistance to be measured lies between 4 - 5 and 46, and 
 is nearer to 4"6 than 45, as two is less than three; 
 that is, the resistance is 4 - 56 ohms. As a further 
 example, suppose 100 ohms to be unplugged on each 
 side of the bridge, and 82 ohms to be plugged in the 
 dials ; on depressing the key, no deflection of the 
 needle is observed. On plugging 81 ohms in the dials, 
 a throw of six divisions to the right is obtained, and on 
 plugging 83 ohms we get the same deflection to the 
 left. We are then amply justified in putting the third 
 figure in the result as 0, and the resistance to be 
 measured is 82'0 ohms. 
 
 Eesistance from "1 ohm to 1 ohm may be measured 
 either with the same bridge coils as are used for resist- 
 ances of 1-10 ohms — viz., 10 ohms in the right side, 
 and 100 ohms on the left, but only to two significant 
 figures, since the 10's dial will be plugged at 0. 
 
 Eesistances of from "1 ohm to 1 ohm may also be 
 measured to three figures by using the bridge coils of 
 1,000 on the left side and 10 on the right, and 
 resistances from 1,000 ohms to 10,000 ohms by rising 
 the bridge coils of 10 on the left and 1,000 on the 
 right. For both of these tests, however, more battery 
 power is required than is provided in the ordinary 
 portable battery supplied. 
 
 For making this test attention may be called to the 
 following points, some of which have been noticed 
 before. 
 
 Except in testing at the extreme range of the 
 instrument — i.e., quantities less than 1 ohm or 
 greater than 1,000 ohms — the galvanometer will be 
 found amply sensitive, and it is better to place the 
 south end of the controlling magnet uppermost, thereby 
 reducing the time of the oscillations of the galvano- 
 meter needle. 
 
 The battery should be in circuit as short a time as 
 
 possible to avoid running down the cells, and it is well 
 to take out one of the battery lead plugs when any 
 alterations are being made in the plug commutators, 
 only replacing it just before pressing the galvanometer 
 key. 
 
 Care should be taken to connect the conductor to be 
 tested very securely to the bridge terminals. This 
 may be done, for very large or stranded conductors, 
 either by soldering to their ends thin brass plates with 
 holes in them of a suitable size to go under the heads 
 of the terminals, or the connection may be made by 
 means of finer wires soldered to the ends of the main 
 conductor. The resistance of these must be inde- 
 pendently ascertained and subtracted from the gross 
 result. 
 
 Measurement of Insulation Resistance. 
 
 This is a measurement of the electrical resistance of 
 the insulating material of a cable to the passage of a 
 current from the inside conductor through the insulation 
 to the lead sheathing, wet yarn, armour, or other 
 outside conducting surface, and the inverse of the 
 insulation resistance is the insulation conductivity, or, 
 as it is generally termed, the leakage. 
 
 This measurement is effected by a method known 
 as that of direct deflections. It consists in passing a 
 current from a battery through a galvanometer into 
 the conductor of a cable whose farther end is free and 
 disconnected, thence through the insulating material 
 to the outside coating or earth, and so back along a 
 temporary conductor to the other end of the battery, 
 the deflection of the galvanometer needle produced by 
 this current being noted. Eeplacing that part of the 
 circuit which was formed by the insulating material of 
 the cable by a standard resistance of known value, we 
 obtain a new deflection of the galvanometer needle. 
 
 The quotient obtained by dividing the deflection 
 produced by the current through the standard resis- 
 tance by that through the cable insulation is a measure 
 of the insulation in terms of the standard. 
 
 Thus, for example, suppose that a given battery 
 produces on the needle of a galvanometer placed in 
 series with the insulation of a cable in the manner 
 described, a deflection of 103 divisions, and that on 
 substituting a resistance of 1 megohm for the insulation 
 we get 42 divisions, we find that the insulation resistance 
 ij ^.^ megohms = 41 megohms approximately. 
 
 The diagram, Fig. 125, shows only those parts of 
 the apparatus and their connections that are used in 
 this measurement, those which relate only to the 
 measurement of conductor resistances being omitted. 
 
 The arrangement, it will be seen, is as follows : One 
 pole of the battery — a battery of 30 Leclanche cells 
 giving an electromotive force of about 45 volts is 
 normally employed — is connected by a conductor 
 ending in an ebonite-headed plug to the lower of 
 the two plugholes marked " insulation." Thence the 
 current passes along a connecting wire to a block 
 marked "shunts," and thence through the galvano- 
 meter to the upper block on the other side ; we may 
 
348 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 observe, in passing, that these two main blocks, one on 
 each side, are practically the terminals of the galvano- 
 meter. If a shunt is plugged, |th, ^th, or xi^th only 
 of the current passes through the galvanometer, the 
 remainder finding its way through the corresponding 
 shunt coil. 
 
 From the upper block on the left-hand side, the 
 current may take two paths, according as the hole 
 marked " insulation," or that marked 10,000 ohms is 
 plugged ; if neither is plugged, the circuit is broken, 
 and no current can pass. This plug forms conse- 
 quently a convenient make and break key. If the hole 
 marked " insulation " is plugged, the current passes to 
 the terminal marked "insulation," so through the 
 insulating covering of the cable to the outside sheathing 
 or earth, back to the terminal marked "earth" and 
 
 produced by the current passing through a known 
 resistance — this is called taking the constant of the 
 galvanometer. It will be found in practice that with 
 the battery of 30 cells supplied, it is necessary to use 
 the highest shunt power available, giving a multiplying 
 power of 100 ; and we may note here that the passage 
 of the current through a galvanometer shunted thus 
 and then through 10,000 ohms, gives the same deflec- 
 tion as passing the whole current through the galvano- 
 meter not shunted, and through a constant of 1,000,000 
 ohms. In short, the deflection thus obtained is the 
 deflection given by the battery through 1 megohm. 
 
 (2) In transferring the plug to the hole marked 
 " insulation," and again noting the deflection obtained, 
 the galvanometer being shunted if necessary to give a 
 convenient reading. 
 
 conductor 
 
 Co -edTtk or skcatking. 
 
 ulahon OatttVii 
 
 Fig. 125. 
 
 the plughole marked E, and then along the lead to the 
 other pole of the battery. If, however, the hole marked 
 10,000 ohms be plugged, the current will pass through 
 a coil of 10,000 ohms, then along a connecting wire to 
 the plughole B, and so back to the battery. 
 
 In beginning this test, the conductor of the cable, or 
 insulated wire, or a temporary lead attached to it, is 
 connected to the terminal of the instrument marked 
 "insulation" and another lead, connected to the 
 outside sheathing of the cable, or the wet soil in which 
 it lies, is attached to the terminal marked " earth," 
 care being taken that these leads are separated, and 
 that no circuit exists between them except through the 
 insulation of the cable. The test then consists in— 
 
 (1) Noting the deflection obtained when the 10,000- 
 ohm hole is plugged— i.e., obtaining the deflection 
 
 On dividing the deflection on the scale obtained when 
 taking the constant by the deflection obtained in the 
 second part of the test, and multiplying the deflections 
 by the shunt or shunts employed, we obtain the 
 insulation resistance of the cable in megohms. 
 
 For example. Suppose that the current from the 
 battery when passed through a constant resistance of 
 10,000 ohms gave a deflection of 42 divisions on a 
 galvanometer shunted to i^v, and that when passed 
 through the cable insulation it gave 23 divisions with 
 the galvanometer shunted to |, the insulation resistance 
 
 would be j3^-j megohms = - 37 megohms approximately. 
 
 For another example : If having found the same 
 galvanometer constant, we obtained from the cable 
 insulation a deflection of 10 divisions with no shunt 
 
MEASUREMENT. 
 
 349 
 
 employed, the insulation resistance of the cable would 
 be yf megohms = 4A megohms. 
 
 It will be observed that when all the holes in the 
 straight commutator near the front of the box are 
 plugged, that the key on the left-hand side, which is 
 used in the bridge test as a galvanometer make and 
 break key, becomes for the insulation test a short- 
 circuit key, and is useful for checking quickly the 
 oscillations of the needle. 
 
 In making this test the following points may be 
 called attention to : 
 
 (1) Too much care cannot be taken in preparing the 
 ends of the cable. Since we are measuring a very 
 small current of electricity passing from the conductor 
 to the outside sheathing, through the insulating 
 covering, it is clear that our results will be entirely 
 misleading if any current be allowed to pass over a 
 dirty surface at the ends where the conductor is 
 exposed. These ends should be looked to before 
 testing, and in the case of indiarubber or other firm 
 material, the section of the insulator should be pared 
 all over with a sharp and perfectly clean knife. 
 
 (2) Care should be taken not to short-circuit the 
 battery, which may easily occur in two ways. One 
 is by allowing the two battery plugs to touch one 
 another, when the other ends of the leads are attached 
 to the battery terminals ; and another is by allowing 
 the lead attached to the earth terminal to touch that 
 attached to the insulation terminal. 
 
 In both cases the battery of small cells will be for a 
 time much overworked, and in the second the needle 
 may become bent or demagnetised. 
 
 (3) Another point that may be noticed is, that in 
 deducing the insulation resistance per statute mile 
 from a test on any given length, the result obtained 
 from a test on the latter is to be multiplied by the 
 length of the piece in miles, and not divided by it. 
 
 For example, if the insulation of a cable 3 miles 
 long be 15 megohms, the insulation per mile will be 
 15 x 3, or 45 megohms ; or, again, if the insulation of 
 a piece of cable, whose length is 350 yards, be 7,520 
 megohms, the insulation per statute mile will be 
 7,5 ^4 85 ° megohms = 1,490 megohms. 
 
 From the switchboard, which, with the apparatus it 
 carries, has now been described, are led the main cables. 
 Of such cables the Silvertown Company says : 
 
 The accurate testing of electric light mains before 
 and after laying, and periodically during use, is a 
 matter of great importance if the efficient working of 
 a system is to be obtained and maintained. 
 
 It is surprising that, while the experience of 
 manufacturers of submarine cables has resulted in a 
 thorough and systematic method of electrical testing 
 being adopted, yet electric light engineers are in many 
 cases content to lay down expensive mains without 
 any reliable tests having been made, and they are 
 therefore completely ignorant of the true value of the 
 insulation resistance, which can only be obtained by a 
 careful test made while the cable is immersed in water. 
 
 As a result breakdowns may occur, due to faults which 
 in all probability existed in embryo at the time of 
 laying the mains, and which are developed under the 
 strain of working. 
 
 New methods of insulating, laying, and jointing con- 
 ductors are being continually devised, with small effect 
 beyond adding to an already dearly-bought experience; 
 yet there undoubtedly exist methods which, when care- 
 fully carried out, have given very satisfactory results 
 when worked under the heavy strain produced by high- 
 tension alternating currents. 
 
 Very frequently when the daily tests show a con- 
 tinually decreasing insulation resistance, the station 
 electrician comes to the conclusion that the cables are 
 giving way, and is inclined to wait for a dead earth 
 to develop before seeking the fault ; he should, how- 
 ever, set to work systematically to find out where the 
 fault lies, and to remove it before the cable becomes 
 unworkable. 
 
 To do this it becomes necessary to carry out tests in 
 a systematic manner to ascertain the true condition of 
 the mains before and after laying and jointing, so that 
 it may be quite certain that, at all events when ready 
 for work, the whole system of insulated conductors 
 may be as perfect as possible. Unfortunately but little 
 can be said with respect to the localising of a fault in 
 distributing mains or cables having numerous ramifica- 
 tions, but it is quite possible to localise a fault in a 
 feeder main or other length without branches. 
 
 For the purpose of accurate testing, each station 
 should be provided with a sensitive high-resistance 
 mirror galvanometer, a shunt-box, a resistance coil of 
 not less than 10,000 ohms (if possible a megohm), a 
 condenser, keys, and a battery of not less than 100 
 Leclanche cells. For tests during laying this battery 
 should consist of 400 to 600 cells. 
 
 The shunt-boxes usually supplied with the mirror 
 galvanometers are fitted with shunts having for that 
 particular galvanometer multiplying powers of 10, 100, 
 and 1,000. For the varying insulation resistances 
 which have to be measured in electric light testing, 
 a resistance-box having coils of units, tens, hundreds, 
 and thousands is most convenient for a galvanometer 
 shunt-box, as it enables work to be done between very 
 wide limits, and the obtaining of the deflection upon 
 any part of the galvanometer scale. 
 
 The instruments should be erected and adjusted 
 before the work of laying or pulling in is commenced, 
 so that tests may be taken on the arrival of the drums 
 of cable and the specified insulation resistance assured. 
 
 The test should be taken with the drums of cable 
 immersed in water wherever possible, but where this 
 cannot be done they must be well wetted. 
 
 The preparation of the ends of the cable is of the 
 utmost importance if the real insulation resistance of 
 the dielectric is to be obtained, the difference in the 
 results given with properly and improperly prepared 
 ends being very great. 
 
 For vulcanised rubber cables the tapes, braiding, or 
 other covering should be removed for at least 6in, 
 
350 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 from the ends, and the tape next the rubber should be 
 carefully removed, the surface of the rubber washed 
 with naphtha, and well scraped to remove any 
 fragments of the tape. The rubber should be tapered 
 off with a pair of clean bent scissors, the length of the 
 taper being lin. to 2in. Next thoroughly dry the 
 whole of the prepared end by means of a spirit lamp, 
 allowing the flame to play round the rubber, care, 
 however, being taken not to burn it. If the test 
 cannot be taken immediately, the prepared end should 
 be well warmed and lapped with clean warm pure 
 indiarubber tape well stretched. 
 
 The most accurate and convenient method of 
 obtaining the insulation resistance is by means of the 
 direct deflection method — i.e., by a comparison of the 
 deflection of the galvanometer needle produced by a 
 given battery power working through a known high 
 resistance with the deflection produced by the same 
 battery power working through the insulation of the 
 cable. 
 
 First, to take the constant — i.e., the deflection when 
 the battery is connected up through the galvanometer 
 and known high resistance. The galvanometer (with 
 its shunt) is connected in series with the battery and 
 known high resistance, and the circuit being completed 
 through a suitable key the shunt is adjusted to give 
 a convenient deflection, and this deflection noted — 
 call it 0. 
 
 If the high resistance is less than half a megohm, it 
 will be found impossible to keep the spot of light on 
 the scale when the full testing battery is connected up, 
 unless a shunt of very low value is used, and this is 
 undesirable. It is therefore necessary to take the 
 constant with a portion only of the battery, the 
 number of cells used depending upon the value of the 
 high resistance ; if a 10,000-ohm coil, one cell will be 
 sufficient. Having obtained a deflection, it is necessary 
 to take the battery ratio — i.e., the ratio between the 
 portion of the battery used for the constant and the full 
 battery to be used for the actual test. For this purpose 
 a condenser having a capacity of one-third microfarad 
 and a discharge key are required. The condenser is 
 first charged from the cell or portion of the battery 
 used to take the constant ; the discharge through the 
 galvanometer is taken and noted — call it /3. Next 
 take a discharge from the same condenser from the full 
 
 battery — call it B — then -^- is the battery ratio, and we 
 
 P 
 multiply the " constant " deflection, 0, by this ratio in 
 order to obtain the deflection which would be given by 
 the full battery through the high resistance. 
 
 The test upon the cable itself is now taken : one 
 pole of the battery being put to earth, the battery 
 itself being well insulated, the other pole is connected 
 through the galvanometer and key to the cable. Of 
 course the tank or drum containing the cable is to 
 earth. A short-circuit key must be connected across 
 the galvanometer terminals, and must always be 
 closed so as to short-circuit the galvanometer at the 
 beginning of each test, otherwise the rush of current 
 
 from a high battery power will probably demagnetise 
 the galvanometer needle. Close the battery key, and 
 after half a minute open the short-circuit key, which 
 may be used to check the swing of the galvanometer 
 needle. At the end of the minute note the deflection — 
 call it D. Still keeping the circuit closed, note the 
 deflection at the end of the second minute, which 
 shows a considerable decrease over that of the first. 
 If the insulating material is high-class indiarubber, 
 the battery should be kept on for 15 minutes, readings 
 being taken every minute. 
 
 When a certain insulation resistance per mile is 
 specified, it should be given by the first minute test. 
 
 The insulation resistance is calculated out as follows : 
 
 ex^xHEx? 
 
 Total insulation resistance = . 
 
 S 
 
 Dx G + S i 
 
 Si 
 
 
 
 where 8 = deflection when taking the constant. 
 
 G = the resistance of the galvanometer. 
 
 S = the resistance of the shunt used when taking 
 the constant. 
 
 Sj = the resistance of the shunt used when taking 
 the test. 
 
 HE is the value of the high resistance — if in 
 ohms, the resulting insulation resistance 
 will be in ohms ; if a megohm, it will be in 
 megohms. 
 
 D = the deflection obtained when taking the test, 
 less the leakage deflection on test leads, 
 instruments, etc., which should always be 
 obtained at the beginning of the test. 
 
 -p, = battery ratio, which equals unity when the 
 full battery is used throughout. 
 
 The expression 
 
 G + S 
 S 
 
 gives the multiplying power of 
 
 the shunt — i.e., the number by which the observed 
 deflection must be multiplied in order to obtain the 
 true value of the deflection. 
 
 When the usual 10, 100, and 1,000 shunts are used, 
 
 .G + S 
 
 those numbers denote the value of ■ 
 
 S 
 
 Example 1. — The deflection obtained when taking 
 the constant with 100 cells through 1 megohm was 
 300 divisions, the shunt having a value of 20 ohms. 
 
 The deflection D obtained when taking the test with 
 100 cells was 22 divisions, no shunt being used. 
 
 The leakage from instruments, testing wires, etc., 
 gave two divisions . • . 22 — 2 = 20, the true value of D. 
 
 The galvanometer resistance was 7,000 ohms. 
 
 300 x 7 '°°Q + 20 x 1 105,300 
 2i\) 
 
 = — 2Q— = 5 ' 2(35 me g° bms - 
 
 the total insulation resistance of the cable, which 
 
MEASUREMENT. 
 
 351 
 
 multiplied by the length in miles gives the insulation 
 resistance per mile. 
 
 Example 2. — When D is obtained with the galvano- 
 meter shunted. 
 
 Let = 300 with a 20-ohm shunt as above. 
 Let D = 50 with a 100-ohm shunt. 
 
 Then 
 
 3Q0x 7,000 + 20 
 
 20_ 
 
 Kn 7,000 + 100 
 50 x — 
 
 100 
 
 105,300 0Q . PC , 
 
 = 2a oo megohms. 
 
 3,550 
 
 Example 3. — When the constant is taken through 
 10,000 ohms, using one cell, the test being taken with 
 the full battery. 
 
 Let 8 = 250 deg. and the shunt used 400 ohms. 
 
 /3 the discharge deflection from the 1 cell = 220deg. 
 
 B the discharge deflection from 100 cells = 150deg., 
 and the shunt used 50 ohms. 
 
 Let D = 20 with the galvanometer unshunted. 
 
 The galvanometer resistance G = 7,000 ohms. 
 
 7,000 + 50 
 7,000 x 400 
 
 150 x 
 
 Then 250 x 4QQ x 10,000 x ■ 
 
 50 
 
 220 
 
 20 
 
 250 x 18-5 x 10,000 x 96-1 000 OQ1 OK ~ , 
 = -~ = 222,231,250 ohms, 
 
 or say 222 meghoms. 
 
 For testing short lengths of highly-insulated cable a 
 battery of least 400 Leclanche cells should be used. 
 
 If the battery is kept on for 1 5 minutes during which 
 readings are taken every minute, it will be noticed that 
 the deflection falls rapidly the first two or three minutes, 
 the fall per minute becoming gradually less, till at 
 the end of 15 minutes the deflection is falling very 
 slowly. It would thus appear as if the insulation 
 resistance rises during the test ; but the chief reason is 
 that when the battery is first connected, and after the 
 first rush of current, the dielectric electrifies or absorbs 
 current, and this absorption naturally becomes less and 
 less as the cable becomes fully charged. 
 
 The great use of this test is that it shows the 
 existence of minute faults in a cable which on a one- 
 minute test would pass the specification, as if there is 
 little or no electrification, or the existing electrification 
 proceeds unsteadily, it is certain the dielectric contains 
 undeveloped faults. 
 
 For this test it is absolutely necessary that the cable 
 should be in water, as no reliable results could be 
 obtained even from a well- wetted drum, and, of course, 
 it cannot be taken on a system connected up to trans- 
 formers, etc., though frequently before connecting up 
 the electrification is very pronounced. 
 
 The method of testing having been explained, we 
 proceed to the system which should be adopted during 
 the laying of the cables, 
 
 All insulated conductors should be tested after at 
 least 24 hours' immersion in water, and none but those 
 giving the specified insulation resistance should be 
 accepted. When laid or drawu in, but before jointing, 
 the section having one end in the station should have 
 its ends prepared, the distant end being well warmed 
 and lapped, and a careful test taken ; if all right prepare 
 section 2, and, having by means of a short piece of 
 fine wire connected it to section 1, test both sections 
 together. If the insulation resistance per mile is found 
 to be correct, the two sections may be jointed, after 
 which they should be again tested, and if not up to 
 the previous tests the joint must be remade. Sections 
 1 and 2 being jointed and passed, section 3 may be 
 temporarily connected and tested, jointed, and the three 
 sections tested, and so on till the entire system is jointed 
 and the final test taken. 
 
 If this method is carried out, the engineer in charge 
 will have a record of tests : first, upon each section in 
 water ; second, upon each section after laying ; third, 
 upon the system as each length is jointed on, and he 
 will feel satisfied that the cables are up to the specifi- 
 cation, that they have not been damaged in laying, 
 and that the joints are all well made. 
 
 With vulcanised indiarubber joints, in a vulcanised 
 indiarubber insulated cable, there is no difficulty in 
 keeping the completed system up to the specified 
 insulation resistance per mile of the cable. 
 
 It is manifestly absurd to lay down high-class mains 
 and by means of inferior joints reduce the insulation 
 resistance to a few thousand ohms. 
 
 If a section cannot be tested from the station, or if 
 jointing cannot be commenced at the station end of the 
 mains, a portable testing set should be employed, and 
 when all sections are jointed up a final test should be 
 taken with the station instruments. 
 
 If the daily tests (which should always be taken) 
 show any decrease in the insulation resistance, the 
 cause should at once be sought for and removed. 
 
 Of course, when transformers are used in the system, 
 and are connected up, there will be a very decided fall 
 in the insulation resistance. This is in a great measure 
 unavoidable, but could be greatly decreased by more 
 careful methods of insulating the transformer, fuses, etc. 
 
 In feeder mains running to a distributing centre it is 
 possible to localise an earth by means of the loop test 
 if suitable instruments are available. The ordinary 
 P.O. pattern Wheatstone's bridge and resistance coils 
 are quite unsuitable, the instruments required being 
 a sensitive low-resistance Thomson's mirror galvano- 
 meter, a few low-resistance cells, together with a finely 
 adjusted set of slide resistances and resistance-box. 
 The conductors should be looped by amalgamating the 
 ends and immersing them in a large mercury cup 
 containing clean mercury. 
 
 The copper resistance of the loop is first taken, by 
 means of the slide resistances and resistance-box, 
 connected up so as to form a Wheatstone bridge. 
 
 Then connect the conductors to the two free 
 terminals of the slide ratio coils, across which the 
 
352 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 galvanometer must be connected. One pole of the 
 battery is put to earth, the other pole is connected 
 to the terminal at the junction of the two ratios. The 
 balance is obtained by varying the ratios. 
 
 Let E = the total resistance of the loop. 
 
 A = the resistance when the balance is obtained 
 of the ratio coils connected to the faulty 
 side of the loop. 
 
 B = the resistance when the balance is obtained 
 of the ratio coils connected to the sound 
 side of the loop. 
 
 Let t be the resistance per yard of the conductor, then 
 
 A 
 
 VA- 
 
 the distance of the fault in yards = 
 
 E '^) 
 
 All connections should be made by means of mercury 
 cups, to which the conductors must be brought ; no 
 leading wires should be employed. If no special 
 instruments are obtainable, the distance of the fault, 
 if an earth, may be found by the fall of potential 
 method, and if carefully taken good results should 
 be obtained. 
 
 The instruments required are a high-resistance 
 Thomson mirror galvanometer, shunt, condenser, dis- 
 charge key, and a resistance capable of carrying, say, 
 30 amperes, and having a resistance of, say, "1 ohm. 
 The wire composing the resistance should be of 
 German silver or platinoid, and of sufficient section 
 to carry 30 amperes without any appreciable rise in 
 temperature. Loop the conductors at the distributing 
 centre. Connect one pole of a battery or accumulators, 
 or, failing these, of a continuous-current dynamo, to 
 earth, the other through an adjustable resistance and 
 the standard resistance described above, to the con- 
 ductor of the cable forming the faulty side of the loop, 
 the end of the side being insulated. 
 
 Now, by means of the adjustable resistance, adjust 
 the current, from the accumulator or dynamo, arranging 
 it to be as high as possible, as if owing to the fault 
 having a high resistance only a small current passes, it 
 will be very difficult to localise the fault. 
 
 If possible use a condenser having a large capacity, 
 one side of which is connected to earth and to one of 
 the galvanometer terminals, the other side being con- 
 nected to the movable lever of the discharge key, the 
 two free terminals of which are connected — one to the 
 remaining terminal of the galvanometer, and the other 
 to a wandering lead, which for the first test is con- 
 nected to the terminal of the standard resistance 
 nearest the dynamo or accumulator. A discharge 
 deflection is taken, call it D. 
 
 Bemove the wandering discharge key lead, and 
 
 connect to the other terminal of the standard resistance 
 to which the conductor of the cable under test is also 
 connected. Take a discharge deflection and call it D,. 
 Eemove the wandering lead and connect it to the 
 insulated end of the sound side of the loop. Take a 
 discharge deflection and call it a. 
 
 Then x = standard resistance 
 
 L^-a 
 
 = distance of the 
 
 fault in yards, where t ■= the resistance per yard of the 
 conductor. 
 
 If no condenser is available, the galvanometer in 
 series with a suitable resistance may be connected 
 between earth and the various points as mentioned 
 above, and the permanent deflection values used in 
 the formula. 
 
 The distance of a contact may be obtained by 
 measuring the resistance of the conductor, first from 
 the station end, the distant ends being insulated, and 
 afterwards from the distant end, the station end being 
 insulated. 
 
 Let E equal the resistance as taken from the station. 
 
 Let E 2 equal the resistance as taken from the dis- 
 tributing station. 
 
 Let Ej equal the conductor resistance of the faulty 
 mains. 
 
 Ei + E — E 2 
 
 resistance of one conductor to the fault, 
 
 Ej + E — E 2 
 
 and 
 
 = distance of the contact in yards, 
 
 when t = the conductor resistance per yard. 
 
 The instruments used for the loop test will be suitable 
 for the above test, which should be taken when there 
 is no traffic on the roads under which the mains are 
 laid. It need hardly be said that as all measurements 
 with a view to localising a fault are greatly dependent 
 on the state of the joints in the system, to ensure a 
 satisfactory test the joints must be sound and reliable. 
 
 It is very desirable that a proper distributing station 
 should be arranged inside some building and fitted with 
 arrangements for disconnecting the distributing from 
 the feeder mains, so that the condition of any feeder or 
 distributing main with its branches can be accurately 
 obtained. Test-boxes in the street should be avoided, 
 as they are of small practical value and a continual 
 source of trouble. 
 
 It is unfortunately impossible to localise earth faults 
 by direct testing in branch wires without disconnecting 
 from the distributing main, and the existence of this 
 difficulty should be the strongest possible reason for 
 using the very best known means of insulating, 
 jointing, and laying underground mains, 
 
ELECTRIC LIGHT MAINS. 
 
 353 
 
 CHAPTER XV. 
 
 ELECTEIC LIGHT MAINS. 
 
 jpHE various methods of laying electric light 
 mains will now be described : 
 
 The Crompton System. 
 The system of underground mains used 
 by Messrs. Crompton and Co., Limited, 
 is that of bare strips of solid copper 
 stretched upon glass insulators in an underground 
 culvert, and strained in air from point to point 
 by means of specially constructed straining appa- 
 ratus. The Crompton system is in use for the 
 low-pressure distribution of electricity for direct cur- 
 rents, with or without the use of accumulators, and 
 is principally used upon the three-wire system of 
 mains, though it is also applicable where desired for 
 ordinary parallel distribution. The system has been 
 in successful operation for about four years, and besides 
 the original Crompton central station at Kensington 
 and Knightsbridge, where it was first developed, and 
 where over 12 miles of mains are now laid, it is in use 
 or is being laid for the Notting Hill Electric Light 
 Company, the "Westminster Electric Company, at 
 Northampton, at Birmingham, and also in France. 
 Nearly 100 miles of mains in all are now in use or 
 are being laid. 
 
 In principle, the system may be considered as a 
 return to the ordinary methods of running telegraph 
 wires, but applied to heavy copper mains, and under- 
 ground. Mains for electric light were originally heavily 
 insulated, afterwards cased in lead, then passed through 
 protecting iron pipes, and finally run in conduits or 
 culverts. To make these culverts such that bare con- 
 ductors of sufficient thickness could be strained in air — 
 which, after all, is the best as well as the cheapest 
 insulator — thus doing away with costly insulation, has 
 been the aim in the Crompton system, and, in point of 
 fact, this system is one of the earliest attempts at a 
 practical solution of the question of underground dis- 
 tribution on a large scale, and its success has given a 
 great impetus to the planning of central distributing 
 systems. 
 
 The development of the Crompton system of electric 
 mains has been, roughly speaking, through the following 
 stages: First, a solid square bar was carried upon a 
 wheel insulator fastened sideways in a culvert, changed 
 afterwards to an upright cup insulator having a slot in 
 its head, carrying first a bar, then modified to a bundle 
 of wires, and again modified to a set of strips placed 
 sideways on their edges. The alteration of the position 
 
 of the strips, placed next on the flat, brought the 
 method to one much as now in use, its later develop- 
 ment being mostly in the mechanical form and the 
 practical method of placing the insulators, and in the 
 system of straining. At first this straining was brought 
 about by the use of a right and left handed screw on 
 the solid copper bars, which drew the lengths together ; 
 the alteration was soon made of having each length 
 strained separately, and connected together by a 
 by-pass of solid copper. As at present used, Figs. 126 
 and 127, each length of copper strip is separately 
 strained, and the overlapping ends are then tightly 
 screwed together by a screw connecting piece, the 
 strain being taken by specially constructed straining- 
 pieces in the manner hereafter described. 
 
 The mains on the Crompton system are made to 
 three standard sizes. These represent (1) the usual 
 three-wire distributing main ; (2) the distributing main 
 with one pair of feeders ; and (3) the distributing main 
 with two pairs of feeders. It is seldom that other sizes 
 than these are necessary; ii such cases arise, two 
 separate culverts would be run for the distance required. 
 
 The principle of the three- wire system of distribution 
 is now sufficiently well known. With a total pressure 
 of say 210 volts between the outside wires, the pressure 
 between the inner and either of the outer wires is 105 
 volts. The wires to the houses are connected alternately 
 in such a manner that the number of lamps on one side 
 balances that on the other, so that no current (or only 
 the slight difference between the two halves) is flowing 
 through the centre wire. To keep the pressure the 
 same over all parts of the circuit, feeder mains, going 
 to certain determined points where the demand is 
 heaviest, are connected direct from the source of 
 supply, and thus the potential is kept constant. 
 
 In laying out a system of mains the position of the 
 generating station is taken as centrally as possible. 
 The streets are laid out with mains in the simplest 
 and most direct manner, and feeder points are then 
 determined upon at various points upon the plan. In 
 the streets, were no feeders are required, a culvert for 
 three wires only is laid ; but where one pair of feeders 
 go alongside the mains, a culvert for five wires (the 
 three wires and one pair of feeders) is necessary. Near 
 the station where feeders in two directions are required, 
 culverts for seven wires (three wires and two pairs of 
 feeders) are laid, and so for other parts of the circuit. 
 These three sizes are shown in the illustrations, Figs. 
 
354 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 128, 129, and 130. The sizes inside are respectively Crompton system is 2ft. 5in. wide, and for the largest 
 
 lft.'3in.,'lft. 8in., and 2ft. wide. culvert 3ft. 2in. wide by 1ft. 9Jin. deep. 
 
 The trenches for the culverts are made, wherever The first process, when actual laying is begun, is the 
 
 possible, under the pavement, and the usual type of digging of the trenches and the construction of the 
 
 Fig. 126. 
 
 boxes are made for this situation. Occasionally, how- culverts. These are laid in concrete, of the required 
 ever, it is necessary to lay the mains in the roadway width of opening, with the walls biassed outward at 
 for some length, or to cross the road from one side to the base for strength ; first the floor is laid in concrete, 
 
 Fig. 127. 
 
 the other, and the culverts and boxes are then made and then the walls of the culvert, which are kept up 
 
 extremely strong to withstand the ordinary traffic or by rough boarding till firmly set. 
 
 even the passage of * steam roller The second part of the ^ is ^ construction of 
 
 Itxe size of trench for the ordinary main on the insulator boxes. At every alternate house along the 
 
ELECTRIC LIGHT MAINS. 
 
 355 
 
 road — roughly, at the distance of every 15 yards — these box covers are previously filled in level with the 
 
 arrangements for an insulator box are made. This box concrete so as practically to form one continuation of 
 
 serves both for allowing inspection of the insulator, the pavement itself. 
 
 and as a junction box for either of the houses. An At certain distances along — from 20 yards up to 
 
 oak baulk, 4in. by 3in., for supporting the insulators, 100 yards if the road is straight — provision is made 
 
 i 
 Fig. 129. 
 
 is imbedded in the concrete, each baulk having round for straining-boxes, Figs. 126 and 131. Two baulks 
 
 holes drilled in its centre for the after reception of the of oak are firmly bedded in the concrete one above 
 
 insulator. These insulator baulks are marked with the the other, and form the fixed parts to take the 
 
 letters IB on the figures. strain of the copper strips. The size of the baulks 
 
 C »_•!» 
 
 ""■ ."■«'.»■.• '.'■•'."■ L" ' _-_;_'l- —it-!., -u.-i--i-.—--.-2.6' ---'-,:'-.-.---"-•""-"" 
 
 ■---* ;CONCR ETE 
 
 '9 : - <,''■ o '.' 
 
 -3 . 2 
 
 Fig. 130. 
 
 Thick York flagstones are now placed along over the is 4in. by 4in. for ordinary mams, 4m. by 5m. for 
 
 culverts, and the surface is filled in and rammed down mains with one pair of feeders, 4m. oy om. o 
 
 and the pavement replaced. At the places for insulators mains with two pairs of feeders. Ihese straimng 
 
 a metal frame is bedded in the concrete. Each of these baulks are marked SB m the illustrations. Alter 
 
 frames has a properly fitting lid or box cover, B C, and they are laid the spaces above are covered in witq 
 
356 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 Fio. 132. 
 
ELECTRIC LIGHT MAINS. 
 
 35*7 
 
 cast-metal boxes and covers of the same type as the on the ground by the workmen's feet, and the length 
 
 between the straining-boxes is measured and cut off. 
 
 PAVEMENT 
 
 ? KERB \ 
 | STONE 
 
 \ \ \ w 
 
 ■ ■= ■ ■'.•'0-; 
 
 .--.-•---• - . -■.b:,'-:-.° -■-'. ••".v.°:v.-.'. -.->•. °- :■-'.'.;;„.•? 
 
 - --••'.:"? , • V ••-"• - >v?-c -"C-O.N-c ft'.E TE P . •■ r-' •■'.■■ 
 
 ^:-^o;-:--: ? ;-^:-v,^^o-:v:o^ P :-:-;^OVv ; ; 
 
 Fig. 133 
 
 Fig. 134 
 
 When the culverts are finished, the next process is In the first place a cord is passed through the culvert, 
 the actual drawing in of the mains These come in and usually for this an ingenious and simple method is 
 
 0" wide for* 2" . 0" Culvert. — 
 8" wide for 1" . 8" Culvert, 
 j« 1 " . 3" wide for I". 3~ 
 
 Culvert.^ , ^-^-.^ — >j 
 
 Fig. 135. 
 
 copper strip lin. wide by Jin. thick, just as rolled from adopted. At the Kensington station may be seen a 
 the mill. They are uncoiled and stamped out straight curly-haired black dog, a permanent official of the 
 
358 
 
 Practical Electrical ElvGitfEERWc-. 
 
 works. This dog, with a cord attached to his collar, is 
 put into the culvert and the top closed in. The men 
 then go to the next box and call the dog, who creeps 
 along, drawing the cord, and so on to the end. The 
 cord is then attached to a chain, and the chain in its 
 turn to the copper strip, which is then drawn in by 
 the men. 
 
 At each insulator box the glass insulators are now 
 pressed into the holes (l^in. diameter) in the timber 
 
 or 
 
 into its position upon the insulator. The three 
 more strips are thus successively drawn in. 
 
 The copper strips having been passed into the 
 culvert are yet lying along the bottom. They have 
 now to be stretched sufficiently taut to be everywhere 
 in air, except where they touch at the insulating 
 supports. A sag of about 2£in. in the mains is usually 
 allowed between the insulators. 
 
 The straining pieces, S, each consist of a bridge or 
 
 Fig. 136. 
 
 baulk. These insulators, I, are of the form shown 
 in the illustrations, Figs. 127, 132, and 134, with 
 five deep rings to give a long length of insulating 
 surface, and thus to guard against leakage by 
 moisture creeping. The insulation is still further 
 increased by each insulator being previously dipped 
 in hot copal varnish. The head of each insu- 
 lator is slotted out of a width to receive the copper 
 strip. As each strip is drawn past, a man stands at 
 each insulator box along the route and lifts the strip 
 
 arch made of gunmetal, of the shape shown in the illus- 
 trations, Figs. 126, 131, and 133, and in detail in Figs. 
 138 and 139, and carefully designed to afford the requi- 
 site stability, both for resistance to the strain of the 
 copper mains and for holding these latter absolutely 
 fixed, while at the same time obtaining the highest 
 possible insulation. The insulation is assured by the 
 use o'f glass insulators with deep grooves, similar to 
 those already mentioned, but more solid, and squat to 
 withstand the strain. 
 
ELECTRIC LIGHT MAltiS. 
 
 m 
 
 In placing them into position, a ring or pad of solid second pad of solid indiarubber is placed over 
 indiarubber, E, Figs. 126 and 131, is first placed each insulator, which are thus evenly bedded and 
 
 
 l 
 
 
 
 1 
 
 i 
 
 i 
 
 
 ) 
 
 f 
 
 ( 
 
 r 
 
 i 
 
 ■\ 
 
 > 
 
 
 < 
 
 r 
 
 i 
 
 ^i 
 
 1 
 
 
 c 
 
 r 
 
 i 
 
 ^ 
 
 i 
 
 
 < 
 
 r 
 
 i 
 
 ^ 
 
 
 i 
 
 
 
 i 
 
 
 K-T7 
 
 | 
 
 ~T"' 
 
 
 
 Fig. 138. 
 
 4 -- t 
 
 Fig. 140 
 
 , bled to withstand the heavy end strain without the 
 
 gainst the straining baulk, upon these are placea e The etal brid stra i mn g 
 
 he glass insulators, I I. Fig. 137, and then a edges cmpp g 
 
 the gl 
 
360 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 pieces, S, are then placed in position, and the copper 
 main, Cu, is passed through the slot in the centre. It 
 should be noticed that this slot is made with the lower 
 side sloped, or wedge-shaped. A loose piece inside the 
 slot screws down tightly upon the copper main, which 
 is thus slightly bent, and becomes firmly fixed or 
 wedged in position, so that by this simple means the 
 more the strain on the main the tighter it is gripped. 
 The pinching screws are made of aluminium bronze or 
 Delta metal, or in dry situations of steel. 
 
 The straining is accomplished in various ways. The 
 most scientific and certain is that often used of a 
 hydraulic jack, made to form a lever to draw the mains 
 up tight. Placed in position at the side of the trench, 
 bedded against the concrete edge, a few strokes of the 
 
 is usually large enough to hold four strips, but can be 
 made to hold six strips if necessary. It is shown in 
 detail in Fig. 140. 
 
 The culverts when going under roadways have to be 
 made of extra strength to carry the heaviest traffic 
 without fear of the covers breaking or the side walls 
 crushing in. Certain regulations have been issued by 
 the London County Council, and the designs of culverts 
 and covers given are constructed fully to comply with 
 these requirements. The trenches in these cases are 
 about 4ft. deep. The illustrations, Figs. 132, 133, 
 and 136, show cross sections and longitudinal sec- 
 tions of a 2ft. culvert under roadway. The smaller 
 culverts are of similar construction, the only 
 difference being in the width. The culverts have 
 
 Fig. 141. 
 
 hydraulic handle tightens up the main, and the 
 pinching screws are then screwed tightly down upon 
 the _ wedge-piece, W, Figs. 126 and 131. The 
 straining for shorter lengths is also often done by 
 the simple means of ropes and pulleys, or by a 
 crowbar strained against a baulk, or sometimes by 
 a right or left-handed screw. The means employed 
 are mainly dependent upon the length of the strip, and 
 thus upon the necessity for greater or less force. 
 
 When the copper strips are properly tightened from 
 straining-box to straining-box, the ends are over- 
 lapped and tightly connected together by a grip-box or 
 connecting-box, C, Figs. 126 and 133, which simply 
 means pressing the ends of the flat strips firmly 
 together by means of pinching screws. This grip-box 
 
 a solid cement bottom, and the walls are of 9in. 
 brickwork set in cement and with an inside coating oi 
 cement to prevent any percolation of moisture. They 
 are also sometimes constructed entirely of cement 9m. 
 thick. Two methods of covering the culverts are 
 adopted according to judgment. They are either 
 covered in with two thicknesses of 2|in. York stones ; 
 or in some cases with cast-iron covers having ribs on 
 the upper surface and filled in with concrete. Both 
 these methods are shown in the illustrations, Fig. 132 
 being a longitudinal section of a culvert, half of which 
 is covered with thick York stones, and the other halt 
 indicating the culvert as covered with ribbed iron plates 
 and concrete. Fig. Ill is a reproduction from a photo- 
 graph of the work of LrenchhiLr. 
 
ELECTRIC LIGHT MAINS. 
 
 361 
 
 The covers to the boxes under the roadway have, of 
 course, to be adapted to the nature of the rest of the 
 roadway surface. Both the insulating and straining boxes 
 are dealt with in the same way. The covers, Fig. 135, are 
 made of a cast-iron frame, hollow on the upper side, 
 fitting into the boxes and easily removable for inspec- 
 tion. The hollows in the upper surface are filled in 
 with concrete, cement, asphalte, or other materials, to 
 match both the colour and the rate of wear of the 
 material of which the roadway itself is constructed. 
 The joints of these covers are made water-tight, and at 
 
 cases heavily-insulated cables drawn into pipes are used 
 for the distance required, each cable having its own 
 separate pipe. 
 
 The ends of these pipes are brought into insulator 
 or straining boxes (Fig. 142) of the same form as those 
 used for the culverts, and the cable is screwed tightly 
 to the copper strips in the manner shown. Similar 
 arrangements are adopted for connecting the house 
 mains to the three-wire mains. They are led to the 
 nearest insulating box, and are connected respectively 
 to the inner main and one of the outer mains. If it is 
 
 Fig. 142. 
 
 the same time noiseless, so as not to vibrate under the 
 traffic, by inserting a serving of tarred rope in the frame- 
 work of the box before the cover is put on. Where 
 wood pavement is in use, the cover is made with ribbed 
 openings of the necessary size to match the rest of the 
 pavement, and hard wood blocks are cemented in, as 
 shown in the illustration, Fig. 134. The arrangements 
 for connecting a culvert under a roadway with that 
 under the pavement are seen in Fig. 133. 
 
 It sometimes happens that the space under the foot- 
 way is so small that culverts cannot be built. In such 
 
 ever necessary to alter the balance of the current in the 
 two halves of the three-wire mains, it is a matter of a 
 few minutes' work to lift the cover, unscrew one end of 
 the house main, and screw it upon the opposite outer 
 main strip. 
 
 The Callender-Webber System. 
 
 Two systems of underground mains for electric light 
 distribution are carried out by Calender's Bitumen 
 Telegraph and Waterproof Company, Limited, of 
 Leadenhall-street, London, These two systems are 
 
362 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 known respectively as the Callender-Webber system 
 and the Callender Solid system. In- both of these, 
 highly-insulated copper cables are used, laid under- 
 
 In the Callender-Webber system the bitumen is 
 moulded into solid casings having holes throughout the 
 length, and through these the insulated cables are 
 
 Fjg. 143. 
 
 m\\\\\\m\\\\\m\>\m; 
 
 SERVICE BOX A 
 
 FEEDING BOX 
 
 Fig. 144. 
 
 ground in trencbes and protected by bitumen. The 
 method of using the bitumen constitutes the difference 
 between the systems, 
 
 drawn in. This gives facilities for the removal of 
 cables and the substitution, if necessary, of others; 
 and for the provision of spare space to be occupied by 
 
other cables as the demand increases, without the 
 necessity for again taking up the streets. 
 
 In the Callender Solid system the insulated cables 
 are laid in an iron trough and the whole filled in solid 
 with melted bitumen, the cables being held practically 
 immovable when once laid. 
 
 The decision as to which of these two systems is the 
 most suitable depends largely upon conditions of cost 
 and of locality. The solid system will in most cases be 
 the cheaper, if the cable is for a certain stipulated 
 transmission of current which will never be exceeded, 
 
 ELECTRIC LIGHT MAINS. 
 
 363 
 
 The general arrangement is well shown in the 
 descriptive section and plan, Figs. 143 and 144, of the 
 roadway, showing six feeding mains, laid under the 
 street among a bewildering array (not at all uncommon 
 in towns) of gas and water pipes and main sewers, with 
 a distributing main on the three- wire system laid under 
 the footpath, feeder boxes and service boxes being 
 constructed flush with the pavement. 
 
 The appearance of the bitumen casing is clearly 
 indicated in Fig. 145. It is made in 6ft. lengths of 
 blocks of bitumen concrete, pierced in the manufacture 
 
 JOINT 
 
 Fig. 115. 
 
 or if the demand for current is known or can be 
 estimated closely, so that necessity for alteration or 
 addition is not likely to occur. 
 
 The Callender- Webber system is most suitable for 
 central station work with feeders where the addition of 
 other mains as the demand increases will be certain to 
 be required. 
 
 Many large contracts on one or other of these plans 
 have been carried out in all parts of the kingdom, both 
 for central station distribution or for feeders at isolated 
 installations, and also in combination with the various 
 systems of bare copper mains for use in special situa- 
 tions. The numerous practical details, which are so 
 necessary for the success of any system of distribution, 
 have thus been worked out as the result of actual 
 experience, and the precautions fully learnt which must 
 be taken to ensure high insulation and safe and 
 uninterrupted supply, as well as to assure ease in 
 laying and in the connection of branch or house mains. 
 
 We will first describe the Callender- Webber system 
 of underground electric mains, this being the system 
 most applicable to networks of new central stations 
 likely to increase in size. The system is the outcome 
 of the joint inventions of the Callender Company and 
 Major-General Webber, E.E., and its essential distinc- 
 tion lies in providing a solid yet insulating protection 
 to the cables, and a separate way or passage for each 
 cable run, 
 
 by a varying number of holes or ways. The standard 
 sizes of these cable cases are shown in Fig. 146, and 
 are for two, three, four or six ways, of either l£in, 
 lfin., 2fin., or 3in. diameter. 
 
 The bitumen concrete is composed of natural 
 bitumen (one of the most durable materials known) 
 mixed with sand and wood fibre specially treated. 
 
 oo 
 oo 
 
 ooooo 
 
 Fig. 146. 
 
 The resulting material is tough and strong, with great 
 power of resistance to crushing, breaking, and tensile 
 strains. It is impervious to water, is not affected by gas 
 or acids found in the ground, and is a non-conductor 
 of electricity. It does not expand or contract, and is 
 capable of withstanding, when embedded in the ground, 
 the weights of heavy traffic. It can be made to any 
 shape, and the cases can be bent up or down, or to 
 curve to the right or left on the job in the work of 
 
364 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 laying, and the process of laying is comparatively 
 simple. When laid each way is a closed passage, com- 
 pletely separated from all the other ways and from 
 earth by an insulating wall of bitumen, and the cables 
 are, therefore, subject to no unfavourable conditions 
 from contact with each other, or from deleterious or 
 conducting substances. 
 
 When constructing the conduit it is easy to lay the 
 casing with a larger number of holes than is necessary 
 for present requirements, the additional cables being 
 easily drawn in as the demand on the network increases 
 without opening the roadways or disturbing the traffic. 
 
 The following table gives the stock sizes, with the 
 size of cable taken : 
 
 Sizes op Callender-Webeeb, Casings. 
 
 Size. 
 
 2 way 
 
 3 way 
 
 4 way 
 6 way 
 
 l£in. ways, 
 
 taking up to 
 
 19 / 14 cable. 
 
 4£in. by 2|in. 
 6|in. ,, 2jin. 
 4£i;i. ,, 4£in. 
 4|in. ,, 6|in. 
 
 l|in. ways. 
 
 taking up to 
 
 19 /n or *V 
 
 2|in. ways, 
 
 taking up to 
 
 "Vis cable. 
 
 3£in. by 6in. 
 3^in. „ 8|in. 
 6in. ,, 6in. 
 6in. „ 9in. 
 
 3in. ways, 
 takes over £in. 
 sect, area, in- 
 sulated. 
 
 4in. by 6Jin. 
 4in. ,, 9i-in. 
 6|in. ,, 6|in. 
 7in. ,, 9|in. 
 
 Sin. by 9in. 
 
 Sin. ,, 13in. 
 
 9in. ., 9in. 
 
 9in. „ 13in. 
 
 In laying mains underground, whatever the system, 
 it is advisable to carry them as near to the houses 
 as the cellars will admit, and as far as possible the 
 trenches are made under the footpath. When the 
 pavement is of a costly nature and the roadway is of 
 macadam, it is better to run the trenches along the 
 road, but usually the pathway, if not already occupied 
 by telegraph mains or large cellars, is best and 
 cheapest. 
 
 Under paved footpaths the depths at which the 
 mains are run need only be sufficient to be just below 
 . the higher gas and water service pipes, the concrete 
 foundation of asphalte, or the flagstones of the pave- 
 ment. In practice, the bottom of the trench is usually 
 12in. to 18in. below the pavement level, but in cases 
 where space is limited the casings can be run with 
 perfect safety in 3in. or 4in. of depth. This adaptability 
 of the system for limited space, combined with its high 
 mechanical and electrical protection and its power of 
 curving up, or down, or to one side, make it extremely 
 convenient in many places were more rigid systems 
 would be inapplicable. 
 
 In the roadway it is customary to have the trench 
 3ft. deep. Under an ordinary macadamised road this 
 depth will be quite safe for all ordinary traffic and for 
 temperate climates. If the roadway is concreted as 
 for wood or asphalte paving, it is sufficient to have the 
 mains laid immediately below the level of the concrete. 
 
 The actual process of laying the Callender- Webber 
 mains is as follows : 
 
 The course of the main is first laid out, and the 
 trench dug to the required depth. The trench is made 
 in as straight a line as possible, and is somewhat 
 carefully levelled, a good foundation being secured by 
 ramming down the soil. Occasionally rough concrete 
 is employed in the trench to bed the conduit, but this 
 
 is necessary only in exceptional circumstances. As 
 long lengths as are permitted by the authorities are 
 opened at a time, arrangements being made to avoid 
 interference to the traffic by bridging the trench where 
 necessary. At crossings, however, the work is done 
 in short sections to avoid blocking the traffic. The 
 surplus earth is everywhere carted away regularly. 
 
 Flo. 147. 
 
 The bitumen casing is next laid upon the bottom of 
 the trench and jointed together. This jointing is 
 carried out length by length to form one solid and con- 
 tinuous casing. Each fresh 6ft. length is placed on the 
 ground, with 2in. to 3in. interval between it and the 
 one already laid. In the meantime, some bitumen 
 concrete is being heated in the large iron cauldron. 
 Mandrels are pushed through each of the holes in the 
 casing, joining the two sets of holes together. Then 
 the bitumen concrete is run in, stamped into position, 
 and shaped into the same shape as the original casing. 
 The joint is thus made absolutely waterproof, and the 
 mandrels ensure the holes being smooth and con- 
 tinuous. The result is practically one solid bitumen 
 casing, and except for the newness of the material a 
 person passing along can hardly tell where the joint 
 has been made. The mandrels are then pulled out, 
 the earth thrown in again, rammed down, and the road 
 repaved. 
 
 Wm 
 
 Fig. 148. 
 
 In certain circumstances, where the mandrels cannot 
 be used, a different method is adopted. The ends of 
 the casing are butted together (see Fig. 147), and saddle 
 pieces of material similar to the case itself are placed 
 at the joint, which is made tight by a little bitunieij 
 seared with a hot iron, 
 
ELECTRIC LIGHT MAINS. 
 
 §65 
 
 The next detail is the construction of the drawing-in 
 boxes. 
 
 These are constructed in a simple manner by the 
 building of a pit in brickwork at the places required, in 
 the manner shown in Fig. 148. In the larger sizes 
 these are made 3ft. by 3ft., and about 4ft. deep. The 
 bottom is not concreted, but left in "ordinary brickwork, 
 so that any moisture that may get in may easily drain 
 away. The walls of the pit are built up, and the 
 bitumen casings are so arranged that they project 
 about 6in. into the pit. Above them the brickwork, 
 overlaid and tapered up, as shown, forms a manhole in 
 
 Three different kinds of boxes are in use : 
 
 1. Feeder boxes, for the connecting of feeders to 
 distributing mains. 
 
 2. Service boxes, for the connection of mains to 
 customers' mains at points where large supplies of 
 current are to be taken off. 
 
 3. Supply boxes, for the supply of householders, put 
 in, if required, when service boxes are too far away. 
 These latter are usually constructed of cement or brick 
 built round the conduit after it is laid, so that the 
 conduit is cut at the point most convenient for tapping 
 the cables. 
 
 Fig. 149. 
 
 the surface of the street, having a cast-iron box, and 
 removable lid. 
 
 The cover plates are generally made 18 by 12, 20 by 
 15, 24 by 15, or 24 by 24 inches, according to the size 
 of the pit. The plate on the footpath is filled in with 
 cement to resemble the paving in which it is placed. 
 In some cases, where access is not often required, it is 
 often found desirable to seal in these boxes. For this 
 purpose the boxes are cast in a special form (see 
 Fig. 149), so that a second lid can be placed on the 
 inner rim there seen under the pavement lid, and the 
 box then made absolutely watertight by sealing in the 
 false lid with red lead or asphalte. 
 
 When the conduit is all laid and the various boxes 
 properly built, the drawing-in of the cable itself 
 follows next. In the smaller sizes of casing a cord is 
 passed in as the conduit is laid. The mandrels are 
 specially made with a hook at the end, to which the 
 cord is fastened as each length of casing is laid and the 
 mandrel withdrawn, the cord is therefore drawn 
 through. With the larger sizes of holes such a pre- 
 caution is not necessary, and when the conduit is laid a 
 jointed rod— something like a sweep's rod— can easily 
 be pushed through from box to box. This is attached 
 to a cord, which in its turn is drawn through and pulls 
 in a rope. The rope is attached to a cable, which is 
 
3(36 
 
 PRACTICAL ELECTRICAL ENGINEERING-. 
 
 pulled in either by a number of men hauling or by the as these would only be used for feeders or for supply 
 aid of a winch. The ways are quite smooth inside, and mains in the country, and for town supply the lengths 
 
 FOOTPATH LEVEL 
 
 Fig. 150. 
 
 without any great difficulty. Lengths such the demand was heavy and service boxes numerous. 
 
'ELECTRIC LIGHT MAINS. 
 
 367 
 
 The cables that are used are Callender cables of high 
 insulation with dielectric of vulcanised bitumen, and 
 are made as follows : The conductor is of stranded 
 copper cable of not less than 98 per cent, conductivity. 
 This is first covered by a solid sheath of bitite (bitumen 
 vulcanised under the Callender process) put on uuder 
 heavy pressure at one operation. This core is then 
 served with tape and insulating compound, and again 
 served with either jute yarn or additional tapes, finally 
 braided with hemp yarn and passed through a bath 
 of asphalte compound. The cables are thus made 
 extremely durable, resisting perfectly themselves all 
 action of moisture or coal and sewer gas. When laid in 
 the solid casing of bitumen, therefore, they are practi- 
 cally indestructible. The bitumen is also, it appears, 
 a thorough protection against rats, for although these 
 mains have been laid now for several years, the rats 
 have never been found to bite either the cables or 
 the casings, although it is known they have made in 
 some cases the spare holes in the casing their mode of 
 travel. 
 
 In connecting the mains to branch or service supply 
 circuits, the course is often adopted of building what 
 are termed fuse-boxes, Fig. 150, in the ground, in 
 which the junction of the service line is safeguarded by 
 the insertion of a solid fuse. The outer pit is made of 
 a depth to suit the position of the casing with its wires. 
 The inner box is made watertight, and is placed inside 
 the pit with air space around it. The bitumen casing 
 projects a few inches inside the inner box, to which it 
 is carefully sealed with bitumen. The upper lid or 
 manhole of the whole street box is set flush with the 
 pavement, and any little moisture that might get in 
 simply drops to the bottom of the pit and filters away. 
 The method of connecting the feeders to the mains 
 is simple and ingenious. It is the same in both 
 systems. 
 
 The Callender Solid System. 
 
 In the Callender solid system of underground electric 
 mains, the cables, heavily insulated, are laid in suitable 
 troughs, generally of cast iron, placed in trenches under 
 the street, and the whole of the vacant space in the 
 trough round the cables is run in solid with refined 
 bitumen. All the bitumen employed is genuine 
 natural Trinidad bitumen, free from any admixture of 
 gas, tar, and pitch. 
 
 The whole forms a solid and compact mass, into 
 which neither gas nor water can possibly penetrate, 
 and which itself forms a very high electrical insulation. 
 There are few substances more durable than this 
 bitumen, and with cables thus embedded there is little 
 probability of any deterioration taking place in mains 
 when once laid, while the thorough mechanical pro- 
 tection which is ensured reduces to a minimum the 
 liability of injury from exterior causes. 
 
 Mains laid on this plan are permanent, and form a 
 sound engineering job. It is of course clear that such 
 mains should only be laid when the conditions do not 
 require that the capacity of the cables should be after- 
 
 wards altered. The Callender solid system is especially 
 suitable for connecting mains for isolated plants where 
 the conditions are settled and well known ; for heavy 
 mains in large cities where the consumption is likely 
 to attain its maximum at once ; for heavy distributing 
 networks ; and for the heavy feeders which are run to 
 various points of the system of three-wire distribution. 
 It is also peculiarly suitable for arc lighting on the 
 series system where long lengths of single or double 
 mains are required to be run. 
 
 The combination of vulcanised bitumen as the 
 insulation for the cable itself, with plain bitumen run 
 around it as a sheath in the trough, is one that ensures 
 an absolutely safe and constant insulation, on which 
 faults are unlikely to appear. Eepairs, when necessary, 
 are easily and rapidly made. Altogether, it forms one 
 of the most trustworthy and satisfactory methods of 
 placing cables underground, and has a certainty of 
 being more permanent and lasting than most of the 
 other parts of an electric fighting plant. The work of 
 laying, once done, is thorough and complete. 
 
 The troughs for the mains are made of cast iron. For 
 town work this is especially preferable, as it offers a 
 permanent protection from the picks of workmen of 
 other companies. In some cases the upper surface of 
 the trough having been filled nearly to the level by 
 bitumen, is covered by a layer of cement. In others 
 the trough is covered in with cast-iron lids. The 
 thickness of the metal is from ^in. to -&va.., varying 
 with the size of the trough and the liability of the soil 
 to disturbance. These iron troughs are made in lengths 
 of 6ft., having socket pieces cast on one end, so that 
 when fitted together the surface inside the trough is of 
 one level. For carrying the mains round corners, and 
 for changing levels at crossings, circular pieces and 
 curved troughs are made, but considerable deviation 
 from the straight line is possible with the ordinary type. 
 
 In country roads, and in places where wood is plen- 
 tiful, troughs of sound timber are often substituted for 
 those of cast iron, care being taken to select wood 
 which will stand underground without rotting. The 
 planks should not be less than fin., and the lid should 
 be of lin. stuff. It is not advisable to use creosoted 
 wood, as the action of the products of coal-tar distil- 
 lation are found to be injurious to nearly every form of 
 dielectric. 
 
 Brick and cement trenches are occasionally used for 
 special situations, but they are expensive, and offer no 
 advantages over those of cast iron. 
 
 The ordinary shapes and sizes of mains on the 
 Callender solid system are shown in Fig. 151, which 
 comprises sections of four-way and six-way insulated 
 mains, laid in bitumen in iron troughs, and covered 
 over at the top with cement. 
 
 A very usual case in a distributing network is where 
 a pair of heavy feeders to a distant point are run along 
 with the distributing mains in the same trough. This 
 arrangement is shown in Fig. 152. In both cases the 
 position of the cables in the trough is kept by bridges 
 of hard bitumenised wood of the shape shown, which 
 
368 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 are first thoroughly dessicated and then saturated with 
 bitumen. These prevent the cables from touching the 
 
 In Fig. 153 is shown a pair of heavy feeders, 91 
 strands of No. 9 wire. These are highly insulated, and 
 
 &> 
 
 c 
 
 Fio. 151. 
 
 bottom of the trough, or of accidentally shifting their each covered with outer lead casing; then laid one 
 relative portions. above the other iu solid bitumen in ft narrQW irQU 
 
ELECTRIC LIGHT MAINS. 
 
 369 
 
 trough, and covered in with cast-iron lid. The joint of While still molten, the bitumenised wood bridges are 
 the trough is shown in section below. pressed into place across the trough at intervals of 18in. 
 
 Fig. 152. 
 
 In actual practice the method of laying is as follows : 
 
 The trenches in the ground having been dug and 
 
 levelled, the iron troughs are laid therein, and are 
 
 jointed together in one length with countersunk bolts 
 
 and nuts, Fig. 154. 
 
 \ 
 
 \ 
 
 Fig. 153. 
 
 apart. When the bitumen is set, the cable, which has 
 been brought on huge drums, is paid out into the 
 
 ''>'<>><wii&< ^ mm 
 
 Fig. 154. 
 
 Bitumen is meanwhile being heated in the cauldrons, trenches, and laid along into the spaces of the wooden 
 ana a little of this run along the bottom of the trough, bridges. The cable is then drawn taut, and melted 
 
370 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 bitumen is poured in between the cables and between 
 the cables and the sides of the trough. 
 
 This is done by moving the bitumen cauldron on 
 wheels along the trench, and by means of a large ladle, 
 pouring a little molten bitumen in at a time. When 
 this first layer is cool, the men come back again and 
 pour in a little more, and so on, for about six times, 
 thus allowing the bitumen to fill up every cavity round 
 the cables. In the case of cables above one another, 
 the whole process is, of course, repeated along the 
 
 In connecting the mains together, two different 
 methods have to be considered : (1) those in which the 
 joint is solid, permanent, and covered in with insula- 
 tion in the same manner as the rest of the cables ; and 
 (2) those in which the connections of feeders to distri- 
 buting mains and of those to house service mains are 
 sometimes required to be altered or disconnected easily, 
 either for change of load or for testing. 
 
 We will first of all take the solid connection, pre- 
 mising, however, that it is now found desirable to use 
 
 ^ ^^\v\s^^^^^^^^^^vv^^^sv'^^ V^V : ^^^^^ s ^^^^^ SSN '^^ : 
 
 Fig. 155. 
 
 length. The bitumen is run in to within an inch of the 
 top of the trough. 
 
 When the bitumen is quite cold a layer of Portland 
 cement from lin. to 3in. thick, according to circum- 
 stances, is put upon the top of the bitumen, and where 
 crossing a roadway a brickwork, cast-iron, or other 
 protection is placed over the top, to guard against 
 damage from heavy traffic. 
 
 We now come to an interesting point, the connec- 
 tion of feeders, mains, and service lines, 
 
 the second or changeable method in very many cases 
 where the solidly insulated junction used to be em- 
 ployed. 
 
 In making the solid connection on the Callender 
 system — as, for instance, where feeders are required to 
 be permanently connected at a certain point to the 
 distributors, or where two distributors are required to 
 be connected in parallel to two other distributors at 
 right angles — the method shown in detail in Fig. 155 
 is adopted. A double cast-iron box is prepared, cast in 
 
ELECTRIC LIGHT MAINS. 
 
 371 
 
 one piece, with an inner and an outer box, shown in 
 plan and section, having their corresponding lids. 
 This double box is pierced on all four sides with 
 openings for the mains to be connected. The two 
 larger mains are now cut to the requisite length to just 
 pass a few inches into the inner box, and their copper 
 ends, E E x , are bared. A solid round copper socket, P, 
 having a flat connecting piece projecting upwards, is 
 securely sweated or soldered on each of these bared 
 
 a line with the second main cable and projecting 
 upward to the same height as the copper pieces 
 thereon. 
 
 All that has now to be done is to place solid lengths 
 of copper bar, L L, across the mains, securely fastened 
 by nuts and bolts, the positive main, E E, to the 
 distributor, D, passing over the other distributor within 
 touching, and the negative to the negative distributor, 
 T>i, passing over the positive without touching. 
 
 j^""^ 
 
 WWWWWWWWWW^ 
 
 vwwxww^x^vww^^wwwwwwwwwwwwwwwwxwwww y 
 
 qfn rr\\ 
 
 pl A©j4i 
 
 itW^ 
 
 OB 
 
 UJJ UJJ 
 
 LUJ LB IL & 
 
 / / / / WV\V\' 
 
 l A^7f/)77777777 V///,//////////////////// ///////M_lJ / ///////Mv/////77^77777/77^ Za 
 
 Fig. 156. 
 
 ends. The second pair of distributor cables, D D 1( are 
 passed right through the junction box without cutting, 
 and a space of some inches is bared to the copper upon 
 each of them, these spaces being not opposite to each 
 other, but opposite to the mains to which they are 
 to be connected. Similar copper pieces, Pj P x , are 
 sweated to each of these bare places, one copper piece 
 being fixed on the main, D M on a line with the ends of 
 the first main, E E, and the other piece being fixed on 
 
 The cables and the connecting pieces are then 
 insulated by semi- vulcanised material, and then pro- 
 tected by tape ; the inner box is filled in with molten 
 fine bitumen, and the cover placed on. The outer box 
 is then similarly filled in with bitumen, thus forming a 
 perfect protection from moisture getting into the joints 
 in the inner box. The last lid is placed in its place, 
 and the whole covered up with earth and finished off 
 level to the road with the same paving as the roadway. 
 
372 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 In certain cases these four-way boxes are left without 
 the inner space being filled in with bitumen to allow 
 junctions of house mains to be easily made. This type 
 of box is, however, designed to be filled in solid and 
 permanent, and, if accessible junction-boxes are re- 
 quired for the junction of house service mains or of 
 a branch distributing main, a solid main fuse is used 
 in a fuse-box with accessible manhole, as shown in 
 Pig. 150. These accessible fuse-boxes are applicable 
 to either system, solid or otherwise, and therefore need 
 not be again described. 
 
 "We now pass to the system of connecting mains and 
 feeders in an ordinary central station network. In an 
 extensive area of supply, the mains, both of feeders 
 and distributors, will spread over a considerable area. 
 Two things may happen to necessitate change in the 
 connections. In the first place, a fault may be found, 
 
 unscrewed, can be swung round off the bar, and thus 
 disconnected at once, without interfering with or 
 touching any of the others. 
 
 The way in which this is carried out is easily seen 
 in the illustration. A double cast-iron box similar 
 to those already described for the simple four-way 
 junction, has leading into it the ends of the main, 
 M Mj. To each of these is sweated or soldered copper 
 socket pieces, P, having a broad solid lug of copper, L, 
 which is pierced for a bolt. The broad solid omnibus 
 bar, O B, is bolted closely to these lugs, having a short 
 piece of similar thick copper, S P, through which the 
 full connection across is made. It can easily be seen 
 that by unscrewing this swing-piece, S P, and swinging 
 it back, the main, M 1( becomes disconnected. 
 
 In the same way the ends of the distributing mains 
 are led into the junction-box, each having its end bared 
 
 Fig. 157. 
 
 in which case tests would require to be made, and a 
 certain length of cable might be required to be cut out 
 of circuit ; and, again, the demand for current may be 
 altered in certain districts, and the points at which 
 the feeders should feed into the distributors may 
 require to be changed into some other point. It will 
 be seen from this that a ready and rapid method of 
 disconnecting mains, and of connecting up others in 
 their place, is desirable on both grounds. 
 
 A simple yet ingenious method is used in the 
 Callender system, shown in plan and section in 
 Pig. 156, which illustrates a single-pole junction-box, 
 having the ends of a network of four subsidiary mains 
 connected to the feeding main. The principle of the 
 arrangement is that of having, on the ends of each 
 main, lugs of solid copper, which can be screwed to 
 a copper " omnibus " bar, and so arranged that each, if 
 
 and a socket piece fastened thereto. Short swing- 
 pieces, S P, are then bolted between the omuibus bar 
 and these distributors. If occasion arises to alter the 
 connection, or to cut off the supply from a certain 
 district, all that is it necessary to do is to unscrew one 
 bolt and then move the swing-piece round clear of the 
 bar. The position of these pieces is arranged so that 
 each can be swung round at its centre without the 
 necessity for touching or removing any other con- 
 nection. 
 
 The accessible junction-boxes are fitted with an 
 indiarubber gasket round the joint of cover, and are 
 screwed down with a phosphor bronze set-screw. 
 
 The buried junction-boxes are used under roadways, 
 and are covered with the ordinary pavement without 
 coverplate, the paving being lifted if it becomes neces- 
 sary to examine the connections, in which case the 
 
Electric light mains. 
 
 373 
 
 bitumen can be easily removed by the use of a heated 
 
 tool. 
 
 Ordinary sizes of cables are : for feeders 91 / n , equal- 
 ling one square inch sectional area, 61 / u , equalling -j^in, 
 sectional area of copper, and both can be laid up to 
 250 yards length ; for heavy feeders 91 / 9 , equalling 1 Jin. 
 sectional area, laid in 150-yard lengths. The distri- 
 butors are usually 87 /i3 cable, Jin. sectional area, or 19 / 14 , 
 T yn. sectional area, laid, according to circumstances, 
 up to 600 or 700 yards. The latter size is usually 
 employed for the house service lines. 
 
 The Callender solid system is particularly applicable 
 to high-tension systems for arc lighting in series where 
 the mains have to be led from lamp to lamp under the 
 streets, and the strength of current remains constant, 
 without necessitating increase of cable section even if 
 the number of lamps is increased. 
 
 =*&= 
 
 years, and has received much favourable commenda- 
 tion, in Great Britain from no less an authority than 
 Mr. Preece, and in America numerous experts have 
 testified to its efficiency, and dating back as far as 
 1877, in which year Mr. Brooks laid some experimental 
 wires in Philadelphia. 
 
 In 1887 a line of \\ miles was laid by Mr. Brooks 
 on the Pennsylvania Railroad. In this cable there 
 were 18 splices and 53 conductors, and a very high 
 insulation value was obtained, and an important 
 feature in reference to telegraph and telephonic 
 circuits was practically demonstrated — viz., its low 
 inductive capacity, which in this case gave a mean 
 value of about '212 microfarad per mile. 
 
 The experience in America of over 10 years, though 
 almost exclusively telegraphic, amply demonstrated 
 the permanence and efficiency of the system. It is 
 
 =Cc 
 
 a2t= 
 
 =£(= 
 
 Fig 159. 
 
 Sections of mains for high tension are shown in 
 Fig. 157, where the sizes of troughs for four and six 
 highly insulated 7 / M cahles are given. The process of 
 laying is exactly the same as that previously described, 
 with the exception that the troughs are always 
 covered with a cast-iron cover, or in particular places 
 with steel covers, to afford absolute security against 
 accidental interference by picks of other workmen. 
 
 claimed in reference to these circuits that while other 
 multiple cables have frequently broken down and deve- 
 loped faults due possibly to lightning and other causes, 
 the Brooks circuits have never given way, and continue 
 to this day to give values as high as when the lines 
 were first put down. 
 
 In the earlier experiments Mr. Brooks employed 
 ordinary fluid mineral oil as insulating medium. This, 
 
 Fig. 160. 
 
 The method of connection of the arc lampposts in 
 series is seen in section and plan in Fig. 158, where a 
 junction-box at a street corner is shown. The cables 
 are laid in the trough with sufficient spare length to go 
 up the lamppost without joint, a junction-box being 
 let in under the pavement and covered with a steel 
 plate lid. These high-tension mains can be easily laid 
 in lengths up to one mile without joint. 
 
 The Brooks Oil Insulation System. 
 
 This system, the invention of the late Mr. David 
 Brooks, of Philadelphia, embodies the use of heavy 
 mineral oil— one of the best of insulators— as the 
 dielectric for high-tension underground mains. The 
 system has recently been brought prominently to the 
 front in England by Messrs. Johnson and Phillips. 
 The system has, however, been well known for many 
 
 however, was found to have a very high penetrating 
 power, and continual care was needed to keep the oil 
 confined. The oil now adopted, produced from the 
 waste products of rosin oil, is a thick sticky mass less 
 liquid than treacle, and has the advantage of being 
 
 extremely cheap. 
 
 Practical details have now been carefully worked 
 out, and it may be confidently stated that the Brooks 
 svstem, well carried out, has the best claims of 
 permanency and efficiency of any underground high- 
 tension system, and that it will probably contrast 
 favourably as to cost with any other system of anythmg 
 like similar insulation and permanency. 
 
 The dielectric is a pure homogeneous thick fluid of 
 extremely high insulation value. It does not oxidise 
 W hea exposed to the air, and when occupying he 
 whole of the space in an iron pipe beyond that occupied 
 
3?4 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 by the conductors and their fibrous covering, possesses 
 all the elements of permanency — in fact, theoretically 
 the system seems to exclude all the ordinary factors of 
 decay, and the permanence of the system seems only 
 limited by the life of the containing pipes, which, in 
 ordinary cases, we believe have been known to last 
 
 Fig. 161. 
 
 over 20 years when used for gas or water, where it 
 must be remembered both the inside as well as the 
 outside of the pipe is exposed to influences tending to 
 cause decay; whereas in the Brooks system it is 
 evident that no decay of the pipes could go on from 
 
 Fro. 163. 
 
 the inside, and any deterioration that took place would 
 be limited to the oxidising and perishing process from 
 the outside only. Hence it is urged with much show 
 of reason that the Brooks system must rank well, not 
 on the score of insulation alone, but on that of per- 
 manency. 
 
 The mechanical details of the Brooks system are 
 seen in the accompanying cuts, larger details of which 
 are given further on. 
 
 The iron pipes are used in the usual lengths made 
 and kept in stock, and have their ends carefully bell- 
 mouthed, and connected together by T-sockets, as 
 shown in the small cut, Fig. 159, and at suitable 
 distances draw-in boxes, Pig. 160, are provided, which 
 may or may not have glands for the cables to pass 
 through, according to circumstances. Branch circuits 
 are taken off by means of junction boxes fitted at points 
 
 Fig. 162. 
 
 determined upon. The junction-boxes are in all cases 
 provided with glands for the cables to pass through, so 
 that the fluid dielectric may be removed from the 
 junction-box for the purpose of making a branch 
 
 Fig. 164. 
 
 connection, or for testing purposes, the glands then 
 preventing the dielectric from flowing into the box 
 from the pipes containing the mains. 
 
 The branch connections connected thereto are pre- 
 ferably lead-covered cables of the ordinary type, which 
 also pass through glands. The junction-boxes may 
 
ELECTRIC LIGHT MAINS. 
 
 37§ 
 
 also contain fuses between the branch circuits and the 
 mains. The terminals are formed of cast-iron boxes 
 of special form, provided with glands with suitable 
 ■vulcanite sleeves, through which the ends of the cables 
 are brought, Fig. 161. At the highest point of the 
 
 prepared and all joints in the pipes carefully made, the 
 cables, which consist of ordinary stranded copper con- 
 ductors, covered to a suitable thickness with raw jute 
 hemp strand, preferably by braiding, and again 
 
 or 
 
 wound over all with a sewing of braiding, are coiled on 
 
 'J 
 
 <> 
 
 ^~ 
 
 & 
 
 <> 
 
 -a- 
 
 e- 
 
 -f 
 
 ■£5- 
 
 & 
 
 ■& 
 
 a 
 
 w- 
 
 ■9- 
 
 -®- 
 
 ' . — \ ' 
 
 
 
 
 
 J) 
 
 
 *■+- 
 
 ^^ *_ 
 
 _.,i.-'' 
 
 
 
 
 
 U 
 
 Fib. 165. 
 
 line a tank or cast-iron box with removable lid is 
 provided, which is occasionally examined and kept 
 charged with the fluid dielectric, so as to ensure the 
 whole system being fully charged. A small trench 
 having been dug and a section of the line having been 
 
 drums and placed in a tank of the fluid dielectric and 
 carefully heated up to about 300 deg. F., and main- 
 tained at this temperature sufficiently long to drive off 
 all moisture. The cable may be of suitable size, and 
 composed of varying number of strands; that shown 
 
sw 
 
 Practical electrical Engineering. 
 
 in section, Fig. 162, consists of four cables of 19 /i 3 , with 
 four pilot wires of No. 12 gauge. The end or ends 
 are attached to a hauling-in strand of wire or rope 
 previously passed into the pipes, and the cable is drawn 
 in direct from the hot dielectric, Figs. 163 and 164, by 
 means of a winch. 
 
 Messrs. Johnson and Phillips have made several very 
 severe tests of the system at their works, and continue 
 to have the strongest convictions that it is a reliable 
 
 During the interval many searching tests of the line 
 have been made, and on many occasions six of the wires 
 have been grouped into two circuits of three wires each, 
 and a Kapp alternator giving 2,000 volts has been con- 
 nected to one end of the line, and a group of incande- 
 scent lamps lighted at the other end for several hours 
 continuously. Measurements made before and after 
 the run showed no fall in the insulation, and it must be 
 remembered that in this case a pressure of 2,000 volts 
 
 rooks Syslr^, r 
 
 Fig. 166. 
 
 and efficient one. A line of 700ft. in length was laid 
 at Charlton, Kent, under the supervision of Mr. David 
 Brooks, who visited England in order to convey 
 to Messrs. Johnson and Phillips his experiences in 
 the carrying out of his method. The cable consisted 
 of seven No. 18 wires, each wire double cotton-covered, 
 and braided with cotton twisted into a cable and braided 
 over all ; this length of cable was drawn into a lin. 
 iron pipe, and the whole system filled with hot fluid 
 dielectric. The line has now been down for about two 
 years, and it is to-day as perfect as when first completed. 
 
 was exerted between wires whose covering of cotton did 
 not exceed ^in., intimately twisted together. 
 
 Another line was afterwards put down at Charlton, 
 one-tenth of a mile in length, containing two cables, 
 each of T /u wires, braided with a fibre to a diameter of 
 about fin. These were drawn into the pipe in two 
 lengths, and the whole system filled with dielectric. 
 This line has frequently had, under the direction of Mr. 
 Kapp, a pressure of 14,000 volts for five hours at a time 
 upon it. The test was made by transforming up from 
 a 2,000-volt Kapp alternator to 14,000 volts, and at the 
 
ELECTRIC LIGHT MAINS. 
 
 37? 
 
 other end transformmg down again to 2 000 volts, and A recent communication from America dealing with 
 
 hen hghtmg 20 100-volt mcandescent lamps in series this system gives particulars of underground mlms 
 
 to full brightness. This potential gave a sparking with liquid insulation laid for the Heisler Electric 
 
 
 -^^- 
 
 S 1 
 
 s 
 
 distance between points of ifin. at the dynamo end of Company, which gives entire satisfaction, having been 
 
 the line, and at the far end a somewhat greater in constant use for nearly a year in circuits using very 
 
 distance. The line tested immediately before and after high electromotive force, varying from 2,400 volts up 
 
 the run showed no change in the insulation value. to 4,000 volts, and the remarkable feature is the fact 
 
3?8 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 that the insulation resistance is higher to-day than 
 when first installed. Eepeated efforts have been made 
 to break down the insulation by subjecting it to various 
 and unusual strains, such as suddenly breaking the 
 circuit when the dynamo was running at its greatest 
 speed, and also subjecting lengths of the line to a static 
 
 jute, and enclosed in an iron pipe, with liquid insu- 
 lation, are used for connecting the church with the 
 overhead high-tension wires. The Brooks main goes 
 under the graveyard to obviate the use of overhead 
 mains. The wires have carried the full pressure of 
 2,000 volts for a period of over 14 months. 
 
 Fig. 169. 
 
 charge equal to 140,000 volts. .They stood perfectly, 
 though this test is sufficient to break through every other 
 kind of insulation, however perfect it may appear by 
 galvanometrical tests. 
 
 In England the Brooks system has been used for a 
 short length in the Keswick central station system. 
 Two single No. 14 copper wires, twice braided with 
 
 They have also been used on low-tension work in a 
 large private installation for Lord Windsor, where 500 
 yards of four strands of 87 / 13 are worked two each in 
 parallel as main conductors under the grounds from 
 the dynamo-room to the house, to entirely do away 
 with the necessity for overhead wires. Various other 
 private installations have adopted the system. Besides 
 
'ELECTRIC LIGHT MAINS. 
 
 379 
 
 this, the system has been for many years in use on 
 certain railways in England for telegraph wires with 
 very marked success. 
 
 Fig. 165 shows the details of the junction box 
 which is laid under the street level. The exits or 
 glands for the cable are arranged so that they can be 
 packed and can also be plugged inside to prevent 
 oil escaping, if the box is required to be left at any 
 
 In this system it is seen that there is no space left 
 for the accumulation of explosive gases, or gases 
 injurious to the dielectric, as might be the case where 
 a solid dielectric is used in conduits or iron pipes ; and 
 Messrs. Johnson and Phillips themselves believe that 
 the only other system that could compare on the score 
 of efficiency and permanency with the Brooks mains, 
 is that in which the cables highly insulated with india- 
 
 -[ 
 
 psss^^ 
 
 3- 
 
 Fig. 170. 
 
 part of the progress of the work. In laying, the cables 
 are drawn through from point to point at these boxes, 
 into which the liquid insulation is afterwards poured, 
 and the lid securely fastened down. 
 
 Fig. 166 shows the terminal-box, which stands 
 directly under tho switchboard at the generating or 
 
 Fig. 171. 
 
 at the receiving end of the line. Glands are provided 
 when required in the case shown for two cables and 
 pilot wires. 
 
 These two illustrations are shown in perspective 
 elsewhere ; they are quarter full size, and being repro- 
 duced from the working drawings, practically explain 
 themselves. 
 
 rubber or guttapercha are drawn into iron pipes kept 
 full of water. 
 
 Experience in America for a period of at least 10 
 years has already demonstrated that, in the Brcoki 
 telegraphic and telephonic mains, the cost per wire in 
 a cable of many Conductors is less than the cost per 
 wire insulated with guttapercha or indiarubber, even 
 including the additional charge of conduit-pipe and 
 boxes, or other device. 
 
 It has also been abundantly proved at Charlton that 
 mains for the distribution of electric light and power 
 will stand the severest tests that are ever likely to be 
 required with even far higher voltages than at present 
 are anywhere employed. 
 
 Probably one of the largest fields of usefulness for 
 the Brooks mains will be in connection with the high- 
 pressure three-wire alternate-current distribution for 
 lighting, and for alternate-current motors, of which the 
 installation to transmit several hundred horse-power a 
 distance of 12£ miles -at 25,000 volts, at the Oerlikon 
 "Works, Zurich, is the first and most noteworthy 
 example. 
 
 The general scheme of the use of the Brooks system 
 for public lighting at Hong Kong is shown in Fig. 167, 
 where the overhead lines are seen coming to tall posts 
 bearing insulators with an arc lamp above the wires. 
 In the crowded streets, where overhead wires would 
 not be advisable, the high-pressure wires for the arc 
 lamps in series, carrying 2,000 volts, are taken down 
 and under the streets in iron pipes filled with the 
 Brooks liquid insulation, are then run along the streets 
 and up a series of lampposts — one of which is shown 
 in the illustration— and then up vertical pipes again 
 to the overhead wires. 
 
 The details of the junction-box and insulator for the 
 top of the posts, where the overhead and underground 
 wires are connected, have been carefully worked out, 
 
§80 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 and are shown in Fig. 168. The wires are pulled 
 through the box, and the ends soldered together and 
 insulated, and a screw cap fixed over them. The 
 overhead line passes to an insulator specially adapted 
 for oil insulation. The underground mains descend 
 an ordinary pipe, and the whole is filled up with oil 
 insulation. 
 
 "We now pass to another and important application 
 of the Brooks system. 
 
 The drawings are almost self-explanatory. The 
 junction-box, shown in Figs. 169, 170, and 171, is 
 for the distribution in a number of directions of the 
 three-phase current. Each pipe contains a three- 
 strand cable, the strands of which on entering the 
 junction-box are separated and led through glands 
 to separate terminals, or omnibus bars, to which the 
 other distributing cables can be connected. Provision 
 is made for fuses, if desired, between the mains and 
 branches within the junction-box. 
 
 Now that the enormous pressures, at one time con- 
 sidered beyond control, are being handled without 
 much greater trouble, save for insulation, than the 
 currents of ordinary voltages, a system such as this 
 seems to be should become more and more valuable 
 to electrical engineers. 
 
 Ferranti Mains. 
 
 Among the various systems of mains for the distri- 
 bution of electricity few are more important, and 
 certainly none are more interesting, than those of 
 Ferranti. Mains for the use of 2,000 volts having 
 been in use for transformer distribution some years, 
 Mr. S. Z. de Ferranti, then chief engineer to the 
 London Electric Supply Corporation, determined to 
 transmit current at the enormous pressure of 10,000 
 volts to London from the company's generating station 
 at Deptford, a distance of 1\ miles. The difficulty of 
 accomplishing this was the more enhanced in that it 
 was necessary to lay these mains to a considerable 
 extent underground, whereas the previous high- 
 pressure distribution had been principally overhead. 
 Several systems of ordinary electric cables were first 
 tried, at 5,000 and 10,000 volts, but none of these were 
 successful. Mr. Ferranti, therefore, undertook to 
 construct his own mains, and, after exhaustive 
 trials, the desired result was obtained in a singularly 
 simple manner, by the proper combination of con- 
 centric copper conductors, covered with a cheap and 
 easily-procured insulating material, together with the 
 perfection of a special method of jointing. Mr. Ferranti 
 determined upon the use of concentric copper tubes as 
 conductors, to be placed one inside the other, and 
 separated by |in. of the most durable insulation that 
 it was possible to find. Mr. Ferranti resolved to 
 manufacture the mains in short lengths of 20ft., 
 necessitating an immense number of joint. 
 
 Scepticism as to the result was rampant, and it 
 is safe to say that no living electrical engineer beside 
 himself would, at the time he started this work, have 
 dared to propose or attempt to carry out such a system. 
 
 The insulating materials finally selected were paper and 
 black ozokerite, or earth wax. It will be interesting 
 here to give Mr. Ferranti's ideas upon the subject of 
 insulation of high-tension mains, a subject to which he 
 has naturally given more than ordinary attention. 
 
 The necessary points to consider are two : durability 
 of substance and the factor of safety. No one can 
 definitely tell how the insulation may change in course 
 of time, but what can be at once tested is the factor of 
 safety of such insulation. No fault will develop in the 
 insulation unless the dielectric is sufficiently strained to 
 produce enough mechanical effect to cause a gradual 
 change ; in other words, unless it is strained beyond 
 the limit of elasticity. Strained below the limit, the 
 insulation will not give way ; strained above its limit, 
 the breakage is but a question of time. The Ferranti 
 mains are made with a very large margin of safety, a 
 margin that has been accurately determined by direct 
 experiment. From these experiments it is found that 
 a thickness of x^-in. of paper saturated with the black 
 wax, is pierced within one hour by 20,000 volts — some 
 specimens will go within 10 minutes, and nearly all 
 within the hour ; above that thickness the insulation 
 is not pierced. This being so, with the present mains, 
 which have £in. of insulation, there is eight times that 
 thickness, and with half the number of volts we have 
 16 as the factor of safety. In the Ferranti mains there 
 are 60 layers of waxed paper wrapped one above the 
 other, and the factor of safety is so large, even if one 
 or two layers were partially faulty, it may be trusted 
 to the remainder to give perfect safety. In point of 
 fact, no failures of tested mains have yet been 
 experienced by reason of direct failure of the insula- 
 tion. Fifteen faults in all were experienced in the 
 thirty miles already laid, mostly of want of continuity. 
 Only two faults were found with 20,000 volts, both of 
 these due to water in the joint at the time of making. 
 
 With regard to the jointing, this is a most important 
 part of the whole system. The mains are constructed 
 ready for laying in the ground. The weight is heavy, 
 and the length is kept to 20ft., so that a few men can 
 handle them. In the thirty miles of mains there are 
 therefore seven or eight thousand joints. To allow 
 this the joint must be absolutely safe when once 
 properly made. This indeed is what renders the 
 whole scheme practicable. To make such a joint in 
 the insulation was an achievement long doubted, but 
 is now proved perfectly possible. The desiderata were 
 a very long jointing surface and perfect contact. 
 These were achieved by making a long slanting cone- 
 like surface, the protruding point of one main fitting 
 into the hollow of the next, these surfaces being 
 accurately and truly turned and polished to gauge, and 
 then heated and forced together by hydraulic pressure, 
 by which means the insulation becomes practically one 
 solid piece. 
 
 Briefly, then, the Ferranti mains consist of tubes of 
 high conductivity copper cut into 20ft. lengths, and 
 straightened. The usual size of tube to carry up to 
 250 amperes is of iin. section, and the size T Vin. inside 
 
ELECTRIC LIGHT MAINS. 
 
 §8i 
 
 diameter, and yfin. outside diameter. Lengths, 20ft. 
 long, of brown paper are cut off from a roll 3ft. wide, 
 and a length of this paper is glued by its edge to the 
 
 bath of hot melted black mineral wax, drawn over 
 rollers and through the air for some distance until dry ; 
 they are then cut into 20ft. sheets and placed for use 
 
 'I 
 
 v, 1va« The copper tube has squared pieces of 
 copper tube. Meanwhile, other rolls of paper are on ' shel ™ S - { ^ end> and is then placed in 
 passed over long iron plates heated below by open fires, wood KnocKe ^ roUer on a table which 
 
 and thus thoroughly dried ; these are passed through a sockets ol a siowiy 
 
§82 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 has at the back a set of rollers, a bath of hot wax, wax is made to flow up and saturate the sheets. When 
 and revolving gear. As the tube revolves, lengths the required thickness is served the wax is made to flow 
 of the prepared paper are inserted between it and back, the insulation compressed still more upon the 
 
 7sy///////////////////////////% 
 
 Fia. 175. 
 
 rm 
 
 r^^ 
 
 i^^^^^^^^^^^^sli 
 
 s^sr 
 
 ft 
 
 Fig. 176. 
 
 A, Inner Conductor ; B, Outer Conductor ; C, Inner Insulation ; E, Outer Insulation ; a, Copper Rod at Joint of Inner Conductor ; D, Iron Sheath ; 0, Iron 
 Sleeve at Joint of D ; F, Copper Sleeve at Joint of Outer Conductor ; H, Hole and Plug In G for running in Compound to finish Joint. 
 
 the brown paper, sheet after sheet, until 60 sheets are copper tube, and a tape is wound spirally over the 
 
 served in. During this time heavy rollers come down whole. The tube covered with its insulation is then 
 
 upon the tube, compressing the paper, and at the same removed, the wooden pieces knocked out, and the 
 
 time by displacement boxes dipping into the bath, the whole slipped into a second tube of copper. Thit 
 
tube is of the same total cross-section as the first 
 
 viz., Jin. — but being larger in diameter is propor- 
 tionately thinner. The size of this tube is lf£in. 
 inner diameter, and ljfin. outer diameter. The tube 
 is left a little larger, so as to slip easily over the inner 
 tube with its insulation, and is then passed through a 
 • die and drawn down upon the insulation. This outer 
 tube is now served with the insulation in the same 
 manner as the inner tube : first a length of brown 
 paper glued by one edge, then several sheets of waxed 
 paper to the thickness of ^-in., compressed and taped 
 
 ELECTRIC LIGHT MAINS. 
 
 fiT'to '« « G 2 '^ 1S C ° ned ° Ut for a distance <* 
 
 ™« 7T\ G I™ 6 taper ' While the iron ^eld is 
 
 for a , 1" W ° f llln - and the 0uter in ^ation 
 
 Z *T\ 0i 1 8m -' Iea ^ the ou ^ conductor 
 exposed for this length. 
 
 The mains are now ready for laying. Each length is 
 separately tested to 20,000 volts, the ends are capped 
 to prevent dirt getting in, and are then sent out to the 
 place required. They are jointed together on the spot, 
 as follows (see Fig. 174) : A tight-fitting copper rod, a, 
 12m. long, is driven into the inner tube, A, of the 
 
 Fiq. 177. 
 
 as before. The whole is slipped into an outer iron tube 
 to act as a protecting shield. Melted wax is forced by 
 a pump through a small hole in the centre beneath 
 this iron casing till the inner space is completely filled. 
 The whole is then sawn off at the ends into exact 20ft. 
 lengths. The section of the main is shown in Fig. 172. 
 The next process is the preparation of the joint. One 
 end of each length is formed into a projecting cone, and 
 the other end into a hollow cone, by means of a special 
 
 hollow coned end. A tight-fitting sleeve of copper, F, 
 is driven for a distance of 8in. on the outer conductor 
 of the main to which it is to be jointed, and this sleeve 
 firmly gripped on by means of a special tool by three 
 or more circular corrugations, as shown. The two 
 cones are then inserted one within the other, the 
 surfaces being previously warmed, and are forced 
 together and driven home by screw clamps, a total 
 pressure of about three tons being employed, and 
 
 Fig. 178. 
 
 hollow spindle lathe. The length is placed on this lathe, 
 the outer iron shield is removed for a distance of 17in. 
 from the end; 14in. of the outer insulation, E, thus 
 exposed is then turned up, leaving bare that length of 
 the outer copper conductor, B ; 6in. of this conductor 
 is now removed, and the insulating material between 
 the two tubes is turned carefully down, forming a cone 
 at the end of the cable, as shown in the illustration, 
 Fig. 173. The interior of the inner tube is then slightly 
 enlarged by drilling for a length of 6in. inside. 
 The length of main is then reversed, and the insula- 
 
 when still under compression the copper sleeve is 
 firmly locked to the other outer conductor by means 
 of circular corrugations as before described. The 
 sleeve, F, and the outer insulation, E, are wrapped at 
 the junction with insulation material until they become 
 of the same external diameter as the iron tube, D, 
 when an iron sleeve, G, 30in. long, is passed over the 
 joint and corrugated down at both ends. In order to 
 fill up any air space in the outer insulation, hot wax is 
 forced in through the boss, H, of the sleeve, G-, the 
 whole being finally closed with a gas plug. 
 
384 
 
 PRACTICAL ELECTRICAL ENGINEERING. 
 
 The laying of concentric mains is thus relatively 
 a very simple matter ; they are supplied ready for 
 jointing together, and may be carted out and laid as 
 gas-pipes are, no cement channels or specially-prepared 
 conduits being necessary. It is usual, however, in 
 crowded streets to lay them in a wooden trough, with 
 wooden separating slips, the trough being filled in with 
 pitch, with an upper layer of concrete for extra protec- 
 tion. When laid in this manner they are subject to 
 variations of temperature, to compensate for which all 
 that is necessary is to give them at certain points a 
 slight wave in laying. To bend a main to go round a 
 corner an ordinary rail bender, as used on railways, is 
 employed ; a curve of 6ft. radius being made in this 
 way with but little trouble. In bending, it is found 
 there is no appreciable drag between the layers of 
 insulation and conductors. 
 
 For making branch connections a special T-joint is 
 employed ; this consists mainly of a cast-iron box with 
 suitably-designed base and cover arranged to fit water- 
 tight. These joints do not, of course, appear on the 
 road surface, being inserted in the run of the mains as 
 required. The joint has three stuffing-boxes through 
 which the ends of the mains are brought in. A screw 
 bolt from the centre of the branch main connects to 
 the inner conductor of the main itself, and the 
 joint is wrapped with paper insulation ; the outer 
 conductors are connected with a gunmetal bridge-piece 
 of the shape shown, Fig. 175. Street boxes are also 
 placed in the run of the mains at distances of about 
 200 yards. These are iron boxes similar in principle 
 to the T-joint boxes, but are placed in small brick 
 chambers, having removable covers flush with the road 
 surface, Fig. 176. The interior is thus accessible for 
 testing and other purposes, while the arrangement of 
 the connections is such that the joints can be easily 
 and quickly connected or disconnected. These joint 
 and street boxes may be filled with rosin oil, by means 
 of which very high insulation is insured at these points, 
 and the full pressure of 10,000 volts may be safely 
 used. While the above description applies more par- 
 ticularly to mains for parallel distribution, the system 
 maybe employed with equal advantage for series work. 
 
 With regard to the resistance of long lengths of 
 these mains, a length of 1\ miles (i.e., 15 miles of 
 lead and return) between London and Deptford was 
 tested by Dr. Fleming, and the actual resistance was 
 found to be 2'20 ohms, while the calculated resistance 
 of a length of copper of that section was 216 ohms, 
 thus showing that the resistance of joints is inappre- 
 ciable. The mains can be touched on the outside and 
 handled with impunity when a current at the highest 
 voltage is flowing, without the possibility of anyone 
 receiving a shock, the metal covering being to earth 
 and acting as a complete discharge shield. In the 
 event of a fault occurring between the inner and outer 
 conductors, the current can only return direct to the 
 machine, where the safety fuses prevent it doing any 
 harm. The fuses employed are also illustrated here- 
 with, The smaller one, Fig. 177, has a 12in. break, and 
 
 is used in houses for the primary circuits of trans- 
 formers ; the larger fuse, Fig. 178, is identical with the 
 first, except that it is adapted for main currents, and 
 has a multiple fuse with a 24in. break. The plugs are 
 arranged for separating the multiple fuse wires. 
 
 The absence of necessity for channels or conduits in 
 the Ferranti system is an item which should be taken 
 into account, and has, further, the immense advantage 
 of avoiding the possibility of explosions from an 
 accumulation of sewer or lighting gas, which have 
 occurred so frequently with both high and low tension 
 systems throughout Europe, and America. Explosions 
 of this kind have already occurred in London. In fact, 
 when a conduit or line of pipes is opened, the presence 
 of gas (which is found, moreover, to impair the insula- 
 tion of cables laid in that manner) is very frequently 
 detected. Such methods are also liable, sooner or later, 
 to dangers which arise when water is present. For 
 instance, when bare conductors are used and water gets 
 access to them, they are liable to be short-circuited or 
 injured by electrolysis. With insulated cables in pipes 
 or conduits, the alternate presence and absence of 
 water affects the insulation, damaging it in time, and 
 when it is present in winter, it is liable to freeze — the 
 ice crushing or piercing the insulation and causing 
 partial or total short-circuitings. This occurred in 
 London during the late winter of 1890-91. 
 
 With regard to safety the following experiments 
 were carried out : In the presence of representatives of 
 the Board of Trade, the Post Office, and local authori- 
 ties, a main which had been running continuously 
 under a pressure of 10,000 volts was submitted to two 
 engineers, who, with a cold chisel and sledge-hammer, 
 managed, after considerable time and much deliberate 
 labour, to cut through from the outer to the inner 
 conductor while 200 h.p. at that tension was being 
 transmitted through it. They did not feel the slightest 
 shock, although they were standing on a large metal 
 plate making earth. A similar trial has been made 
 with a pickaxe. The mains have also been submitted 
 to a lengthened test, after laying, with 30,000 volts, 
 and have given no trouble. It has thus been forcibly 
 demonstrated that this system of conductors offers 
 every possible guarantee of human safety. 
 
 We have referred to the Ferranti mains more par- 
 ticularly, perhaps, in connection with high-tension 
 distribution, but it is apparent that they are also 
 adapted for 100-volt or other low-tension distribution, 
 or where a large mass of copper is required. By their 
 construction the greatest amount of copper conductor is 
 contained in the least space, the cables are buried direct 
 without the use of conduits, and can be carried any- 
 where and under pavements, where there is no room 
 for conduits. The house connections are rapidly and 
 cheaply made, perfect insulation is obtained, there is 
 no fear of any short-circuiting from water, and no fear 
 of explosions. For three-wire distribution the cable 
 is manufactured with three concentric conductors 
 instead of two, and is laid and jointed exactly as 
 above described. 
 
ustdieix: to "v-olttzmzie i. 
 
 A. 
 
 Acceleration, Definition of, 154 
 
 Accumulation in the Inductive Circuit, 28 
 
 Accumulators, see Batteries, Secondary 
 
 Acme Engine Governor, The, 234 
 
 Adiabatic Curve, An, 169 
 
 Adiabatic Expansion and Compression, 168 
 
 Alloys, Resistance of, 9 
 
 Ampere, Determination of the Absolute, 304 
 
 Ampere, the Unit of Current, 9 
 
 Ampere's Laws as to Conductors Carrying Currents, 22 
 
 Ampeke Meters, see also Measurement and Meters : 
 Ayrton and Perry's, 341 
 Calibration of, 341 
 Linesman's Detector, 344 
 Paterson and Cooper's, 342 
 Permanent and Electromagnets for, 342 
 Thomson's Gauge, 323 
 Thomson's Balances, 316 
 Thomson's Magnetostatic, 321 
 
 Ampere-turns, 25 
 
 Amperes, Volts, and Ohms, Their Relation, 304 
 
 Annealing, Its Effect on Resistance, 9 
 
 Arcing at Switches to be Avoided, 265 
 
 Area Method of Graphical Representation, 37, 156 
 
 Areas of Irregular Figures, Determination of, 158 
 
 Areas of Plane Figures, Determination of, 157 
 
 Armatures of Dynamos, see also Dynamos : 
 
 Calculating Heating of Coils, Forbes's and Esson's Formulae, 15 
 
 Determining Diameter of Shaft and Journals, 225 
 
 Graphic Method of Calculating Induced Electromotive Force in 
 
 One-Coil Armature, 33 
 Method of Representing Relation between Speed and Pressure, 40 
 
 Atomic Weights, Table of, 19 
 
 Ayrton and Perry's Meters, 341 
 
 Ayrton's Table of Specific Inductive Capacities, 32 
 
 B. 
 
 B.A. and the Legal Ohm, 7 
 
 Babcock and Wilcox's Boilers, 88 
 
 Balances, Electric, Sir William Thomson's, 316, 319, 323, 326 
 
 Bar Magnet (Permanent), Field of, 24 
 
 Barring Engines for Starting Engines, 249 
 
 Battery, Its Position in Wheatstone Bridge, 307 
 
 Battery Resistance, Measurement of, 329 
 
 Batteries, Secondary, Drake and Gorham's Switchboard for, 278 
 
 Bearinus of Steam Engines : 
 Function of, 223 
 
 of Robey Compound Engine, 223 
 Calculating Friction on Surface of, 224 
 and Journal, friction Between, 224 
 
 Blow-off Apparatus for Boilers, 121 
 
 Boilers, Steam : 
 
 Absolute and Effective Pressure, 61 
 
 Blow-off Apparatus, Use of, 121 
 
 Bourdon's Pressure Gauge, 114 
 
 Brushes for Cleaning Flues, 151 
 
 Check-valve for Feed-pipe, 139 
 
 Chimneys, Construction of, 66 
 
 Classification of, 68 
 
 Cleaning, How Carried Out, 151 
 
 Cocks for Testing Water-level, 114 
 
 Combustion, Determination of Available Heat of, 63 
 
 Combustion of Various Fuels, 62 
 
 Compositions for Preventing Scale, 150 
 
 Compound, Construction of, 78 
 
 Compound Breeches-Flued, Description of, 88 
 
 Compound Cornish : Davey, Paxman, and Co.'s, 86 ; Leeds Forge 
 
 Company's, 87 
 Compound Lancashire: Davey and Co.'s Economic Boiler, 81 ; 
 
 Galloway's, 82 ; Leeds Forge Company's, 84 
 Davies and Metcalfe's Exhaust Steam Injector, 137 
 Deposit, Apparatus for Getting Rid of, 121, 149 
 Deposits, see also Scale 
 Dewrance's Joint for Steam-pipes, 124 ; Renewable Stop-valve, 
 
 126; Scum-cock, 123; Steam-Reducing Valve, 130; Water 
 
 Gauge, 113 
 Dome, Use of, 68 
 Donkey Pumps, 131 
 Drilling and Punching Holes in, 104 
 Dry ana Superheated Steam, 61 
 Economisers, Use of, 144 
 Effect of Conduction on, 163 
 
 Boilers, Steam (continued) : 
 
 Explosions, Their Cause and Remedy, It 2 
 Feed-pipe, Check-valve for, 139 
 Feed-pipe, Duty and Position of, 139 
 Feed-valve, see Valve below 
 Feed-water Apparatus, Description of, 130 
 Feed-water, Heating of, 141 
 
 Feed-water Heaters : 
 Economy of, 141 
 Garrett and Sons', 141 
 Kirkaldy's, 142 
 Musgrave and Son's, 144 
 Robey and Co.'s, 143 
 
 Ferranti's Feed-valve, 140 
 
 Firebox, Construction of, 63 
 
 Fireplace, Parts of, 64 
 
 Foundations for, 54 
 
 Fox's Corrugated Furnace, 78, 84, 87, 111 
 
 Flues : Best Position for, 164 ; Brushes for, 151 ; Different 
 
 Arrangements of, 64 ; Expansion Joint for, 111 ; Protectors for, 
 
 121 ; Rings of, their Construction, 111. 
 Fuel Saved by Heating Feed-water, 141 
 Fuels Used in Different Countries, 63 
 Fusible Plugs, 121 
 Galloway's Safety-valve, 115 
 Garrett and Son's Feed Pump, 131 
 Garrett and Son's Feed-water Heater, 141 
 Grate, Construction of, 64 
 Green and Son's Economiser, 147 
 Gresham and Craven's Injector, 134 
 Gauges, Pressure, 114 
 Gauges, Water, 113 
 Heat Required to Evaporate Water, 61 
 Heat, Waste of, 64 
 Heating Surface, 63, 66 
 Hopkinson s Blow-off Valve, 122 ; Compound Safety-valve, 117 ; 
 
 Feed-v; lve, 140 ; Joint for Steam-pipes, 124 ; Steam Check 
 
 Valve, 27 ; Stop-valve, 125 
 Horizontal Forms of : Breeches-Flued, 73 ; Egg-Ended, 69 ; 
 
 Cornish, 69 ; Galloway Tubes for, 69 ; Galloway's Lancashire, 
 
 73 ; Musgrave and Son's Cornish, 70 ; Lancashire, 73. 
 Hydes and Wigfull's Steam Separator, 129 
 Impurities, Apparatus for getting rid of, 121, 149 
 Impurities, see also Scale 
 Incrustation, How Formed, 148 
 Incrustation, Prevention of, 149 
 Ixji.ctors, Steam : Davis and Metcalfe's Exhaust, 137 ; Exhaust, 
 
 137 ; Gresham and Craven's, 134 ; Lifting, 137 ; Non-lifting, 
 
 137 ; Principle of, 134 
 Jointing, 104 
 
 Joints, Calculating Resistance of, 105 
 Joints, Table of Efficiencies, 107 
 Kirkaldy's Feed-water Heater, 142 
 Lettering on Illustrations, Explanation of, 68 
 Locomotive or Multitubular, Description of, 75 ; Draught, Fuel, 
 
 and Steam, Table showing Relation between, 78 ; Garrett and 
 
 Son's, 77; Robey and Co's., 77 
 Low-Water Alarms, 120 
 M'Neil's Manhole, 112 
 Manholes and Mudholes, 112 
 
 Marine, Illustration of ss. " Brighton's" Boiler, 78 
 Materials Used In and How Worked, 101 
 Minns and Co.'s Boiler Fluids, 150 
 Mumford's Donkey Pump, 131 
 Musgrave and Sons' Feed-water Heater, 144 
 Musgrave and Sons' Steam Separator, 129 
 Non-Conducting Coverings for, 163 
 Paxman's Expansion Flue, 111 
 Pressure Gauges, 114 
 Priming, the Anti-Priming Pipe, 129 
 Pumps, Donkey, 131 
 
 Pumps, Feed, 130 ; Garrett and Son's, 131 ; Mumford's, 131 
 Ramsbottom's Safety-valve, 117 
 Reducing- valves, 130 
 Rhodes's Stop-valve, 125 
 Ri vetting, 102 
 
 Rivets, Rule for Finding Diameter of, 107 
 Rivets, Finding Shearing Resistance of, 106 
 Robey's Feed-water Heaters, 143 
 Robey s Stop-valve, 125 
 Safety Apparatus, 115 
 Safety-valves, see Valves below. 
 Savill's Steam Separator, 130 
 Scale, How Formed, 148 
 Scale, How to Prevent, 149 
 Schaffer's Gauge, 115 
 Scum-pipe, The, 122 
 
11. 
 
 INDEX. 
 
 Boilers, Steam (continued) : 
 
 Sensible and Latent Heat, 61 
 
 Shape of, 68 
 
 Shells of, How Built up, 102 
 
 Shells, Calculating Strength of, 108 
 
 Stays for, 109 
 
 Steam, How Raised in, 61 
 
 Steam, Relation Between Pressure, Temperature, and Density of, 
 61 
 
 Steam- valves, see Valves below 
 
 Steam Injectors, see Injectors above 
 
 Steam-pipe and Fittings, 124 
 
 Steam Separators : Use of, 128 ; Hydes and Wigfull's, 129 ; 
 Musgrave and Son's, 129 ; Savill's, 130 
 
 Strength of, 101 
 
 Test-cooks, 114 
 
 Testing, 151 
 
 Tubes, Calculating Strength of, 110 
 
 Valve, Blow-off, Hopkinson's, 122 
 
 Valve, Check, for Feed-pipe, 139 
 
 Valves, Feed : Use of, 139 ; Ferranti's, 140 ; Hopkinson's, 140 ; 
 Renewable, 140 
 
 Valves, Safety : Board of Trade Rules, 119 ; How to Calculate 
 Size of Opening, 119 ■ Construction of, 115 ; Dead-Weight, 
 115 ; Double-Lever, 117; Galloway's, 115; Hopkinson's, 117 ; 
 Ramsbottom's, 117 ; Relief-valves, 117 ; Rules for Construction 
 of Valve, 119 ; Spring, 116 ; Rules for Calculating Size of 
 Springs, 120 ; How to Find Working Pressure, 117 
 
 Valves, Steam Check : Use of, 127 ; Hopkinson's, 127 
 
 Valves, Steam-Reducing : Use of , 130; Dewrance's, 130 
 
 Valves, Stop or Steam : Dewrance's Renewable, 125 ; Hopkin- 
 son's, 125 ; Robey's, 125 ; Rhodes's, 125 
 
 Vertical, with Flue Tubes : Davey, Paxman, and Co.'s " Essex,'' 
 100 
 
 Vertical, Multitubular : Ransomes, Sims, and Jefferies, 101 
 
 Vertical, with Water Tubes : Garrett and Son's, 95 ; Musgrave 
 and Son's, 95 ; Field's, 98 
 
 Water Alarms, Low, 120 
 
 Water, Heat Required to Evaporate, 61 
 
 Water-Level Indicators, 113 
 
 Water Supply, How Kept Up and Regulated, 130 
 
 Water Tubes : Babcoek and Wilcox's, 88 ; Another Description, 
 94 
 
 Water Used in, Chemical Composition of, 148 
 
 Bottomley's Apparatus for Determining " H," 294 
 
 Bourdon's Steam Pressure Gauge, 114 
 
 Boyle's or Mariotte's Law of Expansion of Gases, 164 
 
 Breeches-Flued Boilers, see Boilers 
 
 Bricks as Building Materials, 57 
 
 British Association and the Legal Ohm, 7 
 
 British Association Committee on Electrical Standards. Its Work 
 
 192 
 Brooklyn Station, Edison's, Plans of, see Plates N and O 
 Brooks's System of Laying Electric Light Mains, 373 
 Browett, Lindley, and Co.'s Engine Governors, 232, 234 
 Brush Electrical Engineering Company's High and Low Pressure 
 
 Switchboard, 289 
 Building a Central Station, see Central Station 
 
 c. 
 
 CG.S. System Explained, 7, 291 
 Cables, Electric Light, see Mains 
 Calibration of Meters, 341, 343 
 
 Callender's Solid System of Laying Electric Light Mains, 367 
 Callender- Webber System of Laying Electric Light Mains, 361 
 Calorimeter, Musgrave's, 206 
 
 Capacity, Electrical : Measurement of, 335 ; Glazebrook and Shaw's 
 Method of Measuring, 337 ; Thomson's Method of Measuring 337 ■ 
 Use of Galvanometer for Measuring, 336 ; Wheatstone Bridsre for 
 Measuring, 338 ; Unit of, 31, 305 
 Cardew's Voltmeter, 343 
 Carnot's Heat Engine, 170 
 Cast-iron, Ultimate Tensile Strength of, 238 
 Central Station, The : 
 Boilers for, see Boilers 
 The Building, 43 
 Engines for, see Engines 
 Foundations oe : 
 
 Bricks as Building Materials, 57 
 
 Characteristics of Various Soils and Rocks, 49 
 
 Concrete, 59 
 
 Designing Footing Courses, 52 
 
 Drainage of, 48 
 
 Draining and Protecting Foundations, 51 
 
 Engine and Machine Foundations, 54 
 
 Estimating ior Excavation, 51 
 
 Excavated Material, Increase in Bulk of, 51 
 
 Excavating for, 51 
 
 Footing Courses, Masonry, 58 
 
 Central Station, The, Foundations or (continued) : 
 
 How Displacement Occurs, 47 
 
 Masonry Footings, Safe Offsets for, 53 
 
 Mortar, 59 
 
 Natural and Artificial, 47 
 
 Settlement of Structures, 47 
 
 Soils, Safe Bearing Power of Various, 59 
 
 Stability of Masonry, 54 
 
 Stones as Building Materials, 55 
 
 Testing Strata, 48 
 
 Underfooting, 51 
 
 Walls, Construction of, 57 
 Mains for, see Mains 
 Meters for, see Meters 
 
 Plans of Edison's Brooklyn Station, see Plates N and O 
 Plans of Installation at Lord Rosebery's Mansion, Mentmore 44 
 
 45, 46 
 The Site, 43 
 
 Switchboards for, see Switchboards 
 Testing Instruments Used in, 341 
 
 Centrifugal Force, 160 
 
 Centrifugal Force, Action of, on Flywheels, 238 
 
 Centrifugal Governor, Principle of, 230 
 
 Characteristic Curves : 
 
 Of Compound Dynamo, 43 
 Of Series Dynamo, 40 
 Of Shunt Dynamo, 42 
 
 Chemical Action of Current, 17 
 Chemical Equivalents, Table of, 19 
 Chimneys for Boilers, Construction of, 66 
 
 Circuits : 
 
 General Description of, 5 
 Conductive : 
 
 Acts as a Magnet, 23 
 
 Ampere's Laws as to Conductors Carrying Currents, 22 
 
 A Channel for Electricity, 5 
 
 Chemical Action of Current, 17 
 
 Conductivities of Various Metals, 8 
 
 Conductors and Insulators, 6 
 
 Current in, How Generated, 11 
 
 Divided, Examples of, 9 
 
 Effect of Annealing on Resistance, 9 
 
 Effect of Length and Sectional Area'on Resistance, 7 
 
 Effect of Temperature on Resistance, 8 
 
 Electrolysis, 17 
 
 Field of a Conductor, 20 
 
 Forms of and Materials for, 6 
 
 Fusibility of Metals, 17 
 
 Heat and Work Units, 14 
 
 Heating Effect of Current, 13 
 
 Heating of Dynamo Coils, 15 
 
 Illustration of Transformation of Energy in Electric Lighting, 
 13 
 
 Illustration of Simple, 6 
 
 Interaction of Fields, 20 
 
 Internal and External Resistance, 10 
 
 Kirchoffs Laws as to Current, 11 
 
 Lines and Loops of Force, 20 
 
 Multiple Arc or Parallel Circuits, 10 
 
 Ohm's Law, Explanation of, 9 
 
 Part Played by the Conductor in, 11 
 
 Phenomena of, 6 
 
 Phenomena Connected with the Exterior of a Conductor 
 Carrying Current, 20 
 
 Phenomena Connected with Wire Carrying Current, 12 
 
 Resistance of Alloys, 9 
 
 Resistance of, What it is, 6 
 
 Resistance of Branch Circuits, 11 
 
 Resistance of, Unit for Measuring, 7 
 
 Resistance of Metals, Table of, 11 
 
 Resistance Increased by Interference with Field of, 22 
 
 Shunt Circuit Explained, 10 
 
 Shunt and Compound Winding Explained, 10 
 
 Specific Resistance, 7 
 
 Transformation of Energy, 13 
 
 Units for Comparing Phenomena of, 7 
 
 Voltameter, 19 
 
 Water-flow and Current-flow Analogy, 12 
 Inductive : 
 
 Accumulation of Pressure in, 28 
 
 Condensers, 31 
 
 Description of Simple, 28 
 
 Distribution of Accumulation in, 30 
 
 Erroneous Notions as to, 5 
 
 A Modification of the Conductive, 28 
 
 Positive and Negative Pressure, 29 
 
 Quantity and Capacity, 31 
 
 Resistance of Dielectric, 29 
 
 Short Circuits in, 31 
 
 Specific Inductive Capacities, Table of, 32 
 
 Specific Resistance, 31 
 
INDEX. 
 
 in. 
 
 Circuits (continued) : 
 Magnetic : 
 
 A Channel for Electricity, 5 
 Compared with Conductive Circuit, 23 
 Effect of Heat on Magnets, 25 
 Electro and Permanent Magnets, 24 
 Inductive Action, 26 
 
 Interactions of Conductors and Magnetic Fields, 27 
 Interactions of Magnetic Fields, 25 
 Leakage in, 25 
 
 Loops of Force and Magnetic Field, 23 
 Magnetic Conductivity, Pressure and Resistance, 25 
 Magnetic Loops, Formula for Calculating Number of, 25 
 Magnetic Loops, On what their number is dependent 24 
 Magnetic Resistance of Dynamo, 25 
 A Perfect Magnet, 23 
 Properties of Magnets, 23 
 Resistance of, How Lessened, 26 
 Clarke's Method of Measuring Electrical Pressure, 334 
 Coils Act as Magnets, 23 
 Coils, How to Calculate Heating of, 15 
 
 Combustion, Determination of Available Heat of, in Boilers, 63 
 Compound Boilers, see Boilers 
 Compound Dynamo, see Dynamos 
 Compound Engines, see Engines, Steam 
 Compound Winding Explained, 10 
 Concrete for Use in Foundations, 59 
 Condensers, Electrical : Description of, 31 ; How to Make 335 • 
 
 How to Measure Capacity of, 336 ; for Increasing Capacity, 335 
 Condensers, Steam : Different Makes of, 227 ; Surface Variety of 
 229 ; Use of, 226 J ' 
 
 Condensing Water, How to Calculate Quantity Required, 223 
 Conduction, Its Effect on Boilers, Cylinders, etc., 163 
 Conductive Circuit, see Circuits 
 Conductivity, Effect of Annealing on, 9 
 Conductivity, Its Relation to Resistance, 7 
 Conductivity, Magnetic, 25 
 Conductivities, How to Compare, 7 
 
 Conductor Carrying Current, Phenomena Connected with, 12, 20 
 Conductors, Electric : 
 
 Carrying Currents, Ampere's Laws as to, 22 
 
 Carrying Currents, Interaction between, 20 
 
 Coiled and Carrying Currents, Act as Magnets, 23 
 
 Effect of Length and Sectional Area on Resistance of, 7 
 
 Effect of Temperature on Resistance, 8 
 
 Equivalent, 7 
 
 Field of, 20 
 
 Forbes's Experiments on Heating of, 14 
 
 Heating Effect of Current on, 13 
 
 Joule's Law for Heating of, 339 
 
 List of, 6 
 
 and Magnetic Fields, Interaction of, 27 
 
 Matthiessen's Table of Conductivities, 8 
 
 and Non-conductors, Properties of, 6 
 
 Part Played by in a Circuit, 11 
 
 Resistance of, Silvertown Testing Set for, 345 
 
 Specific Resistance of, 7 
 
 Conduits for Electric Light Mains, see Mains 
 
 Connecting-rods of Steam Engines : 
 
 Ends with Cotter and Gib, 219 
 
 Ends Marine, 221 
 
 How to Calculate Strength of, 217 
 
 Paxman's, 217 
 
 Robey's, 217 
 Contacts for Switches, 268 
 Convection, Definition of, 164 
 Co-ordinates, Their Use, 37 
 Cosines, Table of, 35 
 Coulomb, Unit of Quantity, 31, 305 
 Couple, Explanation of, 37 
 
 Crank of Steam Engine, Position of, During One Revolution, 242 
 Crank and Connecting Rods, Laws of Motion of, 239 
 Crank-pin, Diameter of, 224 
 Crankshaft of Engine, Duty of, 221 
 Crankshaft, Paxman's, 221 
 Cranked Shafts of Engines, 221 
 Crosby's Visible Drop-Feed Lubricator, 245 
 Cross-head Pin, Diameter of, 224 
 
 Crompton's System of Laying Electric Light Mains, 353 
 Crompton and Co.'s Switchboard, 278 
 
 Current : 
 
 Chemical Action of, 17 
 
 Distribution of in Divided Circuits, 9 
 
 Is Electricity in Motion, 6 
 
 Flow, Analagous to Water -flow, 12 
 
 Heating Effect of, 13 
 
 How Generated, 11 
 
 Ohm's Law for, 9 
 
 Phenomena Connected with, 12 
 
 Pressure and Resistance, Their Relation, 304 
 
 Unit of, 9, 302 
 
 Currents, KirchofFs Laws as to, 11 
 
 Cut-outeS*' ^ ELECTR °- GEAPH1CS and Dynamics 
 
 Cylinders of Steam Engines : 
 Changes of Temperature in, 176 
 Clearance, How to Determine, 197 
 Necessity for Clearance in, 181 
 Details of Construction, 209 
 Expansion and Compression of Gases in, 168 
 ilow bteam Behaves in, 172 
 Indicator Diagrams, see Indicator, Steam 
 Jacketed, 209 
 Lubricators for, 246 
 Non-jacketed, 209 
 Piston, Construction of, 211 
 Piston, Davey, Paxman, and Co.'s, 211 
 Piston, Robey and Co.'s, 212 
 Piston Positions During One Revolution, 242 
 Piston-rod, How to Calculate Strength of, 213 " 
 Piston Speed, How to Calculate, 213 
 Piston Speeds and Flywheel Diagrams, 239 
 Ports, Action of, Explained, 241 
 Slide-valve, Meyer's, 243 
 Slide-valve, Paxman's, 243 
 Slide-valve, Work of, 241 
 
 Table Showing Effect of Clearance on Performance of Eno-ines, 185 
 Variable Expansion Gear for, Section Showing Details of, 243 
 Use of Steam-trap, 210 
 
 Daniell's Cell as a Standard of Pressure, 331 
 Davey, Paxman, and Co., see also under Paxman. 
 Davey, Paxman, and Co.'s Boilers, 81, 86, 100 
 
 Da T. e T' *^ XMAlr > AND Co.'s Engines for Driving Dynamos, 252- 
 Girder Engine, 259 ; Undertype Engine, 253 ; Windsor Engine,' 
 
 Davies and Metcalfe's Steam Injector, 137 
 
 Density, Definition of, 153 
 
 Deposit in Boilers, Getting Rid of, 121, 149, 150 
 
 Deptford Central Station, Mains Used at, 380 
 
 Deviating Force, 160 
 
 Dewrance's Lubricators, 248 
 
 Dewrance's Scum-cock for Boilers, 123; Steam-pipe Joint 124 
 
 Steam Stop-valve, 126 ; Water Gauge for Boilers, 113 
 Dielectric, The, in the Inductive Circuit, 29 
 Dip, Magnetic, Explanation of, 293 
 Dip, Magnetic, Measuring Angle of, 300 
 Donkey Pumps, Description of, 131 
 Drake and Gorham's Accumulator Switchboard, 278 
 Drake and Gorham's Four-way Switch, 278 
 Dynamical Units, 153, 292 
 
 Dynamics, Principles of : 
 
 Acceleration, Definition of, 154 
 
 Areas of Irregular Figures, Determination of, 158 
 
 Areas of Plane Figures, Determination of, 157 
 
 Centrifugal Force, 160 
 
 Density, Definition of, 153 
 
 Deviating Force, 160 
 
 Effort and Resistance, 153, 156 
 
 Energy, Various Forms of, 154 
 
 Foot-pound, Unit of Work, 154 
 
 Force, Definition of, 153 
 
 Force, Deviating and Centrifugal, 160 
 
 Force, Measurement of, 153 
 
 Force, Unit of, 153 
 
 Friction, Work Consumed by, 160 
 
 Gravity, Acceleration of, 154 
 
 Horse-power, Unit of Power, 154 
 
 Inertia; Moment of, Explanation of, 155 
 
 Kinetic Energy, Definition of, 155 
 
 Length, Unit of, 153 
 
 Machine, Efficiency of, 160 
 
 Mass, Definition of, 153 
 
 Mass, Units of, 153 
 
 Mechanical Energy, Kinds of, 155 
 Planimeter, Amsler's Polar, 158 
 
 Planimeter, The Coffin, 169 
 
 Planimeter, Use of, 158 
 
 Potential Energy, Definition of, 155 
 
 Power, Definition of, 154 
 
 Power, Unit of, 154 
 
 Pressure and Volume, 156 
 
 Resistance and Effort, 153, 156 
 
 Specific Gravity, Definition of, 153 
 
 Specific Volume, Definition of, 153 
 
 Tangential Forces, Explanation of, 156 
 
 Time, Unit of, 153 
 
 Velocity, Unit of, 154 
 
 Velocity and Motion, Definition of, 154 
 
 Weight, Its Relation to Mass, 154 
 
 Work Consumed by friction, 160 
 
IV. 
 
 INDEX. 
 
 Dynamics, Principles of (continued) : 
 Work, Definition of, 164 
 Work, Representation of by an Area, 156 
 Work in Terms of Pressure and Volume, 156 
 Work, Unit of, 154 
 Work, Useful and Waste, 160 
 
 Dynamo Coils, Calculating Heating of, 15 
 
 Dynamo and Engine, Ascertaining Loss Between, 14 
 
 Dynamos : 
 
 Armatures of, Graphic Method of Calculating Induced Electro- 
 motive Force in, 33 
 
 Compound Wound, Characteristic Curve of, 43 
 
 Determining Diameter of Shaft and Journals, Example of, 225 
 
 Engines for Driving, Special Types of : Davey, Paxman, and Co.'s, 
 252 ; Robey's, 260 ; Browett, Lindley, and Co.'s, 261 ; Willans's, 
 264 D ; the Windsor, 264 J. 
 
 Formula for Magnetic Resistance of, 25 
 
 Foundations for, 54 
 
 Safe Temperature of, 16 
 
 Series Wound, Characteristic Curve of, 40 
 
 Shunt Wound, Characteristic Curve of, 42 
 Dyne, The, 292 
 
 E. 
 
 Earth, The, a Magnet, 293 
 
 Earth in Electric Light Mains, How to Localise, 351 
 
 Earth, Magnetic Meridian of, and the Tangent Galvanometer, 302 
 
 Earth's Magnetic Poles, Movements of, 302 
 
 Earth's Magnetism : Determination of Horizontal Component of, 
 294 ; Determination of M H, 295 ; Effect of H on Tangent Galvano- 
 meter, 302 ; its Effect on Magnets, 293 ; Ratio of Moment of 
 Magnet to Horizontal Component of, 296 
 
 Eccentrics of Steam Engines, 226 
 
 Eccentric of Steam Engine ; its Relation to the Slide-valve, 239 
 
 Economisers for Boilers, 144, 147 
 
 Edison's Brooklyn Station, Plans of, see Plates N and O 
 
 Efficiency of a Machine, 160 
 
 Effort and Resistance, 153, 156 
 
 Egg-Ended Boilers, see Boilers 
 
 Electric Lighting, see Central Station : 
 
 Illustration of Transformation of Energy in Production of, 13 
 
 Plans of Installation at Lord Rosebery's Mansion, Mentmore, 44, 
 45, 46 
 
 Plans of Edison's Brooklyn Station, see Plates N and O 
 Electricity : 
 
 Circuits for, see Circuits 
 
 Constant in Quantity, 5 
 
 Cannot be Generated, 5 
 
 Conductors and Non-conductors of, 6, see also under Conductors 
 
 An Entity Moved by Electrical Pressure, 5 
 
 Its Composition Unknown, 5 
 
 Manifests Itself in Different Ways, 5 
 
 In Motion, or Current, 6 
 
 Electrical Pressure, see Pressure 
 Electro-graphics : 
 
 Area, Method of Representation, 37 
 
 Characteristic Curve of Compound Dynamo, 43 
 
 Characteristic Curve of Series Dynamos, 40 
 
 Characteristic Curve of Shunt Dynamo, 42 
 
 Co-ordinates, Use of, 37 
 
 Couple, Explanation of, 37 
 
 Curve of Sines, 34 
 
 Graphical Representation of a Single Quantity, 35 
 
 Dr. Hopkinson's Work in, 33 
 
 Induced Electromotive Force in One-Coil Armature, 33 
 
 Method of Representing Relation Between Armature Speed and 
 Pressure, 40 
 
 Method of Representing Electrical Pressure, 40 
 
 Moment of Forces, 36 
 
 Resultant Effect of Forces, 35 
 
 Silvanus Thompson's Work, 37 
 
 Table of Sines and Cosines, 35 
 
 Torque, Explanation of, 37 
 
 Used by Cable-Layers before Electrical Engineers, 33 
 Electro-chemical Equivalents, Table of, 19, 312 
 Electrolysis, Table of Valencies, Atomic Weights, Electro-chemical 
 
 Equivalents, etc., 19 
 Electrolysis, What it is, 13, 17 
 
 Electromagnets : Shunt and Compound Wound, 10 ; How to Calcu- 
 late Heating of Coils, 15 ; Definition of Perfect, 23 ; First Made 
 by bturgeon, 24 ; and Permanent for Meters, 342 
 Electromotive Force, see Pressure 
 Electro-negative Elements, Table of, 19 
 Electro-positive Elements, Table of, 19 
 Electrostatic Voltmeter, Sir Wm. Thomson's, 323 327 
 Elements, Table of Electro-positive and Negative 19 
 Energy, Illustration of Transformation of, 13 
 Energy, Kinetic, Definition of, 155 
 Energy, Mechanical, Kinds of, 155 
 
 Energy, Potential, Definition of, 155 
 
 Energy, Various Forms of, 154 
 
 Engine and Dynamo, Ascertaining Loss between , 14 
 
 Engine and Machine Foundations, 54 
 
 Engine, Heat : Definition of, 170 ; Carnot's, 170 ; Efficiency of, 172 
 
 Engine, Steam, see also Boilers : 
 
 Barring Engines for Starting, Musgrave's, 249 
 
 Bearing Surface, Calculating Friction on, 224 
 
 Bearings, Function of, 223 
 
 Bearings and Journal, Friction between, 224 
 
 Bearings of Robey Compound Engine, 223 
 
 Browett, Lindley, and Co.'s, Governors, 232, 234 
 
 Browett, Lindley, and Co.'s Vertical Compound, 261 
 
 Calorimeter, Musgrave and Son's, 206 
 
 Compound : Davey, Paxman, and Co.'s Girder, 259 ; Davey, 
 Paxman, and Co.'s Horizontal, 252 ; Davey, Paxman, and Co.'s 
 Undertype, 253 ; Robey's, 260 ; Browett, Lindley, and Co.'s 
 Vertical, 261 ; Willans's, 264 D. 
 
 Compound Engine, Definition of, 176 
 
 Condensers, Injection, 227 
 
 Condensers, Surface, 229 
 
 Condensers, Use of, 226 
 
 Condensing Water Required, How to Calculate, 228 
 
 Connecting-rod End, Marine, 221 
 
 Connecting-rod Ends with Cotter and Gib, 219 
 
 Connecting-rod, How to Calculate Strength of, 217 
 
 Connecting-rod, Paxman's, 217 
 
 Connecting-rod, Robey's, 217 
 
 Crank and Connecting Rods, Laws of Motion of, 239 
 
 Crank, Positions of, During One Revolution, 242 
 
 Crank- pins, Determining Diameter of, 224 
 
 Crankshaft, Duty of, 221 
 
 Crankshaft, Paxman's, 221 
 
 Clanked Shafts, 221 
 
 Crosby's Steam Indicator, 189 
 
 Crosby's Visible Drop-Feed Lubricator, 245 
 
 Cross-head, Object of, 215 
 
 Cross-head, Robey's, 216 
 
 Cross-heads, Construction of Various, 215 
 
 Cross-head Pin, Determining Diameter of, 224 
 
 Cylinder, Changes of Temperature in, 176 
 
 Cylinder Clearance, Necessity for, 181 
 
 Cylinder Clearance, How to Determine, 197 
 
 Cylinder Clearance, Table Showing its Effect on Performance of 
 Engines, 185 
 
 Cylinder Ports, Action of, Explained, 241 
 
 Cylinders : Fitted with Variable Expansion Gear, Section 
 Showing Details of, 243; Lubricators for, 246; Meyer's 
 Expansion Gear for, 243 ; Paxman's. Expansion Gear for, 
 243 ; Work of Slide-valve in, 241 ; Details of Construction of, 
 209 ; Jacketed, 209 ; Non-jacketed, 209 ; Use of Steam-trap, 210 
 
 Darke's Steam Indicator, 191 
 
 Davey, Paxman, and Co. 's Compound Horizontal, 252 ; Compound 
 Undertype, 253 ; Compound Girder, 259 ; Windsor, 264 J. 
 
 Dewrance's Lubricators, 248 
 
 for Dynamo Driving, Special Types : Davey, Paxman, and Co.'s, 
 252 ; Robey's, 260 ; Browett, Lindley, and Co.'s, 261 ; Willans's, 
 264 D ; the Windsor, 264 j. 
 
 Eccentric, Its Relation to the Slide-valve, 239 
 
 Eccentrics, Description of, 226 
 
 Expansion Gear, 243 
 
 Flywheel, Its Use, 230 
 
 Flywheels of : Action of Centrifugal Force on, 238; Calculating 
 Rim Velocity and Mass, 237 ; Construction of, 237 ; Deter- 
 mining Thickness of Rim, 238 ; Diagrams, 238 ; Engines for 
 Starting, 249 ; and Governors, Their Relation, 239 ; Irregularity 
 of Speed, How Caused and Regulated, 238 
 
 Foundations for, 54 
 
 Girder Compound, Davey, Paxman, and Co.'s, 259 
 
 Governor, Actuates Throttle-valve, 237 
 
 Governors: Browett, Lindley's, and Co.'s, 232, 234; Centri- 
 fugal, Principle of, 230 ; Formula for Calculating Sensitiveness 
 of, 231 ; Parabolic, 233 ; Paxman's, 234 ; Pickering's, 234 ; 
 Porter's, 231 ; their Stability, 232 ; their Use, 230 ; Watt's, 
 231 ; with Crossed Arms, 234 
 
 Governors and Flywheels, Their Relation, 239 
 
 High-Speed, see Quick-Speed. 
 
 Horizontal Compound, Davey, Paxman, and Co.'s, 252 
 
 Horse-power, Indicated, How to Determine, 202 
 
 Horse power, Actual and Indicated, 204 
 
 Indicated Work, Definition of, 135 
 
 Indicator Diagrams, see Steam Indicator below 
 
 Journals, Neck, 224 
 
 Journals, Overhung, 224 
 
 Journals and Shafts, Determining Diameter of, 225 
 
 Lead, Explanation of, 242 
 
 Leakage and Condensation Losses, How Calculated, 205 
 
 Losses in, Summary of, 207 
 
 Lubricators for : Crosby's Visible Drop-Feed, 245 ; for Cylinders, 
 246 ; Dewrance's Sight-Feed, 248 ; Object and Description of, 
 245 ; Winn's Sight-Feed, 248 
 
 Meyer's Variable Expansion Gear, 243 
 
INDEX. 
 
 Engines, Steam (continued) : 
 
 M'Innes's Steam Indicator, 192 
 
 Musgrave's Driving Gear for Indicators, 193 ; Injection Con- 
 
 denser, 227 ; Surface Condenser, 229 ; Barring Engine 249 
 Pantagraph Driving Gear for Indicators, 192 
 
 9W?"? < j >oss ^ le t ds > 2l5 .' tiovernor, 234 ; Injection Condenser, 
 227 ; Automatic Expansion Gear, 243 
 Pickering's Governor, 234 
 Piston, Construction of, 211 
 Piston, Davey, Paxman, and Co.'s Make, 211 
 Piston, Positions of, During One Revolution 242 
 Piston, Robey and Co.'s Make, 212 
 Piston-rod, How to Calculate Strength of, 213 
 Piston Speed, How to Calculate, 213 
 Piston Speeds and Flywheel Diagrams, 239 
 Porter's Governor, 231 
 
 Ports and Slide-valve, Work of, Explained in Detail, 241 
 Quick-Speed, Browett, Lindley, and Co.'s 261 
 Quick-Speed, Willans's, 264 D. 
 Receiver Compound Engine (Ideal), 178 
 Receiver Compound Engine with Clearance, 183 
 Reciprocating Motion, Laws of, 239 
 Richards's Steam Indicator, 191 
 Robey's Compound, 260 
 
 Shafts and Journals, Determining Diameter of, 225 
 Single-acting Engine with Expansion, 175 
 Single-acting Engine, without Expansion, Description of, 172 
 bingle-Cylinder Engine with Clearance, 181 
 Slide- Valves see Valves below. 
 Starting, Methods of and Engines for, 249 
 Steam, Analysing its Wetness, 206 
 Steam Expansion, Hyperbolic, 179 
 Steam, how it behaves in a Cylinder, 172 
 Steam Indicator Diagram, Theoretical, 238 
 ST ^ M I ? DI0AT011 : Description of, 186 ; Crosby's, 189 ; Darke's 
 191; Determination of Amount of Steam Consumed pel- 
 Indicated Horse-power Hour by Means of, 205 ; Diagrams from 
 Compound Engine, Combination of, 198 ; Diagrams, an Ideal 
 Compared with a Real, 196 ; Driving Gear for, 192 ; Drivi ,g 
 Gear as used on Globe High-Speed Engine, 194 ; Mclnnes's, 
 192 ; Management of, 194 ; Richards's, 191 ; Springs, Maximum 
 Pressures for, 188 ; Tabor's, 187 ; Use of, 185 
 Tabor's Steam Indicator, 187 
 Throttle-valve, The Description of, 237 
 Undertype Compound, Davey, Paxman, and Co.'s, 253 
 Valve, Slide: Use of 239; Description of, 240; Diagram 
 Zeuner's, 240 ; Meyer's, 243 ; Paxman's, 243 ; and Ports Work 
 of Explained, 241 
 Valve, Throttle, Description of, 237 
 Vertical Compound, Browett, Lindley, and Co.'s, 261 
 Watt's Governor, 231 
 Willans's Compound, 264 D. 
 Windsor Type, 264 J. 
 Winn's Sight-Feed Lubricator, 248 
 Woolf's Compound Engine, 177, 182 
 Erg, The, 292 
 
 Esson's Formula for Heating of Dynamo Coils, 15 
 Evaporating Water, Heat Required, 61 
 Excavating Foundations, 51 
 
 Expansion and Compression, Adiabatic and Isothermal, 168 
 Expansion Gear op Steam Engines : Meyer's Variable, 243 ; Section 
 
 Showing Details of, 243 ; Paxman's Automatic, 243 
 Explosions, Boiler, Their Cause and Remedy, 152 
 
 F- 
 
 Farad, Definition of, 31, 305 
 
 Fault in Electric Light Mains, How to Localise, 351 
 
 Feed-water of Boilers, Supply and Heating of, see under Boilees 
 
 Ferranti's Feed-valve for Boilers, 140 
 
 Ferranti's Method of Laying Mains, 380 
 
 Field, Magnetic, see Magnetic Field 
 
 Field's Boiler, 98 
 
 Firebox Construction in Boilers, 63 
 
 Flubs in Boilers : Best Position for, 164 ; Brushes for Cleaning, 
 
 151 ; Construction of Rings, 111 ; Different Arrrangements of, 64 ; 
 
 Expansion, 111 ; Protectors for, 121 
 
 Flywheels op Steam Engines : 
 
 Action of Centrifugal Force on, 238 
 
 Calculating Rim Velocity and Mass of, 237 
 
 Construction of, 237 
 
 Determining thickness of Rim, 238 
 
 Diagrams, 238 
 
 Engines for Starting, 249 
 
 and Governors, Their Relation, 239 
 
 Irregularity of Speed, How Caused and Regulated, 238 
 
 their Use, 230 
 
 Foot-pound-second System, 292 
 
 Foot-pound, Explained, 13, 154 
 
 Footing Courses of Foundations, 52 
 
 Forbes, George, Experiments on the Heating of Conductors, 14 
 
 Force: Centrifugal, 160; Definition of, 153; a Deviating 160 
 Measurement of, 153 ; Unit of, 153 aviating, iou , 
 
 Force, Lines and Loops op : 
 
 Are caused by Current in a Conductor, 20 
 
 Assumptions with Regard to, 20 
 
 Effect of Increase or Decrease of in a given area, 304 
 
 Faraday's Conception of, 20 
 
 How to Calculate Number of, 25 
 
 Illustration of in Magnetic Field, 23 
 
 Induce a Current in a Conductor, 27 
 
 Number Round a Coil, on what Dependent, 24 
 
 force:, MomfnTof 36° rt **** ^ 6 ° US * Mfferent Radii ' 3D2 
 
 Forces, Resultant, Effect of, 35 
 
 Fox's Corrugated Furnace, 78, 84, 87, 111 
 
 Foundations op Central Stations : 
 
 Bricks as Building Materials, 57 
 
 Characteristics of Various Soils and Rocks, 49 
 
 Concrete, 59 
 
 Drainage and Protection of, 48, 51 
 
 Engine and Machine Foundations, 54 
 
 Estimating for Excavation, 51 
 
 Excavated Material, Increase in Bulk of, 51 
 
 Excavating for, 51 
 
 Footing Courses, Designing, 52 
 
 Footing Courses, Masonry, 58 
 
 How Displacement Occurs, 47 
 
 Masonry Footings, Safe Offsets for, 53 
 
 Mortar, 59 
 
 Natural and Artificial, 47 
 
 Settlement of Structures, 47 
 
 Soils, Safe Bearing Power of Various, 59 
 
 Stability of Masonry, 54 
 
 Stones as Building Materials, 55 
 
 Testing Strata, 48 
 
 Underfooting, 51 
 
 Walls, Construction of, 57 
 Friction on Bearing Surfaces, Calculating, 224 
 Friction between Bearings and Journal, 224 
 Friction, Work Consumed by, 160 
 Fuels in Boilers, Combustion of Various, 62 
 Fuels Used for Boilers in Different Countries, 63 
 Fuses, 289 
 
 Fusible Cut-outs, Use of, 289 
 Fusible Plugs for Boilers, 121 
 Fusibility of Metals, Preece's Tables, 17 
 
 G. 
 
 Galloway Boilers, Description of, 69, 73, 82 
 Galvanometers : 
 
 Figure of Merit, How to Calculate, 311 
 The Holden-d'Arsonval, 329 
 as Used for Measuring Battery Resistance, 329 
 as Used for Measuring Capacity, 336 
 as Used for Measuring Pressure, 332 
 for Measuring Resistances by Fall of Pressure, 339 
 Mirror, How to Make, 308, 310 
 Measurement of Resistance of, 313 
 Portable, 344 
 
 Shunts, How to Make, 309 
 In the Silvertown Testing Set, 344 
 Use of in the Wheatstone Bridge, 306 
 
 Tangent : Its Principle Adopted in most Meters, 316 ; Descrip- 
 tion of, 301 ; Determining Reduction Factor of, 311 ; How to 
 Calculate Effect of H on, 302 ; How to Make, 303 ; How to Use 
 303 ; Principle of, 301 
 Garrett and Son's Boiler Pump, 131 ; Boilers, Description of, 77, 95 ; 
 
 Feed-water Heater, 141 
 Gases, Absolute Temperature of, 164 
 Gases, Laws of Expansion of, 164 
 
 Gases, Their Expansion and Compression in Cylinders, 168 
 Gases, Imaginary : and Actual, 165 ; Molecular Heat, 166 ; Specific 
 
 Heat of, 166, 1 67 ; Thermo-dynamics of, 166 
 Gauges for Boilers, see Pressure Gauges 
 Gay-Lussac's Law of Expansion of Gases, 164 
 General Electric Company's Switchboard, 278 
 Girder Engine, Davey, Paxman, and Co.'s, 259 
 
 Glazebrook and Shaw's Method of Measuring Electrical Pressure, 335 
 Glazebrook and Shaw's Method of Measuring Capacity, 337 
 
 Governors of Steam Engines : 
 The Acme, 234 
 Actuate Throttle- Valves, 237 
 Browett, Lindley, and Co.'s, 232, 234 
 Centrifugal, Principle of, 230 
 The Flywheel, 130 
 and Flywheels, Their Relation, 239 
 Formula for Calculating Sensitiveness of, 231 
 The Parabolic, 233 
 Paxman's, 234 
 
VI. 
 
 INDEX. 
 
 Governors of Steam Engines (continued) : 
 Pickering's, 234 
 Porter's, 231 
 Their Stability, 232 
 Their Use, 23C 
 Watt's, 231 
 With Crossed Arms, 234 
 
 Graphic Methods, see Electro-graphics and Dynamics 
 Gravity, Acceleration of, 154 
 Gravity, Specific, Definition of, 153 
 Gresham and Craven's Steam Injector, 134 
 
 H. 
 
 H, How to Determine, 294 
 
 H, Bottomley's Apparatus for Determining, 294 
 
 H, Effect of on a Tangent Galvanometer, 302 
 
 H, Ratio of Moment of Magnet to, 296 
 
 H and M, Determination of, 295 
 
 Heat : 
 
 Adiabatic and Isothermal Expansion and Compression, 168 
 
 In Boilers, Sensible and Latent, 61 
 
 Boilers, Conduction Formula for, 163 
 
 Boiler Flues, Best Position for, 164 
 
 Boyle's or Mariotte's Law of Expansion of Gases, 163 
 
 Conduction, Formulas for, 163 
 
 Convection, Definition of, 164 
 
 Curve, An Adiabatic, 169 
 
 Cylinders, Conduction Formula for, 163 
 
 Cylinders, Expansion and Compression of Gases in, 1 68 
 
 Effect of on Permanent Magnets, 25 
 
 A Form of Energy, 161 
 
 Gases, Absolute Temperature of, 164 
 
 Gases, Expansion and Compression of, 168 
 
 Gases, Laws of Expansion of, 164 
 
 Gases, Imaginary and Actual, 165 
 
 Gases, Imaginary, Molecular Heat, 166 
 
 Gases, Imaginary, Specific Heat of, 166. 167 
 
 Gases, Imaginary) Thermodynamics of, 166 
 
 Gay-Lussac's Law of Expansion of Gases, 164 
 
 How Transferred, 162 
 
 Heat Engine, Carnot's, 170 
 
 Heat Engine, Definition of, 170 
 
 Heat Engine, Efficiency of, 172 
 
 Joule, and the Mechanical Equivalent of, 14, 165, 305, 339 
 
 Latent, Explanation of, 166 
 
 Mechanical Equivalent of, 14, 165, 305, 339 
 
 Non-conducting Materials, 163 
 
 Radiation, Definition of, 163 
 
 Sensible Heat, 166 
 
 Specific Heat, 162 
 
 Steam-pipes, Conduction Formula for, 163 
 
 Temperature, Definition of, 161 
 
 Temperature, Measurement of, 161 
 
 Thermo-dynamics, Theory of, 165 
 
 Units, 162, 305 
 
 Waste of in Boilers, 64 
 
 Heating of Conductors, Forbes's Experiments, 14 
 
 Heating of Conductors, Joule's Law for, 339 
 
 Heating Effect of Current, 13 
 
 Heating-surface of Boilers, 63, 66 
 
 Heating of Switches, How Guarded Against, 267 
 
 Hering's Formula? for Heating of Magnet Coils, 16 
 
 High-Speed Engines, see Engines 
 
 Holden-d 'Arson val Galvanometer, 329 
 
 Hopkinson, Dr. His Work in Electro-graphics, 33 
 
 Hopkinson's Feed-valve for Boilers, 140 ; Steam-pipe Joint, 124 • 
 Steam Check-valve, 127 : Steam Stop-valve, 125. 
 
 Horizontal Boilers, see Boilers 
 
 Horizontal Engines, see Engines 
 
 Horse-power, Electrical, 14 
 
 H onn E " ] r? W f .' Mechanioal : its Equivalent in the Metric System 
 292; Explained, 14, 154, 292; Indicated and Actual, 204 ; Indi- 
 cated, How to Determine, 202 
 
 Horseshoe Magnet (Permanent), Field of, 24 
 
 Hydrogen, Amount liberated by one ampere in one second, 17 
 
 Hydrogen and Oxygen Produced in Voltameter, 19 
 
 Impurities in Boilers, Getting Rid of, 121, 149, 150 
 Inclination or Dip, Measuring Angle of, 300 
 Incrustation in Boilers, CauBe and Prevention of, 121, 148 150 
 Indicated Horse-power and Actual, 204 ' 
 
 Indicated Horse-power, How to Determine, 202 
 Indicator, Steam : 
 
 Crosby's, 189 
 
 Darke's, 191 
 
 D H^ n „ ati0n £ A u 0U IV t 0f Steam Consumed per Indicated 
 Horse-power Hour by Means of, 205 
 
 Indicator, Steam (continued) -. 
 Description of, 186 
 Diagram, Theoretical, 238 
 
 Diagrams from Compound Engine, Combination of, 198 
 Diagrams, Ideal and Real, 196 
 Driving Gear for, 192 
 
 Driving Gear as used on " Globe " High-Speed Engines, 194 
 Mclnnes's, 192 
 Management of, 194 
 Maximum Pressures for Springs, 188 
 Musgrave and Sons' Driving Gear for, 193 
 Pantagraph Driving Gear for, 192 
 Richards's, 191 
 Tabor's, 187 
 Use of, 185 
 
 Inductive Action in the Magnetic Circuit, 26 
 
 Inductive Capacities, Specific, Table of, 32 
 
 Inductive Circuit, see Circuits 
 
 Inductive Resistance, Specific, 31 
 
 Inertia of Magnetic Needle, Determination of Moment of, 296 
 
 Inertia, Moment of, 155, 295 
 
 Injection Condensers, Steam, 227 
 
 Injectors, Steam : Davies and Metcalfe's Exhaust Steam, 137 ; 
 
 Principle of, 134 ; the Exhaust Steam, 137 ; Gresham and Craven's' 
 
 134 ; the Lifting, 137 ; the Non-lifting, 137 
 Installations, Plans of: Mentmore, 44, 45, 46; Edison's Brooklyn 
 
 Station, Plates N and O. 
 Insulation Resistance, Silvertown Testing Set for, 347, 349 
 Insulators, List of, 6 
 Insulators, Magnetic, 25 
 Insulators, Their Use and Properties, 6 
 Insulators for Electric Light Mains: Brooks's, 373; Callender- 
 
 Webber, 363, 367 ; Calender's, 371 ; Crompton's, 358 ; Ferranti's, 
 
 380 
 Iron Filings, How Influenced by Wire Carrying Current, 12, 20 
 Isothermal Expansion and Compression, 168 
 
 Joints for Electric Light Mains : Calender's, 370 ; Crompton's. 353 
 
 361 ; Ferranti's, 380 
 Joule, The, £92, 305 
 
 Joule and the Mechanical Equivalent of Heat, 14, 163, 305, 339 
 Joule's Law for Heating of Conductors, 339 
 Journals of Engines, 224 
 Junction-eoxes FOR Electric Light Mains : Brooks's, 374 ; 
 
 Calender's, 370; Callender-Webber, 367; Crompton's, 355: 
 
 Ferranti's, 384 
 
 K. 
 
 K, Table of Values of, 295 
 Kilogrammetres and Foot-pounds, 14 
 Kinetic Energy, Definition of, 155 
 Kirchoffs Current Laws, 11 
 Kirchoffs Slide-wire Bridge, 307 
 Kirkaldy's Feed-water Heater, 142 
 
 Lancashire Boilers, see Boilers 
 
 Lead in Engines, Explanation of, 242 
 
 Leakage, Magnetic, 25 
 
 Leeds Forge Company's Boilers, 84, 87 
 
 Length, Unit of, 153 
 
 Leyden Jars, Description of, 32 
 
 Lines and Loops of Force, see Force 
 
 Locomotive Boilers, see Boilers 
 
 London Electric Supply Corporation's Mains, Method of Laying, 380 
 
 Low- Water Alarms for Boilers, 120 
 
 Lubricators for Engines : 
 
 Crosby's Visible Drop-Feed, 245 
 Dewrance's Window Form, 248 
 Use of, 245 
 Winn's Sight-Feed, 248 
 
 M 
 
 M divided by H, Determination of Ratio, 296 
 M H, Determination of, 295 
 
 Machine, Calculating Efficiency of, 160 
 M'Neil's Boiler Manhole, 112 
 
 Magnet : Magnetic Moment of, 294 ; Measuring Magnetic Moments 
 of, 299 ; Ratio of Moment of to H, 296 ; a Perfect, 23 ; Short, 
 Relation of Forces Acting on with Coils of Different Radii, 303 
 
 Magnets: Comparing Magnetic Moments of, 299; Effect of Earth's 
 Magnetism on, 293; for Meters, 342; Properties of, 23; .i,e 
 Electromagnets and Permanent Magnets 
 
 Magnetic Circuit, see Circuits 
 
 Magnetic Conductivity and Resistance, 25 
 
 Magnetic Declination, or Dip, 293 
 
INDEX. 
 
 Vll. 
 
 Magnetic Dip, Measuring Angle of, 300 
 
 Magnetic Field : 
 
 Dae to Leakage, 25 
 The, Illustrated, 23 
 of Bar and Horseshoe Magnets, 24 
 Measuring the Strength of, 298 
 Magnetic Fields : And Conductors, Interaction between, 27 ; Inter- 
 actions between, 25 ; of Conductors, 23 
 Magnetic Insulators, 25 
 
 Magnetic Lines and Loops of Force, see Fobce 
 Magnetic Meridian, Determination of, 293 
 Magnetic Meridian and the Tangent Galvanometer, 302 
 Magnetic Pressure, 25 
 
 Magnetism, may be Electricity in Motion, 23 
 
 Magnetism of Earth : Determination of Horizontal Component of, 
 294 ; Effect of H on Tangent Galvanometer, 302 ; Its Effect on 
 Magnets, 293 ; Determination of M H, 295 
 Magnetism, Residual, 227 
 Magnetometer, How Used, 296 
 Mance's Method of Measuring Battery Resistance, 330 
 
 Mains, Electric Light : 
 
 Contact Fault, How to Localise, 352 
 Earth, How to Localise, 351 
 Joint-Testing, 351 
 Methods of Laying : 
 
 Crompton's System, 353 
 Callender-Webber System, 361 
 Callender Solid System, 367 
 Brooks's System, 373 
 Ferranti's, 380 
 Silvertown, Testing Set for, 344 
 Testing, 341, 34* 
 
 Testing Conductor Resistance, 345 
 Testing. How to Carry it Out, 349 
 Testing Insulation Resistance, 347, 349 
 Manholes and Mudholes of Boilers, Their Construction, 112 
 Marine Boilers, see Boilers 
 Masonry, Stability of, 54 
 Mass, Definition of, 153, 291 
 Mass, Units of, 153, 292 
 Mass and Weight, Difference between, 291 
 Matthiessen's Table of Conductivities, 8 
 Measurement of Electrical Quantities : 
 Ammeters, Ayrton and Perry's, 341 
 Ammeters, Paterson and Cooper's, 342 
 Ammeters, Permanent v. Electromagnets for, 34<i 
 Ampere, The Absolute, 305 
 Ampere Balances, Thomson's, 316 
 Ampere Gauge, Thomson's, 323 
 Ampere-meter, Thomson's Magnetostatic, 5^1 
 Ayrton and Perry's Ammeters and Voltmeters, 341 
 Balance, Composite, Thomson's, 319 
 Balances, Electric, Thomson's, 316 
 
 Balances, Electrostatic, Thomson's, 323, 3^6 . W h M t«tonp 
 
 Battery and Galvanometer, Their Position in the Wheatstone 
 
 Bridge. 3/7 
 Battery Resistance, Measurement ot, 3<sa 
 
 Calibration of Ammeters, 341 
 
 Calibration of Voltmeters, 343 
 
 Capacity, Definition of Unit of, 305 
 
 Capacity, Measurement of, 335 
 
 Cardew's Voltmeter, 343 
 
 Clarke's Method of Measuring Pressure, 334 
 
 Condenser, How to Make, 335 
 
 Condenser for Measuring Capacity, 335 
 
 Coulomb, Definition of, 305 
 
 Current, Electromagnetic Unit ot, «« 
 
 Current Meters, see Ammeters, above m 
 
 Current, Pressure and Resistance, their Relation 304 
 
 Daniell's Cell as a Standard of Electromotive 1 orce, 331 
 
 Detectors for Linesmen, 344 
 
 Dynamical Units, 292 
 
 Dyne, The, 292 
 
 Earth, The, a Magnet, 293 
 
 Earth's Magnetic Poles, Movements of, «« 
 
 Earth's Magnetism, ^s Effect on Magnets 293 ient 
 
 Earth's Magnetism, Determination of Horizontal uuwp 
 
 Electro-chemical Equivalents, Table ot, 31<i 
 
 Electromotive Force, see Pressure below 
 
 Erg, The, 292 
 
 Farad, Definition of, 305 
 
 Foot-pound-second System, 292 
 
 Galvanometer : „ , , , ... 
 
 Figure of Merit, How to Calculate, 311 
 The Holden-d'Arsonval, 329 , g 
 
 for Measuring Resistances by Fall of IWuie, 339 
 
 Measurement of Electrical Quantities (continued) : 
 Galvanometer : 
 
 for Measuring Capacity, 336 
 
 for Measuring Battery Resistance, 329 
 
 for Measuring Pressure, 332 
 
 a Portable, 344 
 
 Its Use in the Wheatstone Bridge, 306 
 
 Mirror, How to Make, 3C8, 310 ; Resistance, Measurement 
 
 of, 313 
 Shunts, How to Make, 309 ; of the Silvertown Portable 
 
 Testing Set, 344 
 Tangent : Description of, 301 ; How to Calculate Effect of 
 H on, 302 ; How to Make, 303 ; How to Use, 303 ; Deter- 
 mining Reduction Factor of, 311 ; Principle of, 301 ; 
 Principle of Adopted in most Meters, 316 
 Glazebrook and Shaw's Method of Measuring Capacity, 337 
 Glazebrook and Shaw's Method of Measuring Pressure, 335 
 Gravity, Its Effect on Weight, 291 
 H, How to Determine, 294 
 
 H, Bottomley's Apparatus for Determining, 294 
 Heat, Mechanical Equivalent of, 339 
 Heating of Conductors, Experimental Apparatus for Verifying 
 
 Joule's Law, 340 
 Heating of Conductors, Joule's Law for, 339 
 Horse-power, Its Equivalent in the Metric System, 292 
 Horse-power, Unit of, 292 
 Inclination, or Dip, Measuring Angle of, 300 
 Inertia, Moment of, Tables of Value of, 295 
 Insulation Resistance, Measurement of, 347 
 Insulation Resistance, Silvertown Portable Testing Set for 
 
 Measuring, 347 
 Joule, The, 292, 305 
 
 Joule's Law for Heating of Conductors, 339 
 Joule's Law, Experimental Apparatus for Verifying, 340 
 K, Table of Values of, 295 
 Kirchoff's Slide-wire Bridge, 307 
 Loop Test for Localising Earth, 351 
 Loops of Force, Effect of Increase or Decrease in a Given Area, 
 
 304 
 M H, Determination of, 295 
 M/H, Determination of Ratio of, 296 
 Magnet, Magnetic Moment of, 294 
 Magnet, Measuring Magnetic Moment of, 299 
 Magnet, Ratio of Moment of to H, 296 
 Magnet, Short, Relation of Forces Acting on with Coils ol 
 
 Different Radii, 303 
 Magnets, Comparing Magnetic Moments of, 299 
 Magnetic Declination, or Dip, 293 
 Magnetic Dip, Measuring Angle of, 300 
 Magnetic Field, Measuring the Strength of, 298 
 Magnetic Meridian, Determination of, 293 
 Magnetic Meridian, The, and the Tangent Galvanometer, 30J 
 Magnetic Needle, Determination of its Moment of Inertia, 296 
 Magnetometer, How Used, 296 
 Magnetostatic Current Meter, Thomson's, 321 
 Mass and Weight, Difference between, 291 
 Mass, Standards of, 292 
 Moment of Inertia, Table of Values of, 295 
 Multicellular Voltmeter, Thomson's, 323 
 Ohm, The Absolute, 304 
 
 Ohm's Law, Experimental Verifications of, 313 
 Paterson and Cooper's Ammeter, 342 
 Poggendorf's Method of Measuring Pressure, 334 
 Potential, see Pressure below 
 Poundal, The, 292 
 Power, Definition of Unit of, 305 
 Poynting's Experiments on a Short Magnet and Coils of Different 
 
 Radii, 303 
 Pressure, Absolute Unit of, 304 
 Pressure, Measuring Resistances by h all ot, 33B 
 Pressure, Methods of Measuring, 331 
 Quantity, Definition of Unit of, 305 
 Resistance, Absolute Unit of, 304 
 Resistance Coils for Wheatstone Bridge, 308 
 Resistance of Conductors, Silvertown Testmg Set for Measuring, 
 
 345 
 Besistances, Measurement of by Fall of Pressure, 338 
 of Resistances by Wheatstone Bridge Method. 305 
 Shunts, Theory of, 309 
 Silvertown Portable Testmg Set, 344 
 
 Swinburne On the Want of Practical and Cheap Apparatus, 328 
 Tests of Electric Light Cables, How to Carry Out, 349 
 Testing at Central Stations, Instruments Used for, 341 
 Testing Sets, Portable, 344 
 
 Thomson Sir W., His Measuring Instruments, 316 
 Thomson's Method of Measuring Galvanometer Resistance, 315 
 Thomson's Method of Measuring Capacity, 337 
 Units of C.G.S. System, 291 
 Units Fundamental and Derived, 292 
 
 ^™" S ty^nd Perry's, 342 ; Cardew's, 343 ; Pocket, 
 343 7 Thomson's Electrostatic, 323 ; Thomson's Engine-room, 
 327 ;' Thomson's Marine, 319 
 
VI II. 
 
 Index. 
 
 Measurement of Electrical Quantities (continued) : 
 Volt and Watt Balance, Thomson's, 319 
 Watt, Definition of, 305 
 Wheatstone's Standard Cell, 331 
 Wheatstone Bridge Method, 305 
 Wiieatstone Bridge : 
 
 Method of Measuring Galvanometer Resistance, 313 
 
 Cheap Form Wanted for Engineering Work, 328 
 
 for Measuring Battery Resistance, 329 
 
 for Measuring Capacity, 336 
 
 for Measuring Pressure, 332 
 
 as Used in the Silvertown Portable Testing Set, 346 
 
 Post Office Form of, 305 
 
 Slide-wire Form of, 307 
 Wiedemann's Method of Measuring Pressure, 333 
 Work or Heat, Definition of Unit of, 305 
 Work of B.A Committee in, 291 
 
 Mechanical Energy, Kinds of, 155 
 Mechanical Equivalent of Heat, 14, 165, 305, 339 
 Mentmore Installation, Plans of, 44, 45, 46 
 Meridian, Magnetic, Determination of, 293 
 
 Metals : Are Conductors, 6 ; Conductivities of Various, 8 ; Fusi- 
 bility of, Preece's Tables, 17 ; Resistance of Various, 11 ; Effect of 
 Temperature on Resistance of, 8 
 Meters, sea also Measurement 
 Current : 
 
 Ayrton and Perry's, 341 
 
 Calibration of, 341 
 
 Linesman's Detector, 344 
 
 Paterson and Cooper's, 342 
 
 Permanent and Electromagnets for, 342 
 
 Thomson's Ampere Gauge, 323 
 
 Thomson's Balances, 316 
 
 Thomson's M agnetostatic, 321 
 Pressure : 
 
 Ayrton and Perry's, 341 
 
 Calibration of, 343 
 
 Cardew's, 343 
 
 Pocket, 343 
 
 Thomson's Electrostatic, 323 
 
 Thomson's Engine-room, 327 
 
 Thomson's Multicellular, 319 
 
 Meyer's Slide-valve, 243 
 Molecular Heat, 166 
 Moment of Forces, 36 
 Moment of Inertia, 155, 295 
 Mirror Galvanometer, How to Make, 308, 310 
 Moment of Inertia of Magnetic Needle, Determination of, 296 
 Moments of Magnet, Measurement of, 29J 
 Moments of Magnets, Comparing, 299 
 Mortar for Use in Foundations, 59 
 Motion, Definition of, 154 
 
 Multicellular Voltmeter, Sir William Thomson's, 323 
 Multiple Arc or Parallel Circuits, 10 
 Multitubular Boilers, see Boilers 
 Mumford's Donkey Pump, 131 
 Musgrave's Barring Engine, 249 
 
 Musgrave's Boiler, Description of, 70, 95 ; Calorimeter, 206 ; Feed- 
 water Heater, 144 ; Injection Condenser, 227 ; Surface Condenser, 229 
 
 N. 
 
 New York Station, Edison's, see Plates N and O 
 Non-conducting Materials for Boilers, 163 
 Non-conductors, see Insulators 
 
 o. 
 
 Ohm : The Unit of Resistance, 7, 9 ; the B.A. and the Legal Ohm 7 • 
 
 Determination of the Absolute, 304 
 Ohm's Law, Explanation of, 9 
 Ohm's Law, Experimental Verification of, 313 
 Ohm's, Amperes and Volts, Their Relation, 304 
 Oiling Engines, see Lubricators 
 Opera House, Royal English, Switchboard used at, 286 
 
 P. 
 
 Parabolic Engine Governor, The, 233 
 Paterson and Cooper's Ammeters, 342 
 
 ''mTilSZc^L 2 ^ Crank8haft ' 22l! En ^ ine G ~- 
 
 Permanent Magnets, Description of, 24 
 Permanent Magnets, Effect of Heat on, 25 
 Permanent and Electromagnets for Meters, 342 
 Pickering's Engine Governor, 234 
 Pin, Cross-head, Diameter of, 224 
 Pins, Crank, Diameter of, 224 
 
 Pistons of Steam Engines : 
 Construction of, 211 
 
 Cross-heads for, Object and Construction of, 215 
 Paxman and Co.'s, 211 
 Paxman's Cross-heads, 215 
 Position of, During' One Revolution, 242 
 Robey and Co.'s, 212 
 Robey's Cross-heads, 216 
 Rods for. How to Calculate Strength of, 213 
 Speed of, and Fywheel Diagrams, 239 
 Speed of, How do Calculate, 213 ' 
 
 Plane Figures, Determination of Areas of, 157 
 
 Planimeter, Use of, 153 ; Amsler's Polar, 158 ; the Coffin, 159 
 
 PoggendorfFs Method of Measuring Pressure, 334 
 
 Poles, Magnetic Description of, 23 
 
 Poles, Magnetic, Interaction of, 26 
 
 Poles, Magnetic, of the Earth, Movements of, 302 
 
 Poole and White's Switchboard, 289 
 
 Portable Galvanometers, 344 
 
 Portable Testing Sets, 344 
 
 Porter's Engine Governor, 231 
 
 Ports, Steam, Use and Work of, 241 
 
 Positive and Negative Pressure, 29 
 
 Post Office Form of Wheatstone Bridge, 305 
 
 Post Office Installation : Switchboard used in, 268, 291 ; Switches 
 
 used in, 217 
 Potential, see Pressure. 
 Potential Energy, Definition of, 155 
 Poundal, The, 292 
 Power, Definition of, 154 
 Power, Unit of, 154, 305 
 Poynting's Experiments on a Short Magnet and Coils of Different 
 
 Radii, 303 
 Preece's Fusibility Tables, 17 
 Pressure in Boilers, Absolute and Effective, 61 
 Pressure Gauges for Boilers, 114 
 Pressure and Volume, 156 
 Pressure, Electrical : 
 
 Clarke's Method of Measuring, 334 
 
 in the Conductive Circuit, 6 
 
 . Current and Resistance, Their Relation, 304 
 
 Daniell's Cell as a Standard of, 331 
 
 Electricity moved by, 5 
 
 Galvanometers as used for Measuring, 332 
 
 Glazebrook and Shaw's Method of Measuring, 335 
 
 Graphic Method of Representing, 40 
 
 How it Acts, 12 
 
 in the Inductive Circuit, 28 
 
 in the Magnetic Circuit, 25 
 
 Measurement of Resistances by Fall of, 338 
 
 Methods of Measuring, 331 
 
 Meters, see under Voltmeters, Meters, and Measurement 
 
 Ohm's, Law for, 9 
 
 PoggendorfFs Method of Measuring, 334 
 
 Positive and Negative, 29 
 
 Production of, Water Analogy. 5 
 
 Unit of, 9 
 
 Absolute Unit of 304 
 
 Wheatstone's Method of Measuring, 373 
 
 Wheatstone's Standard Electromotive Force Cell, 331 
 
 Wheatstone's Bridge for Measuring, 332 
 
 Wiedemann's Method of Measuring, 334 
 Priming in Boilers, The Anti-Priming Pipe, 129 
 Pumps, Boiler, see also Injectors 
 
 Use of, 130 
 
 Garrett and Sons, Feed, 131 
 
 Mumford's Donkey, 131 
 
 Q- 
 
 Quantity, Unit of, 31, 305 
 
 Quantity, Graphical Representation of a Single, 35 
 
 Quick-Speed Engines, see Engines 
 
 R. 
 
 Radiation, Definition of, 163 
 
 Ransome, Sims, and Jefferies' Boiler, 1C1 
 
 Rayleigh's Experiments on the Deposition of Silver, 17 
 
 Reciprocating Motion in Steam Engines, Laws of, 239 
 
 Resistance Coils for Wheatstone Bridge, 308 
 
 Resistance and Effort, Mechanical, 153, 156 
 
 Resistance, Electrical : 
 
 of Alloys, 9 
 
 of Batteries. Measurement of, 329 
 
 of Branch Circuits, 11 
 
 of Conductive Circuit, What it is, 6 
 
 and Conductivity, Their Relations, 7 
 
 of Conductor, Increased by Interference with Field, 22 
 
 of Conductors, Influenced bv l.i>n<-rt.h qml sUw>u,>„„i 'a..,,., <7 
 
INDEX. 
 
 IX. 
 
 Resistance, Electrical (continued) : 
 
 of Conductors, Silverbown Testing Set for, 345 
 
 Current and Pressure, Relation of, 304 
 
 Effect of Annealing on, 9 
 
 Effect of Temperature on, 8 
 
 of Galvanometers, Measurement of, 313 
 
 How Affected by Shunt Circuits, 10 
 
 of Inductive Circuit, How Calculated, 29 
 
 of Insulation, Silvertown Testing Set for, 347, 349 
 
 Internal and External, How Calculated, 10 ■ 
 
 Magnetic, 25 
 
 of Magnetic Circuit, How Lessened, 26 
 
 Its Measurement by Fall of Pressure, 338 
 
 Ohm's Law for, 9 
 
 Specific, 7 
 
 Specific Inductive, 31 
 
 Unit of, The Ohm, 7, 302 
 
 of Various Metals, 11 
 
 Wheatstone Bridge Method of Measuring, 305 
 
 Residual Magnetism, 27 
 
 Rhode's Steam Stop-valye, 125 
 
 Riveting and Jointing Boiler Shells, 102 
 
 Roads and Streets, Methods of Laying Electric Light Mains Under, 
 
 BnW's Boilers, 77 ; Compound Engine Bearings, 223 ; Connecting-, 
 rod! 217?F«l-wJtar Heaters, 143; Steam Stop-valve, 125; Com- 
 pound Engine, 260 
 
 Rocks, Characteristics of Various, 49 
 
 Rosebery, Lord, Plans of Installation at his Mansion, Mentmore, 44, 
 
 Ro^alEnglish Opera House InstaUation, Switchboard used in, 286 
 
 s. 
 
 Safety-valves for Boilers: 
 
 Board of Trade Rules for, 119 
 
 Construction of, 115 
 
 Dead-Weight, 115 
 
 Double-Lever, 117 
 
 Galloway's, 115 
 
 Hopkinson's Compound, 11/ 
 
 How to Calculate Size of Opening, 119 
 
 How to Find Working Pressure, 117 
 
 Ramsbottom's, 117 
 
 Relief -valves, 117 . 
 
 Rule for Calculating Size of Springy, 120 
 
 Rules for Construction of Valve, 119 
 
 Spring, 116 
 Scale in Boilers, Cause and Prevention of, 121, 149, 150 
 Schaffer's Steam Pressure Gauge, 115 
 Scum in Boilers, Getting Rid of, Hi 
 Secondary Batteries, see Batteries, Secondary 
 Sna/Area of Conductors, Its Effect on Resistance, 7 
 Shafts and Journals of Engines, Diameter of, lib 
 Shells of Boilers, Calculating Strength of, 108 
 Short Circuits, 31 
 Shunt Circuits, 10 
 Shunt Winding Explained, 10 
 Shunts for Galvanometers, How to Make, 3Ub 
 Sight-Feed Lubricators, see Lubricators „ p „ nntl 17 
 
 Silver, Amount Deposited by One Ampere in One Second, 17 
 Silvertown Portable Testing Set, 544 
 Sines and Cosines, Table of, 35 
 Sines, Curve of, 34 
 
 Site of Central Station, Requisites, 43 
 Slide-valves, see Valves, Slide 
 Soils, Characteristics of Various, 4y 
 Soils, Safe Bearing Power of, 59 
 Solenoid, Illustration of a, 23 
 Specific Gravity, Definition of, loo 
 Specific Heat, 162 ,_ 
 
 Specific Heat of Imaginary Gases, 166 167 
 Specific Inductive Capacities, Table 01, o<S 
 Specific Inductive Resistance, 31 
 Specific Resistance, What it us, 7 
 Specific Volume, Definition of, lbo 
 Springs for Switches, Use of, 266 
 Starting Engines, 249 
 Stays for Boilers, 109 
 
 Steam : 
 
 Analysing its Wetness, 2T6 Leakaae and Condensation, 205 
 
 Calculation of Losses in Engme by Leakage an Horse -power 
 
 Determination of Amount Consumed per Indicatea 
 
 Hour, 205 
 Dry and Superheated, 61 
 How it Behaves in a Cylinder, \l£ 
 Hyperbolic Expansion of, 179 
 Losses in Engines, 207 
 
 Steam Boilers, see Boilers 
 
 Steam Check- valves for Boilers: 
 Use of, 127 
 Hopkinson's, 127 
 
 Steam Engine, see Engines 
 
 Steam Indicator : 
 Crosby's, 189 
 Darke's, 191 
 Description of, 186 
 Determination of Amount of Steam Consumed per Indicated 
 
 Horse-power Hour by means of, 205 
 Diagrams, An Ideal and Real, 196 
 
 Digrams from Compound Engines, Combination of, 198 
 Driving Gear for, 192 
 
 Driving Gear as used on " Globe " High-Speed Engines, 194 
 Mclnnes's, 192 
 Management of, 194 
 Maximum Pressures for Springs, 188 
 Musgrave and Sons' Driving Gear, 193 
 Pantagraph Driving Gear for, 192 
 Richards's, 191 
 Tabor's, 187 
 Use of, 185 
 
 Steam Injector : 
 
 Davies and Metcalfe's Exhaust Steam, 137 
 
 The Exhaust Steam, 137 
 
 Gresham and Craven's, 134 
 
 The Lifting, 137 
 
 The Non-lifting, 137 
 
 Principle of, 134 
 Steam-pipe Joints and Fittings, 124 
 Steam Pressure Gauges for Boilers, 114 
 
 Steam-ports, Use and Work of, 241 , .,. 
 
 Steam-reducing Valves for Boilers : Use of, 130 ; Dewrance s, 1AU 
 Steam Separators for Boilers : Use of, 128 ; Hydes and Y\ igfull s, 
 
 129 ; Musgrave and Son's, 129 ; Savill's, 130 
 Steam-trap in Cylinders, Use of, 210 
 
 Steam -valves for Boilers : 
 Dewrance's Renewable, 126 
 Hopkinson's, 125 
 Rhodes's, 125 
 Robey's, 125 
 Use and Construction of, 125 
 
 Stones as Building Materials, 55 
 
 Strata for Foundations, Testing, 48 M . T ^« 
 
 Streets, Methods of Laying Electric Light Mains Under, see Mains 
 Sturgeon, Maker of First Electromagnet, 24 
 
 Swinburne on the want of Practical and Cheap Measuring Apparatus, 
 328 
 
 Switches : 
 
 Arcing to be Avoided, 265 
 
 Brush Company's High and Low Pressure, 289 
 
 Contacts, Methods of Ensuring Good, 268 
 
 Bases for, 268 
 
 with Double and Multiple Contacts, <i68 
 
 with Ring Contacts, 268 
 
 with Wedge Contacts, 268 
 
 Designing, Points to be Observed, 267 
 
 Drake and Gorham's, Four-way, 278 
 
 Heating of, How Guarded Against, 267 
 
 Best Material for, 267 
 
 Object of, is65 
 
 Post Office Switches, Details of, 277 
 
 Quick Action a Necessity, 265 
 
 Single and Double Pole, Principle of, 265, 268 
 
 Use°of Springs, 266 
 
 SW Tus°h A C D ompany's High and Low Pressure Board, 289 
 Crompton and Co.'s Boards, 278 
 Cut-outs for, 289 
 
 Drake and Gorham's Accumulator Board, 278 
 General Electric Company's Board, AM 
 Object of, 278 
 
 Poole and White's Board, 289 
 Post Office Board, 286, 291 
 in Use at Madame Tussaud s, 289 
 in Use at Royal English Opera House, 286 
 
 T. 
 
 Tangent Galvanometer, see Galvanometers and Measurement 
 S^l2^Sg£S^^U«I of Testing, 3C6 
 
 ?Stinf Set, The Silvertown, 344 
 
 TVstine Sets, Portable, 344 
 Stin| Wheatstone Bridge Method, 3C5 
 Thermal Units, etc., see Heat 
 Thermodynamics, Theory of, 165 
 
ikDEX. 
 
 Thermo dynamics of Imaginary Gases, 166 
 
 Temperature : Its Effect on Resistance, 8 ; Definition of, 161 ; 
 Measurement of, 161 ; of Gases, absolute, 164 ; Changes of in an 
 Engine Cylinder, 176. 
 Thompson, Silvanus, His Work in Electro-graphics, 37 
 Thomson, Sir William : 
 
 His Method of Measuring Battery Resistance, 330 
 His Method of Measuring Capacity, 337 
 His Method of Measuring Galvanometer Resistance, 315 
 His Meters : 
 
 Ampere Balances, 316 
 Ampere Gauge, 323 
 Composite Balance, 319 
 Electrostatic Balances, 323, 326 
 Electrostatic Voltmeters, 323 
 Engine-room Voltmeters, 327 
 Magnetostatic Current Meter, 321 
 Marine Voltmeter, 319 
 Multicellular Voltmeter, 323 
 Throttle-valve, see Valve, Throttle 
 Time, Unit of, 153 
 Torque, Explanation of, 37 
 Tubes for Boilers, Construction of, 110 
 Tubular Boilers, see Boilers] 
 Tussaud's, Madame, Switchboard Used at, 289 
 
 u. 
 
 Underfooting Foundations, 51 
 
 Underground Mains, Methods of Laying, see Mains 
 
 Unit Current, Definition of, 302 
 
 Units for Comparing Phenomena of Circuits, 7 
 
 Units, Various : 
 
 B.A. Committee's Work on 291 
 
 The B.A. and the Legal Ohm, 7 
 
 C.G.S. System, 7, 291 
 
 of Current, 9, 302 
 
 Dynamical, 292 
 
 of Force, 15, 3 
 
 Fundamental and Derived, 292 
 
 u r Meat, 14, 162, 305 
 
 of Length, 153 
 
 of Mass, 153 
 
 The Ohm, 7, 304 
 
 of Power, 154, 305 
 
 of Pressure, 9, 304 
 
 of Quantity and Capacity, 31, 305 
 
 of Resistance, 7, 304 
 
 Thermal, 14, 162, 305 
 
 of Time, 153 
 
 of Velocity, 154 
 
 The Volt. 9, 304 
 
 of Work, 13, 154 
 
 V. 
 
 Valve, Blow-off for Boilers, Hopkinson's, 122 
 Valve, Throttle, Description of, 237 
 
 Valves, Feed, for Boilers : 
 Use of, 139 
 Ferranti's, 140 
 Hopkinson's, 140 
 Renewable, 140 
 
 Valves, Safety, for Boilers : 
 
 Board of Trade Rules for, 119 
 
 Construction of, 115 
 
 Dead- Weight, 115 
 
 Double-Lever, 117 
 
 Galloway's, 115 
 
 Hopkinson's Compound, 117 
 
 How to Calculate Size of Opening, 119 
 
 How to Find Working Pressure, 117 
 
 Ramsbottom's, 117 
 
 Relief- valves, 117 
 
 Rule for Calculating Size of Springs, 120 
 
 Rules for Construction of Valve, 119 
 
 Spring, 116 
 Valves, Slide, of Steam Engines : 
 
 Description of, 240 
 
 Meyer's 243 
 
 Paxman's, 243 
 
 and Ports, Detailed Explanation of Work of, 241 
 
 Relation of Eccentric to, 239 
 
 Use of, 239 
 
 Zeuner's Diagram, 240 
 Valves, Steam, for Boilers : 
 
 Use and Construction of, 125 
 
 Dewrance'B Renewable, 126 
 
 Hopkinson's, 125 
 
 Rhodes's, 125 
 
 Robey'e, 125 
 
 Valves, Steam Check, for Boilers : Use of, 127 ; Hopkinson's, 127 
 Valves, Steam-Reducing, for Boilers : The Use of, 130 ; Dewrance's 
 
 130 
 Valencies, Table of, 19 
 Velocity, Definition of, 154 
 Velocity, Unit of, 154 
 Vertical Boilers, see Boilers 
 Volt, The Absolute, 304 
 Volt, The Unit of Pressure, 9 
 Volts, Amperes and Ohms, Their Relation, 304 
 Voltameter, Construction of, 19 
 
 Voltmeters, see also Measurement 
 Ayrton and Perry's, 341 
 Calibration of, 343 
 Cardew's, 343 
 Pocket, 343 
 
 Thomson's Composite, 319 
 Thomson's Electrostatic, 323 
 Thomson's Engine-room, 327 
 Thomson's Marine, 319 
 Thomson's Multicellular, 323 
 
 Volume and Pressure, 156 
 Volume, Specific, Definition of, 153 
 
 w. 
 
 Walls, Construction of, 57 
 
 Water Analogy as an Explanation of Current Flow, 12 
 
 Water Gauge for Boilers, Dewrance's, 113 
 
 Water, Heat Required to Evaporate it, 61 
 
 Water-level Indicators for Boilers, 113 
 
 Water Supply oe Boilers : 
 
 Chemical Composition of, 148 
 
 Compositions for Preventing Scale, 150 
 
 Davies and Metcalfe's Exhaust Steam Injector, 137 
 
 Feed-pipe and Feed-valves, 139 
 
 Feed-water, Fuel Saved by Heating, 141 
 
 Feed-water, Heating of, 141 
 
 Garrett's Feed Pump, 131 
 
 Garrett and Son's Feed-water Heater, 141 
 
 Gresham and Craven's Injector, 134 
 
 How Kept Up and Regulated, 130 
 
 Kirkaldy's Feed-water Heater, 142 
 
 Mumford's Donkey Pump, 131 
 
 Musgrave and Son's Feed-water Heater, 144 
 
 Robey and Co.'s Feed-water Heater, 143 
 
 Steam Injector, Principle of, 134 
 
 Steam Injector, Exhaust, 137 
 
 Steam Injector, Lifting, 137 
 
 Steam Injector, Non-lifting, 137 
 Watt, Electrical Unit of Work, 14, 305 
 Watt's Engine Governor, 231 
 
 Webber and Calender's Method of Laying Electric Light Mains, 361 
 Weight, its Relation to Mass, 154 
 Wbeatstone's Method of Measuring Pressure 333 
 Wheatstone's Standard Cell, 331 
 
 Wiieatstone Bridge : 
 
 Cheap Form Wanted for Engineering Work, 328 
 
 Kirchoff's Slide- wire Form, 307 
 
 Method of Using, 305 
 
 Measuring Resistances by, 305 
 
 for Measuring Battery Resistance, 329 
 
 for Measuring Capacity, 338 
 
 for Measuring Galvanometer Resistance, 313 
 
 for Measuring Insulation Resistance, 347 
 
 for Measuring Potential or Pressure, 332 
 
 Post Office, Form of, 305 
 
 Resistance Coils for, 308 
 
 As used in the Silvertown Testing Set, 346 
 
 Wiedemann's Method of Measuring Pressure, 333 
 
 Windsor Engine, The, 264 j. 
 
 Winn's Sight-Feed Lubricator, 248 
 
 Wire Carrying Current, Phenomena Connected with, 12 
 
 Wires, see also Conductors 
 
 Wires, Conductivities of Hard-drawn and Annealed, 9 
 
 W r lre ,'l;: U ndel 'S round > Method of Laying, see Mains 
 Woolf's Compound Engine, 177, 182 
 
 W ork : 
 
 Electrical, Unit of, 14, 305 
 
 Definition of; 154 
 
 In Terms of Pressure and Volume, 156 
 
 Mechanical, Unit of, 13 
 
 Unit of, 154 
 
 Representation of, by an Area, 156 
 
 Useful and Waste, 160 
 
 z. 
 
 Zeuner's Slide-valve Diagram, 240