IRLF. 
 
 -' ',.: ' . ::.-: rr 
 
 toSl ^MATURES -'Mi) 
 
 / ; / 
 
*- 
 
 LIBRARY 
 
 OF THE 
 
 UNIVERSITY OF CALIFORNIA, 
 
 Deceived J>J&/^* . 189 
 
 Accessions No.(p/ u// . Class No. 
 
DRUM ARMATURES AND 
 COMMUTATORS 
 
 (THEORY AND PRACTICE). 
 
 A COMPLETE TREATISE ON THE THEORY AND CONSTRUCTION OF DRUM 
 
 WINDING, AND OF COMMUTATORS FOR CLOSED-COIL ARMATURES, 
 
 TOGETHER WITH A FULL RESUME OF SOME OF THE PRINCIPAL 
 
 POINTS INVOLVED IN THEIR DESIGN ; AND AN EXPOSITION 
 
 OF ARMATURE REACTIONS AND SPARKING. 
 
 F. MARTEN V^EYMOUTH. 
 
 ENLARGED AND REVISED FROM A SERIES OF ARTICLES IN "THE ELECTRICIAN. 
 
 LONDON : 
 
 THE ELECTRICIAN " PRINTING & PUBLISHING COMPANY, 
 LIMITED, 
 
 I, 2, AND 3, SALISBURY COURT, FLEET STREET, E.G. 
 I8 94 . 
 
 [All Rights Reserved.} 
 
 Of 
 
 UJTI7BESIT 
 
T/W77 
 
 \A/ 
 
 Engineering 
 Library 
 
 "77 
 
 Printed and Published !>> 
 
 THK KLECTPaCIAN ' PRINTING ANI> PUBLISHING CO., LlMITKI) r 
 
 1, 2, and 3, Salisbury Court, Fleet Street, 
 London, E.O. 
 
CONTENTS. 
 
 PAGE 
 
 INTRODUCTORY xi. 
 
 CHAPTER I. 
 
 THE GENERATION OF CURRENT AND POTENTIAL IN DRUM 
 
 ARMATURE WINDING ... ... ... ... ... 1 
 
 Advantages of Drum over Gramme Winding Inactive Copper 
 Lines of Force Generation of E.M.F. Direction of Current- 
 Plane of Commutation Accumulation of E.M.F. Two 
 Currents Augmenting each other Non-Augmentation of their 
 Potentials Potential Difference between Adjacent Conductors 
 Self-Excitation Distortion of the Field Torque Im- 
 mobility of Armature Poles during Rotation Torque or 
 Magnetic Drag on Coils Peripheral Creeping of Conductors 
 Binding Wires and Projections for Driving. 
 
 CHAPTER II. 
 WIRE DRUM WINDING : SIEMENS ... ... ... 17 
 
 Difficulties with Drum End Winding Siemens' Wire Wrapping 
 " Turns per Section" Potential Difference between Sections- 
 Windings for Large Currents Merits and Demerits of Wire 
 Wrapping Heating with Overflow of Main Current Short- 
 Circuit Difficulty in Unwinding for Repairs Potential 
 Difference between Overlying Wrappings at Armature Ends 
 No Ventilation Suitability for Small Current and High 
 Voltage. 
 
 CHAPTER III. 
 HEAVY WINDING : DIFFICULTIES AND PRELIMINARY CON- 
 
 SIDERATIONS 
 
 "Bars" instead of Wires Considerations Efficiency Economy 
 in Copper and Material Compactness Floor Space- 
 Ventilation Joints Ease of Construction Repairs Driving 
 Centrifugal Force Balance Knife Edges Insufficient for 
 Testing Balance Symmetry Foucault or Eddy Currents 
 Cooling of Large and Small Bars Trailing Horns. 
 
 A 
 
 25 
 
v. CONTENTS. 
 
 PAGE 
 
 CHAPTEK IV. 
 HEAVY WINDING: PREVENTION OF FOUCAULT CURRENTS ... 31 
 
 Two Principle Methods Avoidance by Removal of Cause Cause 
 Remaining : but Effects Prevented Mather and Platt 
 Widened Air Gap under Horns Siemens' Cable Conductors 
 Twisted Copper Tape Cronipton's Method Swinburne's 
 Method Ventilation Insulation and Heat. 
 
 CHAPTER V. 
 HEAVY WINDING: PLATE END CONNECTION. EDISON ... 39 
 
 Necessity for Special Means of End Connections with Rigid Con- 
 ductor Bars Edison's Radial and Circular Bars Want of 
 Symmetry Edison's Disc Connectors Potential Difference 
 between Adjacent Discs Stamped Discs Bars Straight 
 Joints Outside. 
 
 CHAPTER VI. 
 
 HEAVY WINDING : EVOLUTE END CONNECTIONS. CROMPTON 
 
 AND SWINBURNE. ANDERSEN. HOPKINSON ... ... 43 
 
 Crompton and Swinburne Volute and Crank End Winding 
 Variations Two Turns per Section " Westminster" Armature 
 Wire and Copper Tape Wire Crank Rolled Copper Strip 
 Connectors Extra Joints Potential Difference between 
 Adjacent Volutes Alteneck and Hopkinson : Double Volutes 
 Two Turns per Section Solid Bars Driving by Snugs on 
 Bars Rolled Copper Strip Connectors Jointless Double 
 Volute End Connector. 
 
 CHAPTER VII. 
 EVOLUTE WIRE WINDING: EICKEMEYER... ... ... 55 
 
 Eickemeyer : Double Volute Coils Wire Throughout Without 
 Joint Many Turns per Section for High E.M.F. Wires in 
 Parallel for Large Current Coils Formed on Mould Driving 
 Pins Ease of Construction Two Layers of Conductors 
 Potential Difference between Layers Air Insulation between 
 Two Series of Volutes Interior or Exterior Connections to 
 Commutator. 
 
 CHAPTER VIII. 
 
 HELICAL END WINDING : KAPP. CHORD WINDING : 
 
 SWINBURNE ... ... ... ... ... ... 63 
 
 Kapp's Helical or Corkscrew Method Self-contained End Winding 
 Strip Wire Cable Symmetry Potential Difference between 
 Adjacent Connectors Stamped Plates Joints Outside 
 Ventilation Swinburne's Chord Winding Wire Plate End 
 
CONTENTS. V. 
 
 PAGE 
 
 Connectors Potential Difference between Adjacent Plates 
 Theory of Chord Winding Cross - Magnetisation Back- 
 Induction General Theory of Armature and Field Inutility 
 of Chord Winding to Obviate Cross-Induction Two Planes of 
 Commutation. 
 
 CHAPTER IX. 
 
 EXTERIOR END WINDING : CROMPTON AND KYLE. PARSONS. 
 
 FRITSCHE ... ... ... ... ... ... 77 
 
 Crompton and Kyle : Exterior Double Volute End Winding 
 Collision with Magnet Coils to be Avoided Magnetic Circuit 
 of Dynamo Parsons' Winding Hollow Shaft Oil or Water 
 for Cooling Fritsche Winding. 
 
 CHAPTER X. 
 COMMUTATORS. INTRODUCTORY REMARKS... ... ... 85 
 
 General Considerations Conflicting Requirements AVearing and 
 Burning Effect of Insulation on Diameter Size and Propor- 
 tions Brush Contact Surface Elwell-Parker Motor-Generator 
 Flats Grooving and Cutting Brushes Lubrication 
 Carbonisation of Oil Metals for Segments Process of Con- 
 struction Self-Contained when Complete. 
 
 CHAPTER XI. 
 COMMUTATORS. INSULATION ... ... ... ... 93 
 
 Varying Potential Difference between Segments Potential 
 Difference between Segments and Sleeve Insulation 
 Vulcanised Fibre Manufacture of Fibre Absorption of Oil 
 and Moisture Specific Resistances Experiments with Fibre 
 Mica Choice between Fibre and Mica Mica Less Affected 
 by Heat and Moisture Account of Mica Its Characteristics 
 Construction of Mica Cones .for Commutators. 
 
 CHAPTER XII. 
 COMMUTATORS. METHODS OF CONSTRUCTION ... ... 103 
 
 Classification of Designs Depth for Wear Shortness of Length 
 Commutators of Crompton, Mordey, and Alioth and Co. Full 
 Depth for Wear Clear Ends for Brush Contact Easton's 
 Groove Angle of Cones Locking of Nuts Kennedy, " West- 
 minster," Giilcher, and Kapp Commutators Economy in 
 Length Undercut Ends Cones with Two Slopes Circular 
 Arch Principle of Construction Bedding on Sleeve Prime 
 Cost Centreing for Segments Double Bush for Large Com- 
 mutators Air Insulation Inside Kapp. Paterson and Cooper, 
 ' : Westminster," and Hopkinson Commutators Bedding on 
 Sleeve at Ends Only Mica for Bedding Slewing of Segments 
 Driving Methods of Tightening Nuts. 
 
VI. CONTENTS. 
 
 PAGE 
 
 CHAPTEK XIII. 
 COMMUTATORS. CONNECTIONS WITH ARMATURE WINDING 117 
 
 Lugs Cast with Segment, or Separate Copper Strips Forked 
 Outer End Folded Outer End Straight Strip Fitting Saw 
 Cut at Each End Short Cast Lugs with Wire Winding 
 Long Cast Lugs with Bar Winding Sloping Lugs Ventilation 
 Long Cast Lugs, Lightened Out Economy Hopkinson 
 Interior Bar Connection. 
 
 CHAPTER XIV. 
 
 SPARKING AT COMMUTATORS. ELEMENTARY THEORY OF 
 
 ELECTRIC SPARKING ... ... ... ... ... 125 
 
 Sparking at Commutators Wasted Energy Electric Sparking 
 Generally Considered Five Orders of Sparks Cursory Glance 
 at Orders of Sparks Statical Disruptive Spark Surface 
 Electricity Electrodes not Damaged Rupture of Intervening 
 Bodies Lightning Voltaic Arc Circuit No Circuit with 
 Static Discharge Prof. S. P. Thompson on Voltaic Arc Com- 
 mencement with Self-induction Spark- Electrodes Consumed 
 Intervening Bodies Destroyed by Heat Induction Dis- 
 ruptive Sparks Magnetism Coil and Ampere's Law 
 Primary and Secondary Coils Induced Current and Sparks 
 in Secondary Circuit Effect Due to Magnetism Lenz's Law 
 Illustrated with Armature Coil Magnetism and Secondary 
 Coil High, Tension in Secondary of Fine Wire Induction 
 Sparks, Disruptive Transformers Alternating Currents- 
 High Tension Experiments by Siemens Bros, and Swinburne 
 and Co. Sparks in Primary Self-induction. 
 
 CHAPTER XV. 
 
 SPARKING AT COMMUTATORS. SELF-INDUCTION AND ELE- 
 MENTARY THEORY CONTINUED... ... ... ... 139 
 
 Self -Induction Illustrated Magnetic Induction Disruptive Spark: 
 Continuance as Arc Lenz's Law Again " Electric Inertia " 
 an Unsafe Term Absence of Momentum in Electricity 
 Magneto-Electric Generation of High E.M.F. Fifth Order of 
 Sparks High Normal Tension Continuous Disruptive Dis- 
 charge Comparison with Arc Easy Divergence of Arc Ex- 
 periments by Siemens Bros, with Long Arc Intense Proclivity 
 of Disruptive Discharge for Straight Course High Tension 
 and " Shock " Large Current and Heat Smelting by 
 Electricity Back E.M.F. in Arc Minuteness of Time for 
 Disruptive Spark Lightning Flash Flash at Opening of a 
 Switch Compound Nature of Flash at Switch Deflagration 
 Due to Current Destruction of Switch Contacts, Due mostly 
 to Momentary Arc. 
 
CONTENTS. Vli. 
 
 PAGE 
 
 CHAPTER XVI. 
 
 SPARKING AT COMMUTATORS. THE ELEMENTARY PLANES 
 
 THROUGH COMMUTATOR AND ARMATURE ... ... 147 
 
 General Assumption of Armature Coils, Commutator, and Field- 
 Varying Potential Difference between Adjacent Segments 
 Points of Maximum and Minimum Potential Plane of Com- 
 mutation Plane of Coil Dividing Commutator Six Orders of 
 Critical Planes Three Planes Connected with Current : Three 
 with Potential Plane of Commutation : in Armature Only : 
 Controlled by Brushes : Concerned with Current Only Neutral 
 or Non-Sparking Plane Plane of Requisite Reversal Plane 
 of Maximum E.M.F. through Commutator Plane of Maximum 
 E.M.F. through Armature Desirability of Small Angular 
 Distance between Planes of Non-Sparking, and Maximum and 
 Minimum E.M.F., through Commutator Lead to Brushes 
 Apparent and True Lead Plane of Cessation of Force Com- 
 mutation of E.M.F. Normal to Distortion of Field Ulti- 
 mate and Self -Polarisation of Armature Plane of Cessation of 
 Force Slightly Affected by Moving Brushes Conditional 
 Identity of Plane of Maximum E.M.F. through Armature and 
 Plane of Cessation of Force Coincidence of Three Planes 
 through Armature Flow of Current Counter to E.M.F. 
 Oscillation of Normally Fixed Planes. 
 
 CHAPTER XVII. 
 
 SPARKING AT COMMUTATORS. ELEMENTARY CONSIDERATION 
 
 OF THE BRUSH ... ... ... ... ... 161 
 
 Sparking at the Brushes Self-induction of the Armature Coils 
 Reversal of Polarity and Loss of Energy Short- Circuits under 
 Brush Short-Circuit Currents Sparking at Brush similar to 
 Flash at Opened Switch Segment and Brush both Burnt 
 " Shunt " Flash Non-Sparking Conditions : Baulking of Main 
 Current by Counter-flowing Short-Circuit Current Coil in Plane 
 of Requisite Reversal. 
 
 CHAPTER XVIII. 
 
 SPARKING AT COMMUTATORS. THE SHORT-CIRCUITS AND 
 
 CURRENTS ... ... ... ... ... ... 167 
 
 Two Short- Circuited Coils : Effect on each other Two Circuits 
 with Common Conductor Confluence and Augmentation of 
 Current Opposition and Assimilation of Currents Analogy to 
 "Three-Wire System" Conditions at Brushes Short-Circuit 
 Currents and Armature Main Currents Ohm's Law Normal 
 Flow of Currents Effect of Currents on one another Ultimate 
 Flow after Assimilation or Augmentation Diversion of Cur- 
 rent, as result Conditions for " Shunt " Sparking. 
 
 UFIVBRSIT7 
 
Vlll. CONTENTS. 
 
 PAGE 
 
 CHAPTER XIX. 
 
 SPARKING AT COMMUTATOKS. THE SHORT -CIRCUITS AND 
 
 POTENTIALS ... ... ... ... ... ... 175 
 
 Potentials of Assimilated Currents With Differing Resistances, 
 Different Voltages to Produce Similar Current Accumulation 
 of E.M.F. of Generators in Series Effects with Resistances 
 Exterior to Generators Effect with Interior Resistance of 
 Generator only E.M.F. Spent Internally Effect of Potential 
 of one Circuit on the other Circuit Dependent on Position of 
 Resistances External Voltage of Generator Dependent on 
 External Resistance Change of E.M.F. Cause Fluctuation of 
 Current Potentials at Ends of Short - Circuited Coils 
 Accumulation of Voltages of Armature Coils Dependent on 
 Exterior Resistance Energy Lost in Attaining Reversal 
 Requisite for Sparklessness. 
 
 CHAPTER XX. 
 
 SPARKING AT COMMUTATORS. EFFECTS OF MISPLACEMENT 
 
 OF BRUSHES ... ... ... ... ... ... 183 
 
 Sparking Due to Misplacement of Brushes, or to Oscillation of 
 Neutral Plane Positive Brush too far Back : Heavy Forward 
 Sparking Positive Brush too far Forward : Light Back Spark- 
 ing Negative Brush too far Back : Heavy Forward Sparking 
 Negative Brush too far Forward : Light Back Sparking 
 General Remarks Diversion of Main Current from or to Brush 
 Points by Assimilation with Short- Circuit Currents Plane of 
 Commutation with Varying Lead Causation of Back Sparking. 
 
 CHAPTER XXI. 
 
 SPARKING AT COMMUTATORS. NON- SPARKING WITH CARBON 
 
 BRUSHES 195 
 
 Two Causes of Non-Sparking with Carbon Brushes Principle 
 Cause : Resistance of Contact Care usual in Attachment of 
 Carbon to Metal Short-Circuit Currents practically Eliminated 
 Operation of Ohm's Law Different Paths for Current Flow 
 with Resistances Changing with Rotation Resistance In- 
 creasing at Forward Edge of Brush Commutation before 
 Recession Minimum Current at Recession Heating of For- 
 ward Edge Less Current : but Increasing Current Density up 
 to Moment of Recession. 
 
 CHAPTER XXII. 
 
 SPARKING AT COMMUTATORS. CAUSES EXTERIOR TO MACHINE 203 
 
 Indirect Causes of Sparking Oscillation of Normally Fixed Critical 
 Planes Position of Fixed Planes Dependent on Distortion of 
 the Field Distortion Dependent on Ratio of Field Strength to 
 Armature Self-Magnetisation Effects of Varying the Ratio 
 
CONTENTS. IX, 
 
 PAGE 
 
 Bending or Deflection of Planes External Causes of Sparking 
 Gradually Formed Short-Circuit in Outer Circuit Outer 
 Resistance Diminishing : Outer E.M.F. also Dimmish : All 
 E.M.F. Spent Internally Increase of Current : Strengthening 
 of Armature Self -Magnetisation Effect with Separately Excited 
 Dynamo Effect with Shunt-Wound Dynamo Field Lost 
 Sudden Short - Circuit Self-induction of Shunt Coils 
 Hysteresis of Magnets Series-Wound Dynamo Opposite 
 Effects According to Amount of External Resistance Effect 
 Opposite to those with Shunt-Wound Dynamo Change of 
 " Load " similar to Short-Circuit Lamps in Parallel Com- 
 pound Winding. 
 
 CHAPTER XXIII. 
 
 SPARKING AT COMMUTATORS. ARMATURE REACTIONS, 
 
 FIELD WEAK 215 
 
 Causes within the Machine Weakness of Field Back- and Cross- 
 Induction Armature Reactions Armature Self -Polarity as 
 Resultant and Components Forward - Induction Assisting 
 Field Back-Induction Opposing Field Reactions Regarded 
 Optionally as Resultant or Components Effect of Reactions on 
 Horns of Pole-Pieces Self- Polarity of Drum as Compared with 
 Gramme Advantage of Drum Winding Great Inclination 
 Forward of Plane of Requisite Reversal in Weak Field 
 Sparking with Brush Points in M' M' Plane Negative Lead ; 
 Forward Induction ; and the Short-Circuit Currents Under 
 Brush Inutility of Forward-Induction in a Dynamo - Forward- 
 Induction with Positive Lead in a Motor Motor Armature 
 Creating its Own Field Dynamo Armature Creating its Own 
 Field Short-Circuit Currents in a Motor Armature Inutility 
 of Forward-Induction in a Motor Sparking with a Motor 
 Loss of Outside Voltage when Brushes are at Non-Sparking 
 Points, with Weak Field in Dynamo. 
 
 CHAPTER XXIV. 
 
 SPARKING AT COMMUTATORS. ARMATURE REACTIONS. 
 
 MAGNETIC FLOW 227 
 
 Armature Reactions as Components Magnetism as a Current 
 " Free " Magnetism : Likened to Static Electricity Point of 
 Failure in Similitude to Electricity Horseshoe Magnet and 
 Field Magnet in Field Assisting or Opposing Magnet across 
 the Field : Distortion of Magnetic Flow Distortion shown 
 Graphically and Analytically as Components Back-, 
 Forward-, and Cross -Inductions, represented Armature in 
 Field Negative Lead and Forward-Induction Positive Lead 
 and Back-Induction Density of Magnetic Flow Opposition 
 and Confluence of Inductions Effect in Horns with Back- 
 Induction Counteraction of Horns Effect in Horns with 
 Forward-Induction Effect in Necks Non-Elimination of 
 Magnetism Internal Induction Flow Converted into "Free" 
 
. X. . CONTENTS. 
 
 PAGE 
 
 Magnetism Differing Attraction of Pole-Pieces for Foreign 
 Bodies of Iron or Steel Armature Self -Magnetisation become 
 Free Waste Field Air Gaps Sayers' Winding Forward- 
 Induction without Sparking " Commutator Coils " producing 
 Short Circuit Current Counter to Main Current through Brush 
 Point Armature Create its own Field Field Winding not 
 necessary Accumulation of Currents Mutual Effects on one 
 another Potentials of Commutator Coils Energy lost. 
 
 CHAPTER XXV. 
 
 SPARKING AT COMMUTATORS. ARMATURE REACTIONS. 
 
 FIELD ASYMMETRICAL, ... ... ... ... 245 
 
 Asymmetry of the Field Example Possible with Double-ended 
 Magnets Asymmetry Caused by Throttling Induction Pole- 
 Pieces with Narrow Necks Confluence and Opposition of 
 Induction Causes of Weakness of Fields Saturation -Dis- 
 tortion of Field due to Cross-Induction Diverting Effect of 
 Cross-Induction on Main-Induction Throttling of Cross-In- 
 duction ; and Inequality of Diversion Either Side of Field 
 Opposite Horns not Equally Counteracted Deflected Distortion 
 of Field Sparking at One Brush Asymmetry of Free Mag- 
 netism and Waste Field Throttling of Main as well as Cross- 
 Induction Deflected Distortion of Field Sparking at One 
 Brush Double-ended Magnets Thick Necks and Cross-In- 
 duction Indented Necks, and Both Cross-Inductions Throttled 
 Division into Oblique Inductions Fischer-Hinnen's Method 
 of Ketarding Cross-Induction, and Lessening Lead. 
 
 CHAPTER XXVI. 
 SPARKING AT COMMUTATORS. ARMATURE DEFECTS ... 259 
 
 Mechanical Defects in Armature Winding Ever Changing Effects 
 During Rotation Oscillation of Planes Asymmetrical Wind- 
 ing Effects Greater Output from One Half of Winding than 
 from Other Half Asymmetrical Self-Polarisation of Armature 
 Coils Separately Generating Unequal Amounts of Short- 
 Circuit Current Asymmetry Confined to End Winding Coils 
 Unequal in Resistance : Short-Circuit Currents Unequal 
 Oscillation of Planes of Requisite Reversal and Non-Sparking 
 Defective Joints: Divisible under Four Heads Effects on' 
 Short Circuit Currents Varying Opposition of Short-Circuit 
 and Main Currents Sparking Availability of Carbon Brushes 
 Various Causes of Sparking. 
 
 CHAPTER XXVII. 
 THE TAPER OF COMMUTATOR SEGMENTS 269 
 
 Accuracy of Fit Required for Commutator Segments Trotter's 
 Gauge Specification of Taper of Segment Insulation : Its 
 Effect on Diameter to be Reckoned Principle of Gauge 
 Manner of Stating Instructions. 
 
INTRODUCTORY. 
 
 THE first of these Articles (comprised within 
 Chapter I. of this work) appeared in The 
 Electrician about the time of the lapse of Siemens' Drum 
 Armature Patent. Prior to that event Gramme arma- 
 tures had been in very general use. But since then the 
 Drum armature has become a formidable rival ; and 
 though, for some purposes, the Gramme may still con- 
 tinue well suited, yet it is undoubtedly the fact that, on 
 account of its many advantages, the Drum armature has 
 now driven the Gramme very largely from the field. 
 The first chapter will thus be found to relate entirely to 
 the theory of the Drum winding. An explanation, it is 
 hoped with sufficient fulness, is there given of the 
 generation of electromotive force and current within 
 that form of winding ; and questions of magnetism 
 connected therewith are also considered. Succeeding 
 chapters are devoted to the description of various 
 methods in which Drum winding has been carried out 
 in practice, with special reference to what is termed 
 the "end-winding." Subsequent chapters, again, touch 
 upon the mechanical construction of Commutators ; and 
 
 A 2 
 
Xll. INTRO DUOTORY. 
 
 finally, the book closes with some chapters on com- 
 mutator sparking, with which has become necessarily 
 involved the whole subject of what are known as 
 " armature reactions." 
 
 Objection may be taken that these various subjects 
 are here treated without, or almost without, the use of 
 mathematics. But I venture to apprehend that, in this 
 case, the absence of mathematics is far from being un- 
 justifiable. 
 
 In the first place, it is not intended that this book 
 should supplant any existing work, but that it should be 
 a useful guide or introduction to those who may wish 
 ultimately to proceed with the mathematical treatment 
 of the subjects. It is hoped, also, that the non-mathe- 
 matical reader will here find a sufficiently clear, correct, 
 and succinct account of the " action " of a dynamo. 
 
 It is on this word " action " that the whole question 
 of mathematics rests : for my object is simply to render 
 as easy of comprehension as possible the action of a 
 dynamo, especially with a Drum armature, as well as 
 with a Gramme. To this end anything more than 
 an occasional simple mathematical illustration is not 
 necessary. 
 
 It is apprehended that the beginner will read these 
 pages during the early period of his training, while he is 
 studying his mathematics, and so may combine the two 
 together at a later and more advanced stage. 
 
 Inasmuch as frequent reference is made to the Gramme 
 armature, there is added at the end of this introductory 
 
INTRODUCTORY. 
 
 chapter, for simple comparison, a diagrammatic repre- 
 sentation of this form of winding, similar to that of the 
 Drum winding (Fig. i) in Chapter I. The upper brush 
 is negative, and the direction of the current flow is 
 indicated by small arrows. It will thus be observed that 
 the current arriving by the upper brush splits right and 
 left, and so flows in two streams, one in each half of the 
 winding, till they unite again at the bottom, and proceed 
 through the lower brush to the outer circuit. The 
 direction of the circulation of current in each half being 
 then noted, it will be seen that, according to Ampere's 
 law, each half of the winding is exciting a north pole at 
 the top and a south pole at the bottom, as indicated by 
 the small n n and s s. Thin dotted lines represent the 
 core of the armature. 
 
 <U/ 7^1 
 
 
 1 w if-l 
 
 
 ^^ . u / 
 
 1 1 ' 
 
 f 9 
 
 DIAGRAM OF GRAMME WINDING. 
 
 A further understanding may be gathered from 
 Chapter I., in so far as the theory and action of the 
 Gramme is similar to that of the Drum armature. 
 
CHAPTEK I. 
 
 THE GENERATION OF CUEEENT AND POTENTIAL 
 IN DEUM AEMATUEE WINDING. 
 
 As it is well, before considering practice, to understand 
 theory, it is our purpose in this preliminary chapter to inves- 
 tigate the elementary theory of " drum winding." We may 
 thus ascertain exactly in what manner the current and its 
 potential are generated, and the " torque " set up. What is 
 meant by this latter term may also be made clear, and cer- 
 tain points that are inherent with this effect may be touched 
 upon. By the enunciation and application of known laws we 
 shall also find that all these phenomena, under stated condi- 
 tions, are but the natural results of such laws. We can then 
 in subsequent chapters proceed to consider various methods 
 in which theory has been put into practice ; while some of the 
 more complicated detail of theory, from its technical nature, 
 we will leave for discussion till we have reached that more 
 advanced stage. 
 
 Our attention may, however, first be engaged in briefly 
 considering what are the more essential advantages that this 
 method of laying the conductors on an armature possesses over 
 the Gramme winding it is now to so great an extent super- 
 seding. Some of these are more or less obvious. It offers the 
 whole area within the limits of the periphery and ends of the 
 drum for the passage of lines of magnetic induction, as com- 
 
2 DRUM ARMATURE WINDING. 
 
 pared with the portion only of this total area offered by the 
 Gramme ring type. Thus, other things being equal, for any 
 given length and diameter of core, the drum will give the 
 larger output ; and conversely, for any given output, the drum 
 armature, and with it the whole machine, can be made smaller, 
 and so cost less than the other type. A greater radial depth 
 of iron also can be got into the core, still further favouring 
 the passage of lines of magnetic induction a point of great 
 moment. Economy in copper is also promoted, inasmuch as- 
 there is less length of conductor, as compared with the 
 Gramme type, owing to the absence of the interior parallel 
 conductors that exist with the latter. This consideration is, 
 however, much affected by design, a question it is not our 
 present object to discuss. 
 
 The accompanying diagrammatic sketch (Fig. 1) represents 
 in isometric perspective the drum-winding of an armature and 
 commutator, but without spindle or core, except that the last 
 named is indicated by two lightly dotted circles. A winding of 
 eight turns only is depicted, so that all parts may be visible, 
 and the commutator is shown an abnormal distance from the 
 core to avoid the complication of its covering the cross- 
 connections behind it. This sketch represents in essence what 
 is the more modern method of arranging the winding, par- 
 ticularly for heavy currents. In the earlier drum armatures, 
 such as those of Siemens and Edison, it was the custom 
 to connect up the different sections, not one after the other 
 in order round the periphery of the core, but variously. This 
 disorder of connection had the effect of preventing the attain- 
 ment of a minimum difference of potential between neighbour- 
 ing commutator segments, so that sparking at the brushes was 
 not diminished as much as might be. But for constructional 
 reasons also it is now more customary to effect the joining up, 
 as here shown, one section after another in regular order round 
 the armature. 
 
GENERATION OF CURRENT AND POTENTIAL. 
 
4 DRUM ARMATURE WINDING. 
 
 To assist the eye in following up the connections, the 
 parallel conductor-bars are here numbered in the order of 
 their connection. No. 1, for instance, is joined at the right- 
 hand end to No. 2, and No. 2 at the commutator end to No. 3, 
 and so on, till No. 16 is united at the commutator end to No. 
 1 again. The bars with odd numbers alone, it will be observed, 
 are extended to the commutator lugs. Each odd-numbered 
 bar, and an even-numbered bar nearly opposite, with their end 
 connections, form a section of one turn, the end of each section 
 being joined up behind the commutator to the neighbouring 
 odd-numbered bar, and not to its own. An endless circuit is 
 thus formed round the armature, connected up at eight dif- 
 ferent points, corresponding to the eight sections, to eight 
 several insulated commutator segments. Sometimes two turns 
 per section are put in ; only this, with conductors for heavy 
 currents, involves a double series of end connections, which, 
 according to design, may add considerably to the total length 
 of the armature and increase the percentage of inactive 
 copper that is, of copper not actually cutting lines of force, 
 and so not producing current or potential. 
 
 Now this latter reflection is by no means without import- 
 ance. For it bears on all copper that may project at either 
 end beyond the core. This subject may in fact well receive 
 some attention, though it will be out of place here more than 
 just to touch on it. Thus in the illustrations herewith, we 
 have simple armature coils shown revolving in fields, the coil 
 in Fig. 2 being short, and that in Fig. 3 long. Now it might 
 he supposed that the coil in Fig. 3, with its extra length, 
 whereby it might enclose more lines of force than that in 
 Fig. 2, would be preferable to the latter. This argument 
 would have some weight if there were no iron core in the 
 armature. But with such a core it will not hold good. 
 Figs. 4 and 5 will make this clear, which are sectional plans 
 through the field, but without coils. A core is shown in 
 
GENERATION OF CURRENT AND POTENTIAL. O 
 
 Fig. 4. The dotted lines / represent lines of force crossing 
 from the one pole to the other. It will now be noted that 
 with a core present, the lines / outside the field proper are 
 few and sparse, at least as compared with the vast number 
 that will crowd through the core. In Fig. 5, owing to the 
 absence of a core, the total number of lines will be small; 
 
 I" 
 
 It 
 
 bit, 
 
 
 i 
 
 i 
 i- 
 i 
 
 
 
 - IP, 
 
 H 
 
 Hi 
 III 
 
 in 
 
 i 11 
 
 11 
 - jj- 
 
 
 a c 
 
 !;i 
 
 it ' 
 
 fi I 
 ! \\il 
 
 <* & 
 
 FIG. 2. 
 
 FIG. 3. 
 
 FIG. 4. 
 
 FIG. 5. 
 
 but the number outside at / will be much greater in propor- 
 tion to the whole than would be the case in Fig. 4. Hence 
 there might be some justification in Fig. 5 for a long coil. 
 But, returning to Fig. 3, while in the parts of the coil lettered 
 c and d a few stray lines of force might be enclosed, there 
 would be practically none at all caught out at a and b. Now 
 
6 
 
 DRUM ARMATURE WINDING, 
 
 it so happens as a matter of fact, that owing to construc- 
 tional reasons, which will be touched on in future chapters, 
 the few lines at c and d usually are caught. At the same 
 time the very small gain thus accruing would not by any 
 means justify the expense of deliberate enlargement of parts 
 for the purpose. It will thus be seen that, conversely, it 
 is distinctly desirable, on the ground of expense, to keep 
 the amount of copper at the ends of the armature as small 
 as possible. 
 
 For many turns per section of wire winding, it may be 
 necessary to resort to the Siemens' or other methods of carry- 
 ing the wire round the ends. These, however, will be dealt 
 with in subsequent chapters. 
 
 Reverting to Fig. 1, the coils are represented as revolving 
 in a field between two pole-pieces, shown by the heavy dotted 
 lines, the back one of which is south and the front north, as 
 indicated by the large S and N, and turning in the direction 
 of the curved arrow. The lines of force will, therefore, be 
 going from the front towards the back, though with a normal 
 distribution more dense toward the top and bottom, where the 
 poles are closer, than in the middle. This latter inequality, 
 however, is of course much modified by the iron of the core. 
 Each section of the winding will form a loop or circuit 
 revolving in this field ; and when in a vertical position, such 
 as nearly occupied by section 1 16, will enclose a maximum 
 number of lines, as compared with none at all when lying 
 horizontal. Eevolving in this field, an electromotive force 
 will be generated in the conductors. 
 
 First, then, to consider the direction of the current. We 
 know that, according to the law discovered by Faraday, when a 
 circuit revolves in a field such as here represented, the direction 
 of the current will depend on two things. These are, first, 
 according to which side of the loop the lines of force enter, and, 
 second, as to whether, during the revolution, the number of 
 
GENERATION OF CURRENT AND POTENTIAL. 7 
 
 these lines is increasing or diminishing. If the loop revolves so 
 that when the lines are diminishing in number they enter on 
 the side towards which the top bar approaches, then, as viewed 
 from this side, the current will run from left to right in the 
 top bar, and round again from right to left in the lower bar. 
 This is the condition in our illustration, where the point of 
 viow, N, is at the front. 
 
 Now, it is obvious that during one revolution there are four 
 different times when the conditions change. The circle of the 
 revolution of the armature may be divided into four quadrants, 
 denned approximately by the vertical and horizontal centre 
 lines. The four times of change are thus those instants in 
 ~w~hich any bar of a section passes successively from each of 
 :these quadrants to the next. There is a difference, however, 
 between the conditions at two of these instants as compared 
 with the other two. When a given bar at the top passes from 
 ;a quadrant on one side of the vertical to that on the other side 
 ;and back again at the bottom, only one change takes place at 
 each of these two instants, namely, that the enclosed lines of 
 force that were increasing in number as the bar approached 
 the top (or bottom) will now be decreasing in number. But 
 at the two instants when the loop is about horizontal and one 
 given bar passes from an upper quadrant to the one below and 
 rises again on the other side, two changes take place at each of 
 these instants. Firstly, the lines of force that were decreasing 
 in number to zero will now commence to increase again ; and 
 secondly, whereas the lines had been entering the loop on the 
 one side, they will now commence to enter on the other side. 
 It hence arises that when passing the vertical position, where 
 there is only one change of condition, this change reverses 
 the direction of the current round the section ; but when 
 passing the horizontal position the one change from excluding 
 to including lines of force is simultaneously cancelled by these 
 lines commencing to enter on the other side of the loop: 
 
8 DRUM ARMATURE WINDING. 
 
 hence the current is not reversed at these latter two instants- 
 The aggregate result, therefore, is that the current is reversed 
 only as each section passes the vertical (approximately), and 
 is always running from right to left on one side of the 
 armature, and from left to right on the other side. Thus,, 
 in our sketch, the direction of the current is to the right,. 
 in front, and to the left at the back, as indicated by 
 small arrows, while those sections which are nearly vertical, 
 namely, 1 16 and 9 8, are neutral. This approximately 
 vertical plane is, of course, what is known as the plane of 
 commutation, though we shall refer to this again at a 
 subsequent period. 
 
 We have dealt so far with the direction of the current 
 merely. We have now, however, to consider the rise of 
 the "electromotive force," or "E.M.F.," or " potential," as 
 it is variously termed, and without which there would be no- 
 current. This, as is well known, is proportionate to the 
 maximum number of lines of force, or magnetic induction,, 
 passing through the section, or loop, and the speed of the 
 revolutions. The field may be small and intense, or large 
 and weak. Eegarded more intimately as to the speed, the 
 potential varies according to the rate at which the magnetic- 
 lines are included or excluded by the section. Now, as in the 
 case before us, when these lines run horizontally, it is palpable 
 that when the section is about vertical, the rate at which it is 
 including or excluding lines is scarcely more than nominal, as 
 the revolution through a small arc at the top or bottom makes 
 little difference in this respect. Hence the E.M.F. generated 
 here passes through zero value. When nearly or about hori- 
 zontal, the reverse is the case, as the section is in the most 
 advantageous position for this purpose, and excludes and 
 includes the lines as it respectively approaches and leaves the 
 horizontal at its maximum rate. Hence the maximum E.M.F. 
 is gathered when the sections are about horizontal ; but 
 
GENERATION OF CURRENT AND POTENTIAL. 
 
 potential is also being generated in the intermediate sections 
 between the horizontal and vertical. Theoretically, the 
 E.M.F. generated in any one of these varies not with the 
 angle but with the sine of the angle which its plane makes 
 with the vertical plane ; or, in other words, varies directly 
 with the horizontal distance of its bars from the central 
 vertical plane, similarly as the number of magnetic lines en- 
 closed varies with the vertical distance of the bars from the 
 horizontal plane. It must be clearly understood that the fore- 
 going refers only to the generation of the potential in each 
 several section, as the potential that may actually exist in 
 any section at any instant may be very different from that 
 which it is itself at that instant generating. 
 
 We may now consider the course of the current round 
 the winding, together with its accumulation of E.M.F., and 
 its delivery to and return from the outer circuit. Keferring 
 to our sketch, Fig. 1, it will be observed that the brushes 
 touch on the segments of the commutator connected to 
 bars 1 and 9. To commence with bar No. 1, we immediately 
 remark that this, being in connection with the upper brush, 
 and the current in it running from left to right, the current 
 from the outer circuit will here be entering, and the upper 
 brush will be the negative brush of the machine. In this 
 bar 1 there will be the minimum potential, and from its 
 approximately vertical position it can generate practically 
 none. The current running to the right passes across the 
 end to bar 2, thence to the left and across again to bar 3, 
 accumulating E.M.F., and so through sections 4 5, 5 6, 
 where it practically gets its greatest amount of E.M.F., 
 through 7 8, and to the left-hand end of bar 9. This latter 
 being connected to the segment whereon the lower brush 
 makes contact, the current will here pass to the outer circuit, 
 and this brush will form the positive pole of the machine. 
 But the total current and potential difference here, it will be 
 
10 DRUM ARMATURE WINDING. 
 
 observed, are what has been accumulated in the sections 
 represented by bars 3, 5 and 7 only ; and these form but half 
 the winding, not reckoning neutral sections. We have now 
 to account for the other half. Returning to bar 1 again, this, 
 with bar 16, forms a section about vertical, and consequently 
 neutral. The right-hand end of the latter bar (16), however, 
 is connected across to bar 15 on the further side, where the 
 current is in all cases running from right to left. Hence, 
 whereas one-half of the incoming current passed along the 
 length of bar 1 and so to bar 2, the other half will cross to 
 bar 16 at the commutator end, and thence into the series of 
 sections 1514, 13 12, 11 10, gathering potential on its 
 way, on to the right-hand end of bar 9, and so to the lower 
 brush, together ivith the current entering that bar at its 
 left-hand end. Thus the current, entering at the left end 
 of bar 1, has two paths in parallel which it may follow 
 throughout the two series of sections which alternate with one 
 another round the armature to reach the positive brush ; but 
 the potential accumulated in each of these series is the same. 
 Hence, as when two pipes of similar diameter delivering water 
 under the same head (or pressure) unite in one common pipe, 
 the pressure may remain the same, but does not increase ; 
 while the amount of water delivered, or the current, is 
 doubled : so the current delivered at the lower brush is 
 equal to the currents of both series of sections added together, 
 or twice that of one of them ; while the electromotive force is 
 not doubled, but remains only such as is accumulated in 
 either one of the series. Consequently, in calculating out a 
 drum winding it is requisite to obtain the total area enclosed 
 by the several sections of one series only, to ascertain the 
 electromotive force, while the section of the bars need be but 
 half that necessary to carry the whole current of the machine. 
 It will be observed here, in accordance with a remark we have 
 already made, how that, although there will be little or no 
 
GENERATION OF CURRENT AND POTENTIAL. 11 
 
 potential generated in sections 98, 9 10, yet the potential 
 .actually existing in these is the aggregate potential of the 
 series to which they each belong. It may appear paradoxical 
 that in the case of the second series the potential should be 
 rising on its way down the further side in bars 15, 13, 11, 
 and correspondingly up the front side, in a contrary direction 
 to that in which the armature revolves. But this is the way 
 the current goes in this series of sections, and it naturally 
 forms its " head," or point of highest potential, at the end 
 towards which it is tending. 
 
 We have thus far shown how the E.M.F. and current are 
 induced. There are, however, some further points in this con- 
 nection to be referred to. The first which may be mentioned 
 is one of no little importance. It will be remembered that 
 the long bars down the sides of the armature have an in- 
 creasing potential in their order from top to bottom ; con- 
 sequently and inversely, their respective short bars opposite 
 have a potential increasing from bottom to top. It follows 
 that in all cases where the sections are about horizontal, the 
 currents in their respective parallel bars on both sides are of 
 about the same potential ; but at the top the short bars of 
 the high potential come alternately with the long bars of low 
 potential, and vice versa at the bottom, the differences of 
 potential increasing gradually from the sides toward the top 
 and bottom, till at these latter points bars of maximum and 
 minimum potential are lying adjacent. This constitutes an 
 inherent evil of drum winding which cannot be eliminated. 
 Another point to which attention may be drawn concerns the 
 brushes. Should these be slipped round so as to come on the 
 segments of the commutator 7 and 15, instead of 9 and 1, it 
 might appear at first sight as though the current would be 
 reversed in the outer circuit. But it would not be so, as the 
 current would be reversed in bars 8 and 7 and 15 and 16 
 instead. In other words, the sections to which bars 7 and 15 
 
12 DRUM ARMATURE WINDING. 
 
 belong would for the time become neutral, and the current 
 continue to flow out at the lower brush, inasmuch as the 
 pressure or potential in 7 is higher than that in 15. But 
 as, according to the principle already enunciated, whereby 
 the highest potential must be at the bottom of the vertical 
 plane of commutation, the effect of moving the lower brush 
 on to 7 is simply to take it from a place of higher potential 
 to that of a lower, with the further effect of reducing the 
 potential of the outer circuit in proportion. 
 
 We have, when mentioning the plane of commutation, 
 alluded to this as only approximately vertical. It is in this 
 connection that the polarisation of the iron core of the drum 
 may now be considered. According to Ampere's well-known 
 rule, the direction of the current round the horizontal sections 
 of the winding will induce a north pole at the top of the core 
 and a south pole at the bottom. With regard to the sections 
 more nearly vertical, we may assume the plane of commutation 
 to be exactly vertical, and so the bars 1 and 9 in our illustra- 
 tion to be exactly at the top and bottom respectively. We 
 then observe that section 10 9 tends to produce a north 
 pole on the further side of the core, while section 15 16 
 tends to produce a north pole on the front side ; hence these 
 two tendencies counteract each other. Intermediate sections, 
 however, such as 11 10, 15 14, will tend to produce a 
 north pole, the one on the further side of the top and the 
 other on the nearer side. Hence the ultimate result will be a 
 broadening of the poles at the top and bottom. This polarity, 
 with poles in the plane of commutation, is known as the 
 "self excitation" of the armature. In respect of the field, 
 however, it is called " cross induction." But we shall enlarge 
 on this in a later chapter. The effect of the magnets between 
 whose pole-pieces the armature revolves is to induce poles 
 at the sides of the armature, a north pole on the further side 
 and a south pole on the front, at right angles to the self* 
 
GENERATION OF CURRENT AND POTENTIAL. 13 
 
 polarisation of the armature. Experience has shown that it 
 is desirable to magnify this effect, and to make the enclosing 
 magnets so powerful that they shall almost entirely overcome 
 the self -polarisation of the armature. Hence ultimately, by a 
 compromise, the north pole of the core becomes located a 
 little above the horizontal on the further side next the S 
 pole-piece, and the south pole of the core a little below the 
 horizontal on the front side opposed to the N pole-piece. 
 A plane joining these two poles will form a centre plane of 
 the lines of force of the field passing through the core ; and 
 the tilting of this plane represents what is termed the 
 4 'distortion of the field." The joint centre plane of the 
 ultimate polarisation of the core and the field being thus 
 a little tilted, the plane of commutation, which, as already 
 explained, must be normal to the field, becomes tilted also, 
 determining the lead of the brushes. That is, the plane of 
 commutation becomes tilted, not as a direct result of the 
 tilting of the self -excited polarity of the armature, but because 
 of the tilting or distortion of the field. 
 
 A remaining point for discussion in this present chapter 
 is the torque. By this is meant the tangential pull on 
 the armature tending to prevent its rotation, and requiring 
 power to overcome it. In the case of a motor it is the 
 turning moment by virtue of which it revolves, and performs 
 work. 
 
 Now this torque is the result of the attraction between the 
 self-excited poles of the armature and the poles of the field 
 magnets. For, apart from the polarity of the core caused by 
 the armature coils, there would be no torque. The polarity 
 of the core induced by the magnet poles has no tendency to 
 produce torque, as the strains thereby set up are in a line 
 with a diameter through the core ; and the magnet poles 
 have, in fact, no joint tendency to move, or prevent the 
 motion of, the armature in any manner, inasmuch as the 
 
14 DRUM AEMATUEE WINDING. 
 
 opposing poles attract it equally in opposite directions. It 
 may be said that the torque is an effect of the contest between 
 the polarity of the core induced by the magnets and that 
 caused by the armature coils, whereof the result is that they 
 amalgamate, as already intimated. And the extent to which, 
 in their amalgamation in any given case, the self-polarity of 
 the armature distorts the polarity induced by the magnets, is 
 the measure of the torque. 
 
 In different cases, however, with any given torque, the 
 distortion may be large in some and less in others. The 
 strength of the self-polarisation of the armature depends on 
 the total amount of current traversing its coils ; while the 
 strength of the polarity induced by the magnets depends, of 
 course, on the strength of the magnets. Hence, in one case 
 the armature self-excitation, due to a large amount of current 
 in its coils, may be stronger in proportion to the field than in 
 another where the total armature current may be less in pro- 
 portion ; and the distortion in the former will be more than 
 that in the latter, and the brushesjrequire more lead. In any 
 one case, however, the torque may be increased by increasing 
 the armature current, when the distortion will also become 
 greater ; or it may be increased by strengthening the mag- 
 nets, which will decrease the distortion. In the latter case 
 there will thus be the paradoxical result that with more torque 
 there will be less distortion. 
 
 But, though the armature revolves, its poles do not go 
 round with it. This may seem difficult to understand, and 
 may be explained as follows. It will be borne in mind that 
 the polarising effect of each coil is only as detailed above when 
 it is in the particular position referred to. Directly any one 
 coil has moved into the position previously occupied by 
 another coil, it has assumed all the characteristics which the 
 other coil had when in that position. The characteristics, in 
 fact, appertain to the position per se only being affected by 
 
GENERATION OF CURRENT AND POTENTIAL. 15 
 
 those laws which govern the nature of the current as a 
 whole ; and do not follow the coil in its rotation. Now this 
 being so in the case of individual coils, it follows that it is 
 so with all. Hence the polarising effect of the whole that 
 is, the armature poles remains stationary, though the arma- 
 ture itself rotates. The reason of the persistence and con- 
 tinuance of the torque will also now be the more apparent, 
 inasmuch as it is obvious that, if the poles of the armature 
 continue in opposition to the poles of the field magnets in the 
 manner already shown, the mutual attraction between them, 
 which forms the torque, must also remain. 
 
 It is necessary, however, to examine this subject yet more 
 closely. For, although disregarding the effects induced by 
 the field the iron of the core becomes polarised, the coils 
 themselves en masse have a resultant polarity, apart from the 
 existence of, though coincident with the polarity of, the iron 
 core. That is to say, if the iron were removed, the north and 
 south poles of the winding would remain just the same, in 
 mid air, as it were, though reduced in strength. Hence the 
 upshot is that, though the core will experience a torque on its 
 own account, due to its magnetisation, yet the coils also will 
 experience a torque tending to prevent their rotation, apart 
 from that on the core. As a consequence and this is the 
 important point whereas the torque on the core is met by 
 its being firmly keyed to the shaft, provision is also needed to 
 meet the torque on the coils ; and they must be so secured to 
 the periphery of the core that when the latter is revolved 
 they must go with it, without any possibility of circum- 
 ferential creeping. 
 
 This latter contingency, that is, the peripheral creeping 
 of the conductor bars, is by no means so unknown as might 
 be supposed. For especially in the case of motors, if the 
 conductors are held only at the ends of the armature, as, 
 for instance, between teeth on the end plates, they will bend 
 
16 DRUM ARMATURE WINDING. 
 
 in the middle. Binding wires wound round outside may 
 prevent their flying off ; but howsoever tightly they may be 
 bound on, the bars may none the less creep circumferentially 
 tinder them, unless mechanically stopped from so doing by 
 means of projections from the core. 
 
 Having thus far treated of the theory of drum winding, 
 we may in the next and subsequent chapters proceed with the 
 consideration of the various methods in which theory has 
 been put into practice. 
 
CHAPTER II. 
 
 WIEE DRUM WINDING: SIEMENS. 
 
 VAKIOUS methods have been adopted from time to time by 
 different inventors and makers to put drum armature winding 
 into practical shape. Among the many points for considera- 
 tion that occur, perhaps the most prominent is that involving 
 the question as to how to take the cross-connections at the end 
 of the armature round the spindle. It is thus of interest to 
 observe, for one thing, in what different ways this particular 
 difficulty has been surmounted, and " the end-winding," as it 
 is termed, arranged. It may be true that in the case of arma- 
 tures for small currents, and wound with flexible wire, this 
 difficulty may not be very great. But in dealing with heavy 
 currents and conductors of large section, it is different, and 
 many points of mechanical construction arise for considera- 
 tion. It is with these we shall be] mostly concerned in the 
 course of our remarks. We may state here, parenthetically, 
 that though some of the sketches that accompany this 
 chapter are taken from designs actually carried out, the 
 others will merely illustrate principles of design without 
 pretending to be actual representations of any particular 
 machines. Further, it may be added that we shall confine 
 ourselves almost exclusively to armatures for two -pole 
 machines. 
 
DRUM ARMATURE WINDING. 
 
 Dealing for the present with wire winding for light currents, 
 we meet with the most simple and crude of all armature 
 winding in Siemens' wire wrapping. This winding consists 
 fundamentally as shown in Fig. 6 ; and, supposing a single 
 turn per section, in merely carrying the wire from one 
 commutator segment along the periphery of the drum 
 parallel with the axis, across the further end, giving it a 
 slight bend to one side to get past the spindle, and so back 
 again on the other side, repassing the spindle, and then 
 joining it to the next commutator lug from whence it started, 
 or in irregular winding to some other lug. In practice, 
 
 FIG. 6. 
 
 however, the wire may be carried round many times before 
 being joined up to the lug of the next commutator segment. 
 -Each " section " of the winding is thus said to be composed 
 of so many "turns"; and while there may thus be any 
 number (within reason) of "turns per section," there are as 
 many "sections" as there are segments in the commutator, 
 neither more nor less. All the turns in a section may lie in 
 approximately the same plane, carefully insulated from each 
 other, and each section likewise from neighbouring sections, 
 and from the core. There may hence be in a winding of this 
 sort .a large total number of turns of the single wire ; and as 
 the volts depend for one factor on the aggregate of the areas 
 
WIRE WINDING. 19 
 
 enclosed severally within half the number of these single 
 coils or turns, it will be seen that this form of winding is 
 particularly suitable for small currents and a high pressure, 
 for a few amperes but many volts. 
 
 It will now become obvious that in a fully wound armature 
 of this sort, especially when the wires lie in two or more 
 layers on the periphery of the drum, there will be a great 
 accumulation of wire at each end, giving the ends of the 
 armature a more or less hemispherical shape, where the coils 
 overlie each other, as shown in Fig. 7. The current courses 
 round the armature in one section as many times as there 
 
 FIG. 7. 
 
 are turns in it, thence to a commutator segment, whence again, 
 supposing the segment not to be in contact with a brush, 
 to another section. Now, as each turn round the armature 
 means virtually the opportunity of acquiring potential, a 
 considerable difference of pressure accumulates between each 
 section as a whole, and so between each commutator segment, 
 rendering special care necessary as to their insulation. 
 
 But now, instead of each section of the winding being 
 composed of one long length of wire carried round many 
 times, the several turns being "in series" with one another, 
 let us suppose that each "turn" consists of a separate wire, 
 and the " section " a group of wires all carried round 
 
 c2 
 
DRUM ARMATURE WINDING. 
 
 together, with all their ends at one end of the section 
 connected to one commutator lug, and all the other ends to 
 another lug. The several " turns " would now be connected 
 " in parallel." The current, instead of being able only to flow 
 through the turns one after the other, would now flow through 
 them all at once as they lie side by side, and thus would go 
 only once round the core lengthways between any two seg- 
 ments of the commutator. The effect will hence be noted. 
 The aggregate of the areas enclosed within the several "coils," 
 as we may now term the sections, being considerably reduced, 
 other things being equal, the volts will be proportionately 
 reduced ; and as the current is able to flow through all the 
 wires in a coil at once, it meets with less resistance the 
 resistance depending for one thing on the sectional area of 
 the conductor and there will be a large current ; in a similar 
 manner as there would be a greater flow of water under a 
 given head through a large pipe than through a small one. 
 We thus find, in fact, that the conditions as compared with 
 the former case are reversed, and we have many amperes, 
 l>ut few volts. It will be necessary to provide a larger 
 commutator ; for while the number of segments remains 
 unaltered, these must be of a section sufficiently large to 
 carry the now heavy current, and of a length sufficient to 
 afford room for a suitable number of brushes to take it off. 
 
 But to return once more to the former case of many turns 
 per section, we may consider briefly some of the merits and 
 demerits of this particular form of winding apart from its 
 adaptability for heavy currents. As is well known, and fully 
 explained in Chapter I., the current in its course through the 
 armature winding pursues two paths from brush to brush ; 
 in fact, as it might, supposing it were possible to remove the 
 core, and pull the wire all out into one large circular loop 
 with the brushes on the opposite ends of a diameter, when 
 the current would be able to flow along both sides of the 
 
WIRE WINDING. 21 
 
 loop from one brush to the other. Now, with too great a 
 density of current, the conductors are liable to heat. Let 
 us consider this point. Take the case of two pipes conveying 
 water, one, say, 4in. diameter bore and the other 2in., and 
 let us imagine the same delivery through each. Now, as 
 the sectional areas of the two bores or, in other words, of 
 the fluid cylinders in each will vary with the square of 
 their diameters, it follows that the velocity of the water in 
 the small pipe will be 2 2 = 4 times the velocity in the large 
 pipe. Confining ourselves to slow velocities (as with high 
 velocities the analogy will not hold good, for hydraulic 
 reasons), and imagining a long length of piping, we find 
 that besides there being in the first place the necessity of four 
 times the head or pressure to produce the higher velocity in 
 the small pipe, there will be a still further increase of 
 head necessary in the case of the small pipe, on account of 
 the increased friction between the water and the internal skin 
 of the pipe due to the higher pressure. Thus the point to be 
 noted is, that in this small pipe, as compared with the large, 
 we find there is a waste of energy incurred due to this 
 friction. If this be carried to an extreme, and an attempt be 
 made to still further increase the delivery, the small pipe may 
 end by bursting. Now, returning to our electric conductors, 
 we find this analogy holds good. If it is attempted to force 
 more than a certain quantity of electricity per unit of time 
 through a given size of conductor, there will be similar waste 
 of energy ; only, in the case of the electric current, this 
 will show itself by the rise of temperature of the conductor ; 
 and this waste energy, in the form of heat, will be dissipated 
 by radiation or convection. As the heating of the conductor 
 will decrease its conductivity, there will likewise need to be 
 a corresponding increase of pressure or potential. Similarly, 
 also, if matters be carried to an extreme, and it be attempted 
 to make the wire carry an undue amount of current, the 
 
22 DKUM AKMATUEE WINDING. 
 
 wire will simply melt, or even burst, with a sudden increase 
 of current, owing to the inside of the wire melting before 
 the outside, and thus expanding quicker. 
 
 Now, an accident that is liable to happen to a dynamo is 
 what is known as a " short-circuit." The current from the 
 machine under ordinary circumstances has open to it a course 
 only through an outer circuit, of lamps and so forth, offering 
 resistance to its progress. If, by some mischance, however, 
 such as a wire or piece of metal falling across the positive and 
 negative terminals of the switchboard, connecting them elec- 
 trically, a much easier way of getting from the one terminal 
 to the other will be open to the current, instead of its having 
 to go through all the circuit with the lamps ; in other words, 
 the resistance of the outer circuit will be practically eliminated. 
 There will be a rush of current which the armature will not 
 be able to withstand. The armature winding will heat, and 
 may get red hot, and perhaps even melt. The insulation is 
 destroyed, and metallic connections thereby set up which are 
 fatal to the proper action of the machine. The result of a 
 bad short-circuit is, therefore, that the armature is rendered 
 useless, though often only one or two or a few of the sections 
 are burnt, the remainder being uninjured. 
 
 Kecalling our attention now once more to the wire-wrapped 
 drum armature, from which we may seem to have wandered, it 
 will at once be obvious that with the overlay of the wrapping 
 at the hemispherical ends, to renew a few burnt sections, 
 especially if these should be among the innermost, the whole 
 must be unwound, involving expense and trouble. This is thus 
 one of the principal drawbacks of this form of drum-winding, 
 though not altogether peculiar to it. Another drawback lies in 
 the fact that it is not possible, with this form of end winding 
 and so much crossing of the wires, to prevent the close 
 proximity of wires of high and low potential. Hence especial 
 care is necessary as to insulation. Again, the combined heat- 
 
WIRE WINDING. 23 
 
 ing of the wires at the ends where they lie in such a mass is 
 DO!J so easy to prevent, or to mitigate by ventilation, as when 
 arranged differently. It has its merits, however. Each 
 section of the winding, of howsoever many turns, is whole 
 throughout its length, without joints, which involve labour 
 and expense to make, and form possible sources of resistance. 
 It is in some respects easy of construction, dealing only with 
 flexible wire. The fact also of its allowing of many turns 
 per section as compared with most other windings, gives it 
 a distinct advantage for obtaining high volts out of a small 
 machine without the necessity of an inconveniently high 
 speed when a small current only is required. This point 
 may indeed, perhaps, be considered as its most valuable 
 feature. 
 
CHAPTER III. 
 
 HEAVY WINDING: DIFFICULTIES AND PRE- 
 LIMINARY CONSIDERATIONS. 
 
 WE may now, however, proceed to the consideration of arma- 
 tures wound not only for high volts, but also for a heavy cur- 
 rent. The chief difference between these and those we have 
 discussed is that the conductors must be of a much larger 
 section than that of mere wires, and will, in fact, become 
 " bars " rather than wires. Before going on to describe any 
 particular heavy current winding, however, it may be advan- 
 tageous, perhaps, first to consider some of the points that 
 have to be borne in mind when getting out the design of such 
 a winding. These naturally fall under the heads of efficiency, 
 economy, and practicality. The question of efficiency, how- 
 ever, in this connection, is mostly disposed of in the funda- 
 mental calculations of the machine. Granting, then, that the 
 sections of all the conductors are of sufficient size to carry the 
 intended current without undue heating, it is a matter of 
 contrivance so to arrange these on the armature that while 
 the greatest economy in labour and material is aimed at, 
 efficiency shall not be sacrificed. It thus becomes an object 
 in designing an end winding, while retaining the necessary 
 cross section, to aim, for economy's sake, at a minimum total 
 length of copper. Compactness is also required, so that the 
 
26 DRUM ARMATURE WINDING. 
 
 winding may project the smallest amount possible beyond the 
 ends of the core. As these parts of the winding generate no 
 current, there is no object in their enlargement ; while com- 
 pactness assists not only economy in the copper itself, but also, 
 by keeping down the total length of the armature, the total 
 length of the machine as a whole is kept down, together with 
 its weight, first cost, and floor space, minimum floor space 
 being frequently a desideratum, especially on board ship. But 
 while aiming at compactness, ventilation must not be for- 
 gotten, especially in the case of large machines, though there 
 are makers who do not attach the importance to this point 
 that others, probably the majority, attach to it. Again, it is 
 well to have as few joints as possible in the coils; as these, 
 depending on the human factor for their perfection, may be 
 faulty, and so cause resistance. The extra labour and expense 
 incurred in [making them must, moreover, be remembered. 
 An armature should not be too complicated, so that its con- 
 struction may be comparatively easy and straightforward ; and 
 all joints should be accessible. Labour is thus economised ; 
 besides, by avoiding work that is at all " fidgety," dependence 
 on a few special men who may acquire the particular aptitude 
 necessary, and whose services may be lost in a strike or other- 
 wise, is avoided. With this last desideratum is in some 
 degree involved another point of importance, which is, that 
 parts burnt or otherwise damaged may either be removed and 
 replaced with facility, or else that the whole winding should 
 be capable of being easily dismounted and put together again 
 for the same purpose. The conductors, as explained in Chap. I. 
 (pp. 15, 16), require to be well held on by steel wire wound 
 round in bands outside, circumferential slipping being still 
 further prevented by projections from the core and end plates. 
 For it will be borne in mind, that by virtue of the work done 
 in driving the armature, the parallel conductor bars are 
 subjected to a magnetic drag or torque tending to hold them 
 
HEAVY WINDING. 
 
 27 
 
 back from revolving with the core and spindle ; and this drag 
 is no light matter : hence the necessity of fixing the coils well. 
 To this magnetic drag is added the centrifugal force tending 
 to make the coils fly off altogether. 
 
 There are, however, still some two or three other points 
 that require attention. Of these, one is the question of a true 
 balance. It is necessary to put on the winding and form up 
 the ends with such symmetry that the centre of gravity of the 
 whole armature when complete shall be in the axis of the 
 spindle, and not outside. Now, in ordinary practice, the 
 armature, when finished, is tested for balance by being rolled 
 with its spindle at each end resting on two horizontal " knife- 
 edges." If the centre of gravity of the whole be in the axis, 
 it will rest hi any position ; but if not in the axis, the armature 
 will then roll, oscillating backwards and forwards till it rests 
 with the heavy side downwards, when extra weight must be 
 soldered or otherwise be attached to the light side till a balance 
 be attained. But it has to be noted that this is not really a 
 true test. It is a question whether or not armatures are 
 sometimes made having the centres of gravity of their two 
 ends, taken separately, not in the axis, though equi-distant on 
 opposite sides of it. A fault of this nature would not show on 
 the knife-edges. But when revolving rapidly, there would be 
 a tendency in an armature thus faulty for its axis to gyrate 
 round the true axis of revolution in two cones, one at each 
 end, having their apices united at a point within the true 
 axis somewhere between the two misplaced centres of gravity. 
 Hence a greater or less tendency to vibrate will still exist, 
 though the armature might show a perfect balance on the 
 knife-edges. This kind of error might, perhaps, be detected 
 by mounting the armature within a magnetic field, on bearings 
 supported on flexible standards, or, better still, on bearings 
 slung from adjustable springs, and then, with brushes lightly 
 touching the commutators, revolve it as a motor. If the 
 
28 DRUM ARMATURE WINDING. 
 
 balance be true, with the centre of gravity of each end within 
 the axis, the armature, when rotating at a tolerable speed, 
 would evince a tendency to maintain a constant and steady 
 axis in any one position, like a gyroscope. But if the balance 
 be out of truth in the particular manner under discussion, a 
 decided desire to " wobble " would show itself. 
 
 Another reason, however, for being careful as to symmetry, 
 is to secure the perfect roundness of the exterior circumference 
 of the completed winding, with its centre within the shaft axis. 
 The clearance between the armature and pole-pieces may thus 
 be kept to a minimum. Further, by attaining exact similarity 
 with all the coils, in length, section, shape, and position, their 
 resistance and self-induction will also be similar, tending to a 
 steady current and an absence of sparking. 
 
 A remaining, and very important point, however, that 
 arises, especially in connection with heavy-current windings, 
 is the liability of the conductor bars to heat by reason of 
 "Foucault" or " eddy " currents generated within them in- 
 dividually, apart from the main current. These we may ex- 
 plain : In some kinds of dynamos other than those with drum 
 or Gramme armature such, for instance, as those used for 
 alternating currents the coils, instead of being revolved in a 
 field, are translated across a field edgeways. In other words, 
 the coil is so placed that its plane is perpendicular to the lines 
 of force of the field ; being still kept in this plane, it is moved 
 across the lines of force, and so cuts them. A current is thus 
 set up in the coil. Now, if instead of a flat coil, a flat metal 
 plate were used, and were made to cut through the field edge- 
 ways in a similar manner, a current would be caused to circu- 
 late in the plate in a plane perpendicular to the lines of force, 
 this, indeed, being much the principle of Faraday's first disc 
 dynamo. If the plate were turned so that its plane were 
 parallel with the lines of magnetic induction, and then moved 
 through them, no current, comparatively speaking, would be 
 
HEAVY WINDING. 29 
 
 generated within it. This point is important. Now if, again, 
 instead of the plate, a single metal bar be moved sideways 
 across the field in a plane normal to the lines of induction, 
 a current will arise within the bar; and this current will 
 circulate within the bar in planes perpendicular to the lines of 
 force, running along one of the sides parallel with these lines 
 and returning by the other. Thus it is obvious that as a current 
 will produce heat, if the experiment with either the coil, the 
 plate, or the bar, be carried out to any great extent, and they 
 be moved rapidly and continuously through a sufficiently 
 strong field, they may be made red hot ; and the fact remains 
 that if a single insulated bar be laid along the periphery of an 
 armature parallel with its axis of rotation, or held in a frame, 
 and then revolved in an ordinary magnetic field, such as used 
 for drum or Gramme dynamos, it will get hot from these 
 internal currents alone. 
 
 But in dealing with large " bars," however, there may also 
 be borne in mind their capability of cooling as compared 
 with wire, or bars of small section. It will be noted that 
 whereas the transverse sectional area of a bar increases with 
 the square of its diameter if round, or as the square of the side 
 if of square section, the circumference or outer surface in- 
 creases only in proportion to its diameter or thickness. 
 Now, capability of cooling, whether by radiation or convection, 
 varies with the surface. Hence, a bar, say six times the 
 thickness of a small bar or wire of round or square section, 
 while it will have 6 2 = 36 times the sectional area of the 
 smaller, and so may carry about 36 times the current, has only 
 six times the cooling surface. Hence, large bars as compared 
 with small ones, besides being heated by the main current and 
 by eddy currents, are still further handicapped, so to speak, by 
 their greater difficulty in cooling. The eddy currents in wires, 
 it may be mentioned, are so trifling as to be negligible. In 
 ordinary drum or Gramme- wound dynamos, where the field 
 
30 DRUM ARMATURE WINDING. 
 
 is bored concentric with the axis of the armature, these 
 Foucault currents arise principally, if not entirely, when the 
 bars pass under the trailing horns of the pole-pieces, where 
 the induction lines are particularly dense. By " trailing " 
 horn is meant the last horn of a pole-piece which the 
 bars leave or recede from as they revolve. The prevention 
 of these internal eddy currents will form part of the subject 
 of our next chapter. 
 
CHAPTER IV. 
 
 HEAVY WINDING : PREVENTION OF FOUCAULT 
 CURRENTS, AND VENTILATION. 
 
 THE methods for the prevention of Foucault currents may 
 be divided under two heads. Under the first, the difficulty is 
 met by removing the principal cause of the evil : under the- 
 second, the cause remains, but its effects are prevented from 
 occurring by some counter-arrangement. 
 
 The plan adopted by Messrs. Mather and Platt comes under 
 the first head. They avoid the difficulty by making the air 
 space between the armature and the pole-pieces greater as it 
 nears the horns either way, as shown in Fig. 8. Thus the 
 principal cause of the heating is removed, inasmuch as the 
 magnetic induction does not get so intense at the horns as it 
 is wont to do when the air space is of even thickness. 
 
 But it is under the second head that we may look for the 
 most variety. The eddy currents may be prevented by so 
 making or arranging the bars that a current cannot circulate 
 within them. One method, instead of making the bar of one 
 solid piece, is to compound it of several insulated wires or 
 strips, in such manner that there shall be no internal circuit 
 open to the current. Thus, Messrs. Siemens Brothers make 
 their conductor bars of stranded copper wire cable, each wire 
 separately insulated by varnish, and the whole brought into- 
 
.32 DRUM ARMATURE WINDING. 
 
 a rectangular section by hydraulic pressure, or by drawing 
 through a die, or otherwise. As the Foucault currents will 
 only circulate in an unbroken circuit lying in a plane normal 
 to the lines of force, as already explained, it will be seen that 
 no such circuit is open in the cable arrangement : conse- 
 quently, the currents are not set up. 
 
 Another method, somewhat similar, is to make the bar of 
 insulated wires laid parallel, and soldered together at the ends 
 so as to form a compound bar of rectangular section, to give 
 the bar a half twist in the middle, and then to press or 
 
 FIG. 8. 
 
 hammer the twisted portion in an iron mould, so as to bring 
 it to the same section as, and alignment with, the untwisted 
 portions of the length. 
 
 Yet another method is to make the bar of insulated copper 
 strips or tape laid edgeways on the core in radial planes, and 
 soldered together at the ends. The strips, thus laid, being in 
 planes, for the most part approximately parallel to the lines 
 of force they cut, will not, as we have already pointed out, be 
 individually appreciably liable to eddy currents. But as they 
 are in metallic connection at the ends, an internal circuit is 
 open along the side strips and across both ends, which must 
 be broken. Hence, this kind of bar also needs a half twist in 
 
PREVENTION OF FOUCAULT CURRENTS. 
 
 33 
 
 the middle, and to be hammered in a mould, like the last. 
 By means of the half twist, the current coursing along one 
 side of the bar in each separate wire or strip is at the twist 
 conducted to the other side, where it meets an opposing 
 current, and so both are neutralised. It will, of course, be 
 observed with regard to the strips, that if these be laid flat on 
 
 Fm. 9. 
 
 FIG. 10. 
 
 FIG. 11. 
 
 FIG. 12. 
 
 the periphery of the drum, eddy currents could set up in each 
 one separately. 
 
 Further methods, still similar in principle, are some 
 patented by Mr. E. E. Crompton. These are illustrated, 
 Figs. 9 to 12. The conductor bar here is simply made double ; 
 but at the middle of its length each half is so bent that it con- 
 tinues on the centre line of, and in alignment with, the other 
 
34 DRUM AEMATURE WINDING. 
 
 half. A groove is turned in the core to receive the half loop. 
 (Fig. 9.) 
 
 As will be observed, the circulation of a current within the 
 double bar is stopped by the tendencies in each half being 
 opposed at the middle crossing, similarly as in those already 
 described. 
 
 Foucault currents may, however, be stopped by another 
 method, included in one of Mr. James Swinburne's patents, 
 by which any twist in compound bars is rendered unnecessary. 
 This applies in cases where it is possible to connect up the 
 component parts of two bars of a coil, separately, and in such 
 a manner that the tendency to these currents in each bar shall 
 oppose that in the other. This is achieved in the manner 
 indicated diagrammatically in Fig. 13. The pulley end of the 
 armature is here shown, and one coil, each bar of which is 
 composed of four strips laid edgeways. The strips of the one 
 bar must then be connected to those in the other bar, 
 separately, in the manner represented. The coil is, in fact, 
 split into so many parallel planes. 
 
 To understand this, we may imagine the coil to be revolv- 
 ing in the field in the direction indicated by the curved arrow. 
 The upper pole-piece is north, and the lower south. The 
 magnetic lines of force will be thus running from N to S, 
 and will be most dense at the trailing horns, a a. As rotation 
 takes place it will be seen that both bars are approaching 
 dense portions of the field, and the lines of force are running 
 through both bars in the same direction that is, toward S. 
 Hence, as viewed from some point in the same plane as the 
 coil, on one side, but outside the armature, the eddy current 
 would circulate in each bar in the same direction ; or, as seen 
 from S, in the same direction as the hands of a watch. Thus, 
 at the sides w and x of the two bars, the current would 
 approach; while at the sides y and z it would recede, as 
 indicated by the small arrows. The opposition of the two 
 
PREVENTION OF FOUCAULT CURRENTS. 
 
 35 
 
 tendencies is at once obvious. The union of the strips in 
 each bar at the commutator is, of course, of no consequence. 
 
 The explanation can also be put in terms somewhat as 
 follows : Whereas the volts of w z may be greater or less than 
 y x, the sum of the volts of y z will about equal the sum of 
 the volts of w .y. Hence no current will flow. 
 
 In all cases, it may be mentioned, the bars, whether solid 
 or compound, may be wound with cotton or silk tape and var- 
 nish for insulation. Solid bars are sometimes left bare, in- 
 sulated from one another by air gaps, and from the core by 
 
 Willesden or Manilla paper, or other material, and varnish. 
 They will, of course, be more liable to damage when left un- 
 covered ; but with compound bars the tape winding is, for the 
 most part, necessary to maintain their shape. The varnish 
 used is commonly shellac, though some makers use a 
 speciality. 
 
 The heating of the bars should be still further obviated 
 by efficient ventilation, a point we have already mentioned. 
 The cores of large machines are usually made hollow. 
 When revolving, the long lugs of the commutator, if there 
 
 D2 
 
36 DRUM ARMATURE WINDING. 
 
 are any, act as centrifugal fan blades, and draw air 
 through the hollow between the interior of the core and 
 the spindle, and so assist to keep down the temperature. 
 Some makers leave gaps in the core along its length, so 
 that air from the interior may fly outwards, and help to keep 
 cool, not only the core, but also the conductor bars on the 
 periphery, and the pole-pieces. The bars will also draw air 
 round after them, and so ventilate themselves. One of the 
 advantages of the Hopkinson dynamo (Fig. 8) is thus here 
 illustrated, since, owing to the fact that the horns stand away 
 from the armature, air is more easily drawn down between it. 
 and the pole-pieces, and thus can be more effective in keeping 
 the bars cool than when throttled and wire-drawn, as it were,, 
 by the closeness of the horns to the armature. The whole 
 question of ventilation is not one to be passed over as incon- 
 siderable. A phenomenon showing the effect of thorough 
 ventilation may be observed when a dynamo stops after a 
 run. The end winding, owing to its complete exposure and 
 revolving in a thorough draught, may feel cool to the hand at 
 first ; but presently it will become warm, owing to the heat 
 from the interior portions of the winding and armature creep- 
 ing out. This shows at once how great a difference ventila- 
 tion makes. It may be remarked that not only does heat 
 tend to generate in the conductors, but also in the iron core 
 and the pole-pieces, especially in the trailing horns of the 
 latter. Hence, when it is further borne in mind that a high 
 temperature reduces the conductivity of the copper, the desir- 
 ability of good ventilation becomes still more obvious. In 
 fact, other things being equal, a greater output is obtainable 
 from a well-ventilated machine than from one not ventilated. 
 When a machine is run very hot, energy is wasted in the 
 dissipation of the heat, as already pointed out in an earlier 
 passage. Some makers supply auxiliary fans, or air pro- 
 pellers, especially for the purpose of creating a draught 
 
VENTILATION. 37 
 
 through a machine as it runs. But, on the other hand, 
 though thorough ventilation may keep the bars cool, there is 
 yet, perhaps, some degree of falsity in the theory of this 
 method of procedure. Heat is abstracted, and so energy lost, 
 just as when the bars are allowed to heat though it may 
 be not so much owing to the lesser resistance of the cool 
 bars beside the loss of power involved in driving a fan when 
 one is used. Only by good ventilation the heat is drawn off 
 more quickly, and the bars are kept cool, and at their normal 
 conductivity. Hence, as already intimated, the size and 
 therefore the first cost of a well-ventilated machine for a given 
 output should be less than for one not ventilated. In some 
 machines ventilation is further assisted by the magnets being 
 formed of bars, with interstices between them, through which 
 air may pass in and out of the field. A practical limit, how- 
 ever, to heating is the ability of the various materials used for 
 insulation to endure a high temperature without perishing or 
 losing their insulating qualities. To this limit no doubt 
 machines are sometimes run. 
 
CHAPTER V. 
 
 HEAVY WINDING. PLATE END CONNECTION. 
 EDISON. 
 
 WE may now give our attention to some of the different 
 methods in which heavy-current windings have been, and are, 
 constructed, having special regard to the arrangement of the 
 cross connections at the ends of the armature. 
 
 It will be noted that a bar being the reverse of flexible, it can- 
 not be dealt with as a mere wire. To have to bend it would 
 be a matter of considerable trouble. Hence special means have 
 to be adopted at the drum ends. These consist principally, 
 though not in all cases, of separate lengths of metal. These 
 are used to form a connection between the ends of two 
 opposing bars forming the two sides of a coil, and are united 
 thereto by riveting, screwing, or soldering the latter gene- 
 rally together with either of the two former. As we have 
 already pointed out, in armatures for heavy currents there are 
 generally only one, or, at the most, but two turns per section. 
 If high volts are required, the machine must be made larger, 
 or the field stronger, or the speed increased, or all three. 
 Matters are thus much simplified. It is, of course, not 
 mechanically impossible to have more turns per section. But 
 the extra complications involved, combined with many other 
 disadvantages, render the idea in ordinary cases inadmissible. 
 
 Among the earliest of these methods brought out is one 
 patented by Edison. This is illustrated in Fig. 14, which 
 
40 DRUM AEMATUEE WINDING. 
 
 shows an end view of an armature. The rectangular black 
 dots a a are the parallel conductor bar ends. Each of these 
 is let into and jointed to the periphery of a flat radial 
 insulated metal plate, as shown. The radial plates of two bar 
 ends desired to be connected up, such as c and d, are then 
 united by a circular bar, as shown, jointed to each, but insu- 
 lated from the other bars and plates. An electrical connection 
 is thus carried round from one side to the other. It will be 
 understood, in accordance with some of our preliminary 
 remarks, that this sketch merely represents the principles of 
 
 mechanical construction involved, and does not necessarily 
 represent any one actual design. It is probable that the bars 
 in Edison's early machines were connected up in the manner, 
 though not in the order, here shown. The parallel conductors 
 here are bare, and, it will be observed, are a long way apart. 
 This method is practicable indeed for a winding of a few 
 sections only. Though each end of the armature is thus con- 
 nected up, yet at the commutator end, which is here given, 
 alternate radial plates are projected further inwards than the 
 remainder, as shown in solid black at b b. These projections 
 
PLATE END CONNECTORS. 
 
 41 
 
 are each united to a commutator segment. Besides the draw- 
 back already intimated, this end winding has the further fault 
 of not being symmetrical. Hence, a perfect balance and 
 equality of resistance in the coils become often impossible. 
 Some new arrangement thus becomes desirable. 
 
 In Figs. 15 and 16 we find the method adopted by Edison 
 of placing insulated metal discs a a at each end of the 
 core, and of the same number at each end as there are 
 segments in the commutator, and of the same outside 
 diameter as the core. A few of the parallel bars only 
 
 FIG. 15. 
 
 are shown. The two bars of a coil would thus at the 
 pulley end be connected to the approximately opposite edges 
 of the same insulated disc, while at the commutator end 
 they would be jointed to separate discs. Fig. 16 shows 
 one of these discs. A hole through the middle serves for 
 the passage of the spindle and for ventilation, and a tongue b 
 shown across it, cut through at one end, is bent outwards 
 parallel with the spindle for connection with a commutator 
 segment. The four smaller holes c c are for the passage 
 of insulated screws running parallel with the spindle by 
 
42 
 
 DRUM ARMATURE WINDING. 
 
 Which the whole set of discs at one end are bolted to the 
 end plate of the core. 
 
 This arrangement has the advantage of being compact, 
 and has all the joints on the outside and accessible. Hence 
 any burnt or damaged bars can easily be renewed singly 
 without disturbing the remainder. By making the discs 
 rather stout in proportion to the current they have to 
 carry, their resistance can be kept low, and the tendency 
 to heat obviated, rendering their ventilation not so necessary. 
 But they are liable to the disadvantage of having a maximum 
 
 FIG. 16. 
 
 difference of potential between any two that are adjacent; 
 and will thus require more care as to their insulation from 
 each other. A drawback, moreover, to using round discs 
 is that these have to be stamped or cut out from sheet 
 metal, involving waste of metal, besides extra cost in labour 
 and use of machinery. It is a point as to whether or not 
 this cost is counterbalanced by subsequent facility of 
 construction. It will be noted that the bars are straight, 
 and have no bends ; but in both designs, as shown hi 
 Figs. 14 and 15, it will be observed that these are solid, 
 and are liable to heat with Foucault currents. 
 
CHAPTER VI. 
 
 HEAVY WINDING. EVOLUTE END CONNECTIONS. 
 
 CROMPTON AND SWINBURNE. ANDERSEN. 
 
 HOPKINSON. 
 
 THE next example we may take is Messrs. Crompton 
 and Swinburne's evolute end winding. This, as illustrated 
 in Figs. 17 to 22, will be seen to differ entirely from 
 Edison's. Fig. 17 represents an end view of the armature, 
 
 FIG. 17. Crompton- Swinburne Winding. End View of Armature. 
 
 Fig. 18 a longitudinal section, and Fig. 19 an outside 
 elevation, all diagrammatic, and with a very open winding 
 for the sake of clearness. It will be noted that alternate 
 parallel bars run the full length of the armature, and are 
 
44 
 
 DRUM ARMATURE WINDING. 
 
 united to the commutator lugs at one end. The remaining 
 bars are short, at least as to their straight portions, and we 
 will for convenience call them so. A "long" bar and a 
 
 FIG. 18. Crompton- Swinburne Winding. Longitudinal Section. 
 
 " short " one opposite thus form the two sides of a coil. In 
 Fig. 18 it will be seen that the short bars are cranked inwards 
 towards the shaft, as shown at a a. At b b are sections 
 
 FIG. 19. Crompton-Swinburne Winding. Outside Elevation. 
 
 through the evolute connections formed of thin pliable 
 copper strips, indicated more clearly in the end view, Fig. 17. 
 As may be seen, each of these, jointed at its interior end 
 
EVOLUTE END CONNECTORS. 45 
 
 to the crank of a short bar, is carried round in an evolute 
 path till it reaches the desired long bar on the approximately 
 opposite side, and there is jointed to it. It will be noted 
 how the evolute connectors lie neatly within each other, 
 though in practice they fit closely, separated only by mica or 
 other insulation. 
 
 In the arrangement, however, as here shown, the cranked 
 conductors, if burnt, cannot be removed without dismounting 
 the whole winding, or a great part of it. Hence the 
 modification shown (Fig. 20), where the crank is separately 
 jointed to the turned-up end of the " short " bar, as indicated 
 at/. By unsoldering these joints/, or the joints g between 
 the "long" bars and the evolute connectors, any single 
 
 FIG. 20. Crompton-Swiuburne Winding. 
 
 parallel bar can be removed. Further, by withdrawing the 
 spindle (which can be done if the core is carried on a spider, 
 or otherwise self-contained apart from the spindle), the 
 loose crank piece can be drawn in radially, its joint with 
 the interior end of its connector unmade, when both can 
 be removed singly. Thus any one, or a few sections, can 
 be replaced without dismounting the remainder. The total 
 number of joints is, however, increased. It will be noted 
 that these bars are here all solid, and when single can 
 only be used for small currents. For large currents the 
 bars are made double, as illustrated, Figs. 9 to 12; or 
 
46 
 
 DRUM ARMATURE WINDING. 
 
 else are compounded of twisted wires. In Fig. 21 is shown a 
 modification to gain more room for the evolute connectors, 
 the long bars being slightly cranked outwards. 
 
 FIG. 21. Croinpton-Swinburne Winding. 
 
 Thus far we have dealt with but a single turn per section. 
 In Fig. 22 is shown the arrangement for two turns per 
 
 FIG. 22. Crompton-Swinburne Winding. 
 
 section. It will be seen that there are two parallel series 
 of evolutes ; also, the cranks are alternately long and short 
 to suit. In this case the connectors may be narrower and 
 
EVOLUTE END CONNECTOKS. 
 
 47 
 
 thicker than in the former arrangement ; thus in the total 
 mass there will be more copper and less insulation than 
 when the curved plates are thin ; and so space is economised. 
 The spindle within the core is either of triangular or square 
 section, with slightly rounded angles. The core is carried 
 tight on the angles, its interior bore being grooved to fit 
 these, and ventilation is thus allowed for along the flats. 
 Sometimes a spider is used to carry the core, instead of 
 its being built on the shaft direct. 
 
 CL. 
 
 FIG. 23. Andersen Winding. Longitudinal Section through Armature. 
 
 Further developments of this form of evolute end winding 
 are to be found in the " Westminster " dynamos of Messrs. 
 Latimer Clark, Muirhead and Co. This firm, under the 
 supervision of their electrical manager, Mr. F. V. Andersen, 
 have employed conductor bars built up of either wire or 
 tape, and sometimes both kinds on one armature. Fig. 23 
 shows a longitudinal section through an armature of this 
 latter description. The long bars a a are of copper strips laid 
 edgeways, and each strip insulated with varnish. The short 
 bars b b are of wire laid parallel, each wire insulated with 
 
48 
 
 DKUM ARMATURE WINDING. 
 
 shellac varnish, and the whole group cranked at both ends. 
 An end thus cranked is given to a larger scale in Figs. 24 
 and 25. The cranked end is splayed out, so that the wires 
 embrace the inner end of an evolute strip c ; and are the 
 more carefully spread over the end of the strip, that the 
 soldered joint may be as large as possible, and so offer a 
 minimum resistance. A transverse section through the 
 conductor bars on the periphery of the drum is given 
 (Fig. 26), showing their compound formation alternately of 
 copper strip and wire. Each bar is insulated with varnished 
 
 FIGS. 24, 25. Andersen Winding. Cranked Ends. 
 
 tape. The layers a a are of Willesden or Manilla paper, 
 varnished, and b is the usual exterior band of steel wire. 
 The evolutes, duly insulated from each other, run from the 
 inner cranks outwards to the long copper strip bars opposite, 
 as in the case of the Crompton- Swinburne winding, shown 
 in Fig. 17. The conductor bars are here jointed at both 
 ends to other metal, and the several wires or strips in each 
 are consequently in metallic contact at each end. They need 
 to have a half-twist given them in the middle, in the 
 manner already referred to when discussing the subject of 
 
E VOLUTE END CONNECTORS, 
 
 Foucault currents. In Fig. 23, h h are fan blades for drawing 
 
 through the core by airways k k, of which there are three. 
 
 In subsequent machines, however, these blades have been 
 
 omitted, the suction caused by the commutator lugs being 
 
 FIG. 26. Andersen Winding. Transverse Section through the Conductor 
 Bars on the Periphery of the Drum. 
 
 considered sufficient. The winding is held from slipping 
 by a crown of teeth m m on each end plate, between which 
 the cranks lie. A burnt section, however, cannot be taken 
 out without dismounting at least half of the winding. 
 
 IOI 
 I I 
 
 FIG. 27. Andersen Winding. Cranks. 
 
 This firm has made machines^with the conductors formed 
 of copper tape only. As these are necessarily laid edgeways, 
 and will not consequently easily bend, the cranks have to be 
 formed of pieces soldered and rivetted together, in the manner 
 shown in Figs. 27 and 28, the latter being a sectional plan 
 
50 DRUM ARMATURE WINDING. 
 
 on a b in the former to twice the scale. For the greater 
 flexibility of the evolutes c these are here made of two thin 
 strips of copper, instead of one, as is shown clearly in 
 Fig. 28. Although this is of somewhat more simple con- 
 struction than the wire -made cranked bars, inasmuch as 
 these latter have to be carefully formed up in a mould, yet 
 two extra joints are introduced. 
 
 It will be observed that these evolute windings so far 
 represented have one common advantage. They are made 
 entirely of rolled copper in one form or another. Hence^ 
 all desired lengths need but to be cut off, with a minimum 
 of waste, both in material, labour, and use of machinery. 
 
 
 FIG. 28. Andersen Winding. Sectional Plan on a b of Fig. 27. 
 
 On the other hand, comparatively speaking, the subsequent, 
 construction of the end winding is not exactly simple. In 
 one method of procedure, the short bars have first to be 
 made up separately, with cranked ends, and with the 
 connectors, as yet straight, duly attached to each crank, 
 and lying in the same plane with it. These, thus far 
 individually complete, are then laid in position on the 
 core, and temporarily bound on, with the connectors pro- 
 jecting radially all round. By roping and so forth, all 
 the connectors at one end are now forced round in a com- 
 bined mass, each one naturally assuming an evolute curve, 
 until the outer end of each reaches the opposite side of the 
 core end to that from whence it started, as shown in Fig. 16. 
 Being held thus, and the connectors at the other end of the 
 armature having been treated similarly, the long bars are 
 finally added, and severally jointed up to their respective 
 
EVOLUTE END CONNECTORS. 
 
 51 
 
 connectors. Hence, by the magnitude of the operation as a 
 whole, it will be seen that the first-named advantage in the 
 use of rolled copper is in some measure lost. One of the 
 principal points, however, in this winding, as compared with 
 Edison's discs, is, that there is a minimum electric pressure 
 between the evolutes. 
 
 The evolute end windings we have thus far discussed, it- 
 may now be noted, have the peculiarity that each evolute 
 from beginning to end is approximately half a circle. In 
 
 ^A. 
 
 FIG. 29. Hopkinson Winding. Longitudinal Section through an 
 Armature. 
 
 another class of evolute winding, however, each evolute is 
 approximately only a quarter circle, and the crank system is 
 absent. This design was first brought out by Herr Hefner 
 von Alteneck, and is that adopted by Drs. J. and E. Hop- 
 kinson in the manner represented in Figs. 29 to 32. Fig. 29 
 is a longitudinal section through an armature wound with one 
 turn per section. Fig. 30 is an end view, but showing only 
 a portion of the winding for the sake of clearness. The bars, 
 it will be noted, are solid, the heating being obviated in the 
 manner already described, by enlarging the air space under 
 
 E2 
 
52 DRUM ARMATURE WINDING. 
 
 the horns of the pole-pieces (Fig. 8). The parallel conductors 
 in this case also are short and long. But instead of the short 
 ones being cranked at their ends, they are jointed to evolute 
 strips as well as the long bars. There are thus two series of 
 evolute strips side by side, as shown at a and b in the illus- 
 trations. The evolutes in the one, however, run round in the 
 opposite direction to those in the other, as shown in Fig. 30, 
 and only cover an angle of approximately 90 degrees, in- 
 stead of 180 degrees. On each commutator segment a bar 
 c is screwed, which projects within the end winding in a 
 direction parallel with the shaft. To this the interior end of 
 
 FIG, 30. Hopkinson Winding. End View of Armature. 
 
 one connector in each series is screwed or otherwise jointed, 
 thus connecting a long bar on the one side to a short one 
 opposite, as, for instance, d and e, or / and g, as shown in 
 Fig. 30. Figs. 31 and 32 show this arrangement duplicated 
 for a winding of two turns per section, when there are four 
 series of evolutes at each end. 
 
 The driving of the winding is effected by means of snugs 
 formed in some or all of the bars themselves, which fit into 
 recesses in the peripheries of the end plates and a middle disc, 
 as shown at h h in Figs. 29 and 31. At the pulley end of the 
 armature, there being no commutator bar for the strips to 
 
EVOLUTE END CONNECTORS. 
 
 53 
 
 be fastened to, an annular block of insulating material k 
 (Fig. 29) is run on the shaft, and on this bars i i, parallel 
 with the shaft, are screwed for the purpose. 
 
 It will be noted that with this method of two or four series 
 of evolutes, all the parallel bars are straight, without bends, 
 
 FIG. 31. Hopkinson Winding. Two Turns per Section. 
 
 and their joints being all external and accessible any can be 
 removed separately. This latter advantage, however, does not 
 
 FIG. 32. Hopkinson Winding. Two Turns per Section. 
 
 appear to apply to the evolutes, should any of these require 
 renewal. A minimum pressure exists between adjacent 
 evolutes in any one series, as with the method of Messrs. 
 Crompton and Swinburne. Likewise, also, the winding is 
 
DKUM AEMATURE WINDING. 
 
 symmetrical, tending to a good balance, and the separate 
 sections are all the same length round, and consequently of 
 the same resistance. This winding is also constructed of 
 rolled copper strips. 
 
 Instead of making these cross connectors in three pieces 
 that is, of two evolute strips united at their inner ends to one 
 bar, as at c or i, Fig. 29 a simpler plan now in vogue is to 
 
 FIG. 33. 
 
 FIG. 34. 
 
 make them each in one piece. This is illustrated above. A 
 single strip of copper is first folded as shown in Fig. 33. One 
 member b is then bent over into the dotted position ft', so 
 that the whole appears in plan as represented in Fig 34. 
 The two parts a and b will thus form the evolutes as depicted 
 in Fig. 35 ; while the position c becomes radial. 
 
CHAPTER VII. 
 
 EVOLUTE WIRE-WINDING: EICKEMEYER. 
 
 WE now come to another winding, the end connections of 
 ivhich are designed on the principle of a double series of 
 evolutes, running in opposite directions, similarly to the last 
 described. This is Eickemeyer's, of New York. Although 
 like in principle to the Hefner von Alteneck method, it is in 
 other respects altogether different. Each section of coil is 
 composed of wires laid together, and contains no joint in it 
 from one end at one commutator lug to the other end at the 
 next lug. Each section, moreover, is entirely self-contained, 
 and is formed to the right shape and curvature on a mould, or 
 in part on a mould, before being put on to the armature. 
 Thus the coils on any armature are exactly similar, and con- 
 sequently interchangeable. Here, as with the Siemens' wire 
 winding, a high-pressure small current can be obtained by 
 making each section of one wire carried round several times ; 
 or the wires can be united in parallel, each section thus con- 
 sisting of a group of wires carried round once. Or, again, a 
 section composed in the aggregate of many turns of wire may 
 be made up of a group of a few wires in parallel carried a few 
 times round in series. 
 
 Figs. 36 to 39 are views of an armature fitted with a simple 
 winding of two turns per section. Fig. 36 is an elevation 
 
Ob DRUM ARMATURE WINDING. 
 
 Fig. 37 is an end view with the commutator removed, leaving 
 a section on line p q in Fig. 36. Fig. 38 is a plan of one coil, 
 and Fig. 39 an isometric perspective view of the same. The 
 portions lettered b b in Figs. 38 and 39 correspond to the outer 
 parallel bars that lie on the periphery of the core on opposite 
 sides of the armature. The parts c c (Fig. 38) lie across the 
 ends, and correspond to the connectors. It will be observed 
 in Fig. 39 that the downward angles d d are arranged to 
 
 FIG. 36. Eickemeyer Winding. Elevation of Armature. 
 
 approach the shaft ; while the swelling shoulders e c c e on 
 either side will form the evolutes right and left, as shown in 
 Fig. 37. As the current here is small, and the sections not. 
 numerous, the commutator is also small as compared with the 
 armature. Hence, the connections / to the commutator are 
 led off from one of the interior angles </. The pins a a in 
 Figs. 36 and 37 are to drive the winding. The dotted lines 
 gg in Fig. 36 represent the exterior binding wires. In Fig. 38 
 
EVOLUTE WIRE WINDING. 
 
 57 
 
 it will be observed that one side, />, is longer than the other, 
 thus, in fact, forming our "long " and " short " bars. This 
 allows room for the two series of evolutes at each end of the 
 armature ; also, when putting the coils on the core, the small 
 side of one coil can easily be passed within the large sides of 
 other coils, and so allow of their being got into position. 
 
 In Figs. 40 and 41 an arrangement is shown of many turns 
 per section for a small current of high voltage. Two coils are 
 
 FIG. 37. Eickeineyer Winding. End View of Armature with 
 Commutator removed. 
 
 here represented, one of them being dotted in Fig. 40, but both 
 shown in full in Fig. 41. It will be seen here clearly how the 
 small side of the one fits within the large side of the other, 
 leaving the clearance space a a between. The wires in the 
 parallel bars are laid 4x2; while in the convolute portions 
 they are all eight abreast, lying in curved planes parallel with 
 the shaft. The "short" bars, it will be observed, are here 
 intended to form a complete layer by themselves on the core 
 
DRUM ARMATURE WINDING. 
 
 mm, while the " long " bars are arranged outside. The two 
 sides of a coil thus come exactly opposite, instead of only 
 nearly so. As the two layers will be exposed to a maximum 
 
 FIG. 38. Eickemeyer Winding. Plan of a Single Coil. 
 
 FIG. 39. Eickemeyer Winding. Isometric Perspective View of a Single 
 
 Coil. 
 
 difference of potential between them, they need to be well 
 insulated from each other. This insulation, and the taping 
 round the bars, are omitted in the sketches for the sake of 
 
EVOLUTE WIKE WINDING. 
 
 59 
 
 -clearness. In Fig. 41 the section through the bars of one coil 
 is shown in solid black, b b, while through the other coil the 
 wire sections are left clear circles, as shown c c. The con- 
 nection to the commutator is led off from the middle at / 
 (Fig. 40). 
 
 In Fig. 42 this same winding is shown, but with the wires 
 in each coil all united in parallel for a large current. As the 
 commutator will now need to be larger it will be more con- 
 venient to lead off the connections to it from the outside of 
 
 FIG. 40. Eickemeyer Winding. Several 
 Turns in Series per Section. Side View 
 of Two Coils. 
 
 FIG. 41. Eickemeyer 
 Winding. Several Turns iu 
 Series per Section. End 
 View of Two Coils. 
 
 the coils, as shown at g, rather than from the middle. This 
 view, being a longitudinal section through the armature, also 
 gives sections through the two series of evolutes at either end, 
 showing clearly their mutual positions, and the air spaces a a 
 between them effectually insulating one from the other. The 
 semi-circular portions dd are the same as those similarly 
 lettered in Figs. 40 and 41, where it will be seen that they lie 
 in radial planes. 
 
60 
 
 DRUM ARMATURE WINDING. 
 
 A few words on this point of the air insulation. Our 
 remarks will also apply to the Crompton and the Alteneck 
 end-windings. This point, indeed, as has already been 
 hinted at, constitutes one of the chief considerations of these 
 special forms of end-winding. When adjacent coils are 
 approximately in the plane of commutation, the "long " bar 
 of one will be in connection with the commutator segment in 
 contact with one brush, and the " long " bar of the other coil 
 will similarly be [connected to* the other brush. Let the 
 brushes be top and^bottom, and the upper one positive. Using 
 
 F IGt 42. Eickerneyer Winding. Several Turns per Section, all in parallel. 
 
 Longitudinal Section through the Armature. 
 
 Fig. 42 for illustration, we shall now find that the bar b is at 
 its highest pressure delivering current to the brush ; while 
 the bar c, being a side of the coil connected to the negative 
 brush, is at a minimum pressure. Hence the two bars b 
 and c are lying close together, with a maximum difference of 
 pressure between them. On the other hand, as the volts only 
 gradually rise in one long bar after another as they approach 
 in order from the bottom to the top on both sides of the 
 armature, it comes about that there is only a minimum 
 pressure ever between adjacent long bars or adjacent short 
 
EVOLUTE WIKE WINDING. 61 
 
 bars. At the end of the armature, all the short bars are 
 connected to the series of evolutes or radial cranks formed 
 next the core ends ; while all the long bars are united to the 
 outer series. Hence there is only a minimum pressure 
 between adjacent connectors in the same series ; and any 
 connectors proceeding from the ends of two adjacent long and 
 short bars, with maximum volts between them, pass each 
 into different series of evolutes, where they are effectually 
 insulated by the air spaces a a between them. Thus it will 
 be seen, that while a good amount of insulation may be put 
 between the layers on the periphery, where it is required, 
 without much affecting the total size ; among the end con- 
 nectors in either series, on the other hand, where a slight 
 increase to the insulation between the strips would greatly 
 affect the size of the whole mass, very little is needed. 
 Hence, of the total mass of a series of evolutes, a much 
 larger proportion can be allowed for the copper than might 
 otherwise be the case. 
 
 This Eickemeyer winding, in common with the Alteneck 
 and Crompton arrangements, will be observed to be sym- 
 metrical with the attendant advantages. One of the chief, 
 however, among its many claims is the ease with which any 
 damaged coil may be replaced, it being only necessary to 
 undo the joints with the commutator, and remove the outside 
 binding wires, to free the whole. It is economical also in 
 total amount of copper used, and in labour of construction 
 after the coils are made, that is, as forming up the individual 
 coils in the mould must doubtless take time. The absence 
 of joints, however, especially in the case of large currents, 
 remains one of its chief distinguishing characteristics. 
 
 As, both in connection with this latter and the previous 
 windings we have described, we have alluded to the inside 
 or outside union with the commutator, we may, perhaps, 
 with advantage at this juncture also call attention incidentally 
 
62 DRUM ARMATURE WINDING. 
 
 to a result following on either of these two arrangements. 
 In the Edison (Figs. 15 and 16) and the Hopkinson methods, 
 as illustrated, and the Eickemeyer for low currents, this 
 union is interior. In the Crompton and "Westminster" 
 dynamos, and also the modified Eickemeyer for large currents, 
 this union will be observed to be exterior. In the former case, 
 and considering any one coupling, of whatever shape, that 
 unites two opposite bar ends, the connection with the com- 
 mutator is taken off from the middle of such coupling, half- 
 way between the bar ends. In the latter, the connection is 
 made at the end of a coupling, at its junction with a parallel 
 'bar. It will thus be seen that with the interior union, a 
 plane through the axis of the commutator and one of its seg- 
 ments will be approximately at right angles to the planes of 
 the coils to which such segment is connected. In the latter 
 case, a plane through the commutator and such segments 
 would be approximately coincident with the planes of its two 
 coils lying, in fact, between them. Now, in practice, and 
 taking the case of a horizontal field, with pole pieces right 
 and left, the polarity of the armature is brought by the much 
 superior strength of the field to be nearly horizontal likewise : 
 hence the plane of commutation comes about vertical. Thus, 
 consequently, with exterior connections to the commutator, 
 and assuming a horizontal field, the brushes will be top and 
 bottom ; while with interior connections, and a plane through 
 the commutator axis and connected segment perpendicular to 
 the plane of commutation, they will come at the sides of the 
 commutator, right and left. It is to be noted, however, in 
 the case of Edison's early end-winding (Fig. 14), that the 
 connection, though interior, is yet taken off the end of a cross 
 connector, the current running to or from the centre along 
 a radial plate. Hence here the brushes would be top and 
 bottom. 
 
CHAPTER VIII. 
 
 HELICAL END WINDING : KAPP. CHORD WINDING : 
 SWINBURNE. 
 
 A TYPE of drum armature winding, which, especially as 
 regards the end connections, differs entirely from any of the 
 foregoing, is the arrangement of Mr. Gisbert Kapp. Here the 
 cross connections, instead of being laid as convolutes or discs, 
 are arranged helically in fact, like the threads of a many- 
 threaded screw. In the case of copper-plate connectors, a 
 common carpenter's wood screw, with thin, flat, and deep 
 threads, will convey perhaps the best approximate idea. 
 
 Fig. 43 represents a part longitudinal section through an 
 armature thus wound. The thickly placed vertical lines, 0, 
 are sections through the series of helical connectors. These 
 will be better understood from the developed plan shown in 
 Fig. 44. The inner ends of the connectors, it will be observed, 
 are jointed to the short parallel bars, as at a a ; while the outer 
 ends are jointed to the long parallel bars at b b the unions 
 being effected by means of tags T T, formed on the ends of 
 the connectors for the purpose. A single helical connector is 
 shown, Fig. 45. B B are the parallel conductors forming two 
 sides of a coil, T T the tags, and s the commutator lug. The 
 dotted lines S S' show part of a true circular plane, thus 
 
64 
 
 DRUM ARMATURE WINDING. 
 
 making clearer by comparison the helical or spiral form of the 
 connector. 
 
 Solid bar or wire connectors in the forms of insulated copper 
 wire cable may also be used as shown in Fig. 46. A layer 
 thus formed of bar or cable being of little radial thickness, 
 other layers may be superimposed. As, however, the con- 
 nectors in the outer layers must necessarily be longer than 
 those in the inner layers, this latter method would appear not 
 to have the symmetry of the helical plate arrangement. 
 
 FIG. 43. Kapp Winding. Longitudinal Section through Armature. 
 
 Eeturning to the plate connectors, each helix, it will be 
 observed, goes round approximately half a circle (for a two- 
 pole machine). All the helices, insulated from each other, 
 are laid together in an annular frame I (Fig. 48) of wood, or 
 other insulating material, and are held there by taping or 
 otherwise; this frame is in turn held in a metal sleeve K, which 
 is mounted on a shoulder of the end-plate F. Thus this end- 
 winding has the peculiarity of being quite self-contained, and 
 
HELICAL END WINDING. 
 
 65 
 
 so can be made up separately, and not put on the armature 
 till complete in itself. It will be observed that as the helices 
 are attached at their ends to alternate parallel bars, between 
 which there can only be a minimum pressure, there will like- 
 wise be only a minimum pressure between adjacent helices. 
 Only a slight amount of insulation is consequently needed 
 between them. 
 
 VIII1I1IIIII 
 
 -VIIIIIIIHIl 
 
 -Ylllllllllll E 
 
 -ruiiiiiiiii L 
 
 FIG. 44. Kapp Winding. Plan of Helical End Connections. 
 
 An obvious advantage of this system over the convolute may 
 be observed. If a large number of connectors are needed, it is 
 not necessary to have two series, but merely to widen the one 
 series in a direction parallel with the shaft. The "pitch" 
 being thus increased, a greater number of any given thickness 
 can be got in. As to ease of construction and replacement, it 
 will be observed that all the joints are on the outside and 
 accessible. In the core, at intervals, are placed metal discs 
 
DRUM ARMATURE WINDING. 
 
 with projecting horns A for driving the winding, and air 
 spaces D at either side of these discs allow for ventilation. 
 The core has a large central bore, admitting plenty of air, and 
 
 FIG. 45. Kapp Winding. Single Helical Connector. 
 
 is carried on a spider H as shown. It will be observed that, 
 as with the Edison disc, these plate connectors require to be 
 
 FIG. 46. Kapp Winding. 
 
 Single Helical Connector of Copper Wire 
 Cable. 
 
 stamped out of sheet : hence the same remarks as to waste of 
 material and cost of labour, &c., will apply, the question 
 remaining as to whether or not this is made up for by subse- 
 
CHORD WINDING. 67 
 
 quent economy in labour of construction. This end- winding 
 is very compact and symmetrical, with comparatively few 
 joints, and these easily accessible, besides having the other 
 advantages already mentioned, including free access for 
 ventilation. 
 
 The next winding we come to is another of Mr. James 
 Swinburne's. In this case the peculiarity exists that the coils, 
 instead of being wound over diameters, are wound over chords 
 
 cc 
 
 FIG. 47. Swinburne's Chord Winding 
 
 of the circumference of the armature. An application of this 
 principle to a wire-wound armature is shown in Fig. 47. It 
 will here be seen that the end wrappings, lying in chords as 
 at a a or b I, keep clear of the centre, and thus, especially 
 when the chords subtend smaller angles at the centre than 
 here represented, allow room for ventilation through the 
 hollow of the core. 
 
 For large low-pressure currents the method of coupling up 
 across the ends is as shown Figs. 48 to 51. Curved stamped 
 
 p 2 
 
68 
 
 DKUM AKMATURE WINDING. 
 
 copper plates a a form the connectors. These are laid two in- 
 a plane normal to the axis of the armature, and with their 
 edges butting as at / (Fig. 48). Connecting pieces d d are 
 attached to the plates by means of which union with the bars 
 b b is formed. The plates covering more than a third of a 
 circle each, the space c may be kept clear for ventilation. 
 Otherwise the plates may be laid slightly helically, and overlap 
 
 FIG. 48. Swinburne's Chord Winding. End View of Armature, Two 
 Connectors in situ. 
 
 one another. A spider n carries the core, passages o o being 
 left for ventilation. Figs. 50 and 51 show separate views of 
 one plate connector, with the addition of a tag g formed on the 
 interior edge for interior union with the commutator, if such 
 should be desired. 
 
 Foucault currents are dealt with by making each conductor 
 bar of parallel strips or bars, and then connecting these up- 
 
CHORD WINDING. 
 
 69 
 
 -separately across the pulley end, as shown in Fig. 13 (page 35), 
 it being borne in mind that in the case before us these 
 connectors will be taken across a chord and not a diameter. 
 
 It would appear that in this end winding, as with Edison's 
 discs, the close proximity of connector plates of maximum and 
 minimum volts is not avoided. It will be observed, further, 
 that the sum of the areas enclosed by the coils is not so great 
 as would be the case were the winding diametrical. The plates 
 require to be stamped. On the other hand, the construction 
 is simple, and all the joints are outside and accessible. The 
 
 PIG. 49. Swinburne's Chord Winding. Method of Connecting the 
 Connector-plates to the End Plates of Armature Core by Insulating 
 Screws h. 
 
 connector plates are held in a mass to the end plates of the 
 core by means of insulated screws, one of which is shown at 
 h, Fig. 49. These screws of course also act as drivers, 
 obviating the necessity of any special means for that purpose. 
 Ventilation, as already intimated, is well provided for. 
 
 There are, however, some further points in connection with 
 this winding which it may be of interest to consider. For the 
 better elucidation of these, a diagrammatic representation of 
 this system is appended in Fig. 52. The pulley end is here 
 shown. The outer radial lines // y represent the commutator 
 
70 
 
 DRUM ARMATURE WINDING. 
 
 lugs at the further end extended on magnified diameters to 
 
 make them visible. The curved dotted lines indicate the con- 
 
 
 
 nections thereto. The lugs lettered B B are those of the 
 segments on which the brushes rest, the left-hand one being 
 the positive brush of the machine and the right-hand the 
 negative, while the union with the commutator is supposed to 
 be from the middle of a connector. Inspection will make it 
 clear that the current flow is as shown by the small arrows. 
 
 FIG. 50. 
 Swinburne's Chord Winding. 
 
 FIG. 51. 
 Separate Views of a Connector. 
 
 The pole pieces are also indicated, N and S. The small 
 letters n and s show the armature polarity which each pair of 
 parallel coils tends to produce. Thus the coils a and b, c and d, 
 tend to produce between them a vertical polarity, each rein- 
 forcing the other, with a south pole above, and a north pole 
 below. The pairs e f and g h tend to produce each an approxi- 
 mately horizontal polarity parallel with the field, but contrary 
 to each other. These latter thus neutralise each other, with 
 
CHORD WINDING. 
 
 71 
 
 the result that the vertical polarity alone remains. In relation 
 to the armature 'itself, this may be called its primary polarity. 
 In relation to the field this is known as cross-magnetisation, 
 or cross-induction, as it is at right angles to the field. Now 
 
 FIG. 52. Swinburne's Chord Winding. Diagrammatic Sketch of System. 
 
 it will be observed that on the upper and lower sides of the 
 armature, as sketched in Fig. 52, the current is running in 
 opposite directions in adjacent parallel conductors ; while at 
 the sides next the pole-pieces the current runs the same way 
 in all on one side. It is claimed in Mr. Swinburne's patent, 
 
72 
 
 DRUM ARMATUKE WINDING. 
 
 that by virtue of this arrangement of current direction, the 
 cross-magnetisation is reduced. 
 
 In ordinary diametrical drum-winding, as explained in our 
 first chapter, the coils adjacent on one side of the plane of 
 commutation neutralise those on the other side. Thus, in 
 Fig. 53, where xx is the plane of commutation, the coils a a 
 neutralise the effect of the coils b b, in which the current 
 is running in an opposite direction, as shown by small 
 
 Fia. 53. Ordinary Diametrical Drum Winding. Diagram of Current Flow. 
 
 arrow-heads. On the other hand, the polarising effect of 
 the horizontal coils cc is unopposed, and produces the 
 vertical primary polarity of the armature coincident with 
 the plane of commutation x a-, which in relation to the 
 field, as we have already mentioned, is known as cross- 
 magnetisation. 
 
 Let us now suppose the brushes to require a lead. The 
 plane of commutation becomes tilted as shown xx (Fig. 54). 
 Now note the effect. The coils aaaa are brought normal to 
 
BACK INDUCTION. 
 
 73 
 
 the field. Hence these will produce a polarity parallel to the 
 field. In generating dynamos a north pole in the armature 
 becomes opposed to the north pole-piece, and a south to the 
 south, as indicated by small letters n and s, in direct opposi- 
 tion to the field. We thus find this polarity parallel with the 
 field, termed, in the case of generators, the " back induction" ; 
 and its effect is to weaken the field. In the case of motors, 
 wherein we may imagine the N and S pole -pieces to exchange 
 
 I 
 
 i 
 
 FIG. 54. Same as Fig. 53. Brushes Shifted. 
 
 places, the reverse holds good, and this parallel induction 
 reinforces the field : and the further phenomenon arises that, 
 with an armature revolving between non-magnetised pole- 
 pieces, and having a tilted plane of commutation, the south 
 pole of this parallel induction will induce a north pole in the 
 pole-piece opposed to it, and a south pole will similarly be 
 induced in the other pole-piece. Thus with induced pole-pieces 
 only, a motor can be run in a primarily non-magnetised field, 
 
74 DKUM ARMATURE WINDING. 
 
 though it may not pay to do so in practice. While a tilted 
 plane of commutation is on this account not detrimental to 
 a motor, but rather the contrary, it is detrimental in the case 
 of a generating dynamo. Hence, to avoid this evil in the 
 latter case, as has been already explained (see Chap. I.), the 
 field is made so strong as to draw the primary cross polarity of 
 the armature as nearly as possible parallel to itself. The ulti- 
 mate effect is a compromise between the armature polarity and 
 the field ; and in proportion as the disparity is lessened, so 
 great is the " distortion of the field." Thus, in Fig. 54, the 
 lines of force of the field, instead of running straight across 
 through the armature from N to S, will pass through along 
 y y as a centre line of their course, the angle between y y and 
 the horizontal being the measure of their distortion. The 
 peculiarity will thus be noted that the primary polarity of the 
 armature, or cross-magnetisation, is just exactly drawn down 
 into the line y y, along which there is no primary polarity, as 
 already pointed out, owing to the neutralisation of the coils 
 a a and b b. The cross -induction thus diverted does not oppose 
 the field. Assuming the plane of commutation to have been 
 originally vertical, but the field becoming distorted by the 
 armature, the brushes will require to be moved over till the 
 plane of commutation is brought perpendicular to the new 
 direction of the field, guidance as to when the correct position 
 has been found being obtained by the absence of sparking at 
 the commutator. 
 
 It may be further observed, referring to Figs. 53 and 54, 
 that the whole of the armature above the line y y being north, 
 and below, south, the horns g and / are opposed, and e and h 
 reinforced, by the armature primary polarity. It thus comes 
 about, that whether the plane of commutation be tilted or not, 
 the lines of force gather thick at the horns e and h, which are 
 the "trailing" horns, while at the other horns they become 
 comparatively sparse. The Eoman letters N and S indicate 
 
BACK INDUCTION. 75 
 
 the ultimate polarity of the armature. In the case of a dis- 
 torted field, Fig.. 54, the main body of the lines of force 
 running from the lower half of the N pole-piece and through 
 the upper half of the S pole-piece, and these two halves 
 thereby carrying the most lines of force, this further increases 
 the accumulation at the horns e and Jt. The lines being drawn 
 away from the horns g and /, these may become magnetised 
 the wrong way if care be not taken. 
 
 Returning now to Mr. Swinburne's chord winding, and the 
 claim for reduction of cross -magnetisation referred to, we find 
 it difficult to see how this is made good. Referring to Fig. 52, 
 we have found the cross-magnetisation to be due to the coils 
 abed. The mixed bars above and below have but little effect 
 in this direction, owing to the obliquity of the coils to which 
 they belong. The whole of the upper half of the armature is 
 of one pole, without primary opposition, and the lower half the 
 other pole. The coils being on chords instead of diameters, it 
 may be convenient to speak of two planes of commutation here, 
 as represented by .1- x and z z. While these planes are vertical 
 back induction cannot arise, as pointed out. But neither does 
 it in diametrical winding. If the planes are tilted by a lead 
 given the brushes, then back induction immediately sets up, 
 as, for instance, the coils e and / becoming vertical will 
 exercise influence opposing the field. In fact, with a lead to 
 the brushes, a preponderance of the north primary polarity of 
 the armature will be thrown against the north pole-piece, and 
 of the south against the south pole-piece, and all the attendant 
 consequences will follow as in the case of the diametrical 
 winding. Hence it is not clear how this claim is substan- 
 tiated ; and this failing, it follows that this chord system of 
 winding will not compare favourably with the diametrical 
 winding, owing to the much smaller total area enclosed by the 
 coils ; unless, indeed, there be occult reasons in its favour, 
 which have eluded investigation. 
 
CHAPTER TX. 
 
 EXTEEIOK END WINDING: CROMPTON AND KYLE. 
 PARSONS. FRITSCHE. 
 
 A FEW more windings which we may touch on belong to an 
 entirely distinct class, especially as regards their end con- 
 nections. We refer to those in which the connections are 
 made outside the circumference of the armature, instead of 
 within. 
 
 The first of these we may consider is one patented by 
 Messrs. Crompton and Kyle. This is based on the convolute 
 principle, and consists of two series of convolutions running 
 in opposite directions outside the armature ends. Figs. 55 
 and 56 illustrate this arrangement, though with only a 
 portion of the connectors shown. On the outer circumference 
 will be observed jointing plates a a a parallel with the shaft. 
 From each of these there spring two evolutes b b, which 
 proceed, one in each series of evolutes, to the opposing ends 
 of a long and short bar respectively. One complete coupling 
 b a b is thus shown apart in both views. Though cases may 
 arise when this arrangement may be most suitable, yet in 
 some respects it will not compare favourably with an inside 
 end winding. 
 
 This form of winding plainly requires a longer coupling to 
 unite the two ends of a diameter outside than inside the 
 
78 
 
 DRUM ARMATURE WINDING. 
 
 circumference. The joints, however, are all accessible after 
 removing the commutator. But a disadvantage with this 
 design results from the fact that an accumulation of material 
 round the outer ends of the armature militates against the 
 snugness of the machine as a whole. These remarks apply 
 also to Figs. 20 and 21. It is desirable to keep the 
 " magnetic circuit " through the magnets as short as possible, 
 and the field coils need to be brought close up to the pole- 
 pieces to obtain their fullest effect. It will, therefore, be 
 
 FIG. 55. Crompton-Kyle Winding. FIG. 56. Crompton-Kyle Winding. 
 Side View. End View. 
 
 observed that designers generally bring the field coils as 
 near the armature as possible, and, in fact, in the case of 
 vertical machines, rest them on the horns, as shown at a a, 
 Fig. 57. Thus the lines of magnetic induction generated in 
 the magnets within the coils can pervade the pole-pieces with 
 the least possible loss of strength, and the total magnetic 
 resistance is kept to a minimum. In the design of end 
 winding before us, however, the connectors would come in 
 .contact with the field coils, so that these would have to go 
 
EXTERIOK END WINDING. 
 
 79 
 
 farther away from the armature, thus in some degree causing 
 the advantage named to be sacrificed. This difficulty might, 
 
 I 
 
 FIG. 57. 
 
 perhaps, be met by bending all the bars inward a little way 
 toward the axis, as shown in Fig. 58. But bringing all the 
 
 FIG. 58. 
 
 ends thus into a smaller circumference, unless the original 
 spacing on the periphery of the core should be an open one, 
 will tend to jamb them, and so necessitate their being formed 
 
80 
 
 DRUM ARMATURE WINDING. 
 
 to a deeper and narrower section, thereby introducing 
 complications. 
 
 The next arrangement of outside end winding to which we 
 may refer is one patented by the Hon. Charles A. Parsons ; 
 though also independently invented by a young man of the 
 name of White, in the employ of Messrs. Latimer Clark, 
 Muirhead and Co. This is shown somewhat diagrammatically 
 in Fig. 59, and is in principle somewhat related to Siemens's, 
 Eickemeyer's, and Kapp's. With the two former it agrees 
 in that each coil consists of one wire, or group of wires, 
 
 FIG. 59. Diagram of Helical Outside-end Winding. 
 
 without joint from end to end ; with the latter, in that the 
 helical principle is introduced at the armature ends. Thus a 
 coil starting, say, from a commutator lug at , pursues first a 
 helical path through about a quarter circle to b, thence runs 
 parallel to c, then down another helical path and approximate 
 quarter circle to d, is here turned back on itself, and pursues 
 another helical path to e, and so on until it ends at another 
 lug at g. 
 
 The Parsons winding is illustrated in Figs. 60 to 62, 
 showing the two ends separately of the same armature. 
 
PARSONS WINDING. 
 
 81 
 
 There is here thus represented in practical form what is 
 shown diagrammatically in Fig. 59. The parts lettered c, w, 
 .are sections through the inner and outer layers respectively of 
 
 +VK n g^^- 111 ''" 
 
 lXlj^^ UOOUUOOOO 
 
 f 
 
 FIG. 60. Parsons' Winding. Longitudinal Section through Engine end. 
 
 FIG. 61. FIG. 62. 
 
 Parsons' Winding. Longitudinal Section through Section on a I: 
 
 Commutator end. 
 
 the helical end winding : and it will be noted that at the ends 
 the diameter of the armature is reduced, so that the exterior 
 
S2 DRUM ARMATURE WINDING. 
 
 diameter of the winding is the same throughout. Beyond 
 the limits of the iron core proper, bf and c e, the armature is 
 non-magnetic, forming thus what may be termed a dummy 
 core. The discs gg, hh, are of insulating material; and the 
 smaller discs supporting the helical coils are of brass. At ?, 
 Fig. 60, it will be observed that the coils are turned back on 
 themselves as at d, Fig. 59. In Fig. 61, on the other hand, 
 it will be seen that the conductor bar ends, rr, pass through 
 separate holes in an insulating disc, j ; they are then coupled 
 across by means of wire or copper strip and soldering, as- 
 shown at m m, Fig. 62. This latter method is that adopted in 
 all cases when the conductors are too heavy to bend back as 
 at i. The whole core is held on the spindle between two 
 nuts, nn; and the spindle is hollow for the admission of 
 fluid, such as cold water or oil, for the sake of cooling the 
 parts. Two of the bars, r r, it will be observed, run out to 
 the commutator, lettered t. On the core proper, between 
 bf and c e, the bars lie in grooves, and are bound therein by a 
 plentiful wrapping of steel wire, rendered especially necessary 
 in this class of dynamo by the very high speed at which the 
 armature rotates. 
 
 It will be observed, however, in this type of winding, that 
 the helices being in two layers, and all those from what 
 correspond to the " long " bars crossing those from the 
 "short" bars, the evil of the close proximity of conductors 
 of maximum and minimum volts is in no wise avoided ; and 
 the coils would need to be well insulated throughout. This 
 form of winding is well exposed to exterior ventilation. To 
 remove any burnt section, as in the case of the Eickemeyer 
 arrangement, it would be necessary but to undo the connec- 
 tions with the commutator, and remove the exterior binding 
 wires to loosen the whole. All its joints are accessible. 
 But its requiring what we have termed " dummy " cores 
 at the ends to support the helical portions of the winding,, 
 
FRITSCHE WINDING. Oi> 
 
 introduces so much waste weight. In consort, however, 
 with the extremely high speed characteristic of the Parsons 
 machine and steam-turbine, wherewith but a small diameter 
 of armature is needed, this latter objection will not be of so 
 much consequence as with a larger diameter. But this form 
 of helical end winding makes a very long armature, entailing 
 a long spindle and long machine altogether. 
 
 If, however, we take this winding, and imagine all the 
 parallel portion included within the letters b, c, e, f, to be 
 removed, so that b f coincides with c e, we shall then get 
 approximately the last type of winding we propose to discuss,, 
 the Fritsche armature winding. The core of this armature 
 
 ct 
 
 ?- 
 
 FIG. 63. Fritsche Winding. 
 
 is, in fact, wound helically ; the whole winding comes within 
 the field, and it may be said that there is no " end winding." 
 The pole-pieces are placed diagonally in this machine, with 
 the corners cut off, in the manner indicated by the dotted 
 lines in Fig. 63. This, however, makes a bad field ; for so 
 large a proportion of the armature, as at a a, is uncovered 
 by the pole-pieces ; and the sectional area of the field, taken 
 in a plane normal to its lines of force, is very much reduced 
 compared to a field with the pole-pieces placed square in the 
 usual manner. As regards facility for renewals, it is, of 
 course, the same as the last described. On any armature, 
 
 however, of a given diameter, it is not possible to put so 
 
 G 2 
 
84 DRUM ARMATURE WINDING. 
 
 many bars of a given size when laid spirally as when laid in 
 the ordinary manner parallel with the axis comparing it, 
 that is, with a parallel winding of two layers. Much care 
 must obviously be involved in securing symmetry. An 
 ulterior effect following on the diagonal arrangement of the 
 pole-pieces to suit the helical winding is that, for any given 
 output, the machine, as a whole, will need to be large and 
 heavy as compared with others, and so will cost more. The 
 coils, of course, have no joints, and, except those forming the 
 inner layer, are well exposed to exterior ventilation. That 
 the inner half should be without ventilation is a drawback. 
 The commutator of this machine (of which a full account 
 appeared in The Electrician, Vol. XXII., p. 655) is abnormally 
 large, as indicated. 
 
COM MUTATORS. 
 
CHAPTER X. 
 
 COMMUTATORS. INTRODUCTORY REMARKS. 
 
 THAT part of a dynamo which may be said to be next in 
 importance to the armature and magnets is the commutator, 
 or collector, as it is variously termed. For it is here that the 
 current, after having been generated in the armature, is finally 
 collected ; and it is from and to this part that it is led away 
 through the outer circuit of lamps and so forth, and brought 
 back again for the renewal of lost energy or potential. On re- 
 turning into the commutator at one side at minimum potential, 
 the current courses through the armature, and is delivered 
 again at the opposite side of the commutator changed to a cur- 
 rent of maximum potential : hence the term " commutator," 
 from the Latin commute, to change, or alter entirely. 
 
 Our object is now to consider some of the various forms of 
 commutator that have been brought out ; and also some of 
 the points that have to be kept in view when getting out the 
 design for this portion of a dynamo, such as have led more 
 or less to those differences of form to which we allude. 
 
 Without pretending that a commutator is to any very great 
 -extent intrinsically complicated, it yet would be a mistake to 
 suppose that a design may be just " knocked off out of hand " 
 without much thought. As great an amount of consideration 
 
88 
 
 COMMUTATORS, 
 
 within reason is required in designing a commutator as any 
 other active part of a machine. For we shall find in the course 
 of our investigation, that, as in all designing, from that of an 
 ironclad downwards, there are conflicting requirements to be 
 dealt with, both small and great. To these, therefore, and other 
 collateral considerations, we purpose first to give our attention, 
 before viewing any special examples in particular. Our remarks, 
 moreover, must be premised by mentioning that we confine 
 ourselves to commutators for close-coiled continuous-current 
 machines, such as those having drum, Gramme, or flat-ring 
 armatures. 
 
 FIG. 64. 
 
 FIG. 65. 
 
 As is well known, the essential principle in the construction 
 of a commutator is that of a series of segmental bars, held 
 together on a substructure, insulated therefrom and from one 
 another. This, omitting substructure, is illustrated in Figs. 64 
 and 65, the upper half of the latter being in section, so as to 
 afford a full side vjew of one segment. The shoulder a repre- 
 sents the lug whereby connection is made with the armature 
 winding, and through which the current runs either way 
 between the winding and the segment. The thick black lines 
 represent insulation. 
 
INTRODUCTOKY REMARKS. 89 
 
 Now, it is obviously the first consideration that each bar 
 shall be large enough to carry the whole current without 
 undue heating, and shall continue so after a reasonable amount 
 of wear. It must be of sufficient thickness and depth, and 
 have length enough to allow of a proper number of brushes 
 to take the current off. But another consideration immediately 
 follows, in respect of economy, inasmuch as the segments are 
 usually of either copper, phosphor-bronze, or gun-metal, all of 
 which are expensive materials. Hence, with regard to prime 
 cost alone, it is not desirable to increase the dimensions over- 
 much either way. Again, as to the length, this dimension 
 affects the length of the machine as a whole in particular as 
 to shaft and bedplate. This increases the total weight, and 
 adds to the prime cost in respect of cast iron and steel. The 
 diameter of the shaft would need increasing in due proportion. 
 It also, however, means greater floor-space, or, on board ship, 
 deck room. In the latter case especially, where space is limited, 
 if one over-all dimension can be reduced without adding to 
 another, there is an object in doing so. But this, it will be 
 seen at once, is directly opposed to the already mentioned 
 necessary condition, that there shall be a sufficient length of 
 commutator for the brushes. We shall see later on, when 
 dealing with particular designs, what methods have been 
 adopted to meet this combination of requirements. 
 
 Now, the segments, besides being liable to wear from the 
 constant rubbing of the brushes, are also liable to burn away 
 from sparking between them and the brushes, especially if the 
 sparking is such as flies from them to the brushes. This evil, 
 however, is less liable to arise when the commutator is divided 
 into many segments in proportion to the voltage than into a 
 few. A commutator of many segments is thus for this reason de- 
 sirable ; and for the further reason that it would give a steadier 
 current at any given speed than one with fewer parts. This 
 question, however, is one with which the arrangement of the 
 
90 COMMUTATORS. 
 
 armature winding as a whole is concerned, inasmuch as there 
 are the same number of segments in a commutator as there are 
 sections in the winding. Now, with an armature wound in 
 many sections for a high potential and a small current, the 
 bars need not be of large section ; yet they must be of appre- 
 ciable size. But especially owing to the number of bars all 
 requiring to be separated by insulation, this latter material, by 
 its accumulation, forms a considerable factor in determining 
 the diameter. Hence it comes about that the commutator will 
 be of large diameter, sometimes nearly equalling that of the 
 armature itself. But there being only a small current to be 
 taken off by the brushes, the bars do not require to be long, 
 and the commutator will be short in proportion to the diameter. 
 It is obvious, however, on the other hand, that with any given 
 speed, and greater consequent peripheral velocity, there will be 
 more wear of both brushes and segments, and greater con- 
 sumption of energy, due to the friction of the brushes, with a 
 large diameter than with a small. But with a low potential, and 
 large current, and small number of segments, the total amount 
 of insulation between the bars will be reduced in proportion, 
 while the length will need to be greater to accommodate those 
 brushes. Hence the length of the commutator will be greater 
 as compared with the diameter, and the latter considerably 
 less than that of the armature. 
 
 The thickness of the brushes may, however, be a factor 
 in this question. It is obvious, that with any given number 
 of segments, the same peripheral brush contact surface sub- 
 tending any given angle at the centre may be obtained by 
 having either a long thin brush contact surface on a small 
 diameter of commutator, or else a short broad one on a large 
 diameter. 
 
 Though it is a common feature in small high-tension 
 machines to see the commutator nearly as large in diameter 
 as the armature, a peculiar instance where it is so in the case 
 
INTRODUCTORY REMARKS. 91 
 
 of a large machine is that of the Elwell-Parker * continuous- 
 current transformer, or motor-generator. Here there are two 
 commutators, one at each end. That receiving the high- 
 tension current for the " motor " function of the machine is 
 composed of a large number of small segments, and the total 
 diameter is nearly equal to that of the armature ; while at the 
 same time the length is only a small fraction of the diameter. 
 The other commutator, dealing with the low-potential large 
 current, the outcome of the " generating " function of the 
 machine, has many fewer segments, and is of smaller diameter ; 
 while the length is greater than that of the other commutator. 
 Its length is about the same as its diameter, in fact, and a 
 large brush contact surface results. The contact surface being 
 proportional to the current, and not to the potential, much 
 smaller brushes are sufficient on the high tension. 
 
 Returning once more to the subject of sparking, the 
 principal bearing of this matter lies in the fact of its forming 
 a reason for subdividing the commutator into as many parts 
 as possible, as already intimated. The tendency to spark, and 
 the secondary effect of the formation of " flats," are thus 
 minimised. But, as will be seen, much subdivision increases 
 the total amount of mica between the bars, and consequently 
 the diameter also. Hence a compromise has to be effected, or 
 at least too much subdivision avoided. 
 
 A peculiarity, however, in the wear of the segments, apart 
 from their burning, is that, with wire brushes especially, annular 
 grooves are apt to be cut round the commutator. This is 
 mitigated by allowing the armature a slight end play in the 
 bearings when possible. To equalise the wear also, it is usual 
 to adjust the brushes, when there are more than one in a set, 
 so that the gaps between those in one set do not come opposite 
 
 * This machine is illustrated and described in The Electrician, VoL 
 XXVIIL, p. 424. 
 
92 COMMUTATORS. 
 
 the gaps in the other set. For the further reduction of wear 
 and cutting it is commonly the practice to put a very little 
 vaseline on the commutator. This is more necessary, however,, 
 when the bars are of soft copper or other metal. When of good 
 hard metal, and with stencil-plate or gauze brushes, it appears 
 to cause no detriment to run dry, without any lubrication. 
 
 Though under certain circumstances some oils may act as 
 perfect insulators, better even than air, it is not desirable to let 
 the oil from the bearings creep over or into the commutator. 
 With large difference of potential between the segments and 
 the parts of the machine at. zero potential, a streak of oil be- 
 tween would carbonise ; and carbon being a conductor, a short 
 circuit would be set up, leading to leakage. To prevent this, 
 among other reasons, a ridge is generally turned on the shaft, 
 or on one of the nuts retaining the commutator, or elsewhere, 
 from which the waste oil from the bearing is flung off, and i& 
 caught by an overhanging annular hood, or oil-catcher, project- 
 ing from the bearing, from which it is drained away. 
 
 As to the metal used for the segments, besides the gunmetal, 
 phosphor-bronze, or copper, already mentioned, iron is sometimes 
 used. The prevailing tendency is perhaps towards copper, though 
 this somewhat depends on the form of the segment. Copper 
 segments may either be cast, which can be done by alloying with 
 a little silver a method adopted by Messrs. Crompton and Co. 
 or else they 'can be rolled to the required segmental section and 
 so be cut off in lengths. This metal is, of course, a better con- 
 ductor than either phosphor-bronze, or gunmetal, or iron. The 
 copper, when not cast, may be either hard rolled or else annealed. 
 The former would be more durable ; but the latter is better 
 as regards conductivity. As, however, the ratio of the specific 
 resistances of hard-drawn and annealed copper is given only 
 as 1,652 to 1,615, this seems so slight in proportion that the 
 question of durability would have more weight, and so the hard 
 copper is preferable. 
 
INTRODUCTORY REMARKS. 93 
 
 In the process of construction, these bars are filed, or other- 
 wise worked up true, on their flat sides.; then, with thin sheets 
 of mica, usually about one thirty-second of an inch thick, and 
 shellac or other varnish between them, they are all clamped 
 together so as to form a hollow cylinder, as shown in Figs. 64 
 and 65. According to varying practice, the inside may or may 
 not be bored. But while still clamped together they are put 
 in the lathe, and the ends carefully turned to gauge. The 
 substructure consists of a sleeve or bush, usually of gunmetal, 
 though sometimes when large of cast iron. Over this the 
 cylinder of segments is fitted with insulation between, and 
 fastened on by a nut and washer. These latter are either 
 wrought iron or gunmetal, or may be of steel. The whole now- 
 forming a compact structure by itself, the clamps are removed. 
 Put in the lathe once more, the outside is carefully turned to 
 asmooth bright surface. Though we shall deal with variations 
 later on, commutators of this general type such as we are dis- 
 cussing have thus this characteristic in common, that when once 
 built they are complete in themselves, and as such can be put 
 on or off the shaft of the dynamo. 
 
 
CHAPTER XL 
 
 COMMUTATORS. INSULATION. 
 
 Now, though the difference of potential between adjacent 
 &egments may be slight, yet this varies from a minimum at the 
 plane of commutation to a maximum at the sides of the com- 
 mutator, in a plane at right angles to the plane of commutation. 
 For, as explained in our first chapter, assuming the plane of 
 commutation to be vertical, and so the brushes at top and 
 bottom, and the segments approximately in the same plane 
 with the coils with which they are connected, those segments- 
 at the sides connected to coils lying horizontal are collecting, 
 potential at a maximum rate ; while the vertical segments at 
 the top and bottom are connected to coils which are generating 
 practically no potential. The reasoning would be very similar 
 in reference to Gramme-winding. But in consequence, tha 
 insulation between the segments must be sufficient to with- 
 stand this maximum difference of potential that occurs at 
 the sides half way between the brushes. On the question 
 of insulation, however, that between the bars and the sub- 
 structure requires special attention ; for there is the difference 
 between the whole potential of the machine and the zero 
 potential in the substructure to deal with. Here the insulation 
 may be an eighth of an inch or more in thickness, and may 
 consist of vulcanised fibre, mica, asbestos, or other material.- 
 
06 COMMUTATORS. 
 
 Between the segments, it may be mentioned, mica is specially 
 suitable owing to its natural cleavage, which allows of its easily 
 being split into the very thin sheets required. Moreover, it is 
 hard and well adapted to resist the compressive strain brought 
 upon it, and so, with the segments, help them to form a solid 
 mass compact together. 
 
 Now, while on this subject of insulation, and inasmuch as 
 the substance commonly known as " fibre " is so largely in use 
 for this purpose, it may be of interest perhaps to add some 
 remarks concerning it, for which, in the first instance, we v are 
 indebted to a somewhat complete resume on the subject by 
 manufacturers of this material. 
 
 Vulcanised fibre is produced by chemically treating specially 
 prepared vegetable fibre, whereby the exterior portion of 
 each separate filament becomes glutinous. While in this 
 condition, the whole mass is consolidated under heavy pressure, 
 and rendered practically homogeneous. The chemicals are 
 then extracted, and the mass is manipulated, rolled, pressed, 
 and cured by various methods. These operations being of an 
 extremely delicate nature, liable to vary with different condi- 
 tions of atmospheric moisture and temperature, it requires 
 great skill, care, and experience, as in many other manufac- 
 tures, to produce uniformly good results. The fibre is made 
 in two qualities, "hard" and "flexible"; though for purposes 
 of insulation the hard quality alone is applicable. A pecu- 
 liarity is that this substance is produced only in sheet or 
 tube forms. It will not mould ; hence all objects to be made 
 from it require to be cut out of either of these two forms, 
 whichever may be the most suitable. The sheets, in size about 
 40in. by 60in., vary from J^in. to 1 Jin. thick, and are supplied 
 in three colours, red, grey, and black ; while the tubes, pro- 
 duced in lengths approaching two feet, have a thickness of 
 shell not exceeding half an inch. It appears that these latter 
 cannot be manufactured of less than ^-in. inside diameter, and 
 
INSULATION. 97 
 
 -the thicknesSj moreover, is proportioned somewhat to the dia- 
 meter in the small sizes. Being rolled perfectly true and 
 straight, with a good finish, they may be cut off into washers, 
 or be used in lengths for enclosing and insulating bolts, and 
 so forth. Though it has no grain, properly speaking, like wood, 
 this material has none the less a species of cleavage parallel to 
 the plane in which it is rolled. It may thus be split or cut 
 comparatively easily in one direction, while it is hard to work 
 in another. 
 
 As is well known, this material is strong, elastic and durable, 
 even working well when used for the teeth of mortice wheels. 
 This latter, however, depends on the teeth being lubricated 
 with grease and blacklead, and not with oil. Fibre absorbs 
 both oil and water. The former reduces its durability, while 
 the latter would of course be fatal to its quality as an insulator. 
 Such absorption causes it to swell, though it resumes its 
 original size when dry again. Its specific gravity is about 1'3, 
 and a cubic inch weighs about four-fifths of an ounce. 
 
 Originally imported from America, vulcanised fibre has been 
 put on the market in England and the Continent as an insu- 
 lator, and having some insulating qualities, and being easily 
 worked, has been very generally adopted. In this connection, 
 however, we may give a table of specific resistances, compiled 
 from measurements taken by Profs. Ayrton and Perry, and 
 which we quote from Upperiborrfs Electrical Calendar : 
 
 Vulcanised fibre 1/2 x 10 6 megohms (B.A.) at 20C. 
 
 Mica 84-0 20C. 
 
 Gutta-percha 450'0 , 24C. 
 
 Shellac 9000'0 28C. 
 
 Hooper's vulcanised India- 1C ; mo . n ,>/, p 
 
 rubber . ouuu u ^ u - 
 
 Ebonite , 28000'0 46C. 
 
 Paraffin 34000-0 46C. 
 
 Glass Still greater. 
 
 From the above it will thus be noted that fibre ranks last 
 as an insulator. Now, with the India-Ruller and Gutta* 
 
98 COMMUTATOES. 
 
 Percha, (&c., Trades' Journal (May 9, 1892), as our authority, 
 we there find an account of some experiments made as to the 
 specific resistance of vulcanised fibre by Mr. Eugen Miiller, of 
 Berne, the original of which appeared in the Elektrotechnische 
 Zeitschrift (February 5, 1891).* The liability of fibre to absorb 
 moisture has already been intimated. To how great a degree 
 this liability extends, however, will be made clear from the 
 following tabulated results of these experiments. It is to be 
 observed that these were made under three different conditions. 
 
 * We are indebted to the Editor of The Electricianfor the following 
 epitome of further correspondence on this subject : 
 
 " Eeferring to this article, Mr. W. Courtenay, President of the Vulca- 
 nised Fibre Company, writes to point out the rapidity with which ebonite 
 or vulcanite (even when it is of the best quality, which is seldom the case) 
 deteriorates with age, whilst vulcanised fibre actually improves, and after 
 eight or ten years has been found in a better state than when first put.in. More- 
 over, ebonite or vulcanite is almost always adulterated, and in time becomes 
 perfectly perished, and its only claim to being considered an insulator lies 
 in the fact that it is waterproof. Mr. Courtenay entirely denies that vulca- 
 nised fibre contains either jute or oxide of iron, as asserted by Herr Miiller. 
 All Herr Miiller's figures giving comparisons with paraffined wood are, says 
 Mr. Courtenay, quite irrelevant. As everyone knows, paraffin is one of 
 the best non-conductors, and wood being exceedingly porous, the tests given 
 refer in reality to the paraffin, and not to the wood. It is not to be denied 
 that ebonite and vulcanite when pure (which they seldom are) have a 
 higher resistance against high potentials than fibre. Against this must 
 be balanced the mechanical weakness to which they are liable. Herr 
 Miiller is aware that the hygroscopic properties can be entirely done away 
 with by giving it a coating of shellac or other watertight varnish, whenever 
 circumstances would otherwise allow it to absorb moisture. Mr. Courtenay 
 states that it is by no means certain that Herr Miiller had got hold of 
 genuine specimens of fibre, as many imitations of the genuine article had 
 been produced. In answer to this, Herr Miiller, in a later issue of the 
 EleTctrotechnische Zeitschrift, states that the specimens of fibre tested were 
 obtained from the sole agent of the Patent Vulcanised . Jfibre Company,. 
 Herr, Wilfert, of Cologne. To this the latter replies that he has only 
 supplied Herr Miiller with a small sample of fibre, and that nearly two 
 years ago. He also says that since the Vulcanised Fibre Company supply 
 only one sort of fibre, whilst Herr Miiller gives the results of tests on eight 
 different varieties, there must be an error in supposing that the material, 
 experimented on was prepared by the Vulcanised Fibre Company." 
 
INSULATION. 
 
 Firstly, the fibre was previously thoroughly dried by being en- 
 closed in an exsiccator for several weeks ; secondly, experiments 
 were made after its exposure for 24 hours only to the by no 
 means damp atmosphere of the laboratory ; and, thirdly, after 
 exposure for several months. Results with walnut wood are 
 also included. (12 L = Legal megohms.) 
 
 Specific Resistance (at 15 C.} per Cubic Centimetre. 
 
 
 -P. Exposed 
 
 Exposed 
 
 
 
 1Jry ' 24 hours. 
 
 several months." 
 
 
 A. B. 
 
 C. 
 
 1. White vulcanised 
 
 
 
 fibre 
 
 2,500 x 10 6 n L. 200 x 10 6 fl L. 
 
 14xl0 6 nL. 
 
 2. Different kind 
 
 
 3,300 
 
 1,080 
 
 i 
 
 22 
 
 3. Light brown 
 
 
 7,4CO 
 
 580 
 
 
 18 
 
 4. 
 
 
 12,400 
 
 1,002 
 
 
 54 
 
 
 5. Ked 
 
 
 16,500 
 
 245 
 
 
 10 
 
 
 6. Black 
 
 
 20,500 
 
 , 2,OCO 
 
 
 68 
 
 
 7. Eed 
 
 
 35,400 
 
 3,250 
 
 
 54 
 
 
 8. Brown 
 
 
 48,500 
 
 , 3,800 
 
 
 26-3 
 
 
 9. Ordinary walnu 
 
 \ 
 
 
 
 
 
 
 woo 
 
 ft 
 
 99,000 
 
 2,870 
 
 
 53 
 
 
 10. 
 
 
 49,500 
 
 21,000 
 
 
 572 
 
 
 11. Paraffined 
 
 
 620,000 x 10 7 
 
 
 3,690 
 
 
 12. 
 
 
 oo 185,000 x 10 b 
 
 
 11,080 
 
 
 13. 
 
 
 x 11,700 
 
 
 1,380 
 
 
 14. 
 
 X) X> 
 
 830 
 
 
 
 \ 
 
 
 It will thus be seen that though fibre, when perfectly and 
 absolutely dry, has decided insulating qualities, it rapidly 
 deteriorates in this respect with exposure. The loss, in fact, is 
 tremendous, even in the first 24 hours. But in the humid 
 climate of England, wherewith by far the greater part of the 
 year the atmosphere is damp, the first two columns in the table 
 would with the better reason be considered as appertaining to 
 a condition of things contrary to general experience. For safety, 
 the last column (C) is the one on which calculations should be 
 based. We here find the specific resistance ranging from 
 
 H 2 
 
100 COMMUTATORS. 
 
 10 x 10 6 megohms (L) for one specimen of red fibre, to 68 x 10 6 
 megohms for a black specimen. 
 
 The means of the results, however, are given as follows : 
 
 White vulcanised fibre about 18 x 10 6 H L. 
 
 Brown 26 x 10 6 
 
 Red 32xl0 6 
 
 Light brown 36 x 1C 6 
 
 Ordinary dry walnut wood 53 to 572xlO ; 
 
 Paraffined 530toll,OCO x 10 6 
 
 In the first table we quoted it will have been noticed 
 that the specific resistance of mica was given as 84 x 10 
 megohms (B.A.), at 20C. This is considerably better than 
 any results with exposed fibre, though by no means high as 
 compared with the other insulating substances that follow it 
 in the table. 
 
 It is obvious, however, that the whole question of insulation 
 is bound up with the consideration of what is practicable and 
 practical under any given conditions. Thus, in the case of 
 commutators, with which we are more immediately concerned, 
 mica, from its natural cleavage and hardness, is eminently 
 suited for putting between the flat sides of the segments. For 
 this purpose, therefore, it has been largely adopted. If we 
 consider the various substances that are used generally for 
 insulation, such as glass, porcelain, paraffin, ebonite, vulcanite, 
 India-rubber, gutta-percha, besides mica, fibre, and so forth, 
 we find none of the first named are applicable for use in com- 
 mutators or, at least, are not used to our knowledge. India- 
 rubber and gutta-percha are too soft ; ebonite and vulcanite 
 are too friable, and wanting in strength ; glass and porcelain 
 have a certain strength; and though one may be tempted to 
 look at these, the practical difficulties of adapting them to a fit 
 would be great, besides the question of their extreme brittleness, 
 and liability to chip, crack, and break. In a commutator, the 
 insulation has to form part of the general structure, and has 
 
INSULATION. 101 
 
 to take strain in common with other material used. This at 
 once limits the substances that are applicable. For the neces- 
 sary strength, therefore, together with practicability of working 
 and other qualifications, we find ourselves more or less con- 
 fined to vulcanised fibre and mica, though asbestos and other 
 materials have been adopted. 
 
 Now, the excessive hygroscopic properties of fibre, to which 
 we have been drawing attention, and which interfere so 
 largely with its quality as an insulator, cannot unfortunately 
 with any success be diminished by paraffining. For on ac- 
 count of its very inferior porosity, the temperature of the 
 paraffin has to be raised very high (about 180C.) to make it 
 penetrate ; and at this heat the fibre decomposes with efferves- 
 cence, becomes brittle, and is rendered perfectly useless. Its 
 swelling with absorbed moisture, which may amount to 
 J^ in-> or even more in .an inch and, contrariwise, its warp- 
 ing with heat present considerable difficulties. For com- 
 mutators especially in the case, for instance, of motors worked 
 out of doors are exposed to every atmospheric change, on the 
 one hand ; while, on the other hand, they are liable to get 
 warm from causes inherent to the machine. The futility, 
 moreover, of any trust in a coating of varnish is apparent, in 
 consideration of this not being able to prevent the effects of 
 heat. In most cases, also, the fibre having been exposed to 
 the air for some days at least, when being machined to shape, 
 will have absorbed moisture before the application of the 
 varnish, while any attempt to dry the fibre immediately 
 before varnishing would probably cause some change of 
 shape. It is thus for these various reasons that some firms 
 are now abolishing vulcanised fibre altogether from their com- 
 mutators, and are using mica throughout. We hence propose 
 now to give our attention more particularly to this latter 
 material, and its adaptability for use in commutators for 
 the cones especially. 
 
102 COMMUTATORS. 
 
 Touching briefly on the sources whence mica comes, it may 
 be stated that it is found and mined, among other places, in 
 New Zealand, Australia, Canada, Bengal, Madras, and Ceylon. 
 There are various kinds of mica, and the term is applied to a 
 group of crystalline rocks having the peculiar laminated struc- 
 ture well known. While a small specimen may easily be sliced 
 thin through the cleavage with a pocket-knife, on the edge it 
 is about as hard as slate. It is found in slabs sometimes 
 two or three inches in thickness. But it cannot be worked 
 like slate, nor even as though it were hard wood or soft metal. 
 When required of any particular shape, size, and thickness 
 it is necessary to build it up to the approximate dimensions, 
 of thin slices, separately worked, and all cemented together. 
 The mass thus built may then be finished off by subsequent 
 tooling. At the Electrical Exhibition at the Crystal Palace, 
 in 1892, there were two exhibits of this substance, in the 
 gallery, by Messrs. F. Wiggins and Sons, and Messrs. Wake 
 and Sanders. In both of these there was a plentiful display of 
 mica strips cut for use between the segments of a commutator. 
 The former exhibitors, however, added two mica rings built up 
 in the manner described, which were ready prepared for turning 
 down into cones for commutators. These were to be observed 
 in a case containing various very fine specimens of machined 
 work, placed round the foot of a column of turned mica, 
 which latter was doubtlessly unique in the Exhibition, as show- 
 ing what can be produced with this material with proper 
 knowledge and skill in its manipulation. Reverting to the 
 rings, however, these are by no means wanting in strength. 
 For, besides being hard on the edge, mica is also tough. A 
 thin sheet will bend considerably before breaking. But 
 unless built up in the manner described, it has no great 
 strength to resist parting in the cleavage. Some firms build 
 up their own rings, though, as intimated, they may also be 
 had from the makers, leaving only the turning to accomplish. 
 
INSULATION. 103 
 
 Very heavy pressure is needed for making the rings those at 
 the Exhibition, for instance, though only about Sin. in diameter 
 and J x jin. in section, requiring no less a pressure than 
 175 tons. But in this manner cones may be obtained of 
 sufficient strength, and as regards insulation and other quali- 
 ties, considerably superior to vulcanised fibre. 
 
CHAPTER XII. 
 
 COMMUTATORS. METHODS OF CONSTRUCTION. 
 
 HAVING thus briefly reviewed some of the different points 
 that- occur in connection with the design and construction 
 of commutators generally, we may proceed now to the con- 
 sideration of definite examples. In so doing, it may be 
 helpful to adopt some kind of classification. Thus a general 
 division may be drawn between those designed to give the 
 greatest depth of wear for the brushes, and those in which 
 shortness of length has been made a special object. Figs. 66 
 to 69 herewith illustrate designs of the first class, each showing 
 a longitudinal section through the upper half of a commutator 
 only, similar to the upper half of Fig. 65, as with other figures 
 that follow. Figs. 66 and 67 show two forms of commutator as 
 made by Messrs. Crompton and Co., Limited ; Fig. 68 is that 
 of Mordey's "Victoria" dynamo; and Fig. 69 shows the com- 
 mutator of the "Helvetia" dynamo, constructed by Messrs. 
 Alioth and Co., of Basle. It should be mentioned, however, in 
 reference to these and succeeding figures that they are not 
 scale drawings from specific examples ; and though suited for 
 the purpose of this work, no guarantee is intended that they 
 all illustrate present day practice. Few parts of a dynamo 
 probably are more affected by change than the commutator ; 
 
106 
 
 COMMUTATORS. 
 
 and it is our purpose to discuss the reasons that lead to such 
 changes, rather than merely to illustrate ultimate results. 
 
 Now, the leading principle of these four examples is that 
 of endurance. For there is no limit to the depth to which 
 
 FIG. 66. Crompton. 
 
 FIG. 67. Crompton. 
 
 FIG. 68. Mordey's " Victoria." 
 
 1 
 
 FIG. 69. " Helvetia." 
 
 the bars may be worn down. They may, in fact, be worn 
 right through an occurrence which is not unknown ; though 
 allowing a commutator to run so long instead of renewing at 
 an earlier period is perhaps scarcely commendable. This is, 
 
METHODS OF CONSTRUCTION. 107 
 
 however, a leading point in these designs. Regarding them 
 more closely, we note that they are all mounted on a sleeve or 
 bush a a, which in Figs. 66, 67 and 69 has a mushroom head, b. 
 In Fig. 68 the head is flat, and is placed at the contrary end to 
 that where the shoulder,/ for the lugs occurs, as compared with 
 the other figures. The segments are held on by a nut and 
 washer cd: and in Figs. 66 and 67, the whole commutator is 
 held on the shaft by a nut and lock-nut fe. In Fig. 69 only 
 one nut is used for this latter purpose. The solid black 
 represents insulation. 
 
 Having regard now more particularly to Fig. 66, it will be 
 noted that the washer d is of the same diameter as that of the 
 combined segments. Thus it differs from the other designs in 
 that these latter have a clear shoulder at the left end, as shown 
 lettered i in Fig. 67. It may perhaps be said, without its being 
 stated as a set rule, that Fig. 66 is more especially suited for 
 small sizes. For this, referring to the nuts and washers, would 
 be on the general principle that parts should never be too small 
 absolutely. On a large scale certain parts might be minute 
 in proportion to the whole, but not so absolutely or in other 
 words to the size of a man. The same parts oa a smaller 
 scale would need to be of increased size in proportion to the 
 whole, to avoid being too small absolutely. Beyond this con- 
 sideration, however, a clear shoulder as at i allows of an un- 
 hampered peripheral surface for wearing down. A groove 
 turned at the other end as shown dotted at h, Fig. 67, leaves a 
 surface clear at both ends. In Fig. 66 the dotted line shows how 
 the wear is confined and hampered at both ends> and illustrates 
 the difficulty, if not impossibility, of keeping the wear even all 
 along, so as to avoid sparking. The surface ih in Fig. 67 has 
 not these difficulties. In a patent for this groove h taken out 
 by Mr. J. W. Easton, of New York, it is recommended that 
 it should be kept filled with some easily removed insulating 
 material, to prevent its filling with metal dust from the wear 
 
108 COMMUTATOKS. 
 
 of segments and brushes, which might otherwise cause short 
 circuits between the bars. Though familiar with the idea of 
 the annular slot some time before the date of this patent, we 
 are not aware to what extent, and with what success, this 
 method has been adopted in practice. 
 
 As to the washers and heads, it will be noted that in 
 Fig. 68 there is one of the former at each end, on account of 
 the flat head. A point, however, to which attention may be 
 drawn is the angle with the axis under the head and washer, 
 as at m and n, Fig. 66, assumed in different designs. This, as 
 will be seen further in subsequent sketches, varies from about 
 60 and 50, to 45, 30, and even less. A- great difference thus 
 exists. The principal point to be attained, however, beyond 
 holding the segments securely against the centrifugal force 
 tending to make them fly off, is to hold them so firmly together, 
 considering the large number of parts forming the whole, that 
 they shall not easily be put out of truth, after leaving the 
 lathe, by any accidental blow or jar. With the larger angle,, 
 the best effect, of course, is to hold the segments tightly 
 between the washer and head ; while the smaller angles have a 
 best effect in holding the segments in toward the axis. It has 
 to be considered, however, that insulating materials have not 
 the same density and solidity as metals. They, moreover, vary 
 in nature, and, as has already been pointed out, fibre is apt 
 to absorb moisture and swell thereby, and to shrink again 
 when dry and hot. Now, this shrinkage would naturally have 
 more effect in loosening the bars diametrically with the large 
 angles than the small. Thus a small angle with the axis 
 might seem the best with fibre. But a substance less liable to- 
 shrink, such as mica, may be preferable for large angles. 
 Conversely also, in consideration of its laminae being in planes 
 normal to the axis of the shaft, a large angle would be better 
 for the mica, when so used. With a small angle, and assuming 
 the possibility of there being any give in the insulation between 
 
METHODS OF CONSTRUCTION. 109 
 
 the bars, and assuming further that these are tightly fitted 
 on the cylindrical body of the sleeve, with hard insulation 
 between, then on the nut being tightened up, and the slope of 
 the washer having its best effect diametrically inwards, there 
 would be a tendency for the ends of the bars within the 
 washer to become jambed against the sleeve, the slope of the 
 washer in a manner acting as a wedge. The result of this 
 would thus be, that the segments dragging on the sleeve 
 would not be driven home so tight under the head as under 
 the washer. In Fig. 68, where the angle is very small indeed, 
 any possibility of this eventuality is met by the fact that the 
 washers at each end lie flat against the nut and head respectively, 
 and also against the bar ends ; and thus the bars are driven 
 directly by the nut. There does not seem much allowance 
 here, however, for any give of the insulation enclosed between 
 the washer and the projecting heels or snugs of the segments. 
 A practical point is to allow a thickness of sleeve where the 
 nut rides for a sufficiently large thread ; and a length parallel 
 with the axis, such that the nut may not be too thin: thus, 
 tight screwing up may be effected without the thread stripping, 
 or the nut bursting, which is of course a matter of primary 
 importance in securing the solidity of the whole mass. This 
 nut, it should be mentioned, requires to be locked to prevent 
 possible slacking back. One method is to put radial screws 
 through it, generally three, with their points taking into the 
 sleeve. Another method is to put one or more screws 
 parallel with the shaft through the nut, with their points 
 taking into the washer, as shown in Fig. 77. 
 
 But now we may turn our attention to commutators of the 
 second kind in our classification. We find these exemplified in 
 Figs. 70 to 73. Of these, Fig. 70 is a Kennedy design ; Fig. 71 
 an example from a " Westminster " dynamo by Messrs. Latimer 
 Clark, Muirhead, and Co. ; Fig. 72 a Giilcher ; and Fig. 73 an 
 early type of Kapp commutator. Here, as was intimated, a 
 
110 
 
 COMMUTATORS. 
 
 special object has been to secure a short total length in 
 proportion to the length of brush contact surface. The 
 difference between these and the first four is at once apparent, 
 in that they are undercut at the ends, and the washer and 
 
 FIG. 70. Kennedy. 
 
 FIG. 71. "Westminster.' 
 
 FIG. 72. Giilcher. 
 
 FIG. 73. Kapp. 
 
 mushroom head more or less let into the ends of the 
 segments. This is especially the case with Figs. 70 and 71 ; 
 while with the former in particular the total length scarcely 
 exceeds that of the segments. A further difference, how- 
 
METHODS OF CONSTRUCTION. Ill 
 
 ever, will be seen, in that while the segments of the first 
 four were all bedded on the sleeve with insulation between, 
 in these latter there is no insulating material here, an air 
 space between being left instead. Moreover, whereas the 
 washers and heads of the former had only one slope, tending 
 to force the segments inwards, in these latter, excluding 
 Fig. 71, there are two slopes. Thus the effect of tightening 
 the nut is to maintain the segments at a constant diameter or 
 pitch, instead of only forcing them inwards. This forms an 
 essential difference in construction. The angles of the reverse 
 slopes are 45 with the axis of the shaft. Now in Fig. 71 we- 
 find this peculiarity that air spaces are left, lettered II, between 
 the inner side of the undercut and the outer periphery of the 
 insulation ; while at the same time there is no bedding on the 
 sleeve, and the washer and head have only one slope. Thus 
 the bars are all simply pressed inwards by tightening the nut,, 
 and held together like the voussoirs of a circular arch or culvert ; 
 the case would be analogous in Figs. 66 to 69 were the interior 
 insulation round the sleeve omitted or but loosely fitted or 
 packed in. If, however, returning to Fig. 71, the spaces II 
 were filled up solid, then the segments here also would be 
 maintained at one set diameter or pitch, as in the case of Figs. 
 70, 72, and 73. Assuming, however, the single slope, without 
 there being any direct support for the bars radially from the 
 centre, then, especially when the commutator is divided into a 
 large number of parts, the taper of each segment on its cross 
 section is so slight, that an accidental blow on the outside, on 
 any one or two of them, might drive them in, in spite of th& 
 tightness of the nut ; thus the whole would be put out of 
 truth, necessitating a return to the lathe. Hence a direct 
 support both inwards and outwards would appear preferable, 
 either by means of the single slope under the head and washer, 
 and bedding on the sleeve, or else by double slopes on washer 
 and head, as in Figs. 70, 72, and 73. 
 
112 COMMUTATORS. 
 
 In these four designs, Figs. 70 and 71 especially, it will be 
 observed that the brushes cannot wear right through, as they 
 were open to do in Figs. 66 to 69, for they would reach the 
 washer and insulating ring before doing so. They all have a 
 clear shoulder, as at i, Fig. 67. It might appear in reference 
 to Figs. 70 and 71 that on much wear taking place the sections 
 of the segments at r would become so reduced as to choke the 
 current flowing to and from the lug. But as so little wear is 
 allowed at all, renewals would be needed before much trouble 
 in this direction could arise, at least, in the case of Fig. 70. 
 Excessive wear may, of course, lead to difficulties here, as else- 
 where. In the other designs the undercut is not nearly so 
 great. It may be noted that there was a slight undercut 
 at the lug end in Fig. 69. 
 
 Regarding the matter now, however, from the point of view 
 of prime cost, we find these latter four examples compare 
 favourably with the former. Thus a slight gain accrues at 
 once from the shorter sleeves. But from the fact that there 
 is no insulation between the bars and the sleeve, and pro- 
 vided the air space thus left is not less than (say) T \in., 
 there is no necessity either to bore the interior of the seg- 
 ments or to turn the outer cylindrical surface of the sleeve ; 
 and so machining is saved. In Fig. 72 a further saving in this 
 same direction is effected by chambering the interior cylin- 
 drical surface, lettered s s, thus leaving only the ends to bore 
 smooth to fit the shaft, while the chambered portion need have 
 only a roughing cut taken off it, if any. Moreover, as the 
 length of the segments themselves in the latter four does not 
 exceed what is required for the brushes and lugs, whereas in 
 all the other designs we illustrate there are projections at 
 both ends beyond these limits, metal is saved in this respect. 
 Economy is also effected, of course, in insulation. 
 
 A distinctive feature in Fig. 73 is that nuts are eschewed, 
 and two sleeves are adopted, the one screwing into the 
 
METHODS OF CONSTRUCTION. 113 
 
 other, as shown. As in Fig. 68, however, it has two washers, 
 one at each end, and both from the same pattern. We may 
 note a further characteristic which these hold in common with 
 the washers in Figs. 68 and 69. This is that they are not 
 centred on the sleeve, excepting in so far as in Figs. 68 and 
 69 the washers there force the segments on to the interior 
 insulation, and so become centred thereon themselves. In 
 the other designs before us the washers are bored to fit the 
 sleeve, and so are centred on it ; and they are thus the better 
 adapted to assist in maintaining the truth of the commutator. 
 But in Fig. 73 they are not centred on anything whatever, 
 and entire dependence for maintenance of truth is apparently 
 placed on the friction of the grip between the sleeve heads. 
 
 An advantage, however, in connection with these latter four 
 designs is that for larger diameters of commutator, while 
 assuming the same diameter of shaft, it is only necessary to 
 increase the diameters of the washers and nuts, simply leaving 
 a larger air space between the segments and the bush. For 
 very large diameters the bush may be double, being com- 
 posed of an inner cylinder riding on the shaft, and an outer 
 cylinder carrying the segments, united by radial ribs to the 
 inner one. By this arrangement passage for ventilation is 
 afforded by the space between the cylinders. 
 
 We now come to four more designs, Figs. 74 to 78. Of 
 these the first two, it should be stated, we borrow from 
 Prof. Silvanus P. Thompson's "Dynamo Electric Machinery." 
 Fig. 74 is another design by Mr. Gisbert Kapp, and Fig. 76 
 one by Messrs. Paterson and Cooper. Fig. 77 is another 
 ** Westminster " design and Fig. 78 is a design adopted by 
 Dr. Hopkinson, though with this latter we are not able to 
 show any insulation. Fig. 74 will be seen to be similar to 
 Fig. 73 in having a double sleeve : otherwise it is totally 
 unlike. It is, however, similar to Fig. 78 in simplicity of end 
 slope, and in having very large angles. Fig. 75 represents 
 
114 
 
 COMMUTATORS. 
 
 Mr. Kapp's latest form of commutator, which we borrow from 
 the very fine collection of plates at the end of the fourth 
 edition of Prof. Thompson's book above referred to. It will 
 be observed that the double sleeves are abolished, only a very 
 fragmentary idea of a second sleeve being represented by a 
 prolongation of the nut. Comparing with Fig. 71, it will be 
 observed that the gaps there lettered I are here inside the 
 
 FIG. 74. Kapp, 
 
 %2%g%%^^ 
 
 FIG. 75. Kapp. 
 
 insulating cones, instead of outside. But this apparently small 
 transposition of the gap is not without effect. It will be seen 
 that it makes all the difference between the circular arch prin- 
 ciple of support for the segments, already alluded to, and a 
 solid central bedding. For with the gap inside the insulation, 
 as shown in Fig. 75, the segments are directly supported by 
 the cones at either end, and so cannot fall inwards toward the 
 
METHODS OF CONSTRUCTION. 
 
 115 
 
 centre, and are, in fact, held at a certain rigid pitch or diameter, 
 in the same manner as with Figs. 70 and 72. The washer 
 rides on a shoulder of the nut, and so is centred on it. It may- 
 be noted further how that, in accordance with previous remarks, 
 owing to the adoption of an air space for insulation between 
 the segments and the sleeve, and to the supporting of the seg- 
 ments at the ends only, a large diameter may be, and is, 
 
 FIG. 76. Paterson and Cooper. 
 
 FIG. 77." Westminster." 
 
 I 
 
 FIG. 78. Hopkinson. 
 
 reached by simply increasing the diameter of the washer and 
 head. 
 
 Figs. 76 and 77 will be seen to have similar segments, and 
 both differ from the remainder in that they have insulation 
 under the ends of the segments only. They thus effect a com- 
 promise in this respect between the first four designs and the 
 second four we have illustrated. Bedding on the sleeve is 
 secured without the expense of carrying this the whole length 
 
116 
 
 COMMUTATORS. 
 
 of the segments, and so saving in both machinery and insula- 
 tion is effected. This method of bedding may be effected 
 with mica, folded round in narrow strips, and in very thin 
 layers : for when very thin, mica is extremely pliable. In 
 Fig. 76 a small washer of triangular section is used, which 
 rides, however, on a shoulder on the nut, and so is 
 centred. In Fig. 77 the washer rides on the sleeve, and is 
 prevented from turning by a pin let into the sleeve, as shown 
 at t. The object of this is two-fold, inasmuch as it prevents 
 any possibility of slewing the segments on tightening the nut ; 
 and also it gives security for locking the nut by means of the 
 screw u put through both parallel with the shaft as shown. 
 Another simple method for securing the same ends is to cast a 
 snug on the inner periphery of the washer, in such a manner as 
 not to interfere with its being turned in the lathe, which shall 
 engage between snugs or in a slot in the sleeve. For the evil 
 of slewing the segments, caused by the washer going round 
 with the nut (when not prevented), and taking the ends of the 
 segments with it, is not an imaginary one ; though many 
 makers rely for its prevention on the tight clamping of all the 
 parts together during the process of tightening up the nut. 
 A further distinctive feature about Fig. 77 is that it is put on 
 the shaft without being held there by a nut or nuts at the end. 
 A sufficiently stout steel screw is put in at the end of the 
 sleeve projecting beyond the nut, as shown at v, experience, so 
 far as we are aware, justifying this as quite enough. For as 
 there is no conceivable force, at least such as could be antici- 
 pated, apart from mere vibration, with a tendency to separate 
 the commutator from the armature, there seems no reason why 
 the simple screw should not be sufficient. This method is, 
 however, not peculiar to the one type we illustrate. 
 
 It is essential with all commutators that they be driven 
 direct from the shaft. Hence we find in Figs. 76 and 78 that 
 the sleeve is keyed to the shaft ; while in Figs. 66, 67, 71, 
 
METHODS OF CONSTRUCTION. 117 
 
 and 77 a steel pin is driven into the shaft, as at g, for this 
 purpose. The nuts, whether of gunmetal, wrought iron, or 
 steel, are usually circular, with radial holes, or else rectangular 
 notches in the rim, for the use of a tommy to tighten up with. 
 But, as shown in Fig. 76, horizontal holes parallel with the 
 shaft may also be adopted ; while two flats, one on each side 
 the nut, for the use of a spanner, is another method. 
 
CHAPTEK XIII. 
 
 COMMUTATORS. CONNECTIONS WITH ARMATURE 
 WINDING. 
 
 THIS chapter we propose to give to the consideration of some 
 remaining items more or less independent of the interior con- 
 struction of commutators. Some of the various methods in 
 which the lugs are formed, and connection with the armature 
 winding achieved, may thus engage our attention. 
 
 As already intimated, the commutator lugs may be either 
 cast in one with the segments, or they may be separate pieces 
 soldered or brazed in. This latter has of necessity to be the 
 method adopted when the segments are of rolled copper, except, 
 perhaps, when, as in cases previously alluded to, the armature 
 is wire wound, and the commutator large in proportion. The 
 lugs when separate are frequently made of two strips of copper 
 riveted together, opening out into a fork at the outer end 
 to receive the winding end. This is shown herewith, Figs. 79 
 and 80. The nitch in the segment receiving the inner ends is 
 cut out with a small circular saw, which accounts for the 
 rounded corner a. 
 
 Figs. 81 to 83 are other variations. Here one strip only is 
 used, bent over at the top as shown, so as nearly to enclose the 
 winding end soldered into it ; and the inner end, instead of 
 being received in a nitch, lies in a recess at the side of the 
 
120 
 
 COMMUTATORS. 
 
 segment, which may, however, be cut with a saw the same as 
 the nitch. It may be riveted to the segment as well as 
 
 FIG. 79. 
 
 FIG. 80. 
 
 soldered. A very common way, instead of enfolding the 
 winding end when this is a bar, is to make a saw cut in the bar 
 
 FIG. 81. 
 
 FIG. 82. 
 
 FIG. 83. 
 
 end to receive the outer end of the lug, as shown in Figs. 84 
 and 85. The lug is then simply a straight strip of copper, 
 
CONNECTIONS WITH WINDING. 
 
 121 
 
 received and soldered or otherwise fastened into a saw cut at 
 both ends. In all cases, the outer end of the lug, beside being 
 soldered to the armature bar, may also, as already intimated, 
 be either screwed or riveted thereto. 
 
 FIG. 84. 
 
 CD 
 
 FIG. 85. 
 
 A case where rolled segments may be combined with wire 
 winding without the use of additional lugs is illustrated in 
 
 FIG. 86. 
 
 FIG. 87. 
 
 Figs. 86 and 87 herewith. Quasi lugs are formed by turning 
 a groove d in the segments. 
 
 FIG. 88. 
 
 FIG. 89. 
 
 Turning our attention now to cast segments, we find a 
 common form in Figs. 88 and 89. Here very short lugs only 
 are necessary, with wire winding, when, as in the previous 
 
122 
 
 COMMUTATORS. 
 
 figures, the wire can easily be bent down to the smaller dia- 
 meter of the commutator. In the case, however, of bar-wound 
 armatures for heavy currents, where the bar ends cannot be 
 
 FIG. 90. 
 
 FIG. 91. 
 
 brought down, the lugs must then extend to the full diameter 
 of the armature, as illustrated in Figs. 90 and 91. This looks 
 very neat, the lugs being all turned up bright as well as the 
 
 FIG. 92 
 
 brush contact surface. Fig. 92 is a variety where the lugs are 
 sloped toward the armature. This has the advantage of allowing 
 more room for the brush-holder, so that the brushes can bear 
 
CONNECTIONS WITH WINDING. 
 
 123 
 
 right up to the roots of the lugs. It also practically elimi- 
 nates the short piece of single conductor bar that passes from 
 the armature end winding to the lugs. For this piece, when 
 the brush contact surface is not wider than one segment, 
 
 FIG. 93. 
 
 FIG. 94. 
 
 has to carry the whole current when in series with the brush, 
 and being only of a size adapted properly to carry but half 
 the current, it would then offer a resistance. Generally, how- 
 
 J L 
 
 FIG. 95. 
 
 ever, this resistance is considered negligible owing to the 
 shortness of the piece ; and the bar does not heat as it 
 otherwise might, owing to the fact that it revolves in the 
 open air, and so is kept cool. But another variety of the 
 
124: COMMUTATORS. 
 
 same thing is shown in Figs. 93 and 94. Here, it will be 
 seen, the lugs, instead of being carried out solid to the outer 
 periphery, are considerably lightened, and open spaces left 
 between them. This latter arrangement, especially when 
 there is much difference between the diameters of the armature 
 and commutator, must be decidedly more economical. For 
 there is less metal and less insulation than in the former, and 
 labour is saved in not having to machine or file up the sides 
 of the lugs. Ventilation, moreover, is completely stopped in 
 Fig. 91, but not in Fig. 94. A good example, however, of the 
 method shown in Figs. 90 and 92 occurs in the case of the two 
 commutators of the Elwell-Parker motor-generator, or con- 
 tinuous-current transformer, already once alluded to. For here 
 the commutators, especially the high-tension, are abnormally 
 large in diameter. Fig. 95 illustrates a further method of 
 connection, used in the case of the Hopkinson adaptation of the 
 Hefner- Alteneck end winding for drum armatures. Here union 
 is made with the interior ends of the cross connections, instead 
 of to the exterior. Hence a bar, screwed to the segments, and 
 soldered or brazed or otherwise fastened to the ends of the 
 involutes i i, is all that is required for the purpose. 
 
CHAPTEK XIV. 
 
 SPARKING AT COMMUTATORS. ELEMENTARY 
 THEORY OF ELECTRIC SPARKING. 
 
 THE sparking that takes place at the brushes on a commu- 
 tator is a subject that may now engage our attention in this 
 and succeeding chapters. It is far, however, from being a 
 matter of mere scientific interest to find out what may be 
 the precise cause or causes of this peculiar phenomenon. For, 
 as seen from a practical, not to say a mercantile standpoint, 
 these scintillations, so brilliant to the eye, are none the less a 
 manifestation of so much wasted energy, destruction of material, 
 and loss in capital or dividends. A double object, therefore, 
 each of weight, will attend any investigation into such a sub- 
 ject ; and it is hence, with this purpose in view, that we may 
 now proceed. 
 
 At the outset of our task various preliminary questions 
 occur, which will have presented themselves in times past, and 
 doubtlessly still present themselves to the inquiring mind in 
 detail. Thus, what are the exact reasons that brushes will spark 
 in one position on a commutator and not another ; and why 
 worse in one position than another ? Why do some dynamos 
 always spark, but not others ? Or, again, why do dynamos 
 sometimes spark, and yet at other times give no trouble in this 
 
126 SPAKKING AT COMMUTATORS. 
 
 direction whatever 1 While a yet further problem of some pre- 
 sent day interest relates to the comparative absence of spark- 
 ing in connection with the use of carbon brushes. 
 
 A better solution of these and many other similar queries 
 will be arrived at if we endeavour in the first instance to 
 attain some clear understanding on the subject of electric 
 sparking generally, apart from that in connection with com- 
 mutators in particular. For it is evident that, without a 
 proper comprehension of the fundamental laws and facts, no 
 really good and satisfactory understanding of more complex 
 portions of the subject ultimately to be arrived at, will be 
 possible. 
 
 Thus a first necessary condition for any sparks is that there 
 must be a difference of potential or electric pressure between 
 two bodies, for a spark to pass from one to the other ; and 
 such a spark will then flash from the body of higher potential 
 to the lower. 
 
 We may now, however, submit electric sparks to a sub- 
 division into five distinct orders. These are as follows : 
 
 (1.) Statical disruptive sparks. 
 
 (2.) Induction disruptive sparks. 
 
 (3.) Self-induction disruptive sparks. 
 
 (4.) Continuous flame, or voltaic arc. 
 
 (5.) Continuous disruptive discharge. 
 
 The above classification, it will thus be observed, is based on 
 differences, partly in the effects and partly in the causes. We 
 may briefly review these orders separately, before more fully 
 studying them in a greater or less measure conjointly. 
 
 The first is distinguished by the fact that it occurs with 
 statical electricity : that is, the electricity which is produced by 
 friction between various substances, such as glass and silk 
 or fur, and resides on the surface of the charged bodies. 
 Two bodies thus charged respectively with positive and 
 
ELEMENTARY THEORY OF SPARKS. 127 
 
 negative electricity, or one of them not charged at all, being 
 brought near together, a disruptive spark will flash between 
 them without their being in contact. 
 
 A spark of the second class takes place in connection with 
 current electricity, such as is generated by a chemical battery 
 or a dynamo. It is of high voltage and is disruptive, and, like 
 the first, requires no previous contact of the objects between 
 which it passes. 
 
 The third, or self-induction spark, is also the result of 
 current electricity, and a high electromotive force between 
 the terminals. It is likewise disruptive, but only follows after 
 previous contact. 
 
 The fourth in our classification, or voltaic arc, is a further 
 phenomenon in connection with current electricity. It is 
 preceded by a spark of the third order ; but is then depen- 
 dent on the electrodes not being parted by too great a gap 
 in proportion to the strength of the current and its potential, 
 principally the latter. This latter condition being observed, it 
 has the peculiarity that it continues over a gap too great for 
 the normal potential to send a spark across, in the well-known 
 manner exemplified by the ordinary arc lamp, and contrariwise 
 to the complete cessation of their being, and the infinitesi- 
 mally small fraction of time occupied by a spark of either the 
 first three orders. 
 
 The continuous discharge forming our fifth order is a simple 
 effect in current electricity, where there is a normal very high 
 potential difference between the electrodes. It is disruptive,, 
 and requires no previous contact. 
 
 We may now, however, subject the above to a more complete 
 analysis, taking the various orders into consideration, not only 
 separately, but as intimated, also in conjunction with one 
 another. Thus, statical electricity, besides residing on the 
 surface, for instance, of jars or brass knobs and so forth, is of 
 extremely high potential it may be of hundreds or thousands 
 
 ^^^ 
 
 TSM 
 
 IW71ESXTT] 
 
128 SPARKING AT COMMUTATORS. 
 
 of volts ; while the current, if it may be said to have any 
 in passing from one object to another, may only be a minute 
 fraction of an ampere. A peculiarity, moreover, of sparks of 
 this class, even when of great length, and so of great potential, 
 is that they have but little effect on the electrodes they 
 play between, such as the brass knobs of a frictional electrical 
 machine, which only become slightly tarnished. But they are 
 far, however, from being necessarily harmless : for these sparks 
 have this further peculiarity, that they are destructive to any 
 dielectric substance that may be between the electrodes, in the 
 sense that they can pierce or break and rupture any such object. 
 Thus a spark may be made to pierce not only the air, but, 
 with a sufficiently high potential difference between the elec- 
 trodes, also glass, paraffined wood, ebonite, or other such 
 materials. Hence the term "disruptive," as applied to this 
 kind of spark. It is in this direction, indeed, that the de- 
 structive effect of lightning is probably to be found. For while 
 2 cloud may be a body heavily charged with statical elec- 
 tricity, say positive, the true negative will be the damp interior 
 of the earth. Objects projecting above the ground, such as 
 trees and tall chimneys and so forth, are for the most part 
 dielectric, though perhaps not entirely so, especially in so far 
 a,s they contain moisture ; but in so far as they are dielectric, 
 the large disruptive spark known as a " flash of lightning " has 
 a destructive effect on them in its endeavour to pass by their 
 means to its true negative in the body of the earth ; while a 
 good conductor, such as a copper rod extending from the 
 summit of the erection down well into the earth, will not en- 
 counter the disruptive qualities of the discharge, and so will 
 escape damage. 
 
 Passing the second and third of our orders for the time being 
 in this more particular review, we may now glance at our fourth 
 order the arc. As a clear conception of the production and 
 continued generation of this form of electrical flame will tend 
 
ELEMENTARY THEORY OF SPARKS. 129 
 
 to a good understanding of our subject as a whole, we may con- 
 sider it with more or less thoroughness. In the first instance, 
 it has to be borne in mind that the carbons in an arc lamp form 
 part of what is termed a " circuit." That is, the electric cur- 
 rent, starting from some generator, is conveyed by a conductor 
 (say, a copper wire) to one carbon; it thence passes to the other 
 carbon, and back by another conducting wire to the negative 
 terminal of the generator. In this case, moreover, at least 
 with " continuous " currents, the electricity passes through the 
 interior of the conductor, in contradistinction to the manner 
 of the statical electricity, which lies on the surface. With 
 the latter, a " circuit " as above described is not necessary ; 
 for there it is a pure case of two bodies charged at dif- 
 ferent potentials, of which, when they are brought sufficiently 
 near, or are connected, the one higher charged will discharge 
 into the lower one, so that both become of equal potential. 
 This may be likened to the case of two tanks of water at dif- 
 ferent levels ; when connected by a pipe the water in the higher 
 one will flow into the lower till both are of the same level. But 
 there is nothing of the nature of a circuit j and when the water 
 in both is of the same level, the action ceases. Current elec- 
 tricity, on the other hand, may be compared to the case of 
 hydraulic distribution of power, wherewith the water is pumped 
 by engines into a series of pipes which convey it to various 
 points where its energy may be utilised, as for working lifts 
 and hoists, and thence bring it back to the suction of the 
 pumps once more. It may be brought back to the pumps, that 
 is. For after having been used, it can be run to waste, or re- 
 turned to the earth instead, and fresh water be pumped into 
 the pipes. The earth may then be said to be the return half 
 of the water circuit. For as the used up water is returned to 
 the earth, so is the fresh water drawn originally from the earth. 
 It is thus also with electric currents, though in practice only 
 with weak currents, such as are used for telegraphing purposes* 
 
180 SPARKING AT COMMUTATORS. 
 
 The return half of the circuit may be the earth instead of 
 another wire or "lead." 
 
 Reverting, however, to the arc, we find the explanation of 
 this phenomenon, which we take from Prof. Silvanus Thomp- 
 son's " Elementary Lessons in Electricity and Magnetism " 
 (though we use our own words), to be as follows: On the 
 carbons being duly connected up in a circuit as described, 
 they are first allowed to be in contact at their points. 
 A current, which experience shows should not be less than 
 5 amperes as a minimum, or at a less pressure than 40 or 
 50 volts, is started through the circuit. The carbons are 
 then drawn slightly asunder. At the first instant a self- 
 induction spark of the third order of our classification, to be 
 dwelt on later, flies from the positive carbon to the negative. 
 This is accompanied by great heat due to the current, which 
 volatilises into gas some particles of the opposing carbons. If 
 now the carbons be drawn not too far apart, this gas will form 
 a conductor for the current from one electrode to the other. 
 The precursory self-induction spark being spent, and the cur- 
 rent now flowing continuously through this gas, intense heat 
 is developed owing to the high resistance of the latter. This 
 causes . the points of the carbons to become white hot, 
 whence arises the light. They are thereby also slowly con- 
 sumed,; and it is to be noted that the positive carbon, or that 
 from which .the current flows, burns down at twice the rate 
 of, the negative. The current flowing continuously through the 
 gas from one carbon to the other, in appearance like a blue 
 ambient, flame, thus forms what is known as the voltaic arc. 
 
 A marked difference will thus here be noted between the arc 
 and the disruptive spark already enlarged on. For, contrary 
 to the case of the spark, the electrodes are here distinctly sub- 
 ject to destruction. A body between the electrodes that is, 
 within the arc will also be destroyed, however, though not by 
 any form of disruption as with the spark. But the intense 
 
ELEMENTARY THEORY OF SPARKS. 
 
 181 
 
 heat of the arc will burn, melt, or volatilise, almost if not quite 
 every known substance. 
 
 Now, as explained, the arc is immediately preceded by, and 
 has for one of its own pre-existent causes, a spark of the third 
 order. To this latter, and the second order, we may now give 
 our attention. Induction and self-induction sparks, like those 
 of the first order, are disruptive. But they differ from the 
 first, not only in being effects of current electricity as inti- 
 
 FIG. 96. Elementary CoiL 
 
 FIG. 97. Induction Coil at " Make." 
 
 mated, but further in that they have their origin in magnetism. 
 This we may now proceed to make more clear. 
 
 It is well known that if a coil of wire be made as shown in 
 Fig. 96 herewith, and a current of electricity be sent through 
 it, a magnetic field will be formed within the coils. If the 
 direction of the current be as indicated by the arrow heads, 
 then, according to Ampere's well-known law, the polarity of the 
 field thus formed will be as also shown by the letters N and S. 
 Should the current be reversed, the poles will also be reversed. 
 
 K2 
 
132 
 
 SPARKING AT COMMUTATORS. 
 
 Moreover, on an iron bar being enclosed within the coil as de- 
 picted, the magnetic effects will be still greater, owing to the 
 iron having a much greater magnetic permeability than air. 
 
 In Fig. 97 are shown two coils, one within the other, 
 having a common centre line, and also enclosing an iron core. 
 Now, especially when an iron bar is enclosed as represented, 
 a further effect may here be realised. This is, that if a sudden 
 current be sent through one of these coils, say the interior 
 
 Fia. 98. Induction Coil at "Break." 
 
 one, it will induce a sudden current running in the opposite 
 direction in the other coil. Further, on suddenly breaking the 
 primary current in the interior, a sudden current will again 
 be induced in the outer or secondary coil, but this time in 
 the same direction as the primary current, as shown in Fig. 98. 
 In Fig. 97 the induced secondary current tends to produce a 
 polarity opposed to the polarity of the primary coil, as shown 
 by the small and large letters ; while in the case of Fig. 98, 
 the secondary polarity augments the primary. But the point 
 
ELEMENTAEY THEORY OF SPARKS. 
 
 133 
 
 immediately to be arrived at, is that this secondary current 
 is purely a magnetic effect that is, a result obtained through 
 the intervention of magnetic induction. On this we may yet 
 further enlarge. 
 
 In Figs. 99 and 100 is shown a simple armature coil in a 
 field between two poles N and S, which may be revolved 
 about an axis x y. Now, in Fig. 99, where the coil is shown 
 quite vertical, and so with its plane normal to the lines of 
 force running from N to S, if the coil be rotated in either 
 direction, by which motion the number of lines of force it 
 encloses is being diminished, a current will run round it 
 
 FIG. 99. Lenz's Law Illustrated in Elementary Dynamo CoiL 
 
 as shown by arrow heads. Its polarity also (as in Fig. 96) 
 will be as indicated by the small s and n. In the case of 
 Fig. 100, however, the coil is represented by the curved arrow 
 as rotating from a horizontal position to a vertical. It is thus 
 increasing the number of horizontal lines of force that can pass 
 through it. The result is that a current will course round it, 
 but in the opposite direction to that in Fig. 99. Now, a point 
 to be noted is that in Fig. 99, the n and s poles of the coil are 
 attracted by the respective N and S poles of the field that they 
 oppose, and hence will offer opposition to the rotation of the 
 coil. In Fig. 100 also, it will be seen that in conjunction with 
 
134 
 
 SPARKING AT COMMUTATORS. 
 
 the direction of rotation, the s and n poles of the coil are being 
 brought nearer to the S and N poles of the field, from which 
 they are naturally repelled. Hence here also there is opposi- 
 tion to the revolution of the coil. The opposition to the 
 rotation, which in these cases is the necessary cause of the 
 current in the coil, and so of the polarity of the coil, thus 
 illustrates Lenz's law, that any primary cause for an induced 
 electric current produces effects ivhick tend to eliminate such 
 primary cause. 
 
 We may now return to Figs. 97 and 98. It should be mentioned 
 first that the iron core is supposed to be without magnetism 
 
 W \L 
 
 FIG. 100. Lenz's Law Illustrated in Elementary Dynamo Coil. 
 
 when there is no current flowing round it. In Fig. 97 we find 
 the sudden influx of the primary current causing the lines of 
 magnetic induction within the core to rise in number from 
 zero up to a certain maximum. This increase in the lines of 
 magnetism is now the real cause of the current generated within 
 the outer secondary circuit, running round in the opposite 
 direction to the primary, as shown by the arrow heads; pre- 
 cisely as an increase is a cause of such an effect in Fig. 100. 
 The result of this, again, is that the secondary current tends 
 to produce poles s and n, which are contrary to the primary : 
 
ELEMENTARY THEORY OF SPARKS. 135 
 
 that is, to eliminate the very influx of magnetism by which 
 the secondary current is itself called into existence. 
 
 Now, while the primary coil may be of a few turns of stout 
 wire, the secondary may be of a large number of turns of fine 
 wire. The potentials of all the turns of the secondary being 
 accumulated, forms an enormously high total electromotive 
 force between the ends of the secondary, brought near together 
 as at g. This distance not being disproportionate to the 
 electromotive force, a disruptive spark will strike across in 
 the direction of the arrow heads at the instant of the sudden 
 increase of the magnetism of the core. The primary current 
 continuing steady, and there being no further alteration of the 
 magnetism of the core, there will be no continuance of current 
 in the secondary, and nothing more will happen except that 
 the full strength of the primary current and the polarity of the 
 core become established after the cessation of the resistance 
 thereto offered by the secondary. But if the primary circuit 
 be now suddenly broken, there will be an immediate decrease 
 of magnetism in the core, from maximum to zero. This will 
 cause a current to flow round in the same direction as the pri- 
 mary, as shown in Fig. 98, contrary to the manner in Fig. 97. 
 A disruptive spark will fly across at g again, but in the opposite 
 direction to that in Fig. 97. The secondary here will cause a 
 polarity identical with that of the primary. It hence tends to 
 maintain the primary polarity, while it is to the very fact of 
 the decreasing of this magnetism that it owes its own existence. 
 We thus see in these two cases a still further exemplification of 
 Lenz's law as above enunciated. 
 
 Now, the sparks that take place at the gaps g in Figs. 97 and 
 98 are induction sparks pure and simple, such as belong to our 
 second order. They are disruptive, and would do more damage 
 to a dielectrical substance in the gap than to the terminals at 
 either side. As explained, their electromotive force would be 
 very high, and the current extremely small. The potential 
 
136 SPARKING AT COMMUTATORS. 
 
 difference, however, between the terminals will depend, as in the 
 case of an armature winding, on the number and on the rate of 
 the increase or decrease of magnetic lines, and on the number 
 of turns in the secondary coil. Thus, with a given rate of 
 making or breaking the primary circuit, a compound coil, large 
 in diameter so as to enclose many magnetic lines, and with 
 a great number of convolutions in the secondary coil, will 
 obviously generate a much higher electromotive force than one 
 of smaller size, and with fewer turns in the secondary. Trans- 
 formers, or converters, as they are variously termed, exemplify 
 this kind of induction. A low tension alternating large current 
 in the primary coil, consisting of a small number of turns of 
 thick wire or copper bar, induces an alternating high tension 
 small current in the secondary, having a large number of turns 
 of fine wire ; or vice versa. The resistance caused by the fine 
 wire, it will be noted, keeps the secondary current small, 
 though its electromotive force may be high. In the high-ten- 
 sion experiments by Messrs. Siemens Bros, and Messrs. Swin- 
 burne and Co., at the Exhibition at the Crystal Palace in 1892, 
 wherewith pressures of 50,000 and 130,000 volts respectively 
 were attained, it was in this manner that such results were 
 arrived at. Powerful disruptive sparks were emitted between 
 electrodes ; while owing to the presence of an appreciable cur- 
 rent also, enormously long arcs, 2ft. or so in length, could be 
 maintained. It is further to be noted here also, that owing to 
 the normal potential being already so high, preliminary contact 
 between the carbons was not required. The precursory spark 
 of the arc was of our fifth order, and not our third order, as is 
 usual with the arc. For here, on the carbon points being 
 brought within about 3 inches of each other, the current would 
 immediately leap across, after the manner of that order, with 
 which we shall deal in due course. 
 
 As we have stated, however, the above refers to induction 
 sparks only, which are caused in a different circuit to that of 
 
ELEMENTARY THEORY OF SPARKS. 137 
 
 the primary. But sparks similar in nature, and based on the 
 same laws, may be obtained in the primary that is, within 
 any circuit itself. Such sparks, with cause and effect taking 
 place within the same circuit, are the result of what is called 
 the self-induction of that circuit, and we may hence style 
 them self-induction sparks. These, forming the sparks of our 
 third order, may thus now engage our attention. 
 
CHAPTER XV. 
 
 SPARKING AT COMMUTATORS. SELF-INDUCTION 
 AND ELEMENTARY THEORY CONTINUED. 
 
 FIG. 101 shows diagrammatically a complete circuit, including 
 a coil with iron core. A generator is indicated at a, and a 
 gap which can be closed at g, while for the present the dotted 
 line c c may be disregarded. Now, a current flowing steadily as 
 represented by the arrow heads, will cause the same polarity 
 of the coil here as in Fig. 96 ; though this would of course be 
 only when the gap g was closed, as otherwise the circuit would 
 not be complete. Let it now be imagined that in the first in- 
 stance the gap g is open. There will then be no current flowing, 
 and so no lines of magnetism within the coil. Suppose, then, 
 the gap to be suddenly closed. We may note the immediate 
 effect. A sudden generation and increase of lines of magnetic 
 induction will take place within the coil, which will exercise a 
 'counter tendency to make the current flow round the coil in 
 the opposite direction to that shown. It will thus momentarily 
 retard the full establishment of the normal main current. But 
 the full current becoming at length established, let the gap g 
 be suddenly opened. The result will be a cessation of the 
 main current, with consequent decrease of magnetism within 
 the core. But a sudden decrease of magnetic lines will cause a 
 
140 
 
 SPARKING AT COMMUTATORS. 
 
 current to flow round in the same direction as the main cur- 
 rent. Hence at the moment of breaking contact, the main 
 current is reinforced with more or less strength and difference 
 of potential, according to the dimensions of the coil and the 
 rapidity of the break. In consequence, a disruptive spark, due 
 to the sudden and great increase of potential between the oppo- 
 sing ends, will take place at g, leaping across in the same direc- 
 tion as the main current, as in Fig. 98. Yet further, inasmuch 
 
 FIG. 101. Diagram Illustrating Self-induction. 
 
 as we are dealing here with the main current itself, and not 
 a secondary, if the points at g (Fig. 101) are not pulled too 
 far asunder, the current may continue crossing the gap as an 
 arc, in the manner already described. It is to be noted that in 
 the cases of Figs. 97 and 98 these self -inductive retardations 
 do actually take place in the primary coils ; and while 
 induction sparks are produced in the secondary, self-induction 
 sparks may be produced at the moment of break in the 
 primary. 
 
SELF-INDUCTION. 141 
 
 We thus find in the case of a self-induction spark the 
 effect of Lenz's law again. For the momentary counter cur- 
 rent tends to retard the cessation of the main current, although 
 it is the very act of this cessation to which the counter current 
 and the sparks owe their existence. But, as already intimated, 
 these induction and self-induction effects are entirely due to the 
 intervention of magnetism. It is thus that the phrase " elec- 
 tric inertia " is not a safe one to use instead of " self-induction." 
 It is not the inertia of the current per se that causes it to fly on 
 after a break of contact, and cross the gap in a spark. A high 
 electromotive force may accumulate at the positive electrode 
 at the break ; but this is not, correctly speaking, analogous 
 to the high-pressure that may occur behind a valve in a pipe 
 containing water flowing under pressure, when it is suddenly 
 closed. In this latter such pressure is due to the momentum 
 of all the water behind the valve, which is suddenly brought 
 to a stop. Momentum implies weight. But there appears to 
 be nothing whatever of the nature of weight in the electric 
 current. Hence the cause or causes of effects manifested by 
 an electric current that might seem due to something like 
 momentum must be looked for in an earlier state of things. 
 This is made clear by the fact that if the wire circuit, as in 
 Fig. 101, be pulled out so as to form a simple circuit with- 
 out any coils in it, and with the iron core also absent, there 
 will be scarcely any self-induction left; and on breaking the 
 circuit at the gap there will be little or no spark, although the 
 actual total length of the wire is unaltered. Thus the absence 
 of any accumulation of electromotive force at the positive elec- 
 trode in this latter case, shows that there is not anything akin 
 to momentum in the electric current, as with the flowing water 
 suddenly stopped by the valve. With the coils it is a pure 
 case of a sudden generation of an extra electromotive force in 
 accordance with the laws of magneto electricity, already dwelt 
 on. This newly-generated electromotive force, impelled by the 
 
142 SPARKING AT COMMUTATORS. 
 
 magnetic forces to which it owes its origin, rises to so high a 
 potential in the positive electrode at the moment of break it 
 may be to some hundreds or thousands of volts that it is able 
 at the instant to send a disruptive spark through the air and 
 across the gap in the manner described. 
 
 We may turn now to the fifth order of sparks which we 
 have assumed. It has been stated that the sparks of the first 
 three orders are the effects of a high potential difference 
 between the electrodes ; though in the second and third orders 
 this will be higher than normal. Our classification is, however, 
 based on the causes rather than on the effects themselves. 
 Thus, our fifth order of sparks are those that result from a 
 high potential of the normal or main current itself, without 
 any magnetic or other augmentation. It is again an effect of 
 current electricity. Two electrodes, for instance, in a circuit 
 of either continuous or alternating current electricity, of say 
 10 amperes and 20,000 volts pressure, if held a few millimetres 
 apart, would have a continuous disruptive discharge passing 
 between them from positive to negative, and that loithout pre- 
 vious contact. This latter characteristic thus alone would dis- 
 tinguish it from an arc. But their brilliancy, as compared 
 with the low luminosity of the arc, that is, of the arc itself, 
 disregarding the intense light from the white hot carbons, and 
 the absence of the intense heat of the latter, together with a 
 difference of the sound produced, would form further distin- 
 guishing features. This fifth order, moreover, is not to be con- 
 founded with the stream of sparks that may be obtained from a 
 Ruhmkorff or other induction coil. For these latter belong to 
 our second order, having for their cause magnetic generation. 
 
 This continuous disruptive discharge of our fifth order, and 
 indeed all disruptive discharges, have however a yet further dis- 
 tinct difference from the arc. This is, that they take the most 
 straight and direct course to the negative. An arc, especially 
 of any length, is easily made to take a circuitous path ; as, for 
 
ELEMENTARY THEORY OF SPARKS. 143 
 
 instance, in the experiments by Messrs. Siemens Bros, at the 
 Crystal Palace, already alluded to. For here the electrodes, 
 instead of being held the one vertically over the other in the 
 usual manner, were held level with each other in the same 
 horizontal plane. Being also about a foot or ISin. apart, the 
 arc then rose from the one ISin. or 2ft. upwards, coming down 
 again to the other electrode, thus forming the shape of a horse- 
 shoe or arch. Yet this great divergence from the straight line 
 to the other electrode was chiefly caused by the air draught 
 due to its own heat and the mutual repulsion of its 
 different parts. The arc, moreover, in this case, was not 
 of a smooth regular form like, for example, a rainbow, but 
 waved and oscillated, and behaved in a manner that may be 
 better likened to an arched blue flame of fire. The rupturing 
 character, on the other hand, of the disruptive discharges, is 
 entirely due to their intense proclivity to adopt a straight 
 course. Hence, rather than be turned aside by any obstacle 
 that may be in the way, they prefer to break it up or pierce it, 
 and so attain their desired consummation. 
 
 Our attention may now be directed to some special points. 
 We have spoken of disruptive sparks and arcs, of current, 
 and electromotive force, or potential. Some effects in parti- 
 cular, however, may be noted. When dealing with the for- 
 mation of the arc, we mentioned the great heat as being 
 due to the current. A high voltage, with or without much 
 current, produces on communication what is called a " shock," 
 best made analogous perhaps to a vibration more or less 
 terrific, according to its strength. But the current burns. 
 A large current at a very low voltage will produce scarcely 
 any "shock,' 7 but may produce great heat. This latter, 
 however, depends on there being a resistance to its flow. A 
 resistance being offered, the work done in overcoming it is the 
 immediate source of the heat that will be manifested. If there 
 be little or no resistance, there will likewise be little or no heat. 
 
144 SPAEKING AT COMMUTATORS. 
 
 Now there is always some heat with every kind of spark; 
 but the point we wish to arrive at may be made more 
 clear perhaps thus. Suppose an animal is to be killed by 
 electricity. It may be put in a circuit with a high voltage 
 and small current, say 20 amperes at 2,000 volts. On switching 
 on the current, it is struck dead instantly by the " shock " 
 due to the high potential ; while the body subsequently may 
 show no signs of burning, for it will not have offered suffi- 
 cient resistance for the small current to take effect. On the 
 other hand, if intense heat be desired to burn or melt any 
 otherwise refractory substance, large currents are required. 
 Thus, we find in the case of electric smelting furnaces, wherein 
 the intensest heat is obtained by means of the arc, that 
 extremely large currents of even 12,000 or 14,000 amperes are 
 used : while there may be only an electromotive force of 30 to 
 60 volts. Thus we observe that it is to the presence of current 
 that heat is due in an arc, with all its concomitant effects, 
 such as burning, melting, and volatilisation ; and it should be 
 noted that the resistance that gives rise to the heat is not that 
 due to the substance being acted upon, but to the arc itself : 
 hence the comparative lowness of the electromotive force. 
 
 We have already dwelt on the destruction of the elec- 
 trodes of an arc. This, however, is not the only source of 
 expenditure in either material or energy. For it is supposed 
 that a back electromotive force is set up in the arc, which 
 impedes the flow of the normal main current. Hence, besides 
 the energy necessary to overcome the resistance of the gas 
 which is the product of the volatilisation of the carbon, energy 
 is also spent in overpowering this back electromotive force. A 
 complete theory of the arc would, however, appear to be want- 
 ing. But at the British Association meetings in Edinburgh 
 in 1892, a Paper on the subject was read by Prof. Silvanus 
 P. Thompson (see The Electrician, Vol. XXIX., p. 460), should 
 any feel interested in knowing more. Further, however, when 
 
ELEMENTARY THEORY OF SPARKS. 145 
 
 considering the arc and its generation, we have observed 
 that it is commenced by a disruptive spark, which may 
 be of either our third or fifth orders, accompanied by great 
 heat, due principally, as we now find, to the presence of the 
 current : for a statical electric spark, wherewith there is 
 practically no current, does not cause much heat. A further 
 point, however, to be noted, is the extremely small amount of 
 time occupied by a disruptive spark. This is, in fact, infini- 
 tesimal ; the time taken by so great a disruptive spark as a 
 flash of lightning being at the most only about the hundred 
 thousandth part of a second. 
 
 We have now reached that stage in our progress, however, 
 when we may bring the above to a practical bearing. Thus we 
 may consider what happens on the opening of an electric switch 
 such as is in common use. When this is done, usually a flash 
 is seen. Moreover, the switch may so be held that an arc is 
 maintained between the moving tongue or contact piece and 
 the fixed contact piece. But allowing it to be opened quickly, 
 even by the means of some spring snap arrangement, the open- 
 ing will still take a very appreciable amount of time com- 
 pared to the extremely small period required for the spark. 
 No switch could endure, whereof the moving parts moved with 
 a speed even approaching that of a flying bullet. 
 
 The upshot, hence, in the case of a flash seen at the opening 
 of a switch, is that, whereas this may have for its commence- 
 ment a self-induction disruptive spark of our third (or fifth) 
 order, the remainder of the flash will be a momentary arc due 
 to the current ; and the greater the current, the greater will be 
 the deflagration.* 
 
 * This point of the brilliancy of the flash, or the deflagration, depending 
 on the quantity of the current, and also that of the intensity of the shock 
 depending on the E.M.F., may be found fully discussed in the earlier part 
 of Chapter IV. of Prof. J. A. Fleming's book on " The Alternate Current 
 Transformer," Vol. I. 
 
 L 
 
146 SPARKING AT COMMUTATORS. 
 
 Now we found when examining the arc between two carbons, 
 that the precursory spark volatilised some particles of carbon ; 
 that is, heated them so intensely as to turn them to gas. 
 Further, we noted that on the establishment of the arc both 
 carbons were burnt, owing to their oxidisation in the air ; but 
 the positive carbon at twice the rate of the negative. Now r 
 metals will volatilise more easily than carbon ; and moreover, 
 they also oxidise when heated and exposed to the air, even 
 when not fused. Hence, in the case of the flash seen with 
 the switch, and drawing our inference from the arc between 
 two carbons, we find that both contacts are exposed to destruc- 
 tion. Firstly, some of the metal will be volatilised by the 
 precursory spark ; and secondly, the positive contact will be 
 burnt and oxidised away at twice the speed of the negative 
 by the following momentary voltaic arc. This, however,, 
 assumes an ordinary current, such as may be large enough to 
 maintain an arc, and an ordinary electromotive force of but a 
 few hundred volts at the outside. In the case of very high 
 voltages, and a current too low practically for an arc, the flash 
 on the opening of a switch would be a momentarily continuous- 
 disruptive discharge of our fifth order. Both contacts may con- 
 sume away somewhat, owing partly to their oxidisation on being 
 heated and partly by volatilisation. But returning to the case 
 of a switch in a circuit where there is an appreciable current,, 
 we find this peculiarity, that, inasmuch as in point of the time 
 taken by the flash by far the larger proportion is occupied by 
 the arc, it is the arc that works the greater destruction, chiefly 
 to the positive electrode. 
 
 Having thus far entered into the subject generally, we ma}*, 
 now perhaps consider ourselves the better prepared to examine 
 with some closeness what is our main subject the sparking 
 that takes place at the brushes on a commutator. 
 
CHAPTER XVI. 
 
 SPARKING AT COMMUTATORS. THE ELEMENTARY 
 PLANES THROUGH COMMUTATOR AND ARMATURE. 
 
 COMMENCING now to give our attention to the sparking that 
 takes place at commutators particularly, it may be well if at 
 the outset we consider what are the conditions that usually 
 obtain with a commutator of a well-made dynamo. For 
 simplicity's sake, we will refer only to the case of an ordi- 
 nary two-pole machine, of a closed-coil type. Such a case 
 is represented diagrammatically in Fig. 102 herewith. It will 
 be observed that a horizontal field is shown, with the north 
 pole-piece on the right-hand side, and the direction of rotation 
 of the armature as indicated by the curved arrow ; and, to 
 avoid confusion, we will assume these conditions throughout. 
 The commutator segments are denoted by the letters c c, and 
 the radial lines b b indicate the coils or sections of the armature 
 winding, which may be either drum or Gramme, and short 
 intervening radial lines represent the lugs connecting the 
 winding to the commutator. 
 
 Now, it was explained in connection with the armature 
 winding in our first chapter, that in the course of the genera- 
 tion of current and electromotive force under the above con- 
 ditions, the coils of the armature winding which are nearly 
 
 or quite horizontal, as at w x, generate the most potential; 
 
 L 2 
 
148 
 
 SPAKKING AT COMMUTATOKS. 
 
 for they are here including or excluding the horizontal 
 lines of force of the field at the greatest rate. Hence each 
 segment of the commutator connected to coils in this position 
 will have its potential raised above that of the one below, by 
 an amount equal to the potential gathered by the coil or 
 section of winding that connects them. Thus there will be a 
 maximum potential difference between adjacent segments at 
 the sides iv and x. But the coils that lie in planes nearly or 
 
 FIG. 102. Diagram showing Commutator, Armature Winding and Field. 
 
 quite vertical, as at y z, are practically not including or 
 excluding lines at all : hence they generate no potential. Con- 
 sequently, there will be a minimum, or practically no poten- 
 tial difference between their commutator segments; and the 
 potential difference between any two adjacent segments will 
 increase as the plane of their connecting armature coil leaves 
 the position normal to the lines of force of the field and 
 approaches a position parallel to them ; and vice versd. Never- 
 theless, on one side of the commutator, say at the top, there 
 
THE ELEMENTARY PLANES. 
 
 149 
 
 will be a maximum potential, and at the lower side, a minimum 
 potential. 
 
 Now, a plane through the axis of rotation, also containing 
 
 FIG. 103. Plane of Coil between FIG. 104. Plane of Coil Cutting- 
 two Commutator Segments. through one Commutator Segment: 
 
 the coil, or section of the winding, connected to the segments on" 
 which the brush rests, is known as the "plane of commutation."; 
 For it is herein, as the armature coils successively arrive into it* 
 
150 SPARKING AT COMMUTATORS. 
 
 that they change the direction of the current within them, or 
 are " commuted." It needs to be clearly understood, however, 
 that this plane appertains to the armature winding only. The 
 segments of the commutator may lie in or about the same plane 
 as the coils to which they are connected, as in the case of 
 Gramme winding ; or with drum winding (approximately), when 
 the parallel conductor bars on the armature run straight out 
 to the lugs of the commutator. Or the segments may be 
 at right angles to their connecting armature coils, as with drum 
 winding, when the connection with the commutator segments is 
 made from the interior of the end-winding that is, from the 
 middle of the cross connectors, instead of from their exterior 
 ends. This interior connection was exemplified by the Hop- 
 kinson and Eickemeyer end-windings illustrated and described 
 in Chapters VI. and VII. Figs. 103 and 104 will, however, 
 make the matter yet more clear. In both of these a b is an 
 armature coil seen endways, extended in a plane x y. But in 
 the former we find this plane cuts exactly between the segments 
 r and s to which the coil is connected: while in Fig. 104 it cuts 
 through the segment s. The latter is perhaps more usual in 
 dynamo construction. 
 
 It is from this point, however, that we shall now need some- 
 what to enlarge. Having thus far made allusion to what is 
 termed the plane of commutation, it will be necessary to 
 expatiate more fully in this direction.* For in order to arrive 
 at a good understanding of our subject as a whole, we may 
 assume five other such planes. These, together with the plane 
 of commutation, may be enumerated as follows : 
 
 * This subject is discussed in Prof. Silvanus P. Thompson's ''Dynamo- 
 Electric Machinery " (pp. 82 to 86 : fourth edition). We have ventured, 
 however, to carry the analysis still further, and to put the subject, as it 
 were, under a greater power of the microscope. Points, we trust of interest, 
 may thus be made to appear, not otherwise discernible. 
 
THE ELEMENTARY PLANES. 151 
 
 (1) The plane of commutation. (Cn P in the diagrams.) 
 
 (2) The neutral plane, or plane of non-sparking. (N P.) 
 
 (3) The plane of requisite reversal. (R R.) 
 
 (4) The plane of maximum E.M.F. through the commutator. 
 <M' M'.) 
 
 (5) The plane of maximum E.M.F. through the armature. 
 <M M.) 
 
 (6) The plane of cessation of force. (C F.) 
 
 We thus discover here six different planes each of which 
 possesses its own special characteristics. So many may perhaps 
 appear confusing, yet their elucidation is not difficult, and may 
 at once be undertaken. 
 
 At the outset, hence, it is to be remarked that the first three 
 planes are entirely concerned with the current, and the re- 
 maining three with the potential, alone. One in each set of 
 three, (2) and (4), appertains to the commutator, and the other 
 four to the armature. In the accompanying diagrams, Figs. 
 105 to 108, these planes are indicated by lines having the 
 letters assigned to each in the above enumeration ; and for the 
 sake of simplicity, the assumption in these diagrams is that, as 
 in Fig. 103, the plane of the coil cuts the edge of the receding 
 segment r. We may now proceed briefly with their several 
 definitions. 
 
 The first, or plane of commutation, already slightly touched 
 on, may thus be more precisely described. Inasmuch as the 
 act of the reversal of current in a coil cannot take place that 
 is, from full strength in one direction to full strength in 
 the other direction without some expenditure of time, however 
 small, this plane of commutation may be considered as occu- 
 pying that position wherein the coils, as they arrive into it. 
 have an infinitesimally small amount of the reverse current 
 commencing to flow in them. Practically, this is, of course, 
 tantamount to saying that it occupies a position where there is 
 no current in the coils. Further, however, it has to be under- 
 
152 
 
 SPARKING AT COMMUTATORS. 
 
 stood that this plane is not a fixed element in the dynamo ; 
 for its position is governed by the brushes, which are movable. 
 This is illustrated in Fig. 106, which shows the planes as they 
 become arranged when the brushes are shifted back. It will 
 be observed that the Cn P plane follows the brushes, though 
 with some degree of elasticity, inasmuch as it now passes 
 
 FIG. 105. Armature Planes. Positive Lead to Brushes. 
 
 through the points instead of the heels. This latter incident 
 will, however, be explained in a future Chapter (XX.). : 
 We have italicised the word current above, for indeed this com- 
 mutation of current is not to be confounded with a change of 
 electromotive force, which is a distinctly different matter. This 
 latter phenomenon is a characteristic of the sixth plane, to 
 be dealt with presently, and, as we shall see, the current in a . 
 
THE ELEMENTARY PLANKS. 
 
 153 
 
 coil may flow quite counter to the electromotive jorce of that 
 coil. 
 
 Now, whereas the plane of commutation had to do with the 
 armature coils, the neutral or non-sparking plane, the next in 
 our order, has to do with the commutator. This is indicated 
 by the letters N P in Figs. 107 and 108. As its name implies, 
 
 FIG. 106. Armature Planes. Negative Lead to Brushes. 
 
 it is that plane through the axis of the commutator, which 
 contains the line on each side, where, if the brush points 
 are resting there, no sparking occurs. Unlike the last, it is an 
 invariably fixed element in a dynamo that is, normally and 
 with a steady load. It does not move with the brushes, but 
 contrariwise, has to be found by shifting the brushes to and fro, 
 and so by experiment noting the position where no sparking 
 
154 
 
 SPARKING AT COMMUTATORS. 
 
 takes place. Of course, in some dynamos there is no such 
 position, or at the best, only a point of minimum sparking. 
 But to this we shall return. 
 
 The third plane, or plane of requisite reversal, though dis- 
 tinct from, is yet intimately connected with, the neutral plane. 
 For when a brush is at the non-sparking point on the com- 
 
 JNA 
 
 FIG. 107. Commutator Planes. Brushes at Non-Sparking Points. 
 
 mutator, this third plane is that occupied by the coil which 
 connects the segment that is in the act of receding from the 
 brush point with the immediately following segment on which 
 the brush rests. Similarly to the two foregoing, the current 
 alone is here concerned; and it will be perceived that this 
 plane is an element of the winding. It is also a normally 
 fixed plane, and lettered R R in the diagrams. As regards its 
 
THE ELEMENTARY PLANES. 
 
 155 
 
 fixity, in fact, it accords with the neutral plane of the commu- 
 tator : and, depending entirely on construction and the method 
 of connecting up the armature winding with the commutator, 
 it might coincide with the neutral plane. This latter, however, 
 would only occur when, as in Fig. 103, the plane of the coil 
 divides the two segments to which its ends are connected, or, 
 
 Ml I 
 
 IM 
 
 FIG. 108. Diagram showing all Planes, with Brushes at the sides. 
 
 more accurately, when it passes through the edge of the re- 
 ceding segment r, as shown. Otherwise, the two planes will be 
 distinct, though at a mechanically fixed and invariable angle 
 with each other, as illustrated in Fig. 104, where wz may 
 represent the neutral plane, passing through the edge of seg- 
 ment r, and at a fixed angle with the plane of the coil x y ; and 
 also exemplified by the angle ft in Fig. 108. It is necessary 
 
156 SPARKING AT COMMUTATORS. 
 
 to postpone a further explanation of this third plane, however, 
 until we can bestow more ample attention on it in the next 
 chapter. 
 
 We now come to the consideration of the last three planes, 
 which have to do simply with the potential. The first of 
 these, No. 4 in the enumeration, is an element of the commu- 
 tator, as already intimated. It is represented in Figs. 107 
 and 108 by the line M' M', and is that plane through the axis 
 of the commutator which contains the points on either side 
 where there are respectively the maximum and minimum 
 potentials. When the brushes are at the non-sparking points, 
 then this fourth plane would lie behind, as shown in the dia- 
 grams. It is normally a fixed element not movable directly with 
 the brushes, and with a steady load, should not vary in position. 
 
 We now arrive at the fifth in our order, the plane of 
 maximum electromotive force through the armature. This 
 applies to the armature, as the fourth applies to the commu- 
 tator. It is indicated by the letters MM in the diagrams. 
 When the brushes are at the non-sparking points on the 
 commutator, this plane would come more or less behind the 
 plane of commutation, and consequently still further behind 
 the plane of requisite reversal, as shown in Figs. 105 and 108. 
 As with the second and third planes, so the coincidence of this 
 fifth plane with the fourth is a pure matter of construction, 
 and would, in fact, take place when the second and third 
 coincided, as in Fig. 103 : or otherwise the two planes would 
 be at a mechanically fixed angle with each other, as lettered 
 a in Fig. 108, or as illustrated in Fig. 104. The condition 
 depicted in Fig. 103, however, being assumed in these diagrams, 
 it will be observed that if Fig. 107 be superposed on either 
 Figs. 105 or 106, the lines NP and M'M' would coincide 
 respectively with R E, and M M. 
 
 It is thus that when the brushes are resting on segments in 
 the M'M' plane of the commutator they will gather the highest 
 
THE ELEMENTARY PLANES. 157 
 
 E.M.F. ; or, in other words, there will then be the maximum 
 potential difference between the negative and positive brushes 
 attainable with the machine. But sparking would thereupon 
 ensue ; and hence, on moving the brushes forward to the non- 
 sparking points in the neutral plane, a loss of E.M.F. is sus- 
 tained. It is therefore desirable that the angular distance 
 of the NP plane should be as little as possible beyond the 
 M'M' plane, which is to be achieved, as we shall find, by 
 making the field very strong in proportion to the armature self- 
 excitation. 
 
 This is, of course, another way of intimating that in a good 
 machine the brushes will require very little lead. On the other 
 hand, as pointed out in connection with Figs. 103, 104 and 108, 
 the M'M' plane itself may be inclined, and so not normal, to the 
 field, as a simple result of mechanical construction, so that the 
 large lead given in such a case to the brushes would be more 
 apparent than real; while further, by the same means, just 
 the opposite results may be obtained, and the machine made 
 to appear as though its brushes required little or no lead. 
 This is doubtless the case in some machines, wherein the 
 armature winding is so arranged that the coils are in advance, 
 in the direction of rotation, of the segments to which they are 
 connected. The armature planes will be ahead of the commu- 
 tator planes, and the N P plane becoming at or about normal 
 to the field instead of being in^front, the appearance of little or 
 no lead will result. What inference may be drawn, however, is 
 simply that the position of the brushes at their non-sparking 
 points, as seen by the eye, is no criterion of the real lead. 
 A distinction in fact may be drawn between apparent lead and 
 true lead. But the position of maximum E.M.F. can be deter- 
 mined with a voltmeter, and the advance beyond that to the 
 non-sparking position will then give the true lead. 
 
 We have now, however, to consider the sixth plane, which we 
 have termed the plane of cessation of force. This is, indeed, 
 
158 SPARKING AT COMMUTATORS. 
 
 the plane of commutation for the electromotive force of a coil, in 
 contradistinction to the plane of commutation for the current. 
 But as the term " plane of commutation " is universally used 
 only in connection with the current, we adopt another term 
 where reference is intended to the E.M.F. 
 
 This plane of cessation of force, therefore, may be regarded 
 as a plane that is invariably normal, or at right angles, to the 
 lines of magnetism traversing the core. It is thus the plane in 
 which the coils, as they successively arrive into it, neither 
 include nor exclude lines of magnetism, and consequently, cease 
 to generate, or are not generating any E.M.F. though they may 
 be charged with the full voltage of the machine (see p. 9). 
 The instant they rotate beyond, they commence generating the 
 reverse E.M.F., whose function is to impel the current coursing 
 in the other half of the winding (see Chap. I). 
 
 In the diagrams, Figs. 105 and 106, this plane, lettered CF, 
 is shown at 90 with a line yy. This latter line represents 
 the centre plane of the induction lines through the core, and 
 passing through the ultimate poles of the armature, represents 
 its ultimate polarity. Another line, x x t represents the centre 
 plane of the field. The angle 6 between them thus indicates 
 the "distortion of the field" alluded to in Chap. I. (p. 13). 
 
 Now, it is evident that as the plane C F is invariably normal 
 to y y, if any cause affects the inclination of the latter plane, 
 then CF must be similarly affected and made to oscillate. 
 Thus, as was explained in Chap. I. above mentioned (pp. 12 and 
 13), the ultimate polarity of the armature is a compromise 
 between its own self-polarisation and the polarity of the field ; 
 and in practice the latter is made much stronger than the 
 former, so that a line (yy} through the ultimate poles will 
 only form a small angle (0) with the centre line (xx) of the 
 field. But it has to be observed that the self-polarity of 
 the armature, depending on the circulation round it of its own 
 current, can of necessity possess its two poles only at the two 
 
THE ELEMENTARY PLANES. 
 
 regions where the current respectively enters and leaves the 
 winding. In other words, inasmuch as the current can only 
 flow in and flow out where the brushes permit, the armature 
 self-polarity is directly governed by, and must follow the 
 brushes. Therefore, if the armature self-polarity is swayed 
 by a movement of the brushes, the ultimate polarity must be 
 proportionately affected, though of course in a smaller degree. 
 Consequently, C F will be affected also, equally with yy. A 
 comparison between Figs. 105 and 106 will reveal this change 
 of position of C F, concomitant with the shifting of the brushes. 
 In a strong field the difference between the C F and M M planes 
 may be small, as shown. But with a weak field, to be discussed 
 in a later Chapter (XXIIL), a considerable angle may exist 
 between the two planes, with much lead to the brushes. 
 
 We thus find that whereas the plane of commutation (1) 
 moves directly with the brushes, and the other four planes 
 were absolutely fixed, and the two through the commutator 
 to be determined by experiment, this plane of cessation of 
 force occupies an intermediate position in this respect, and is 
 what we may describe as semi-fixed. 
 
 It is to be noted, however, that for any given position of the 
 brushes, this C F plane is also the then plane of maximum and 
 minimum potential for the armature. For as the coils at the 
 positive end of each half of the winding occupy positions of 
 successively smaller angular propinquity to this plane, 
 whether approaching it, or receding therefrom (see p. 11, upper 
 paragraph), so they generate less, but become more fully 
 charged with potential; and so, when actually in the plane, 
 though generating nothing, they are charged to a maximum. 
 At the negative end of the winding, of course, the inverse holds 
 good ; and the coils in the plane are not only generating no 
 E.M.F., but are also charged with a minimum potential. 
 
 The normally fixed MM plane (5) may thus be further 
 understood to be that plane which is occupied by the movable 
 
1GO SPARKING AT COMMUTATORS. 
 
 C F plane, when the E.M.F. of the coil or sectton of winding 
 in the latter is the maximum E.M.F. of the machine. As re- 
 flection will make clear, this also can only be when the Cn P 
 plane, movable with the brushes, is coincident with the C F 
 plane. Thus, when the brushes are at the points of maximum 
 and minimum potential on the commutator, the three planes, 
 C F, Cn P, and M M, are one. In other words, the brushes only 
 gather the maximum E.M.F. of the machine when the planes 
 of the coils in series with them are normal to the main flow 
 of induction through the armature, which can only occur when 
 the brushes are in one particular position. 
 
 Now we have stated that when the brushes are in the plane 
 of maximum E.M.F., that is, the M' M' plane of the commu- 
 tator, they will spark. When the brushes are put forward to 
 the non-sparking position, the C F plane, ceasing to coincide 
 with the M M plane, will occupy an intermediate position, as 
 shown in the diagrams. A fall of potential difference between 
 the brushes will result as stated ; and it will be observed that 
 the current in the coils lying between the C F and the Cn P 
 planes, in order to reach the brushes, will have to flow against 
 a reverse, or counter E.M.F. For the commutation of the current 
 in the Cn P plane is taking place in advance of the reversal oj 
 the electromotive force in the C F plane. 
 
 With regard, however, to the sixth plane once more, as this 
 is necessarily normal to yy, it is palpable that any causes, 
 besides, or additional to, that of moving the brushes, that may 
 alter the inclination of y y, and vary the angle 6, will produce 
 a corresponding oscillation of the C F plane. Such other 
 causes, however, may likewise affect all the normally fixed 
 planes. But these we must leave for subsequent consideration. 
 
CHAPTER XVII. 
 
 SPARKING AT COMMUTATORS. ELEMENTARY 
 CONSIDERATION OF THE BRUSH. 
 
 HAVING thus far ordered and distinguished between these 
 various planes, we may now commence more closely to consider 
 their application to the question of sparking at the brushes of 
 a dynamo. Fig. 109, herewith, represents the top of a com- 
 mutator ; B is the brush resting on two segments r and s, and 
 the letters d d and c c show the coils bringing up the current 
 from either side of the armature, as indicated by the arrow- 
 heads. 
 
 New, normally, the current, in order to get from series cc to 
 series d d t or vice versa, would flow through the coil b, which 
 we have shown thicker than the others. But this coil is for 
 the moment " short-circuited " by the brush : for a current 
 could pass between c c and d d through r and s and the point of 
 the brush, which would be a path offering much less resistance 
 than the passage through the coil in question. This latter coil 
 may consequently seem to be idle. So far from this being the 
 case, however, it is the behaviour of this coil that is chiefly 
 concerned in the question of sparking at commutators. For 
 it has but just belonged to the series dd and is about to join 
 the series c c, wherein the current is running in the contrary 
 direction. Hence, it is just in or about this position in relation 
 
162 
 
 SPARKING AT COMMUTATORS. 
 
 to the brush that a reversal of current takes place within the 
 coil b. Thus the plane wherein this coil lies at that instant is 
 termed the " plane of commutation," or plane of change. 
 
 Now, in Chapters XIV. and XV., when discussing the subject 
 of self-induction, we pointed out that that was purely a magnetic 
 effect ; and that the changes of magnetism were the causes of 
 the electrical effects. Thus it is here : for this coil has a 
 
 FIG. 109. Diagram illustrating Elementary Theory of the Brush. 
 
 polarity of its own, which is enhanced by the iron of the 
 armature core within it. Hence, it is in reversing this polarity 
 that energy is lost. This reversal of polarity may not at first 
 seem obvious, inasmuch as the polarity of the iron within the 
 coil is not actually reversed at all at that moment. But the 
 polarization of the iron of the core is the sum total effect of all 
 the polarizing influences brought to bear on it. The coil in 
 
THEORY OF THE BRUSH. 
 
 163 
 
 question exercises its share in these influences, and it is thus 
 this share that has to be reversed. 
 
 It may be noted further that the coil b, though short- 
 circuited by the brush, is still in series with the c c and d d 
 coils. Or should there be any tendency for an independent 
 current to circulate within b, it is free to do so ; and, the 
 
 FIG/ 110. Diagram illustrating Elementary Theory of a Spark at the 
 
 Brush. 
 
 resistance of the one coil, or section of winding, being com- 
 paratively small, a small electromotive force may thus create 
 within b a large " short-circuit " current. 
 
 We may now, however, transfer our attention to Fig. 110. 
 It will be observed here that the segment r is shown just 
 leaving the point of the brush just the time, in fact, for a 
 spark to occur between r and B if the conditions admit. Now, 
 
 M 
 
164 SPARKING AT COMMUTATORS. 
 
 the coil b has hitherto had the dd current direction and 
 polarity within it, as indicated by the arrowhead. But the cc 
 current, immediately on r receding from the brush, has to flow 
 through b in order to reach B. It is evident, however, that if 
 the coil b offers any sort of resistance to the passage of the c c 
 current, with an opposing current or otherwise, and supposing 
 for the moment that this resistance be absolute, so that no c c 
 current can traverse b at all, then the break of contact between 
 the segment r and the brush will be precisely similar to the 
 opening of the gap g in Fig. 101. For the coil b offering no 
 path, the current is broken. Hence, a spark due to the self- 
 induction of the cc coils, will flash across from the segment 
 r to the brush, as indicated by the small curved arrow z. But 
 the parting of this segment r and the brush is furthermore 
 to be likened to the opening of a switch ; for the conditions- 
 assumed are precisely similar. 
 
 In discussing this latter, we showed that the flash then seen 
 was a compound effect, being composed of a self-induction 
 disruptive spark, as just alluded to, immediately followed by a 
 momentary arc, assuming a current of not less than five or ten 
 amperes, and an E.M.F. not exceeding some few hundred volts y 
 which are ordinary conditions. We mentioned, also, the ex- 
 ceedingly small space of time occupied by the precursory spark. 
 Now, however rapid in practice the peripheral velocity of the 
 commutator may be, the flash at the brush will be such that 
 by far the larger proportion of the time occupied by it will be 
 taken up by the arc. Further, when discussing the arc, we 
 showed that the electrodes were consumed by it, the positive 
 at twice the rate of the negative. Hence we now perceive 
 that in the flash between the segment r and the brush, the 
 arcing will preponderate. Both will be burnt ; but the segment 
 here, being positive, will be burnt the most. 
 
 If, however, the resistance of the coil b (that is, any impeding 
 cause or causes, whether due to the simple resistance of the 
 
THEORY OF THE BRUSH, 
 
 165 
 
 coil itself, or that, with adverse inductive influences added) is 
 only partial, the compound flash will be reduced in strength. 
 Some of the current passing through the coil b, there will not 
 be so much to break at the brush point, and the flash will be 
 what we may term a "shunt" flash. This is illustrated in 
 Fig. 101 (page 140) : for if on the gap g being opened, some 
 of the current could still pass by way of a small lead, c c, shown 
 
 l\ 
 
 FIG. 111. Non-Sparking Position. Short- Circuited Coil b in Plane of 
 Requisite Reversal. 
 
 dotted, it is evident that the reduction of the magnetic induction 
 in the core of the coil would not be so great as though there were 
 no such other path. Hence the self-induction E.M.F. at the 
 positive end of the gap will be diminished, and consequently 
 the violence and length of the initial disruptive spark and the 
 magnitude of the subsequent momentary arc. This " shunt " 
 
166 SPARKING AT COMMUTATORS. 
 
 
 
 effect, however, as we shall find later, would only occur close to 
 the non-sparking position on the commutator. 
 
 We may now turn our attention to Fig. Ill, showing the 
 conditions necessary for sparklessness. As may be perceived, 
 this represents the top of a commutator, three of whose seg- 
 ments are lettered respectively r, s, and t ; the armature coils 
 are represented between c and d t two of them being lettered 
 b and/; and three of the lugs are lettered g, h, and *. The 
 " neutral plane " of the commutator is assumed coincident with 
 the "plane of requisite reversal" of the armature, and these are 
 jointly indicated by the line R R. The brush is shown thicker 
 than that in the last diagram, and here covers more than one 
 segment. Two of the coils, b and /, are short-circuited by the 
 brush, and so are shown heavier ; and the short-circuits through 
 the brush point are indicated by the thick dotted lines mno. 
 The large arrowheads show the main current arriving from the 
 two halves of the armature winding, c and d, and the small 
 arrowheads represent the short-circuit currents. It will be 
 noted that the b and/ coils have just left the series d } and are 
 passing into the series c. Having passed the C F plane, the d 
 current that was circulating within them has first been stopped 
 as they each arrived into the plane of commutation, and then 
 a reverse short-circuit current has been set up within them by 
 the c electromotive force. At the moment shown, this short- 
 circuit, or as for brevity we may term it, the s.-c. current, in 
 coil 6, exactly equals the c current. Meeting each other in the 
 lug g, these two currents neutralise each other : hence there 
 is no current flowing either way between the brush and the re- 
 ceding segment r. There thus being no current to break at 
 the recession, there is no sparking ; and the coil b is just pre- 
 cisely in the "plane of requisite reversal" represented by the 
 line R R. The positive brush being here shown, for a repre- 
 sentation of the non-sparking conditions at the negative brush, 
 it is simply necessary to reverse all the arrowheads. 
 
CHAPTER XVI1L 
 
 SPARKING AT COMMUTATORS. THE SHORT-CIRCUITS 
 AND CURRENTS. 
 
 IT is now our purpose more narrowly to investigate the subject 
 of the short-circuits under the brushes ; for without this, no 
 thorough understanding of the theory of the sparking at com- 
 mutators is to be attained. 
 
 On reference once more to Fig. 1 1 1 in our last chapter, it 
 will be observed that in consequence of the thickness of the 
 brush, two coils were short-circuited, as was mentioned. The 
 question, however, arises, as to how these two coils will affect 
 each other in respect of the reverse, or short-circuit, current 
 that will be generated within them. In this connection we 
 may direct our attention to Figs. 112 and 113. We here find 
 in each diagram two circuits P P and Q Q, having a common 
 conductor x y. But in the former the two currents issuing 
 from generators w and v augment each other in x y, while in 
 the latter they oppose each other. It is thus this latter case 
 which we may now consider. For if the two currents are equal, 
 there will be no current in x y ; while if they are unequal, x y 
 will carry the difference. This, in fact, illustrates the essence of 
 what is termed the " three-wire system," x y representing the 
 " neutral " main. It will be noted further that the P current 
 
168 
 
 SPARKING AT COMMUTATORS. 
 
 leaving w, and the Q current leaving ?;, checking each other in 
 x y, they each continue in the other's circuit. More accurately, 
 it may now be said that there is only one circuit with the two 
 generators in series; and, in fact, the two currents may be 
 said to be assimilated. None the less, if the Q current be, 
 say, 250 amperes, and the P current, say, 200, the Q current 
 will split at X, 50 passing down to y, and 200 only continuing 
 by w. Uniting at y, 250 will enter the negative at v. Thus, 
 
 FIG. 112. Diagram showing Two Circuits with a Conductor common to 
 both, in which the currents are united illustrative of augmentation. 
 
 though assimilated, the two currents, taken separately, retain 
 their original values. This does not apply in the case of 
 Fig. 112. For provided x y is large enough to carry both cur- 
 rents, they will wholly take that path, and there will be 
 nothing approaching the nature of assimilation. It may be 
 noted that in Fig. 113 the generators in series cannot each 
 augment the current of the other in the outer circuit iv y v x, 
 because their inherent resistances are also in series. 
 
THE SHORT-CIRCUITS AND CURRENTS. 
 
 169 
 
 Now in Fig. Ill there were two short-circuits bhnmg and 
 fionh, of which the coils b and / formed respectively a gene- 
 rator in each. Inspection will make it clear that these two 
 circuits are similar to those in Fig. 113, and that hn is a con- 
 ductor common to them both, in which their two currents are 
 opposed. We hence perceive that the two short-circuit currents 
 are assimilated with each other, and that h n will merely carry 
 their difference ; and in so far as the coils b and / may be 
 
 1 
 
 n< 
 
 tl< 
 
 FIG. 113. Diagram showing Two Circuits with a Conductor common to 
 both, in which the currents are opposed illustrative of assimilation. 
 
 regarded as in series with each other in the circuit bfionmg, 
 their internal resistances will of course be in series also. 
 Hence, whatever current each may generate, these do not 
 reinforce each other, and each current retains its own value 
 within its own circuit. 
 
 We may now direct our attention to Fig. 114. This may 
 be recognised as a combined diagrammatic representation of 
 Figs. Ill, 112 and 113. The segments r, s and t, the lugs g, h 
 
170 
 
 SPAKKING AT COMMUTATORS. 
 
 and i, and the coils b and /, together with the brush shown 
 dotted, may be observed. On either side the two series of 
 coils v and w represent the two halves of the armature winding. 
 It will be noted that there are here four distinct circuits, P P, 
 Q Q, P'P', and Q'Q', each with a generator, w, b, v and / 
 respectively. These do not include the outer circuit, which is 
 comprised within the line shown partly dotted between n and y. 
 The arrowheads indicate currents, that is, the currents the 
 
 FIG. 114. Diagram illustrating Theory of Short Circuits in a Commutator. 
 
 four electromotive forces tend to maintain for as we shall see, 
 the actual currents will depend on conditions that may be 
 assumed. Of these currents, the P P and P'P' being the main 
 armature currents, may be regarded as invariable both as to 
 direction and quantity ; while the Q Q and Q'Q' currents, 
 contrariwise, will be variable, according as to whether the 
 brush may have too much positive or negative lead. The 
 negative brush is at y, whence the current from the outer 
 
THE SHORT-CIRCUITS AND CURRENTS. 171 
 
 circuit divides right and left through the two halves of the 
 armature winding, as indicated. 
 
 Now a development of Ohm's law, namely, E.M.F. = cur- 
 rent x resistance, is that a current at a given potential having 
 two or more paths open to it, will split itself among these in 
 inverse proportion to their several resistances. Thus the path 
 offering the least resistance will carry the most current, and 
 vice versd. It will now be noted on reference to the diagram, 
 that in the course of their circulation the P main current needs 
 to get from g to y, and similarly, the P' main current from i 
 to y. The combination of the effect of Ohm's law on the one 
 hand, and the interposed effect of the short-circuit currents on 
 the other hand, thus forms a subject which we may now 
 consider. 
 
 The P current arriving at g has three paths before it by 
 which to reach y, namely, g m n, g b k n and gbfion. Similiarly, 
 the P' current has three paths before it, starting from i. It will 
 be obvious, however, that the resistance of the lugs g and i, will 
 be considerably less than that of the coils b and /. Hence, 
 according to Ohm's law as above enunciated, by far the larger 
 proportion of the P current will travel by g m n ; a very small 
 amount may go through b k n, while the flow of P through / 
 will be quite fractional. For the same reason most of the P' 
 current will flow through ion, and small quantities only 
 through / and b. But this, it will be understood, is only the 
 normal flow of the P and P' currents, disregarding the effect 
 of the Q and Q' short-circuit currents. At these latter we may 
 now glance. 
 
 The short-circuit currents being variable, may be assumed to 
 flow in either direction through b and /, and to be equal to, 
 greater, or less, than the P and P' currents. A direction parallel 
 with P is thus assumed in the diagram, as indicated by the 
 arrowheads. It will now be observed, that the Q and Q' cir- 
 cuits having a common or "neutral " conductor hn, like xy in 
 
172 SPARKING AT COMMUTATORS. 
 
 Fig. 113, if the two currents are equal, h n will carry no 
 current, or if they are unequal, it will carry the difference. 
 In actual practice the latter would be the case, and in this 
 instance, where an amount of forward lead is assumed, b would 
 generate a higher E.M.F. than /, and so, the resistance being 
 similar, would impel a larger current. Thus the lug k will 
 simply carry a small inflow of Q current, equal to the difference 
 between Q and Q' ; and these two currents will be what we 
 have termed " assimilated," each flowing in the other's circuit, 
 but retaining its original value in its own circuit. Further, 
 it is to be observed that there will be a normal outflow of Q 
 current in the lug g, and a normal inflow of Q' current in the 
 luge. 
 
 We may now return to the main P and P' currents. It will 
 simplify matters if, ignoring the small quantities of these cur- 
 rents that normally run in b and /, we assume that they flow 
 wholly through the lugs g and i respectively. Hence whereas 
 h n was a neutral between Q and Q', it is now to be observed 
 that g m n is similarly a common path for the P and Q currents, 
 wherein they oppose each other. Thus we here perceive finally 
 that the actual current in gmn will be the difference, if any, 
 between P and Q, whichever may predominate. If they are 
 equal, and the segment r just leaving the brush as depicted, 
 this will be the condition for no sparking as already touched on, 
 as there will be no current to be broken at the moment of 
 recession. But the result now is that as the P current baulks 
 the Q current in g m, it becomes assimilated with it, precisely 
 as Q and Q' are assimilated. The P, Q, and Q' currents become 
 in fact all three assimilated, though each retains its original 
 value in its own circuit apart. Thus the great point to be 
 noted here, is that the P current arriving at g, does not augment 
 the Q current in b, as at first might be supposed. For baulking 
 the latter in g m n it baulks its entire flow throughout its cir- 
 cuit (Q Q Q), and assimilation takes place instead, as explained. 
 
THE SHORT-CIRCUITS AND CURRENTS. 17B 
 
 Now, however, to consider the P' current at e, we find ion 
 also forms a path common to the Q' and P' circuits. But here 
 the two currents do not oppose each other, as did P and Q in 
 gmn. Consequently they in this case augment each other, and 
 both flow together in one large current through ion, always 
 provided that the lug * is of sufficient section to carry the 
 increase of current without becoming heated, and thereby 
 offering an increased resistance. 
 
 It will be advantageous, however, if we assume cases. Thus 
 suppose P and Q be equal, each being 200 amperes. P' would 
 be 200 also, making 400 amperes for the outer circuit com- 
 prised in ny. Q' may be, say, 80 amperes. There will thus 
 be no current whatever in. gmn, and only the difference between 
 Q and Q' = 200 - 80 = 120 amperes flowing inwards in the lug h. 
 The 200 P will be assimilated with the Q and Q' currents ; and 
 the Q' current of 80 is flowing down lug i. Thus the P 
 current of 200 amperes, assimilated within the Q and Q' 
 circuits, may be said to divide at h, 120 going down h n, and 
 the remaining 80 travelling by / i o n. But the P' 200 
 amperes are running unimpeded in i o n. Thus, in this latter 
 path, there will be a total current of 200 P' + 80 assimilated 
 P, Q and Q' = 280 amperes. Being joined at n by the 120 
 amperes down h n, this thus makes a total of 400 in n y ; while 
 there is no flow in nmg. 
 
 From the above, it may hence be perceived, that in the non- 
 sparking position, the effect of the short-circuits is to divert 
 the flow of the current from the point of the brush towards 
 the heel. It should be noted, however, that this is the effect 
 of the electromotive forces of the short-circuited coils. As 
 prime causes these remain unaltered, while their effects, as 
 in the case of the current in g m n, may be entirely eliminated. 
 
 If, however, while P and P' remain each 200, we assume 
 Q to be 100, and Q' say 10, we shall find P - Q = 200 - 100 = 
 100 amperes of the P current flowing through g mil. In lug h 
 
174 SPARKING AT COMMUTATORS. 
 
 Q - Q' = 90 will inflow ; while the total flow down the lug i 
 will merely be 10 assimilated P, Q, and Q' + 200 P' = 210 amperes. 
 At n there will thus unite, 100 from gm, 90 down hn, and the 
 210 down io, making 400 again in ny. But this current of 
 100 in mn would be broken at the recession of segment r from 
 the brush ; and there would occur, in fact, what we have termed 
 "shunt" sparking (p. 165), as only part of the P current is 
 broken, the remainder being assimilated with Q. The sparking 
 will result from the self-induction of the half of the armature 
 winding represented by iv, as described in connection with 
 Fig. 101 and the dotted line c c therein. This condition for 
 shunt sparking is brought about, it will be observed, by the 
 near approach of the brush point to the N P plane, so that 
 the Q current in b is running in the same direction that P 
 would take through b but the brush is not far enough for- 
 ward for Q to equal P, as would be the case were b in the 
 R R plane. It will thus be seen that this shunt effect occurs 
 just behind the neutral plane, and there only. 
 
CHAPTER XIX. 
 
 SPARKING AT COMMUTATORS. THE SHORT-CIRCUITS 
 AND POTENTIALS. 
 
 ^" \ 
 
 IN our last chapter we dealt almost solely with the currents of 
 the short-circuits. It is now, however, the potentials that we 
 propose to discuss. For a point to be noted in connection with 
 the four adjoining circuits represented in Fig. 114, is that 
 whereas to suit the external resistance comprised in ny the 
 E.M.F. of the P and P' main currents needs to be high, say 
 500 volts ; in the case of the Q and Q' circuits, on the other 
 hand, wherein the resistance is extremely small, the E.M.F. 
 will be very low to produce the same current. Thus the latter 
 E.M.F. may be only a fraction of one volt. This minute 
 voltage may thus appear of no account in contention with that 
 of the main currents : while, on the other hand, the question 
 might be as to whether, as a matter of fact, this small electro- 
 motive force of the short-circuits would in any way affect that 
 of the main current. 
 
 In Fig. 115 herewith are two circuits, PPP and Q Q Q, 
 having a common conductor xy^ in which the currents oppose 
 each other, as in previous diagrams. The letters w and v 
 indicate the respective generators. It is to be noted, however, 
 that the P circuit is represented as having a resistance of 
 5 ohms, while the Q circuit has a resistance of only *05 ohm. 
 
176 
 
 SPAEKING AT COMMUTATOKS. 
 
 Thus to get the same current in each circuit, here shown 
 200 amperes, we find needed a potential difference between the 
 terminals of 200 x 5 = 1,000 volts in the P circuit, but only 
 200 x -05 = 10 volts in the Q circuit. 
 
 Further observations, however, require now to be made. 
 First, we note that the generators w and v being in series with 
 each other by way of a b, without resistance (neglecting the 
 resistance of the conductors), their electromotive forces will 
 
 '1000 
 
 1000 
 
 200 
 
 PJ 
 
 o v - 
 
 200 
 
 HWWfti^ 
 
 -SW- 
 
 l 
 
 w-= 
 
 200 
 
 iooo 
 
 1010 
 
 200 
 
 FIG. 115. Two Circuits of Different Potential with currents assimilated, 
 both having resistance exterior to generators. 
 
 accumulate. Thus the negative terminal of v becomes raised 
 to 1,000 volts, the same as the positive of w ; and the positive 
 of v t having the extra 10 added, becomes raised to 1,010 volts. 
 Secondly, we note that x y being assumed without resistance, 
 this conductor is at 1,000 volts throughout and at both ends. 
 
 It will now be observed that as there is a difference of 
 10 volts between the v positive and ^, this 10 volts will impel 
 a current of 200 amperes through the '05 ohm resistance. 
 
THE SHORT-CIRCUITS AND POTENTIALS. 
 
 177 
 
 This current would now naturally, by Ohm's law, flow through 
 y x, as offering the easiest path to its negative at v once more. 
 The P current arriving at a), however, would also, by the 
 same law, flow in the opposite direction entirely through xy, 
 inasmuch as this would be a path of no resistance as compared 
 with the resistance in the Q circuit. Hence, assimilation of 
 the currents will take place : and no current will flow in x y^ 
 as indicated in the diagram by the opposing P and Q arrows, 
 
 1000' 
 
 1000 
 
 200 
 
 PJ 
 
 200 
 
 -mm 
 
 -2S3- 
 
 Q j 
 
 200 
 
 iooo 
 
 Q{ 
 
 200 
 
 10CO" 
 
 116. Two Circuits of Different Potential with Currents assimilated, 
 with no exterior lesser resistance. 
 
 and the numbers struck out by lines drawn through. It will 
 hence be noted, that as here arranged, the voltage of the P 
 circuit is not affected by that of the Q circuit. 
 
 If, however, the terminals of the generators be reversed so 
 that the current runs in the opposite direction, there will at 
 once be an alteration in various respects. The " neutral " will 
 drop to only 10 volts, the same as v ; and the only high poten- 
 tial portion of the circuits would be between w and the 5 ohm 
 
178 SPARKING AT COMMUTATORS. 
 
 resistance, where there would be a pressure of 1,010 volts. But 
 in the diagram, the conductor xy lies between the w positive 
 and the 5 ohm resistance ; and all parts lying between the w 
 positive and this high resistance, including connected parts 
 such as the whole of the Q circuit, will become charged with 
 the high voltage. 
 
 We may now, however, turn to Fig. 116. The difference 
 between this and the last diagram will be observed to be that 
 the 0'05 ohm resistance of the Q circuit is here represented as 
 being the internal resistance of the generator v itself. In 
 Fig. 115 we neglected the internal resistances of the generators. 
 If, however, we assume an internal resistance for w, in the case 
 before us, of, say, 0'2 ohm, this, in order to admit of a current 
 of 200 amperes, will need a pressure of 0-2x200 = 40 volts. 
 Thus the total E.M.F. of w will be 1 040 volts ; but the odd 40 
 being spent internally there will be only the 1,000 for external 
 use.* 
 
 A further peculiarity of the external voltage, however, is 
 that it will depend for its existence on the resistance of the 
 outer circuit. If the outer resistance be diminished while the 
 internal resistance is unaltered, the total resistance of the 
 circuit is lessened. Consequently, a larger current will flow ; 
 and a greater proportion of the total E.M.F. (assuming this 
 to be unaltered) will be spent in forcing the increased current 
 through the internal resistance ; and a smaller potential differ- 
 ence will appear between the terminals externally. Thus if 
 the 5 ohms of the P circuit be reduced to 1 ohm, while the 
 internal resistance of w, as previously assumed, is 0*2 ohm, the 
 total resistance of the P circuit will become 1*2 ohms. With 
 the total E.M.F. of 1,040 volts, we find the current will now be 
 
 * Tliis and succeeding points may be found dealt with mathematically 
 and geometrically in Prof. Silvauus P. Thompson's " Dynamo-Electric 
 Machinery," pp. 226 to 230, fourth edition. 
 
THE SHORT-CIRCUITS AND POTENTIALS. 179 
 
 1,040 -r 1-2 = 866-7 amperes. But the E.M.F. will be re-distri- 
 buted ; and we find that the amount used internally will be 
 866-7 x 0-2 = 173-3 volts as compared with the previous 40 
 only ; while externally the potential difference will be 866-7 x 1 
 = 866-7 volts instead of 1,000; and the total E.M.F. is summed 
 up thus, 866-7 + 173-3 = 1,040 volts as before. 
 
 It will thus now be obvious, that if the resistance is wholly 
 internal, and there is no external resistance, there will likewise 
 be TIO external voltage ; and the E.M.F. will be spent entirely 
 in overcoming the internal resistance. This is the case assumed 
 with the Q circuit in Fig. 116. The resistance here being 
 wholly internal, there is no potential difference externally 
 between the terminals. 
 
 Similarly, however, to the conditions depicted in the previous 
 diagram, the two generators are in series ; and x y and the Q 
 circuit lie between the w positive and the 5 ohm resistance. 
 Consequently these parts are all charged at the 1,000 volts, 
 included both the terminals at v, now without potential differ- 
 ence between them; and the E.M.F. of the P circuit is un- 
 affected by the E.M.F. of the other circuit. If the E.M.F. of 
 the v generator should rise or fall, this would merely cause the 
 Q current to fluctuate ; and there would still be no potential 
 difference between the v terminals externally. 
 
 We may thus return once more to the diagram Fig. 114. 
 The case of the four circuits there is analogous to the case 
 of the circuits in Fig. 116. For the Q and Q' generators have 
 likewise only their own internal resistances and self-induction 
 to overcome, so far as the short-circuits are concerned, and 
 assuming the resistance of the remainder of the short-circuits 
 to be so small as to be negligible ; while the P and P' 
 generators (w and v) have the resistance of the outer circuit in 
 ny to contend with. We may assume similarly that v and w 
 each generates a total E.M.F. of 1,040 volts, of which 40 are 
 spent in forcing the current of say 200 amperes through its 
 
 UNI7ERSITT 
 
180 * .7 SPARKING AT COMMUTATORS. 
 
 own resistance, leaving 1,000 volts to rise at g and i, for which 
 the negative will be at y. It will then be seen that the Q and 
 Q' circuits, lying between g and i, and the point n where the 
 external resistance commences, are at 1,000 volts throughout ; 
 and that however much the E.M.F. of the coils b and /may 
 vary, there will yet, in respect of the short-circuits, be no 
 potential difference between the points g, k and i, assuming 
 the paths h n, and i o n m g, to be without resistance. 
 
 The actual potential difference between the points g, h and i 
 is entirely dependent on the relations between the E.M.F. of the 
 armature winding, as represented jointly by all the coils, w, b,f 
 and -V, and the exterior resistance in n y. In the diagram, the 
 brush is assumed forward: therefore, the point of highest potential 
 will be behind the brush, where the sum total voltage of w, b- 
 and/, and perhaps some more coils toward k, will have accu- 
 mulated. Consequently, the point i will have a higher poten- 
 tial than h, and h than g. But this accumulation of voltage 
 is entirely due to the exterior resistance. For we know that if 
 the latter be eliminated from the outer circuit, supposing it 
 possible, an enormous current would flow, and there would be 
 no potential difference between the positive and negative 
 brushes, showing that the E.M.F. of the successive coils of 
 the armature had ceased to accumulate. Hence in the diagram 
 there would then be no accumulation of the voltage of u; b 
 and /, and so the points g, h and i would all have the same 
 potential. Yet the coils would be still generating E.M.F. 
 (assuming other conditions unaltered) ; but this would be 
 used now internally only, in impelling the current against 
 their own interior resistances. It is thus that the short- 
 circuit currents will continue to circulate, independently of 
 what changes may be wrought on the potential at g, h and i 
 by exterior influences,; and of themselves alone, these Q and 
 Q' generators would cause no potential difference between 
 these three points, except such as might be due to the 
 
THE SHORT-CIRCUITS AND POTENTIALS. 181 
 
 extremely slight resistance of the short-circuit paths, h n 
 and ionmg. 
 
 It must be noted, finally, that the energy required to 
 commutate, and to requisitely reverse the current in the 
 short-circuited coils, is a deduction from the efficiency of the 
 machine. This is evidenced by the fact that the non-sparking 
 points are not in the M' M' plane, but ahead of it. The pro- 
 duct of the potential difference between the M'M' and NP 
 planes, and the short-circuit current, will indicate the power 
 lost in bringing the current in the short-circuited coils to a 
 state of requisite reversal. 
 
CHAPTEE XX. 
 
 SPARKING AT COMMUTATORS. EFFECTS OF 
 MISPLACEMENT OF BRUSHES. 
 
 WE may now consider cases where sparking takes place. 
 The most simple will be those when the brushes are shifted 
 too far forward or backward : or, again, which amounts to the 
 same thing, if the brushes are at their right places, a cause of 
 sparking may be the oscillation of the neutral plane. For 
 the present, however, we may confine our remarks to the 
 simple displacement of the brushes. 
 
 Fig. 117 herewith shows an armature revolving in a field, 
 and a commutator, together with brushes. These latter are 
 shown in positions A and B too far back, and at C and D too 
 far forward. We now propose to deal with these four positions 
 in detail, and to consider the nature of the sparking that occurs 
 in each. Figs. 118 to 121 are thus four diagrams showing 
 each brush separately. For the better elucidation of the 
 subject, a total current output for the machine of 400 amperes 
 is assumed, as indicated in the diagrams. It is necessary, how- 
 ever, to note that in these diagrams the conditions represented 
 are those that are broken by, and so are not the results of, 
 the recession of the segments r from the brushes. Short-circuit 
 (or s.-c.) currents of 500 and 450 amperes are assumed in the 
 
184 
 
 SPAKKING AT COMMUTATORS. 
 
 coils b and /, according as the one or the other is in the better 
 position for generating E.M.F. The main armature currents 
 c and d, of 200 amperes each, are indicated by the large 
 arrowheads, and the s.-c. currents by the small arrowheads, as 
 before. The double arrowheads represent resultant currents, 
 whether by difference or addition ; and it will be noted that 
 the two amounts which together produce a resultant, are struck 
 
 N M 
 
 Cn\ 
 
 /Cn. 
 
 FIG. 117. Diagram showing Brushes with too much Positive or Negative 
 
 Lead. 
 
 out by a line drawn through, in a manner similar to that 
 adopted in Fig. 116. 
 
 It will now be seen, on reference to Figs. 117 and 118, that 
 the brush A being on the d side of the commutator, up to 
 the moment of the recession of segment r, there has been 
 nothing whatever to reverse the electromotive force of the 
 coils b and /. Hence their s.-c. currents will circulate in the 
 
MISPLACEMENT OF BRUSHES. 
 
 185 
 
 same direction as the d current. As indicated by the large 
 arrowheads, the latter still flows in these coils ; though this is 
 to anticipate. The two s.-c. currents, as explained in connec- 
 tion with Figs. 113 and 114, will become assimilated with each 
 other, their difference only, 50 amperes, entering by lug h : there 
 will be a normal outflow of 500 s.-c. amperes through lug t, 
 
 200 
 
 FIG. 118. Positive Brush too far back. Heavy Forward Sparking. 
 (Position A in Fig. 117.) 
 
 and a normal inflow of 450 s.-c. amperes by lug g ; and it will 
 be noted that in this case the s.-c. currents are traversing the 
 coils b and/ in the opposite direction to that shown in Fig. 114. 
 We may now consider the d and c main currents, and may 
 take the former first. This would naturally attempt an inflow 
 
186 SPAEKING AT COMMUTATORS. 
 
 through lug i. It is, however, met there by the normal 7 
 s.-c. 500 outflow. The resultant 300 outflow is thus what 
 remains. But the whole f s.-c. 500 is reduced by this same 
 act. The 200 d, blocked back in i, takes the place of the 
 missing 200 in /, or, in other words, becomes assimilated in 
 that coil ; and the b and / s.-c. currents being already assimi- 
 lated, it becomes assimilated with both these currents, like the 
 P and Q and Q' currents in Fig. 114. 
 
 Reverting to the c main current, we find this inflows through 
 lug g in the same direction as the b current. These thus aug- 
 ment each other, forming a united or resultant current of 650 
 amperes. To analyse this amount accurately, we observe that it 
 is composed of 200 c, plus the d assimilated with the 450 s.-c. 
 current. At the point n this amount is further augmented by 
 the inflowing 50 down h, thus making a total of 700 from n to o. 
 Of this total, 400 go up the brush, being the sum of the c and 
 d currents, while the remaining 300 form the 300 s.-c. current 
 in lug i. We thus find the cycle completed. 
 
 Now, from the above we note that the actual current that is 
 broken is more than the main output, being 650 amperes instead 
 of only the normal 400. Furthermore, we observe that the 
 spark which will occur at the recession will be due to the self- 
 induction of both the c and d series of coils plus the additional 
 effect produced by the short circuits that is, to the self-induc- 
 tion of the whole of the armature-winding. Hence the spark- 
 ing will be heavy, and will be in the same direction as the 
 main current, as indicated by the thick arrows zz. A large 
 current, moreover, being broken, there will be considerable 
 burning with the arcs, especially of the segment r, which is 
 here the positive electrode. 
 
 Fig. 119 may now engage our attention. This represents in 
 detail the brush in position C (Fig. 117), where it is shifted too 
 far forward of the N P plane, which latter is indicated in both 
 diagrams. It will first be noted that the coil b here carries the 
 
MISPLACEMENT OF BRUSHES. 
 
 1ST 
 
 500 s.-c. current instead of /, as in the last case. A resultant 
 50 amperes of the b s.-c. enters by lug h ; and it will be per- 
 ceived that the s.-c. currents here are traversing the b and / 
 coils in the same direction as in Fig. 114. There is thus a 
 normal 450 assimilated s.-c. current entering through i, and a 
 
 FIG. 119. Positive Bt ush too far forward. Light Back Sparking. (Position 
 C in Fig. 117.) 
 
 normal 500 s.-c. current flowing out through lug g t The'c 
 main current, arriving at g, is blocked by the outflowing 
 b s.-c. 500, and becomes assimilated with the latter in the 
 coils b and / instead, in the same manner as the d currentjn 
 the previous case. We thus find the 450 entering by lug i 
 
188 SPARKING AT COMMUTATOKS. 
 
 to be composed of the c, and the s.-c. currents assimilated. To 
 this latter is now added the 200 d main current, making a 
 total resultant current of 650 amperes entering by this lug. 
 Arriving at o, however, 400 amperes split off to ascend the 
 brush, thus leaving only 250 to continue to n. At this latter 
 point, this current is joined by the 50 b s.-c., making the 300 
 out-flowing through g. Thus again is the cycle completed. 
 
 Now here there are various points to be noted. First, it is 
 not the main d and c currents that are broken, for these have 
 both a free entry into the brush through its heel instead of 
 the point ; and the heel continually makes contact with the 
 segments, one after the other, instead of breaking contact, as 
 compared with the case of the brush at A. It is thus the 
 coils behind the brush that are reversed, and not those just 
 leaving the point. The effect will then be that the momentary 
 check due to the self-induction of the coil /will cause the 
 reduction of the 650 in lug i to 50, as it arrives at position h, 
 to be gradual instead of sudden. We shall return to this, 
 however, in connection with the line CnP. The point to which 
 we would now draw attention is thus, secondly, that it is only 
 a short-circuit current of 300 amperes that is broken at the 
 recession; and further, that this is running in the opposite 
 direction to the main current flowing up through the brush. 
 We have here, hence, what has been termed " back-sparking," 
 as indicated by the arrows zz in Figs. 117 and 119. Being due, 
 moreover, to the self-induction of the coils b and / only, the 
 sparking is slight ; while the current broken being but 300 
 instead of 400, or 650 as in the last case, the arcing is not so 
 heavy. The brush, however, being positive, will incur most of 
 the burning. 
 
 We now come to Fig. 120. This represents the lower brush 
 in position B. As it would have been confusing to invert the 
 diagram to make it accord with the actual position of the brush 
 on the commutator, we do not do so, and merely indicate the 
 
MISPLACEMENT OF BRUSHES. 
 
 189 
 
 main current entering by the brush instead of leaving. It will be 
 seen here that the brush is too far back, like brush A ; while of 
 the two short-circuited coils, / carries the 500 amperes, being 
 nearer the plane yy. The first point that maybe noted is 
 that the difference between the 6 and / s.-c. currents, that is,. 
 
 200 
 
 FIG. 120. Negative Brush too far back. Heavy Forward Sparking. 
 (Position B in Fig. 117.) 
 
 the resultant 50, instead of flowing inwards in the lug h, is 
 here flowing outwards. A normal 450 s.-c. current outflows 
 through lug g ; and a normal 500 s.-c. current enters by t. 
 But now the main current from the outer circuit arrives by 
 the brush at o. Its natural course being to split and send 
 
190 SPARKING AT COMMUTATOKS. 
 
 200 amperes to the c coils, and 200 to the d coils, the first- 
 named 200 endeavour to flow out by the lug i. Here, however, 
 this amount is baulked by the normal 500/s.-c. current. The 
 latter hence becomes reduced to a resultant 300 in io, as 
 indicated. Consequently, the first 200 become assimilated 
 with the 500 /amperes in on ; and being joined by the other 
 200 entering by the brush, a total of 700 proceeds to n. Here 
 the 50 pass up nh. The remaining 650 run on to g t at the 
 outer end of which 200 diverge into the d series of coils, while 
 the 450 run through b. This latter current, joined by the 
 resultant 50 up h, makes 500 amperes in the coil /. At the 
 outer end of i, 200 separate into coils c, leaving the 300 to 
 continue down lug i, as we found at the commencement. 
 
 What we may here note is that the current broken at the 
 recession is the whole outer main current, plus 250 amperes of 
 the short circuit current. There is the whole self-induction of 
 the outer circuit, together with the added effects of the short 
 circuits, thus brought into play. As the outer circuit may 
 include the field magnet coils, the sparking may here be very 
 heavy ; and with the large current, considerable burning will 
 take place, principally, as it will be seen, at the expense of the 
 brush. 
 
 Fig. 121 indicates the last of these four positions. The 
 coil b here carries the 500 s.-c. current, and a resultant s.-c. 50 
 flows out through the lug h, as in the previous case. A normal 
 500 s.-c. current inflows through g t and a normal s.-c. 450 out- 
 flows through lug i. The 400 amperes now arrive at o by the 
 brush. Of these, 200 would pass to the d series by lug g, but 
 are met by the inflowing 500 b s.-c. current, and reduce that to 
 300 instead. They become assimilated with the s.-c. current, 
 and so form part of the 450 s.-c. current outflowing through 
 lug i. The remaining 200 of the outer current augment this 
 450, thus making a total of 650 finding exit by lug i. At 
 the outer end of this lug, 200 split off to enter the c series of 
 
MISPLACEMENT OF BRUSHES. 
 
 191 
 
 coils, while the 450 continue through /. This latter amount 
 is joined by the 50 s.-c. current from h, making 500 amperes in 
 coil b. Of these, 200 diverge into the d series, leaving the 300 
 to continue down lug g. Arriving at n, the 50 pass up n h, 
 leaving only 250 to proceed to o. So the cycle of operations is 
 once more completed. 
 
 FIG. 121. Negative Brush too far forward. Light Back Sparking. 
 (Position D in Fig. 117.) 
 
 It will be remarked that we have back-sparking here, as in 
 the case of the brush in position C ; for it passes in the contrary 
 direction to the main current. Moreover, as in the latter case, 
 there are only 300 amperes broken, and these of short-circuit 
 current. The sparking therefore will be merely that due to 
 
192 SPARKING AT COMMUTATORS. 
 
 the self-induction of the short-circuited coils b and /. It will 
 hence be slight ; and with the small current the arcing will not 
 be great, though the segment will suffer the most. As with 
 the brush C, the main current passes through the heel of the 
 brush, and not the point, and so escapes being broken. 
 
 We may now notice some general results and conclusions. 
 Thus we observe that, as shown in connection with the diagram, 
 Fig. 114, all these effects are produced by the assimilation of 
 main current with the short-circuit current. Yet it is only 
 one-half of the main current that is so assimilated, and hence 
 diverted ; and, inasmuch as by Ohm's law one half of the main 
 current would flow through the point of the brush and the 
 other half through the heel, it is simply a question finally as to 
 which half of the main current shall be thus caused to diverge 
 At the non-sparking point this divergence by assimilation is 
 just precisely achieved, and all flow of current in either 
 direction withdrawn from the brush point. But there is too 
 much of the diverting s.-c. current when the brushes are too 
 far forward, so that its surplus is broken and causes sparks. 
 Whereas, when the brushes are too far back, not only is the 
 divergence of main current in the wrong direction that is, 
 from heel to point but there may be (and is, in the cases we 
 have assumed) a surplus of the diverting s.-c. current also : 
 hence, the grand total, composed of the whole of the main 
 current and the surplus of s.-c. current, is broken, with violent 
 sparking due to the self-induction of the main circuit as a 
 result. 
 
 Another point that may engage our attention, concerns 
 the position of the plane of commutation in relation to the 
 variously misplaced brushes. On an earlier page we stated 
 that this followed the brushes, but with some degree of 
 elasticity. It will be observed that this plane, lettered Cn P 
 in the diagrams, is shown in Fig. 117 (lettered Cn only) with 
 each pair of brushes. At brushes A and B it runs through 
 
MISPLACEMENT OF^BRUSHES. 193 
 
 the points, while at C and D it passes through just within 
 the heels of the brushes. 
 
 If we thus examine Figs. 118 and 120, showing brushes A and 
 B, we shall observe that on the recession of segment r, and the 
 breaking of the assumed [current of 650 amperes, a sudden 
 reversal is forced to take place in the coil b, by its being 
 thrown completely among the series c in the former, and d in 
 the latter. The coils b, in fact, in both these cases are thus 
 suddenly commutated at that exact instant ; and the current 
 in them, subsequently, will run counter to their E.M.F., until 
 the C F plane is reached, as explained in Chap. XVI. (p. 160). 
 Hence the plane they are then in is the plane of commutation 
 for those positions of the brushes ; and if the plane of the coil 
 runs through the points of the brushes, as shown in the dia- 
 grams, so also will the plane of commutation. 
 
 To turn now, however, to Figs. 119 and 121, showing brushes 
 C and D, we find here that it is not the forward coils b that are 
 first reversed, but the hindmost coils, / ; and in the diagrams 
 this reversal is assumed to have taken place, and b and /are 
 both prepared to be thrown among the other series of coils they 
 are approaching, excepting in so far as they are carrying a cur- 
 rent which is too great. It will be noted that directly on the 
 two segments t making contact with the heels of the brushes, 
 the coils /will have become short-circuited. But up to that 
 instant they have had within them the current direction and 
 polarity of the coils behind the brushes. This has now to be 
 reversed. But it has to be further remarked, that the E.M.F. 
 in the coils between / and the N P plane, has already been 
 reversed by the fact that, with brush C, they are on the c side 
 of the C F plane, and with brush D are on the d side of C F. 
 The reversal of the current is thus so far prepared for ; and it 
 is merely a case of overcoming the self-induction of coil / which 
 is now accomplished by the proper E.M.F., already acquired. 
 As we have already pointed out, this reversal is not here neces- 
 
194: SPARKING AT COMMUTATORS. 
 
 sarily sudden, for it takes place [after a making of contact 
 between the brush heels and segments t, and not at a break. 
 Hence the exact position of the plane of commutation will be 
 that plane occupied by the coils / when they have an infini- 
 tesimally small amount of the reverse current from b com- 
 mencing to run in them with which, it will be borne in mind, 
 the current in the series of coils forward of the brushes, is 
 assimilated. And this position, with the plane of the leading 
 short-circuited coil cutting through the points of the brushes, 
 we have assumed to be just forward of the heels, as indicated 
 by the lines On P in the diagrams [of the brushes C and D 
 under discussion, and lines Cn in Fig. 117. 
 
 We have alluded to the factjthat in the cases of the brushes 
 at C and D, the currents broken at the recession of the seg- 
 ments r are merely short-circuit currents. A question might 
 arise as to exactly in what manner the self-induction of these 
 short-circuits takes effect, so as to produce sparking. In- 
 spection, however, will show that this is due to the sudden 
 reduction at the moment of recesion, of the 500 amperes in 
 the coils b, to the normal 200 of the series of coils among which 
 they are being thrown. For by the operation of Lenz's law, 
 already dwelt on, so sudden a reduction will be resisted, in 
 like manner, though not to the &ame degree, as a sudden 
 reversal. Hence at the recession of r, a momentary very high 
 E.M.F. will accumulate at the point of the brush, causing a 
 discharge of disruptive sparks into r, followed by arcs, in the 
 case of brush C ; while the discharge will be from the segment 
 r to the brush, in the position D. Both being contrary in 
 direction to the main current, there is thus formed what has 
 been termed back-sparking. 
 
CHAPTEE XXL 
 
 SPARKING AT COMMUTATORS. NON-SPARKING 
 WITH CARBON BRUSHES. 
 
 IT will not be inappropriate if we here devote a chapter to 
 the subject of the comparative absence of sparking that 
 attends the use of carbon brushes, to which reference was 
 made on an earlier page. 
 
 It may thus be assumed that with carbon brushes there 
 are two causes in particular that tend to prevent sparking. 
 Of these, the first is the higher resistance of the substance as 
 compared with metal : while the second is the resistance of the 
 contact between the brush and the segments of the commutator. 
 It is the latter which we shall endeavour to show is by far the 
 more important in accounting for the absence of sparking; 
 though the former may also have its effects. 
 
 Now, when carbon is used in any shape to form part of an 
 electrical circuit, it is an understood thing that great care is 
 always necessary in securing a good contact between it and any 
 metal connection. Thus, in the case of carbons for a cell, we 
 find the upper ends carefully encased in lead, and so a large 
 surface of the actual carbon put in contact with that metal, to 
 which latter terminals may be attached. Or again, as with the 
 
 brushes now under discussion, it is considered distinctly advis- 
 
 o2 
 
196 SPARKING AT COMMUTATORS. 
 
 able to thoroughly copper-plate the outer ends that go into the 
 brush holders ; and not merely the ends, but a great part of 
 the whole brush. The conclusion is thus negatively deducible, 
 and indeed generally recognised, that the contact between 
 carbon and metal, even when the contact surfaces are as bright 
 and smooth as they can become, is yet not so efficient for the 
 passage of current, as the contact between two bright metal 
 surfaces. 
 
 An immediate effect to be apprehended as a result of the 
 combined non-inductive resistances of the contact between the 
 carbon brush and the segments of the commutator, and of the 
 carbon itself, will be the almost complete cessation of the short- 
 circuit currents. For, as pointed out, these being impelled by 
 but small electromotive forces, a small resistance only may 
 reduce them to nearly or quite negligible amounts. We thus 
 find that the theory of non-sparking dependent on short-circuit 
 currents will not apply in the case of carbon brushes. 
 
 Now we have mentioned that, apart from the interposed 
 effects of the short-circuit currents, the main current will 
 simply follow Ohm's law in its flow to and from the brushes. 
 We also called attention to the fact (page 171) that as an out- 
 come of this law, the current, having two or more paths in 
 which to flow, would divide itself among these in inverse pro- 
 portion to their several resistances. But we then concluded 
 that (with the metal brushes), as two of the paths traversed 
 coils having resistance, most of the current would flow through 
 the one path of practically no resistance ; and the flow through 
 the coils might be considered negligible. In the case under 
 discussion, however, this conclusion requires to be modified. 
 
 Fig. 122, herewith, illustrates part of a commutator as in 
 previous diagrams, rotating in direction indicated by the curved 
 arrow. The hatched portion represents the contact surface of 
 a carbon brush, extending equally over the segments r and 
 s. One coil only, 6, is thus short-circuited ; and what short- 
 
CARBON BRUSHES. 
 
 197 
 
 circuit current may be generated is indicated by small arrow- 
 heads, as heretofore, the brush being assumed at position A in 
 Fig. 117. 
 
 Now it is obvious that, taking the simple resistances alone 
 into account, and with the c and d currents opposing each 
 other equally in the coil b, the whole of the c current will pass 
 through lug g, and the d current through lug h. But two 
 
 FIG. 122. Carbon Brush. Short- Circuited Coil not reversed. 
 
 circumstances will here require attention. First, whereas with 
 the metal brush the disparity between the resistance of the 
 one path through the lug g, and the resistance of the path or 
 paths through the coils b and / was considerable ; with the 
 carbon brush, on the other hand, inasmuch as an additional 
 resistance due to the contact is inserted in each path, this 
 
198 
 
 SPARKING AT COMMUTATORS. 
 
 disparity between the resistances of the different paths is 
 diminished ; and, as further, these inserted resistances are of a 
 varying nature owing to the advance of the segments, the dis- 
 parity may at times even be reversed, so that the total 
 resistance of a path through a coil, may be less than that 
 through the lug g. Secondly, whereas in the case of the metal 
 
 FIG. 123. Carbon Brush. Short-circuited Coil commutated. 
 
 brush at A, a large short-circuit current would be generated in 
 fc, which by assimilation would divert the entire flow of the d 
 current from h into g, and so both the c and the d currents 
 would flow through g\ with the carbon brush, owing to the 
 smallness of the short-circuit current, the divergence of main 
 current by that means would also be extremely small. We 
 thus find that, whereas with the metal brushes the second 
 
CARBON BRUSHES. 199 
 
 count, that is, the question of short-circuit currents, was all 
 important, with the carbon brush, on the other hand, it is the 
 first count, touching on the varying resistances of the contacts 
 between the carbon and the segments, that is of the greater 
 consequence. 
 
 Turning now to Fig. 123, wherein a slight rotation of the 
 commutator beyond the position shown in the last diagram is 
 assumed, we find that the brush contact on segment r is but 
 little more than half that on segment s. Hence the resistance 
 of the contact area w x y z will be nearly double that of the 
 contact on s and the sum "total of resistance by g r into the 
 brush, will be much greater as compared with that by b h s, 
 than it was in the last diagram. Consequently, a much more 
 considerable proportion of the c current will tend to flow by 
 way of b h s. But the d current obeys the same laws as the c 
 current : it has one path of minimum resistance by h and the 
 contact on s, and another path of greater resistance by b g r. 
 Taking c and d, however, jointly, and ignoring for the moment 
 the resistance of 6, we find that of the total flow, about two- 
 thirds will pass by h s and one third by g r, in proportion to 
 the contact areas respectively on the two segments. Hence a 
 flow of main, that is of c, current, will pass through b ; and 
 unless the short-circuit E.M.F. generated singly by the coil or 
 section of winding 6, operating against the resistance of both 
 the contacts on r and s and the resistance of the carbon, is able 
 to impel a s.-c. current of greater strength than which would 
 also be in the opposite direction to the main c current in 6, 
 then the s.-c. current in b will be eliminated ; and in fact the 
 coil b will be commutated. It has to be noted, however, that 
 the resistance of b will have a damping effect on the passage of 
 Current through it, whichever the direction of the flow. If the 
 resistance should be very great in comparison with that of the 
 contacts on r and s, the flow of d current would obviously be 
 confined almost entirely to the lug h, and the c current to the 
 
200 
 
 SPARKING AT COMMUTATORS. 
 
 lug g, up to the moment of recession. But further, the self- 
 induction of b will offer opposition to a rapid commutation of 
 its original d current direction and polarity to those of the c 
 current. It will, however, thus be perceived that if both the 
 resistance and the self-induction of the short-circuited coil or 
 section of winding 6, are small as compared with the resistance 
 
 FIG. 124. Carbon Brush. Diversion of Current from Forward Edge of 
 Brush prior to recession, owing to Resistance of Contracted Contact 
 on Receding Segment. 
 
 of the contacts of the carbon brush on the segments r and s, 
 an early commutation of b may take place, soon, in fact, after 
 the area -on s has become greater than that on r, as represented 
 in the diagram. 
 
 The effect of the alteration to the conditions depicted in 
 Fig. 124 may now be considered. We observe that, owing to. 
 
CARBON BRUSHES. . 201 
 
 the contraction of the area wxyz with the rotation of the com- 
 mutator, the resistance of the path by lug g and segment r into 
 the brush is greatly increased ; while, on the other hand, 
 another entry into the brush has opened up by the lug i and 
 segment t. The coil / is short-circuited as well as 6, though 
 not as yet commutated ; and the d current may enter the 
 brush by the paths it or fh s. But b being commutated, and 
 the resistance by g r becoming considerable, the c current will 
 continue to flow more by b h s and less by g r, up to the moment 
 of recession. Hence at the recession finally of segment r from 
 the brush, there will be little or no current flowing between 
 them, and so little or no sparking. 
 
 It has now to be remarked that the above line of argument 
 will apply to all positions of the brush, as at A, B, C, and D, 
 Fig. 117, or any position between. For a positive or negative 
 lead given the brushes simply affects the direction of the short- 
 circuit currents ; and these we have shown to be of but little 
 importance with carbon brushes. 
 
 The forward edge of the brush, w y, is found liable in practice 
 it would appear, to get very hot. This is thus to be accounted 
 for by the fact that, owing to the resistance offered by the self- 
 induction of the coil or section of winding, b t to the continued 
 increase of the c main current through it, the current density 
 through the area wxyz is abnormally high, as compared with 
 that through the rest of the contact. But further, inasmuch 
 as the ratio of the area w y x z to the remainder of the brush 
 contact area, diminishes, not at a constant rate, but at an 
 increasing rate : so also the rise of c current through b is not 
 uniform, but takes place at an increasing rate, especially 
 increasing at the last instant up to the actual recession. Thus 
 the self-induction of 6, tending to retard the increase of c 
 current through itself, acts more powerfully as the moment 
 of recession approaches, or in other words, as the area, iv x y z, 
 becomes infinitesimal. Hence, as this area narrows to a mere 
 
202 SPARKING AT COMMUTATORS. 
 
 line^the current density through it increases up to a maximum 
 at the moment of recession. The edge w y gets heated ; while, 
 nevertheless, the actual total amount of current passing is too 
 small to produce sensible sparking. 
 
 It would appear that carbon brushes have been used with a current- 
 density approaching 200 amperes per square inch. But 40 or 50 amperes 
 per square inch of contact surface is given as a practical maximum, and 
 30 amperes as a safe working density, as the result of experience. Discus- 
 sion and correspondence on this subject may, however, be found in The, 
 Electrician, Vol. XXIX., pp. 192, 233, 265, 270, 287, 288, 437. 
 
CHAPTER XXII. 
 
 SPARKING AT COMMUTATORS. CAUSES EXTERIOR 
 TO MACHINE. 
 
 WE are now about to enter on another distinct branch of our 
 subject. Hitherto we have dwelt principally on the direct 
 causation of sparking at the commutator. But there are other 
 points which, though seemingly direct causes of sparking, are 
 nevertheless more correctly perhaps, to be regarded as indirect 
 causes. To these we have already made some slight allusion ; 
 and we refer to those faults or accidents in a machine or the 
 circuit, which lead to the oscillation or deflection of the normally 
 fixed planes. For, as we have pointed out, the oscillation, or 
 partial rotation, backwards and forwards round the axis of the 
 shaft, of these planes, amounts virtually to a shifting to and 
 fro of the brushes in relation thereto. In the case of such 
 oscillation, therefore, it is the disagreement between the 
 positions occupied respectively by the brushes, and the normally 
 fixed planes, that is the ultimate direct cause of the sparking 
 that would ensue. 
 
 The stability of all the normally fixed planes, such as the 
 planes of maximum and minimum potential, of requisite 
 reversal, and of non-sparking (see Chapter XVI.) is dependent 
 on the fixity of the centre line or plane of distortion lettered 
 
204 SPARKING AT COMMUTATORS. 
 
 y y in the diagrams, and consequently on the invariability of the 
 angle 9 between yy and xx in the same diagrams. An ex- 
 ception that may be taken in respect of the plane of requisite 
 reversal, is that inasmuch as this forms an angle with the C F 
 plane through the armature, which latter plane is invariably 
 perpendicular to y y, the angle 9 may open to the same 
 extent that the angle between RR and CF may close; and 
 vice versd. Hence in this latter case, the plane of requisite 
 reversal, and consequently the non-sparking plane through 
 the commutator, will not oscillate, although the angle 9 may 
 slightly vary. But apart from this contingency, the position 
 of y y affects that of all the normally fixed planes ; and the 
 slope of y y, or in other words, the extent to which the 
 angle 9 may be open, depends on the ratio of the strength of 
 the field to the strength of the self-polarisation of the armature: 
 It is thus this ratio with which we have ultimately to deal. 
 
 This question may obviously be affected by any of the 
 following causes. The strengthening of the self-polarisation 
 of the armature, and the weakening of the field, one or both, 
 will open the angle 9 while conversely, the weakening of the 
 armature self-polarisation, and strengthening of the field, either 
 or both, will close the angle. A tendency for the normally 
 fixed planes to become bent or deflected at the axis of rotation, 
 is also such as may be subsequently considered. 
 
 These causes may now be internal and inherent to the 
 machine ; or they may be altogether external. As the former 
 are those that will require the most consideration, we may 
 dispose of the latter first, which will hence form the subject of 
 this present chapter. 
 
 External causes of sparks and flashes at the commutator 
 (due directly to the oscillation of the normally fixed planes) 
 may be a " short-circuit " between the positive and negative 
 leads or terminals, "earths," or changes of "load." The 
 effects, however, apart from such other consequences as the 
 
EXTERIOR CAUSES. 205 
 
 burning of insulation or of the armature winding, throwing off 
 of the belt, or pulling up of the engine, may vary according to 
 the manner in which the machine is wound, and as to whether 
 the cause is gradual or sudden. 
 
 Figs. 125 to 128 illustrate variously, separately excited, 
 shunt, series, and compound wound dynamos. The outer 
 circuit in each is represented by the dotted lines a b and c d ; 
 while the line a c will represent a " short-circuit " between the 
 positive and negative leads, whereby the whole outer circuit is 
 cut out. At E and E' are indicated leakages to earth. The 
 effect of leakages into the ground will depend on circum- 
 stances. Ordinarily the earth is a bad conductor for large 
 currents ; and it does not answer to use it as a return lead. 
 Damp or wet earth, and especially the near presence of any 
 iron piping or other metal conductor running from one leakage 
 to the other, would then, of course, render the effect approxi- 
 mate, more or less, to that of a short-circuit. The short- 
 circuit, as shown at a c, may be of more or less gradual growth, 
 such as might arise from the perishing of some insulation, 
 taking some little time to happen : or it may be quite sudden, 
 owing to some accident, such as a metal bar falling across, and 
 so electrically connecting, a positive and negative terminal on 
 a switchboard. 
 
 To consider the gradually growing short-circuit first which 
 may, however, be only a matter of a minute or so, or but a few 
 minutes we find that as the fault increases, the current will 
 flow more and more through it, instead of through, and against 
 the resistance of, the outer circuit c d b a ; for as the term 
 " short-circuit " implies an absence of resistance, so the path 
 ac will offer ultimately practically no resistance. Now as we 
 explained in Chapter XIX., when a machine has an outer circuit 
 between the brushes offering little or no resistance, the outer 
 voltage, or potential difference between the positive and 
 negative brushes, will correspondingly vanish. The whole 
 
206 
 
 SPARKING AT COMMUTATORS. 
 
 E.M.F. of the machine is used internally, and is employed in 
 impelling current through, and against the resistance of, the 
 armature winding only. This is precisely the effect produced 
 by the short-circuit ; and hence an enormous current will 
 flow. Apart from other consequences, this will at once cause 
 an immense increase in the strength of the self-polarisation of 
 the armature. 
 
 c 
 
 K 
 
 E 
 
 FIG. 125. Separately-excited Machine, with Short-Circuit and Earths. 
 
 Now, in Fig. 125, where the magnet coils are on an inde- 
 pendent circuit, it will be observed that the field is not 
 directly affected by the increased current. Therefore, the 
 armature polarisation being increased, while the strength of 
 the field is unaltered, the angle 6 will be widely opened, and 
 all the normally fixed planes will oscillate forward to suit. 
 This, it will be noted, is the same as though the brushes 
 were shifted too far backwards. There will hence be "for- 
 
EXTERIOR CAUSES. 
 
 207 
 
 ward sparking" of violence concomitant with the backward 
 position, as explained in Chapter XX. ; and the flash will be very 
 great owing to the immense flow of current. For the arc of 
 the flash will be large, and the heat great, and the brilliancy 
 will be due to the burning metal of the brushes and commu- 
 tator segments, the presence of any zinc in the metals im- 
 parting a green tinge to the flame. If, further, the brushes- 
 
 FIG. 126. Shunt-wound Machine, with Short-Circuit and Earths. 
 
 should have had lead, permitting back-induction, this latter 
 effect would be intensified by the rush of current through 
 the armature. Hence, the field would now also be weakened, 
 and thus cause the angle 9 to become still further opened. 
 
 In Fig. 126 the field is directly affected by the gradual 
 short-circuit. For, owing to the disappearance of the 
 exterior voltage, there is now no potential difference between 
 the two ends of the shunt, and so no current will flow 
 
208 
 
 SPARKING AT COMMUTATORS. 
 
 through it. Consequently, owing to its dependence on the 
 shunt, the field will be reduced in strength to zero, or 
 nearly so. The total E.M.F. will fall. But what little 
 remains will send current through the armature, causing 
 the strength of its self-polarisation to become abnormally 
 high in proportion to the (remaining) strength of the field. 
 Hence 6 will open, with forward sparking as a result. 
 
 Fig. 127. Series-wound Machine, with Short-Circuit and Earths. 
 
 The above cases, however, assume a gradually formed 
 short-circuit. If this were instead to occur suddenly, the 
 effects, especially in the case of the shunt wound machine, 
 would be different. For, owing to the fact that the induc- 
 tance, or self-induction, of the shunt coils would be considerably 
 greater than that of the armature winding, and also on 
 account of the hysteresis of the magnets, the field would not 
 become reduced so quickly as the current in the armature 
 
EXTERIOR CAUSES. 
 
 would increase. If anything, indeed, by the laws of self- 
 induction (as explained in Chapter XV.), at the first instant 
 the field would be strengthened by the extra current through 
 the shunt that would immediately precede the ultimate 
 reduction of its main current. The total E.M.F. would, 
 therefore, not immediately fall, and would possibly rather 
 rise for an instant. In any case, however, the whole E.M.F. 
 
 FIG. 128. Compound- wound Machine, with Short-Circuit and Earths. 
 
 being used for overcoming the armature resistance only, much 
 the same effects must follow, as described above in connection 
 with the separately excited machine, Fig. 125, and will open. 
 To take the case of a series-wound dynamo, as illustrated in 
 Fig. 127, we observe that a short-circuit in the outer circuit 
 would result in a large current flowing not merely in the 
 armature, but also through the field windings, assuming these 
 not to be cut out. The field will strengthen almost proportion- 
 
210 SPARKING AT COMMUTATORS. 
 
 ately with the strength of the armature self-polarisation : that 
 is, if the rush of current is not too sudden. Hence, the angle 
 may not sensibly vary ; and there will be little or no sparking, 
 It is to be noted that, owing to the resistance of the field 
 magnet windings, some E.M.F. will be maintained outside the 
 armature between the brushes, and the great increase in 
 strength to the field due to the increased current in the 
 magnet winding will cause the total E.M.F. to rise. Most of 
 this will be used in overcoming the resistance of the magnet 
 coils, which will be greater than that of the armature. But 
 as the major part of the outer circuit is short-circuited, a 
 larger proportion of the total E.M.F. will be used up in the 
 armature than would otherwise be the case. A fact bearing 
 on this point, however, in the case of series machines, is that 
 if the short-circuit is only partial, or at least not absolute, so 
 that some resistance in the exterior circuit between the brushes 
 still remains, then, up to a certain limit of decrease of exterior 
 resistance, the exterior voltage, or potential difference between- 
 the brushes, will rise. This is owing to the increasing current 
 through the field coils. But with a decrease of outer resistance 
 beyond this limit, the want of exterior resistance commences 
 to have a direct effect in reducing the outer voltage, in the 
 manner already explained, as with a shunt machine.* 
 
 With the same reasoning as before, it will be observed that 
 a suddenly occurring short-circuit, with a series winding also, 
 will not develop its full effects at once. For owing to the 
 self-induction of the armature winding being considerably less 
 than that of the field magnet coils, and also on account of 
 the hysteresis of the magnets, the armature polarity will be 
 affected much sooner than the field. Hence, at the first instant, 
 
 * Reference, for further explanation on this point, may here be made to 
 Prof. S. P. Thompson's " Dynamo-Electric Machinery," page 264 (fourth 
 edition) more especially to a diagram there numbered Fig. 164. 
 
EXTERIOR CAUSES. 211 
 
 the angle will open, causing a flash of forward sparking at 
 the brushes. It will be remarked further that, in all cases, 
 a merely momentary short-circuit would be unable, from want 
 of time, to develop all the results that would be permitted by 
 a longer endurance. 
 
 It may now be noted that, in the above instances, a gradual 
 or sudden change of load will but have the same effects, though 
 perhaps less in degree, to the gradual or sudden short-circuit. 
 If it be attempted, for instance, to use these machines for 
 incandescent lamps in parallel, a number of lamps switched 
 on would be tantamount, more or less, to introducing a short- 
 circuit. More current would flow ; in the separately excited 
 and shunt machines the angle 6 would open, with consequent 
 forward sparking : while a similar effect on the angle would 
 follow in the case of a series dynamo, should the introduction 
 of the lamps be sudden. The brushes may be advanced to 
 their new non-sparking positions. Suppose, now, the same 
 lamps were switched off again. This would be equivalent to 
 introducing a resistance into the outer circuit, instead of a 
 short-circuit ; and opposite effects would arise. The angle 6 
 would close, with back sparking as a result : and whereas the 
 exterior voltage would have fallen when the lamps were 
 switched on with the separately excited and shunt machines, 
 and would have risen with the series machine, the contrary 
 effects in this latter respect also, would arise on the lamps 
 being switched off again. 
 
 It is thus that in order to achieve the successful running of 
 incandescent lamps in parallel, wherewith a steady and unvary- 
 ing exterior voltage between the brushes is necessary, howsoever 
 the current may vary, the method of compounding the above 
 systems of winding has been adopted. Thus, shunt and series 
 machines being affected in opposite ways as regards the outside 
 voltage, by the same changes of load, a machine may be " com- 
 pound wound" on both these systems; it may be shunt wound, 
 
 p 2 
 
212 SPARKING AT COMMUTATORS. 
 
 and also have some series coils taking the main current round 
 the magnets. This is illustrated in Fig. 128, where both 
 windings are shown, the main by a thick line, and the shunt 
 by a fine line. Hence, on a number of lamps in parallel 
 being switched on (or a small short-circuit occurring), the 
 extra flow of current will also go round the magnets in the 
 few series coils. By strengthening the field, this will cause an 
 increase to the total E.M.F. We have already observed that 
 an increased current will require a higher E.M.F. to force it 
 through the armature, and further, that with a given total 
 E.M.F., and a decrease of resistance outside, a larger propor- 
 tion of this total E.M.F. will be used within the machine as 
 a fact, to drive the larger current through the armature, at the 
 expense of the outer E.M.F. Thus, the compound winding may 
 be so arranged that, when the current increases, the total E.M.F. 
 will also rise, so that the extra voltage will always be suffi- 
 cient to drive the increase of current through, and against 
 the resistance of, the armature, and the series coils on the 
 magnets. In other words, the additional E.M.F. is entirely 
 used within the machine ; consequently, the external voltage 
 remains unaltered, as required. It will thus be observed 
 further, that the shunt current being led off from the brushes, 
 and the potential difference between these now being constant, 
 a constant and unvarying current will flow through the shunt. 
 Thus the latter will steadily maintain its share of the total 
 electromotive force of the machine, and will maintain the 
 exterior constant voltage required for the variable current for 
 the lamps, according to the number that may be in circuit ; 
 while the series winding will overcome the internal resistance 
 of the machine. 
 
 But these conditions, moreover, will subserve to the reduc- 
 tion of the oscillation of the normally fixed planes, especially 
 the plane of requisite reversal (see Chapter XVI.). For, re- 
 ferring back to Fig. Ill, when the c and d armature currents 
 
EXTERIOR CAUSES. 21 8 
 
 are increased, and it becomes necessary that the I short-cir- 
 cuited coil shall have had its current made equal to the c 
 current, on its arrival into the same position, at the same angle 
 with the vertical, as before, we find that the strengthened 
 field may enable this to be accomplished. That is, the plane 
 of requisite reversal will oscillate but little, or not at all ; and 
 the neutral plane, or plane of non-sparking through the com- 
 mutator, being either coincident, or at a mechanically-fixed 
 angle, with the R R plane, the non-sparking points will remain 
 equally stable. Consequently, it is to be observed that in the 
 ideal compound dynamo, under any change of load within 
 wide limits, while the exterior voltage may be maintained at 
 a constant value, the sparking will be practically eliminated. 
 
 The reader desirous of entering more deeply into the .subject of com- 
 pound winding may be referred again to Prof. Silvanus P. Thompson's 
 " Dynamo-Electric Machinery," pp. 263 to 265, and 289 to 300 (fourth 
 edition). 
 
CHAPTER XXIII. 
 
 SPARKING AT COMMUTATORS. ARMATURE 
 REACTIONS : FIELD WEAK. 
 
 CASES may now be considered in which the oscillation, or the 
 permanent deflection, of the normally fixed planes may result 
 from causes internal and inherent to the machine, and thus 
 lead to more or less incurable sparking. These causes may be 
 various in character. Oscillation of the planes may be brought 
 about by a want of symmetry, or by a defective joint, in the 
 armature winding. A permanent deflection through the axis 
 of rotation may be caused by a want of symmetry, or asymmetry, 
 in the field ; or, again, the normally fixed planes may have an 
 undue inclination toward the centre plane of the field, owing 
 to the strength of the self-polarisation of the armature being 
 too powerful in proportion to that of the field. It is this last 
 cause which we propose to discuss in this present chapter. 
 
 In Fig. 129 herewith are represented diagrammatically 
 some of the effects that may arise from undue weakness of the 
 field. The armature winding, which may be considered as of 
 either the drum or the Gramme type, is indicated by the 
 radial lines a a and b b. For the sake of simplicity the com- 
 mutator is omitted ; and the brushes are shown resting directly 
 on the armature. 
 
216 
 
 SPARKING AT COMMUTATORS. 
 
 Now in Chapter VIII., in connection with Figs. 50 and 51, 
 we descanted on "back-induction," and "cross-induction," pro- 
 duced respectively by the coils approximately normal to, and 
 parallel to, the lines of force of the field. The same remarks 
 
 FIG. 192. Diagram illustrating Armature Reactions (with Brushes at 2, 2) 
 and Sparking due to a Weak Field. 
 
 will apply in the case before us, where we find, consequently, 
 that the coils a a a a are exercising back-induction, and the 
 coils bbbb exercising cross-induction, or cross-magnetisation, 
 as it may otherwise be termed. These effects are what are 
 
A WEAK FIELD. 217 
 
 known as "armature reactions"; and, in proportion as the 
 armature self-magnetisation is strong, and the field weak, so 
 will their influence preponderate. As we are assuming a weak 
 field, it is these armature reactions thus that we now find are 
 of special relative importance. 
 
 It has to be observed, however, that these two inductions, 
 normal to each other, are in reality only the components of the 
 armature self -polarity, which latter may consequently be re- 
 garded as their resultant. This may be rendered more clear 
 by Figs. 130, 131, and 132. These show an armature revolving 
 in a field, with brushes ; but with the latter in three different 
 positions. The self-polarity, which we have already shown 
 (Chap. XVI., pp. 158, 159) must follow the brushes, is repre- 
 sented by the thick line sn, which letters may also signify 
 south and north respectively. Thus, in Fig. 130, where the 
 brushes have a forward lead, the lines s c and a n will indicate 
 the back-induction, and the lines c n and s a the cross-induction. 
 Hence, lines, one of each of these inductions, for instance, an 
 and en, may be regarded as components of s n. In Fig. 132, 
 the brushes are shown with a negative lead ; and it will be 
 observed that, while c n and s a still act vertically in the same 
 direction as in Fig. 1 30, though they have exchanged positions, 
 s c and a n are reversed : and although c n and a n still continue 
 components of s n, yet the induction s c and a n represent is now 
 in the same direction with that of the field, which of course 
 runs from N to S. Hence the latter has been termed forward- 
 induction, inasmuch as it assists the field, in contradistinction 
 to the 6ac&-induction, which opposes the field. In Fig. 131, 
 ns being vertical, the cross-induction alone remains: that is, 
 the whole self-polarity of the armature is exerted in this 
 direction ; and there is no resolution of sn into components, as 
 in the other two cases. 
 
 Thus, in investigating the armature reactions in a field, it 
 may be convenient sometimes to regard the resultant sn alone, 
 
218 
 
 SPARKING AT COMMUTATORS. 
 
 \ \ 
 
 13 
 
 O o 
 
 o 
 
 
 
A WEAK FIELD. 219 
 
 and at other times to consider its components instead. For 
 our immediate purpose, it will be better if we view these 
 reactions in the light of their resultant only the armature 
 self-polarity. 
 
 To examine more closely, therefore, the conditions of the 
 armature itself, we know that its actual, or ultimate, polarity, 
 is a compromise between two influences. The first of these 
 is the armature self-excitation, tending to produce the self- 
 polarity sn just enlarged on: while the second is the field 
 induction, tending to induce a different polarity, as indicated 
 by s' n. These two influences being thus at variance with 
 each other, the ultimate polarity, indicated by S and N, 
 is a compromise between them. But it is the former of 
 these, the self-polarity, s and n, movable with the brushes, 
 which may none the less continue chiefly to engage our atten- 
 tion. 
 
 With a lead given the brushes as shown, it will be observed 
 that these two poles of the armature come exactly under the 
 horns E and H ; s opposed to an S horn, and n opposed to an 
 N horn. Consequently, we find the full strength of the self- 
 magnetisation of the armature directed to the reduction of the 
 normal polarity of these horns. According to the mutual 
 strengths of the opposed poles, the effects will be as follows : 
 either the horns E and H will have their normal magnetism 
 weakened ; or they may be entirely demagnetised ; or if the s 
 and n poles are very strong, the horns may even have their 
 polarity reversed, and the horn E will become north, and H 
 south. 
 
 This latter eventuality, however, is more liable to occur with 
 a Gramme winding than a drum. For, as pointed out in 
 Chapter L, p. 12, and illustrated in Fig. 133, the self-excited 
 poles (s and ?i, Fig. 129) of the drum armature are broadened 
 and spread so as to occupy generally two opposite sides of 
 the core : whereas with the Gramme winding, wherein each 
 
220 
 
 SPARKING AT COMMUTATORS. 
 
 half of the armature tends to have its self-excited North pole 
 exactly at n, and its South pole exactly at s, the two self- 
 excited poles of the armature as a whole become much more 
 acutely localised or confined to two mei;e lines, or narrow 
 strips of surface, on opposite sides of the core parallel with the 
 axis, as indicated in Fig. 134. Hence, with the latter, the full 
 strength of its self-magnetisation is brought to bear with much 
 
 FIG. 133. Diagram representing Self -excited Polarity of Drum 
 Armature. 
 
 greater effect on the horns it may oppose, than would be- 
 the case with drum-winding. This is, hence, another respect 
 in which the drum armature has a distinct advantage ; and 
 inasmuch as a large armature will have more powerful 
 self-excited poles than a small armature, a larger size of drum- 
 armature may be safely used without risk of reversing the 
 horns, than would be the case with a Gramme. 
 
A WEAK FIELD. 
 
 221 
 
 In the diagram, Fig. 129, it is assumed that the horns 
 E and H are simply demagnetised. In consequence of the 
 shifting forward of the self-excited poles s and n with the 
 brushes, the angle will open ; and inasmuch as the poles s 
 and n are abnormally strong, this angle will open wider than 
 would otherwise be the case ; and thus the field is greatly dis- 
 torted. The CF plane, invariably perpendicular to yy (see 
 
 FIG. 134. Diagram representing Self-excited Polarity of Gramme 
 Armature. 
 
 Chapter XVI.), is also tilted as shown ; and an effect to be noted 
 is that the induction lines flowing in and out from the pole- 
 pieces, as shown dotted,* become very cramped or dense towards 
 
 * It should be stated that the curvature given to the ends of these 
 clotted lines is only suggested, inasmuch as it is a departure from the more 
 usual method of showing them arranged symmetrically either side of an 
 imaginary straight centre line through the axis, as y y. The assumption is 
 
222 SPAKKING AT COMMUTATORS. 
 
 and at the horns G and F, which hence, contrary to the case 
 with the other horns, become strongly magnetised. 
 
 We may now notice some further effects that will arise with 
 a weak field. Referring hence again to the diagram, Fig. 129, 
 the plane of maximum E.M.F. will be observed, represented 
 by the line M M. If the brush points thus are placed 
 exactly in M M, there will be the maximum potential difference 
 between them. It is to be observed, however, that the s and n 
 poles being then brought, together with the brushes, nearer 
 to z Zj they will not so directly oppose the horns E and H ; 
 and these will consequently not be so much weakened. The 
 horns G and F, on the other hand, will be strengthened, by 
 the nearer opposition of the unlike poles s and n respectively. 
 But bearing in mind, that, to secure an absence of sparking, it 
 is necessary that the short-circuited coils, before receding from 
 the brush points, should not merely have been commutated, 
 but should also have their polarity and current made equal to 
 those of the coils they are about to join, we find that as there 
 are only a few lines of force passing between the armature core 
 and the horns E and H, no sufficient reverse E.M.F. can be 
 generated to reverse the current in the short-circuited coils 
 when the brushes are anywhere near the top and bottom. It 
 is only when they are moved far forward, as shown dotted, in. 
 
 that with a rapidly rotating armature, the tendency of the iron, due to. 
 hysteresis, not immediately to respond to the magnetising forces acting on 
 it, though possibly overpowered in a strong field, may yet, perhaps, take 
 effect in a weak field in some such manner as shown. The point may seem 
 of little consequence ; but its importance will depend on the extent to 
 which, under any given circumstances, the effect may take place. This, 
 however, trenches on a subject on which it is not practicable to enlarge 
 within the scope of these present remarks. Headers desirous of studying 
 hysteresis may be referred to Prof. J. A. Swing's book " Magnetic Induc- 
 tion in Iron and other Metals," or to the same work in The Electrician, 
 Vol. XXVII., more especially to pages 517, 518, 546, 547 and 602. But it 
 should be understood that there is no intention to imply here that the 
 latter writer is in any way directly responsible for the above suggestion. 
 
A WEAK FIELD. 228 
 
 positions 1, 1, so that the coils may be excluding not only a 
 sufficient number of lines, but excluding them also at a rate 
 sufficient to generate the necessary E.M.F. and current (see 
 Chap. I., p. 8), that the sparking will cease. Thus the plane 
 of requisite reversal, lettered R R, is very far forward. Con- 
 sequently, when the brushes are in the M M plane, or anywhere 
 near the top or bottom, as at positions 2, 2, or 3, 3, being 
 behind the R R plane, violent forward sparking will occur, a& 
 at the brushes in positions A and B, Fig. 117, and indicated 
 by the curved arrows. 
 
 To consider the armature reactions in the aspect of the com- 
 ponents, that is, of the two inductions normal to each other, 
 as illustrated in Figs. 130 and 132, we find that, whereas, when 
 the brushes are forward in positions 2, 2, in Fig. 129, back- 
 induction results ; if the brushes should be moved back toward 
 the horns G and F, as at A and B in Fig. 117 just alluded to, 
 /orwarc?-induction will ensue : for the coils occupying the 
 positions a a a a would have current circulating in them in the 
 reverse direction to that indicated in the diagram. 
 
 Now, as forward-induction strengthens the field, this might 
 appear an advantage. A point, however, in this connection 
 may be touched on, bearing reference to the diagram, Fig. 117, 
 in Chapter XX., which was not then dealt with. It will be 
 observed, that when the brushes are shifted back to positions 
 A and B, and the s and n poles are respectively opposed to the 
 horns G and F, unlike poles being then opposed, these horns 
 are strengthened, and the lines of force will gather about them 
 yet more densely than when they are not so directly opposed 
 by the armature poles. Hence we perceive that when the 
 brushes have a negative lead, the short-circuited coils are 
 traversing a denser field than they would traverse when the 
 brushes are forward, even in the case of a normally strong 
 field ; and consequently they will generate a more powerful 
 E.M.F., which will impel a larger s.-c. current, than when the 
 
224 SPARKING AT COMMUTATORS. 
 
 brushes are forward. Thus, when the brushes have a negative 
 lead, not only is the s.-c. current circulating in the wrong 
 direction for sparklessness, and diverting all the current flow 
 to the brush points, as shown in Figs. 118 and 120, but it is 
 also stronger, and has more efficacy in accomplishing the evil 
 results. Hence any apparent gain from forward-induction is 
 entirely counterbalanced by the evil results that also attend 
 that particular reaction. With a very strong field a difference 
 of density of lines of force under the horns, according as to 
 whether the brushes are back or forward, would, of course, not 
 be so marked as with a weak field. It may be noted that it 
 is for this reason that, in the diagrams Figs. 117 to 121, the 
 brushes are shown with a greater positive lead (as at C and 
 D) than a negative lead (as at A and B), so that the same 
 strength of s.-c. current may be assumed in each position. 
 
 In the case of a motor, the conditions being assumed the 
 ^ame as in Fig. 129, as regards N and S, and the position of 
 the brushes, the current direction would be reversed, and the 
 self-excited poles s and n would consequently exchange places. 
 The horns E and H would then be strengthened by being 
 opposed by unlike poles. If the pole pieces were not normally 
 magnetised, yet the magnetism induced within the horns E and 
 H by the armature would pervade those regions of the pole- 
 pieces, so that they would attract the poles of the armature, 
 and thereby cause the latter to rotate. The inverse of this 
 holds good also that is, the case of the dynamo brushes 
 shifted too far back. For, as pointed out, the horns G and F 
 would be strengthened by the dynamo armature ; and so, if 
 not normally magnetised, magnetism would be induced in them 
 by the armature poles. Hence, the armature would create its 
 own field. But the forward sparking would still take place, 
 involving loss of energy and destruction of material. 
 
 When the motor brushes are forward and the motor armature 
 is creating its own field by induction, it has to be borne in 
 
A WEAK FIELD. 225 
 
 mind that the short-circuited coils will behave like the coils of 
 a dynamo armature, and will similarly be generating a short- 
 circuit E.M.F. and current. Thus the s.-c. current will circu- 
 late in the same direction as in the coils b and/ in Fig. 119. 
 But with the motor the main current will be flowing in the 
 opposite direction. Hence Fig. 120 will represent the motor 
 brush at position C in Fig. 117 ; and it will be observed that 
 heavy forward sparking will occur, as with the dynamo brush 
 too far back. 
 
 Reverting, however, to the weak field of a dynamo, as repre- 
 sented in Fig. 129, it has to be noted that the plane of maxi- 
 mum electromotive force being represented by the line M M, 
 the further the brushes are shifted from this plane the less will 
 their potential difference, that is, the outside voltage of the 
 machine, become. For as the positive brush descends on one 
 side below the point of maximum potential, so the negative 
 rises on the other side above the point of minimum potential. 
 Hence on advancing the brushes in order to attain a non-spark- 
 ing position, the efficiency of the machine is sacrificed. It is 
 thus that a field, too weak in proportion to the self-magnetisa- 
 tion of the armature, presents one of those cases wherewith 
 sparking is practically hopeless. 
 
 NOTE. It is necessary to observe that the word demagnetised, as it 
 appears on pages 219 and 221, should be understood to refer only to the 
 stoppage of the flow of induction lines from the pole-pieces into the arma- 
 ture, or vice versa. The word, as here used, is not intended to imply that 
 there may be no "free" magnetism from the parts under consideration 
 outside the field. In subsequent Chapters the word " counteracted " will 
 be used where this particular meaning is alone intended. 
 
 Q 
 
CHAPTER XXIV. 
 
 SPARKING AT COMMUTATORS. ARMATURE 
 REACTIONS: MAGNETIC FLOW. 
 
 IN our last chapter we touched on the fact, in connection with 
 Figs. 130 to 132, that the self-excited polarity of the armature 
 might be resolved into two component inductions, respectively 
 parallel and normal to the field. We also remarked that the 
 armature reactions might thus be studied either in the light of 
 the self-polarity merely, or else by considering the separate 
 components. 
 
 It so happens that this latter method affords special advan- 
 tages for the elucidation of some of the reactions that take 
 place in the field. We have already to a slight extent treated 
 our subject in this manner. But it is our purpose now, 
 however, in this present and succeeding chapters to adopt this 
 method of investigation yet more thoroughly, and so to pre- 
 pare the way to practical conclusions not otherwise, perhaps, 
 so easily attainable. With this object in view, we shall 
 thus find it necessary to consider somewhat more closely than 
 heretofore, the precise nature of magnetism itself. 
 
 Now, it will be well understood that, ivithin properly pro- 
 scribed limits, lines of magnetism may be regarded as forming 
 a current, similar to electricity. If an intact ring of iron is 
 
 magnetised by a coil and electric current round a part of it, 
 
 Q2 
 
228 
 
 SPARKING AT COMMUTATORS. 
 
 FIG. 135. 
 
 FIG. 138. 
 
 Fio. 136. 
 
 FIG. 139. 
 
 FIG. 140. 
 
 Diagrams illustrating Magnetic Flow. 
 
MAGNETIC FLOW. 
 
 229 
 
 the induction appears to circulate through the iron of the 
 ring without there being any "field" outside, as shown Fig. 
 135, where the dotted circle cc indicates the magnetic flow; 
 and it will not attract any foreign magnetic body. But 
 now let the ring be severed on one side, and the two ends 
 drawn apart, so as to approximate the shape of a horse-shoe, 
 as illustrated Fig. 136. We then find that the induction 
 has apparently flown to the surface of the iron at the two 
 ends, and, if possible, it will leap the gap between the ends, 
 so as to continue its circulation. One of these ends will be 
 north and the other south ; and the nearer they are to each 
 
 FIG. 137. Diagram illustrating " Free " Magnetism. 
 
 other, the greater will be the number of induction lines crossing 
 between them. If, again, the ring be opened out into a more 
 or less straight bar, the magnetism will still be at the ends 
 (Fig. 137), forming north and south poles; but the magnet- 
 ism will now be what is termed "free" magnetism and will 
 resemble static electricity, positive and negative, at either end. 
 It is at this point, however, that the similitude of magnetism 
 to electricity fails. For, whereas a charge of the latter can 
 be drawn off by the contact or near approach of another body, 
 magnetism cannot so be caused to discharge itself ; and 
 however much a piece of magnetised iron or steel may be put 
 
SPARKING AT COMMUTATORS. 
 
 in contact with other bodies of iron or steel, and may exert 
 attractive force on them, its own magnetism nevertheless 
 remains. It may be noted, however, that a circulation of 
 "lines of force" is still maintained, through space, or air, 
 or other non-magnetic matter, as indicated by the dotted 
 lines cV, Fig. 137. 
 
 Returning to the horseshoe, it is further known that if 
 the gap or field is bridged by a bar of iron or steel placed 
 across, as depicted, Fig. 138 (d\ the induction lines will crowd 
 through it, finding thereby an easier passage than through 
 air, owing to the high magnetic permeability of those 
 
 ^^ 
 
 ^-^ \\\l ,, 
 
 i//, _^ 
 
 ""*""'$ 
 
 =-= 
 
 ^ N 
 
 * i 
 
 i\ 
 
 
 r 
 
 \ 
 \ 
 
 
 
 X 
 
 II 
 
 S "N 
 
 S 
 
 
 //" M[|\\\ ^--V' 
 '/III|O e 
 '/III** 
 
 FIG. 142. Distortion of Magnetic Flow, shown graphically. 
 
 metals. There will be less " free " magnetism, and the 
 induction keeping more within the metal, the field will be 
 reduced to a minimum. Or, again, if a shorter bar were 
 .inserted as at d, Fig. 139, so as just to leave very small gaps 
 between its ends and the two poles of the magnet, then a 
 much denser crowd of induction lines will traverse these 
 narrow interstices, and along the intervening bar, than would 
 be the case were the bar absent. If the inserted bar be 
 itself a magnet, and its north pole be opposed to the south 
 pole of the horseshoe, and its south to the north of the 
 horseshoe, as shown in Fig. 140, the flow of induction lines will 
 
MAGNETIC FLOW. 
 
 231 
 
 be accelerated. On reversing the bar end for end, so that south 
 is opposed to south, and north to north, the opposing " like " 
 poles reduce each other, and the flow of induction lines is 
 retarded as represented in Fig. 141. Further, if the small 
 magnetised bar be placed in the gap with its polarity across 
 that of the field, as shown in Fig. 142 (representing the same 
 gap as before, but to a larger scale), we then find that the main 
 flow of induction is diverted, and so passes through those 
 regions of the small interstices, as at e e, where unlike poles are 
 opposed, but evinces a tendency to escape as " free " magnetism 
 where like poles are opposed, as at //. 
 
 Fig. 143 may now, however, engage our attention. "We 
 discover here the same conditions as in the last diagram ; 
 
 FIG. 143. Distortion of Magnetic Flow, shown analytically. 
 
 but the magnetic flow is represented differently. Fig. 142 
 indicated the resultant or ultimate flow. But in the diagram 
 now before us this ultimate flow is analysed. It will thus 
 be noted that there are here shown three separate circuits. 
 Of these, the dotted line c c indicates the main induction of 
 the whole ring ; and there are two small magnetic circuits, 
 ehfg, which appertain to the polarity of the small inserted 
 block, and are due to the iron on either side. Observing now 
 the various directions of magnetic flow, as denoted by the 
 arrowheads, we may perceive that at the regions // the 
 direction of flow in the small circuits is contrary to that of 
 the main induction, while at ee there is confluence of induc- 
 tion. Hence in the latter regions the lines gather together 
 
SPAKKING AT COMMUTATOKS. 
 
 and become dense, each induction assisting the other ; but 
 at the former, the main flow is baulked, and there is escape 
 of free magnetism, as intimated. It may be remarked here 
 that in Fig. 140 "forward-induction" is represented; in 
 Fig. 141, " back-induction " ; and in Figs. 142 and 143 
 likewise " cross-induction " is illustrated, in which respect 
 these diagrams may be compared with Figs. 130 to 132 in 
 the last chapter. 
 
 
 i! 
 
 
 'f.t 
 
 
 i i 
 
 
 i ; j 
 
 * h * 
 
 /' 
 
 sCV-v 
 
 ,X' 
 
 I ^^^ ^ 
 
 '->^r--*-~ N 
 
 " i i 
 
 y/s^ 
 
 v 1 ' 
 
 
 N 1 
 
 X \ 
 
 \ ^ 
 
 ' 
 ) I | 
 
 \ ^ 
 
 t B 
 ^^*=|= 
 
 ^ F 
 
 / N 
 
 N^TL-* 
 tf - 
 
 / fj 
 
 i ' i > / 
 
 F H 
 
 FIG. 144. Armature Reactions shown analytically : Negative Lead and 
 Forward-induction. 
 
 We may now proceed to the consideration of the diagrams, 
 Figs. 144 and 145. These show once more an armature 
 revolving in a field. In the former the brushes are represented 
 with negative lead, thus giving rise to forward-induction ; and 
 in the latter they have positive lead, with consequent back- 
 induction. The thick line sn indicates the armature self- 
 polarity as heretofore. 
 
 Examining these diagrams now more closely, and comparing 
 them with Figs. 130 to 132 and 140 to 143, we observe that 
 
ARMATURE REACTIONS. 
 
 283 
 
 the self-polarity s n is resolved into its two component in- 
 ductions, and we find that the cross-induction is represented as 
 having two magnetic circuits d ef and g k I, similar to those in 
 Fig. 143; and the forward and back-inductions are represented 
 by the lines p q r and p q r. The main induction of the field 
 is indicated by the thick lines a b c and a b' c. It will thus be 
 observed that the inductions parallel with the field divide, so 
 
 FIG. 145. Armature Reactions shown analytically : Positive Lead and 
 Back-induction. 
 
 that they flow partly through the upper, and partly through 
 the lower, half cf the armature. Were the brushes to be with- 
 out lead, and placed exactly top and bottom, there would then, 
 of course, be neither forward nor back-induction. But with 
 lead assumed, the presence of these inductions needs to be 
 taken into account. 
 
 A point that may be remarked here is that the strength 
 of any magnetic flow is measured by the number of imaginary 
 
234 SPARKING AT COMMUTATORS. 
 
 lines of magnetism that may flow together through any given 
 sectional area of metal, matter, or space in other words, by 
 their density. Moreover, inasmuch as in Fig. 129 we assumed 
 that the armature self-excited poles were equal in strength to 
 the horns they opposed, and so counteracted them, we may 
 take the same assumption here. Hence, in Fig. 145 the horns 
 E and H may also be regarded as being counteracted by the 
 opposing s and n armature poles ; and, consequently, that the 
 density of the magnetic lines issuing from n, equals the density 
 of those that would be issuing from the horn H ; and the two 
 opposing densities of s and E would also be equal. 
 
 We may now consider the various directions of magnetic 
 flow in these latter diagrams. Thus, at the outset, it is to be 
 observed that in Fig. 144 the inductions a be, ab' c, and rqp, 
 rq' p, are all shown confluent with one another through the 
 magnets, field and armature, concomitant with the negative 
 lead to the brushes ; whereas, in Fig. 145, the field-inductions 
 a b c, a b' c, are shown opposed throughout to p q r and p q' r, in 
 accordance with the positive lead. But it has to be noted in 
 Fig. 145, that of the three lines of flow through the horn H r 
 and with the assumptions as above, the strength of p q' r plus 
 that of g I h will together equal the strength of the counter- 
 flowing magnetism represented by the line a b' c. It will be 
 convenient if we further assume that glh &nd p q r are each 
 respectively half of the back and cross-inductions, and are equal 
 to each other ; hence, either one of them will equal half a b' c, 
 which latter is itself half of the main induction. 
 
 Having regard now to horn G.in the same diagram (Fig. 145), 
 it will be perceived here that the cross and back-inductions 
 glh&udpqr oppose each other, and being equal, they thus 
 eliminate each other's flow. Consequently, the half-field in- 
 duction a b c is alone left. We hence find that, inasmuch as 
 there is no flow either way through the horn H, the difference 
 in strength between this horn and the horn G may be said to 
 
ARMATURE REACTIONS. 235 
 
 be infinite. But, of course, with other assumptions, if for 
 instance the back-induction were not so strong as just 
 supposed, then the field-induction flowing from the horn H 
 would not be so equally opposed, and H would not be quite 
 counteracted; and though some of the cross-induction would 
 now be left to assist the field-induction in horn G, and so cause 
 this horn to be stronger than above supposed, yet the horn H 
 having now some strength, the difference in strength between 
 the two horns would no longer be infinite. 
 
 Reverting now, however, to Fig. 144, we find that at the 
 horn H similarly to the first assumption in G, Fig. 145 
 the forward and cross-inductions oppose each other, thus 
 leaving the half-field induction flow a b' c. At horn G, on the 
 other hand, all three inductions are confluent. Thus, beside 
 the fact that in Fig. 144 neither horn is counteracted, the 
 point now to be noted is that, whereas the difference in strength, 
 on the first assumptions, between the horns G and H in Fig. 145 
 was infinite, in Fig. 144 G is merely twice the strength of H. 
 Hence, in Fig. 144 the field is not so distorted, nor the angle 9 
 so wide open, as in Fig. 145. If the brushes are moved to the 
 middle positions, so .that neither back or forward-induction 
 exist, we then find that, with the same assumptions of strength, 
 the horn G will be three times the strength of H ; and the 
 line yy, representing the distortion of the field, will occupy 
 a position intermediate in its inclination between the two 
 extremes depicted. 
 
 In the necks of the pole-pieces on the line xx much differ- 
 ence of effect in the various cases will be observed. Thus in 
 Fig. 144, with the negative lead, we perceive that in the N 
 pole-piece the armature reactions oppose each other, thereby 
 leaving the field-induction to flow alone; but in the S pole- 
 piece, on the other hand, all the inductions are confluent. On 
 the line x x, Fig. 145, we find that as the armature reactions in 
 the N pole-piece both oppose the field-induction, and are 
 
236 SPARKING AT COMMUTATORS. 
 
 together assumed equal to the latter, there will here be no flow 
 at all : and in the S pole-piece the field- induction will flow 
 alone. 
 
 It needs to be well understood, however, that in these 
 oppositions of flow, absolute elimination of magnetism does 
 not occur. For it is only in the sense that the magnetic in- 
 duction approaches the nature of a current that such flow may 
 be stopped. The magnetism still remains, but, as intimated 
 in connection with Fig. 142 (//), it becomes "free." Hence 
 experimentally in practice, the lower end of the N pole-piece 
 in Fig. 145 would be found to attract foreign bodies of iron or 
 steel much more powerfully than would the S pole-piece in 
 Fig. 144, where the induction has an unopposed internal flow, 
 and consequently does not become " free." 
 
 Now magnetism at a south pole becomes free just as much 
 as at a north pole, as shown Fig. 137. Hence we observe that 
 in Fig. 145, the armature self-magnetisation being opposed 
 $n bloc at both the s and n poles, the whole of this also tends to 
 become free. In Fig. 144 the contrary holds good; and we 
 find that the self-magnetisation of the armature has an un- 
 impeded flow in and out of the pole-pieces. The proportionate 
 flow of lines of magnetic force between the pole-pieces and 
 the armature is represented by the small lines crossing 
 the intervening air-gap. Thus, though the total amount 
 of magnetism, so to say, may be the same in each case, 
 yet there is much more waste by magnetism become free 
 with the back-induction than with the forward-induction. But 
 it is equally obvious, however, that with little or no lead to the 
 brushes, there will be less waste by free magnetism than when 
 there is positive lead giving rise to back-induction. The 
 inutility of the forward-induction has already been expatiated 
 upon in the last chapter. It will be noted that Fig. 129 
 depicted the ultimate results of the reactions shown more 
 analytically in Fig. 145. This "free" magnetism forms part 
 
ARMATURE REACTIONS. 287 
 
 of what is termed the " waste field " of a dynamo. It needs 
 to be well borne . in mind, however, that, in the above re- 
 marks, the effect of the air gaps between the armature and 
 pole-pieces has been disregarded. In actual practice the 
 waste field is considerable ; and for this the gaps are chiefly 
 responsible. Were there no gaps, and the armature core 
 fitted tightly between the pole-pieces, then the magnetic 
 circuit within the magnets and armature would be entirely 
 " closed," as in Fig. 135. There would be, practically, no 
 free magnetism and, consequently, no attraction for foreign 
 bodies of iron or steel. This latter remark, however, will apply 
 more justly to the case represented in Fig. 144 than to that in 
 Fig. 145. For, in the latter, the juxta-position of "like" 
 poles at the horns E and H would cause the formation of a 
 joint " consequent " pole at each of those regions, from whence 
 would issue lines of free magnetism, as represented at ff r 
 Fig. 142. With the presence of the gaps, the flow of main 
 induction is retarded in the sense of its resemblance to a 
 current, and thus causes a large proportion of it to fly to the 
 surface and become free. Hence, as shown at eV, Fig. 137, it 
 seeks a path from N to S through the air on all sides, thus 
 forming " waste field " as intimated, to which that produced 
 by the armature reactions above discussed is additional. 
 
 On May llth, 1893, a Paper was read by Mr. W. B. Sayers 
 before the Institution of Electrical Engineers, describing a 
 method of armature winding by which the forward-induction 
 may be utilised without sparking at the brushes. This may be 
 explained with the assistance of the diagram herewith (Fig. 146). 
 It will be observed that between the main winding c d of 
 the armature and the commutator there is inserted another 
 
238 
 
 SPARKING AT COMMUTATORS. 
 
 series of coils it v. These latter are termed " commutator coils" 
 (by Mr. Sayers), and are interwound on the armature with the 
 F 
 
 FIG. 146. Diagram of Sayers' Armature Winding, showing sparkless 
 running with negative lead to the brush. 
 
 main winding c d, although for clearness here shown separate. 
 Put shortly, and comparing this with Fig. 118 showing the posi- 
 
SAYERS' WINDINGS. 239 
 
 tive brush, in the same position, it will be seen that by means of 
 the "commutator coils" a resultant short-circuit current is 
 caused to flow out at the lug g, thus baulking the main current 
 from the armature that would otherwise be flowing in at g. 
 The sparking such as indicated in Fig. 118 is thus obviated. 
 Hence, with this form of winding an armature may be per- 
 mitted to create its own field by inducing the necessary 
 polarity in the pole-pieces of the field magnets, as intimated 
 on page 224, and thus field magnet coils are not required. 
 
 It may, however, be of interest to analyse this method of 
 winding, and its theory, more closely. Thus, an output may 
 be assumed, as represented in the accompanying diagram, of 144 
 amperes up the brush, composed of 72 amperes from each 
 half of the main winding c and d respectively. Now, the brush 
 being at position A, Fig. 117, it will be noted that the coils 
 toward the right at d and v are including lines of force at a 
 greater rate than those to the left, and, consequently, are 
 generating a higher E.M.F. With the same resistances in 
 each case, larger currents will therefore be set up by the coils 
 toward d and v than towards c and u. The numbers on the 
 diagram all representing amperes, it will hence be observed 
 that the coils z, y, x, /, and b are represented as generating 
 currents equal to 360, 198, 99, 99, and 39 amperes respectively. 
 It will be noted, however, that these are respectively only 
 currents that would circulate were there no opposing currents 
 in the same circuits as was also the case in Figs. 118 to 121. 
 All these five coils are short-circuited by the brush ; and it has 
 to be observed that here also the conditions depicted are those 
 that are broken by the recession of the segment r from the 
 brush, and so are not the results of the recession. The analysis 
 of the effects may now be accomplished by considering in detail 
 the course pursued by the current from each short-circuited 
 coil individually and in order ; after which the mutual effects 
 of these various currents one on the other may be observed, and 
 
240 SPARKING AT COMMUTATORS. 
 
 the result noted. On the lugs g, h, and i are thus denoted by 
 the arrowheads, numerals, and letters, the currents which will 
 be tending to flow in them from each of the five coils respec- 
 tively. The 99 amperes at coil /, for instance, leaving at k 
 and needing to return at I, has two paths which it may pursue, 
 namely, bjxgmnoizl and kyhnoizl. In the former, the 
 resistances of the coils b, a-, and z have to be overcome, but in 
 the latter only that of y and z (neglecting the resistance of the 
 intermediate conductors). Coil z being common to both paths, 
 we find that the current will divide itself between these two 
 paths inversely according to their resistances that is, as two to 
 one (see page 171). Hence, in the path wherein the resistance 
 of the two coils b and x is encountered we find only 33/ 
 amperes will flow down, and where the one coil y is encoun- 
 tered 66/will flow; while the undivided current of 99/will 
 outflow, on its return to the negative end of coil /, through lug 
 i. The current from b being analysed in the same manner, we 
 perceive that while the whole 39 6 descends through lug g t 
 this will rise in currents of 26 and 13 amperes respectively in 
 h and i. So also will the other currents divide and flow, as 
 indicated; and it will be noted that the 198 amperes from coil 
 y t having return paths right and left of equal resistance, 
 returns in two equal currents of 99^ each. Cancelling now 
 all opposing short-circuit currents, we ascertain, as indicated 
 at m n o, that a resultant 48 amperes of short-circuit current 
 will tend to outflow through lug g ; and similarly, 68 amperes 
 through h ; while 116 amperes will enter through i ; and 
 116 = 68 + 48. 
 
 But to consider now the main armature currents c and d t we 
 observe that 72 c amperes will arrive at j, and 72 at I. Each 
 of these currents will have three paths to the brush (up to the 
 moment of recession), passing through x t y and z respectively, 
 of which paths consequently the resistances will be as one, two 
 and three. The c current, for instance, will encounter in one 
 
SAYERS' WINDING. 241 
 
 path the resistance of coil x only ; in the second path that of 
 b and y ; and in the third path that of 6, / and z. We thus 
 perceive that the main currents will divide themselves through 
 these three paths in the proportions denoted at j, Ic and I. 
 These form totals of 48 amperes each, as indicated below the 
 points g, h and i. But in g we now find that the 48 c d current 
 is met by the opposing 48 short-circuit current. Hence, at the 
 moment of recession, there will be no current whatever flowing 
 3 in the path m g xj, and so no sparking. In h we observe that 
 a difference of 20 amperes will outflow, as indicated ; but in lug 
 i the 48 c c? will augment the confluent short-circuit 116, form- 
 ing a total of 164. This latter thus divides at o, sending the 144 
 up the brush, leaving the 20 to flow in the circuit n h yfz o ; 
 and the double arrowheads will indicate the currents that will 
 actually and ultimately be flowing in the circuits. 
 
 It will thus be remarked that by the process of assimila- 
 tion of currents (see pp. 168, 169, 192, especially the latter) the 
 action of the resultant short-circuit current due to the electro- 
 motive forces of the short-circuited commutator coils has the 
 effect of diverting the flow of current from the point to the 
 heel of the brush, as in Figs. 119 and 121, such effect being, 
 in fact, almost entirely due to the hindmost coil z. 
 
 With regard to the potential, inasmuch as the coils #, y, and z 
 are in series between the main winding c d and the resistance of 
 the outer circuit, then (as pointed out on pp. 180, 181) their 
 electromotive forces will accumulate with, and become added 
 to, the E.M.F. of the main winding. Further, it will be noted, 
 that taken separately, coil z will generate a higher E.M.F. than 
 y t and y than x. But in the commutator, apart from the effect 
 of the " commutator coils," the M M plane being toward c and , 
 segment r will have a higher accumulated E.M.F. than s, and 
 s than t. It will thus be observed that the commutator coil x 
 of least potential is in series with the segment r of greatest 
 potential ; and inversely the highest charged commutator coil z 
 
242 SPARKING AT COMMUTATORS. 
 
 is in series with the lowest charged segment t. The potentials 
 of the commutator coils, added to the potentials of the seg- 
 ments, thus tend to equalise the latter in this respect. Hence 
 concomitant with the absence of current flow between n and m, 
 segments r and s may be assumed of equal potential. But, as 
 represented in the diagram, it may be assumed that the E.M.F. 
 from z has made segment t of slightly higher E.M.F. than s, as 
 would be concomitant with the small current of 20 amperes 
 flowing from o to n, and up n h. If y generated a higher E.M.F. 
 than here imagined, it would impel a larger short-circuit current. 
 So that with the increase of E.M.F. of s over r t there would also 
 be a flow of current from n to m, and consequent back-sparking 
 at the moment of recession. This would occur, it will be per- 
 ceived, if the negative lead to the brush were increased. Con- 
 trariwise, less negative lead, by letting the proportional E.M.F. 
 of y drop, will cause forward-sparking. With regard to these 
 short-circuit currents, the 360 amperes in coil z, the 198 in y, 
 99 in x, 99 in /, and 39 in b, it has to be noted that the pro- 
 ducts of these and the electromotive forces of their respective 
 coils, that is, the E.M.F. generated in each coil at the moment 
 of short-circuiting not the total absolute E.M.F. of the machine 
 will represent a certain amount of lost energy. This, how- 
 ever, does not necessarily mean energy that has been created 
 and then lost by being misspent. For it is rather a loss by the 
 prevention of the creation of energy which might otherwise be 
 added to the total efficiency of the machine, but here is not 
 added. 
 
 This point may be made clearer. Reference to Fig. 118 
 (p. 185) will show how the forward-sparking that occurs 
 with ordinary drum or Gramme winding, when the brushes 
 have negative lead, is due to the breaking, not only of the main 
 current, but also of short-circuit current. Reverting to the 
 diagram accompanying these remarks, it will be seen that, 
 besides the main current of 48 amperes entering at g, there 
 
SAVERS' WINDING. 243 
 
 are also the short-circuit currents, in part or wholly, from the 
 coils /, b, and x, also entering by lug g, forming a total much 
 larger than the main current alone, similar to the case in 
 Fig. 118. It is thus this whole total, in the case assumed 
 33/+396 + 99# + 48cd=219 amperes, that has to be pre- 
 vented from entering through lug g, which is accomplished 
 by a counter-current of the same quantity. 
 
 Argument might be adduced that as these opposing currents 
 prevent each other's flow, therefore the energy of which they 
 would otherwise form factors is not spent. But, as intimated, 
 this is not the point. Analogy to the case may be found thus. 
 If two steam engines, with full steam turned on, are geared 
 together so that each prevents the other's motion, no work is 
 done, and so no energy is spent. Or, again, if two dynamos be 
 coupled up, positive to positive and negative to negative, on 
 being run, their currents will oppose, and so baulk each other ; 
 and if their outputs are equal, no current at all will flow and 
 so, here again, no work will be done. But the point now is 
 that, with both the engines and the dynamos, these being 
 rendered in each case assumed mutually inoperative, they 
 represent so much capital outlay lying practically idle, and 
 producing no dividends. It is thus that the loss due to all 
 the opposition of currents in this winding may be viewed and 
 estimated; for we find here, that the products of these currents 
 and their several electromotive forces represent so many watts 
 of power opposed to poiver, which, for want of motion, mutually 
 prevented, is not converted into energy or work, which might 
 otherwise become added to the general efficiency of the machine. 
 This loss, in fact, may be regarded as the price of the sparkless- 
 ness attained by this arrangement of winding, to be put against 
 its various advantages, such as the non-requirement for coils on 
 the field-magnets. 
 
CHAPTER XXV. 
 
 SPARKING AT COMMUTATORS. ARMATURE 
 REACTIONS: FIELD ASYMMETRICAL.* 
 
 HAVING thus far discussed the subject and theory of magnetic 
 flow, we may now proceed with our general disquisition into 
 the causes of sparking, in which our last chapter was some- 
 what of the nature of an interpolation. The source of sparking 
 we propose now to consider was among those mentioned at the 
 commencement of Chapter XXIII., and we allude to a want of 
 symmetry, or asymmetry, in the field that is, a field stronger 
 in one half than in the other half ; or, in other words, having 
 regions on the opposite ends of any diameter which are not of 
 equal strength. It is in this direction, therefore, that we may 
 find that particular method of investigating the reactions of an 
 armature in the field, illustrated and described in our last 
 chapter, to be essentially of service. 
 
 The case of a field excited by four magnets in parallel, as 
 illustrated Fig. 147 herewith, affords an example wherein such 
 absence of symmetry might be produced, by one coil not 
 developing its full exciting power, owing, perhaps, to a bad 
 
 * This and the chapter preceding are based largely on a Paper read by 
 Mr. Jas. Swinburne before the Institution of Electrical Engineers, and 
 published in The Electrician, Vol. XXIV., 1890. (See Feb. 14th, pp. 374 
 375 for particular portion to which reference is made.) 
 
 
246 
 
 SPARKING AT COMMUTATORS. 
 
 joint. Thus one side of the field would have fewer lines 
 of force than the other. 
 
 But in all cases of an asymmetrical field, it may be remarked, 
 a drum winding will be less affected by this manner of fault than 
 a Gramme. For the coils of the former embrace the whole 
 field with all its inequalities ; while two coils of the latter at 
 opposite ends of a diameter may be traversing portions of 
 the field unequal in strength, and so will not generate the 
 same E.M.F. 
 
 FIG. 147. Four Magnet Windings in Parallel. 
 
 Apart from the possible cause mentioned above, however, 
 there are two other sources from whence may arise this particular 
 trouble, which are of a nature, perhaps, not quite so evident. 
 To explain these, reference may be made to Figs. 148, 149 and 
 150. In the two former of these diagrams it will be observed 
 that the neck of the pole-piece, as on x x, is only half the 
 section of what it is higher up, as at it ; while in the latter, 
 the section on x x is considerably less than half that above. 
 Thus, in the last case, we find an asymmetrical field may be 
 caused simply by the fact that, of the lines of induction 
 
ARMATURE REACTIONS. 247 
 
 descending, say, through 1 1, the half of the induction lines 
 required for the lower part of the field have not room to pass 
 down through the section vv, and to rise again through ivw. 
 The lines are choked or throttled in the necks of the pole- 
 pieces. Hence the upper part of the field will be stronger than 
 the lower part. 
 
 But, before further discussing this latter diagram, which 
 we must leave for the present, it will be better to consider 
 Fig. 148. This shows a field and armature similar to those 
 in the last chapter ; and it will be observed that, in like 
 manner, three magnetic circuits are indicated respectively by 
 a b b' c, d ef y and g k I, representing the main and cross- 
 inductions. The brushes are shown touching in the vertical 
 centre line, so as to preclude any back or forward-induction ; 
 for either of these latter, if present, would be merely a 
 superadded effect, not necessary for the present argument. 
 The direction of the flow in the three circuits will be as 
 indicated by the arrowheads ; and it will be noted that, 
 as in foregoing diagrams, in S the cross-induction flow from 
 / to <?, and the main-induction from b' to c, are confluent; 
 and in N the inductions from a to b' and I to h oppose each 
 other. Similar confluence will be observed at the horns G 
 and F, and opposition at the horns E and H. 
 
 Now a field may be wanting in symmetry, yet without being 
 absolutely weak. There may be plenty of exciting power in 
 the magnet coils, but the results are not equally distributed 
 through the field. For our present purpose, however, it will 
 be convenient to assume that the field before us is weak as well 
 as asymmetrical, so that the effects it is our object to render 
 evident may be the more conspicuous. 
 
 A field may be rendered weak by two causes. These may be 
 either that the magnets are too small, or else that the field- 
 magnets are not sufficiently excited, owing to a deficiency of 
 ampere turns in their coils. In the former case the magnets 
 
248 
 
 SPARKING AT COMMUTATORS. 
 
 and pole-pieces may be carrying all the induction lines possible, 
 or in other words, they may be " saturated ;" but in the latter 
 case they will not be saturated. 
 
 Of these two causes, hence the former may be assumed to be 
 the case in Fig. 148; and consequently it is understood that 
 both the magnets and pole-pieces are saturated. It should be 
 noted, however, that this means that these parts are saturated 
 by the main induction along abb' c alone. We have now to 
 
 FIG. 148. Asymmetry of Field due to Throttling of Cross-induction. 
 Sparking at one Brush only. 
 
 consider the superimposed effects of the cross-induction, in- 
 dicated by the smaller circuits ghl and def. 
 
 In the first place, it needs to be observed that, inasmuch as 
 s n is vertical, the opposition of the armature self-excited poles 
 to the horns E and H will not be so direct as in the case repre- 
 sented in Fig. 145. Back-induction being absent, we may in 
 fact adopt the exaggerated assumption, with the assumption of 
 
ARMATURE REACTIONS. 
 
 249 
 
 a weak field, that the cross-induction is by itself equal to the 
 field-induction. Thus now to direct our attention to these cross- 
 inductions, we observe in the first place that these particular 
 reactions of the armature do not strengthen the field, but that 
 rather they are the direct cause of its distortion. For in so far 
 as these magnetic inductions may be regarded as currents, it 
 will become palpable on reflection that these cross-induction 
 flows will merely act on the main induction flow in much the 
 
 FIG. 149. The same as previous diagram, but with Polarities reversed. 
 Sparking at one Brush only. 
 
 same manner as on an earlier page we demonstrated that the 
 short-circuit currents at the commutator took effect on the 
 main current, namely, that they diverted the flow of the main 
 current from one course to another. Thus it is in the case 
 before us. We find that the effect or tendency of the g hi 
 induction is to transfer the exit of the main induction from N 
 as much as possible from the bottom to the top of the field ; 
 
SPARKING AT COMMUTATORS. 
 
 and the tendency of the clef induction is to transfer the entry 
 of the main induction into S from the top to the bottom ; and, 
 as illustrated by the dotted lines in Fig. 129, the whole body 
 of induction lines leaving N will now crowd out as much as 
 possible toward the horn G, and passing through the upper 
 and lower halves of the armature, will thus tend to impinge 
 for the most part on the lower half of the pole-piece S. 
 
 It will be noted, however, that though the cross-induction 
 may have this diverting effect on the main flow, it nevertheless 
 does not augment the latter through the armature and field. 
 The flow down g, for instance, will merely equal the strength 
 of the cross-induction ; and the strength along b' will continue 
 to be just that of half the main induction ; and the flow 
 through the horn H will be the difference between that at g 
 and b', here assumed as nothing. 
 
 But it is evident, referring again to the short-circuit cur- 
 rents at the brushes, that if the lugs, for instance, were not 
 large enough to carry an augmentation of main and short- 
 circuit current, the latter could not take effect its very 
 existence in fact being imperilled. Thus it is in the case 
 before us. For we find that the cross-induction from / to e 
 plus half the normal main inductions are wanting to ascend 
 through the section on xx. But the latter induction alone, 
 under the circumstances, saturates this section ; therefore, 
 there is not room for both inductions to pass. Consequently, 
 excepting in so far as S may become super-saturated, the 
 def induction cannot have the effect in diverting the entry 
 of main induction into S from E towards F that it would other- 
 wise have. Hence we find that the entry of main induction 
 into S will extend in area much more fully towards E, than 
 its exit from N extends toward H. 
 
 The upshot, therefore, as may be perceived, is that the field 
 is rendered asymmetrical. For in the first place, as the 
 induction lines will leave N in the regions towards G in a 
 
ARMATURE REACTIONS. 
 
 251 
 
 dense condition, but cannot correspondingly enter densely 
 the regions of S toward F, these diametrically opposite parts 
 of the field are unequal in strength ; and secondly, as already 
 explained, lines are passing between the armature core and S 
 quite up to E, but not between the core and N toward H ; 
 and consequently the horn E is not counteracted like the horn H. 
 
 FIG. 150. Asymmetry of Field due to Throttling of Main and Cross- 
 induction. Sparking at one Brush only. 
 
 Further effects are now to be observed. Inasmuch as the 
 lines are entering the armature more thickly toward the horn G, 
 than they are leaving it toward F ; and the outflow is more 
 evenly spread on the side toward S ; therefore, the ultimate 
 pole S of the armature is higher above x x than the N pole 
 is below it ; and the centre line of distortion y y will be seen to 
 be deflected through the axis as shown. 
 
252 SPARKING AT COMMUTATORS. 
 
 Moreover, as the lines of force between the pole-piece and 
 armature are denser toward E than H, it is obvious that the 
 RR plane above will not be so far forward as that below. 
 Hence the upper brush will not require so much lead to attain 
 a non-sparking position as the lower brush ; and it will not be 
 possible, with the brush points diametrically opposite each 
 other, for the brushes both to be in non-sparking positions at 
 once. It will be observed, however, that this deflection of the 
 R R plane refers only to Gramme winding. For it is obvious 
 that with a drum winding, wherewith each coil embraces the 
 whole field, some one angular position in one plane through the 
 axis will be found where the requisite reversal takes place. 
 
 Now we have remarked that when lines of magnetic induc- 
 tion cannot find a circuit, as at the two ends of a bar, the 
 magnetism will fly to the surface at the ends, and become 
 " free " magnetism radiating into space. Thus, continuing to 
 refer to Fig. 148, we perceive that as of the combined inductions 
 b' and d flowing to meet each other, all cannot pass up through 
 the neck of S, a remainder, equal to the d cross-induction, 
 will be left without a circuit. This remainder will, therefore, 
 become "free." It may be said in fact that the d cross- 
 induction, finding no circuit, but yet not being eliminated, 
 becomes free at both ends, above and below. At the horn G 
 both the main and cross-inductions are flowing unimpeded, so 
 that neither becomes free. But at F, on the other hand, 
 where the inductions oppose each other, both become free, their 
 difference alone, if any, remaining as an internal flow. These 
 remarks of course, it will be observed, are made without refer- 
 ence to the air-gap between the armature and pole-pieces, 
 alluded to in the last chapter, as this would not affect the pre- 
 sent argument. 
 
 Fig. 149 is merely a reproduction of the last diagram, but 
 with all the polarities reversed. It will be seen, however, that 
 though the inductions are reversed, yet their confluences and 
 
ARMATURE REACTIONS. 253 
 
 oppositions of flow are in the same locations as in the foregoing 
 case, and, consequently, the lower brush will still require the 
 most lead. 
 
 Reverting to the question of the free or waste magnetism, it 
 will be observed that in both these diagrams, Figs. 148 and 149, 
 this effect occurs at the horns E, F, and H, but not at G (neg- 
 lecting the effect of the gap). There is thus a want of sym- 
 metry in the waste field produced, concomitant with the want 
 of symmetry in. the field proper. In these cases we have, of 
 course, as intimated, assumed a main induction of exaggerated 
 weakness, at least in respect of the horn H. With a main 
 induction a l> considerably stronger than the cross-induction 
 g I, as would be usual in practice, their difference would con- 
 tinue along b' c ; while the remainder, that is, g I, and a part of 
 a V equal to g l t would become free, and so form waste field 
 seeking an air circuit to S. This will apply also to the horn 
 E, where there is likewise opposition of flow. 
 
 Now all the above results, it will be perceived, are simply 
 due to a want of section in the necks of the pole-pieces or 
 rather, of one pole-piece, as at S, Fig. 148, so as to enable it 
 to carry both the main and cross-inductions. In Fig. 129, it 
 may be assumed that the pole-pieces are not thus wanting in 
 dimensions, and that, therefore, there will be no choking or 
 throttling of cross-induction in the neck, and so no asymmetry ;. 
 and the weakness of field in that case will be attributable to a 
 deficiency of exciting power in the coils of the field magnets. 
 
 At the commencement of this chapter we alluded to an 
 asymmetry of field that might be caused by the throttling of 
 the main-induction itself in the necks of the pole-pieces. We 
 thus find an example of this in Fig. 150. For, as previously 
 pointed out, the sections here at w w and v v are less than half 
 that at 1 1, and consequently are unable to pass the full number 
 of induction lines that is, half the total number to the 
 lower portion of the fiel* 1 . 
 
254 SPARKING AT COMMUTATORS. 
 
 Now, it is assumed in this case that there is no deficiency of 
 strength, at least in the upper half of the field, and that, 
 therefore, there is no falling away of main induction under the 
 horn E, as in the foregoing cases. But, inasmuch as the lower 
 part is weak, the line x x will cease to represent the centre 
 plane of the magnetic lines of the field ; and this plane will now 
 be higher up, as shown at x' x. The tendency of the field will 
 consequently be for the pole-pieces to induce poles on the 
 armature which shall both be on the line x #', south on the 
 right and north on the left, as indicated by s' and n ; and a 
 resultant polarity will eventuate between n and ri at N, and 
 between s and s' at S. 
 
 We thus find again, in the instance of the Gramme ring here 
 depicted, that y y becomes deflected through the centre at a j 
 and that, owing to the induction path through the upper part 
 of the armature becoming much shorter than that through the 
 lower part, the magnetic resistance of the upper will be much 
 less than that of the lower path, and consequently, as shown, 
 the induction lines through the armature will be more dense 
 above than below. 
 
 Moreover, as there is no absence of induction under the 
 horn E, we note that the core will commence parting with lines, 
 and the coils will commence excluding lines, on their arrival at 
 this point. Hence we may assume the upper R R plane to be 
 where indicated by the line R a and the upper brush, being 
 given that amount of lead, will not spark. 
 
 Having regard now to the horn H, this is normally weak, 
 owing to insufficiency of main induction. It may thus be 
 assumed to be counteracted by the self-excited n pole of the 
 armature. There will, consequently, be a falling away of in- 
 duction for some degrees above H, and lines may not commence 
 entering the core from N till about where shown in the 
 diagram. Thus the lower R R plane will come at or about 
 a R' ; and the lower brush, as in the other cases of asymmetry 
 
ARMATURE REACTIONS. 255 
 
 already discussed, will require more lead than the upper brush; 
 and though the upper may not spark, the lower brush will 
 spark persistently. 
 
 We thus discover how an asymmetrical field may be pro- 
 duced by the throttling of the main induction. The cause 
 now described may, of course, operate simultaneously with 
 that previously discussed. It is to be noted, however, as already 
 remarked, that all asymmetry such as we have so far touched 
 on, can be prevented in single horseshoe magnets, by simply 
 making the pole-pieces of sufficient section in the necks to 
 carry all the induction. 
 
 Before closing this chapter some remarks may be added 
 concerning the cross-induction with double-ended magnets, as 
 shown Fig. 147 ; for the conditions here are not quite the same 
 as with the single horseshoe magnets we have usually assumed. 
 Fig. 151 is thus a reproduction of part of the latter diagram, 
 showing the field and armature only. The lines a b c and a' V c' 
 indicate the two main inductions, one from each horseshoe, 
 and the circuits d efg and h ij k represent the cross-induction. 
 
 It will now be observed, in the first instance, that there is no 
 normal now of main induction through the necks at N and S, 
 and that, further, the necks are of appreciable thickness. 
 Consequently, both cross-inductions have an unimpeded course, 
 and they will both be able to exercise their full effect in 
 diverting the main-inductions, and so in distorting the field ; 
 though, being the same on both sides, they will not produce 
 any asymmetry. For by a process somewhat analogous to 
 what on a previous page, when discussing electric currents, we 
 termed assimilation, it will be evident that the. a' b' c flow, 
 being opposed in the horn E by the defg induction, will 
 become diverted, and will flow in the d efg circuit. In like 
 manner, some of the a b c main induction will become assimi- 
 lated in the kijk circuit. It will, however, be of course 
 understood that this view of the subject is only permissible 
 
256 
 
 SPAEKING AT COMMUTATOKS. 
 
 within the limits of the extent to which magnetic flow may be 
 regarded as a current. Outside these limits it will be noted 
 that there will be an escape of free magnetism at the horns E 
 and H, where the inductions are opposed, and these latter 
 horns will be weakened, while the horns F and G, owing 
 to the confluence of the inductions through them, will be 
 strengthened. 
 
 V 
 
 FIG. 151. Four-magnet Field, with (7ross-induction. 
 
 But we may now consider Fig. 152. Here the necks at S 
 and N have been reduced, the effect of which is to impede 
 the flow of cross-induction. A slight flow of that reaction may 
 take place, but this may be disregarded. All armature 
 reactions, however, are not eliminated. For a result of the 
 suppression of the cross-induction per se, is that this induction 
 on either side is now, as it were, divided into two smaller 
 
ARMATURE REACTIONS. 
 
 257 
 
 o6fo'2"w<?-inductions, as indicated by the circuits d e f ; and 
 whereas the line x x represented the equator of the polarities 
 of the c?*oss-inductions, we now find that the two inclined 
 lines x' x' and x" x" will represent the equators of the polarities 
 of these smaller oblique-inductions and the latter will, in fact, 
 be the outcome of the armature coils that are in the planes 
 quite or about parallel respectively with x x and x" x". 
 
 FIG. 152. Four-inagnet Field, with Oblique-induction. 
 
 Inspection will now make it evident that at all the points e e 
 there is opposition of flow, and at d d confluence of the main 
 and oblique-inductions. Hence the ultimate flow will be denser 
 than normal towards the points d d, and less dense towards e e ; 
 and at these latter regions free magnetism will escape. It 
 will be observed that the horns E and H are weakened, and 
 F and G strengthened, as in the last diagram. But, inasmuch 
 
258 
 
 SPARKING AT COMMUTATORS. 
 
 as the induction in any one of these small oblique circuits is 
 the outcome of a smaller number of armature coils than would 
 be the case with the cross-induction proper, an oblique-induction 
 will not be able to exercise such an effect in diverting the main 
 flow as the cross-induction would be able to exercise. Hence 
 we find that the horns E and H in Fig. 152 will not be so 
 weakened, nor the horns F arid G so strengthened, as in the case 
 represented, Fig. 151 ; and, other things being equal, the 
 distortion will also be lessened, so that not so much lead will 
 be required for the brushes to attain absence of sparking. 
 
 These later remarks are suggested by the discussions on Papers read 
 before the Institution of Electrical Engineers by Mr. W. B. Esson and 
 Mr. Jas. Swinburne, to the last of which reference was made in the 
 foot-note on page 243. See The Electrician. Vol. XXIV., pp. 526-527, 
 and more especially observations made by Prof. George Forbes. 
 
 On the subject of obviating armature reactions, attention may be called 
 to an abstract from an Article or Paper by Mr. J. Fischer-Hinnen in The 
 Electrician for May 26, 1893, pp. 104 to 107. In a diagram accompanying 
 
 this article, a method of counteracting the flow of cross-induction in the 
 necks of the pole- pieces is illustrated. By means of small coils surrounding 
 the necks, a counterflow of induction is set up which baulks that of the 
 cross-induction. Thus the distortion of the field is minimised ; and little 
 lead is required for the brushes, which may consequently rest nearly or 
 quite in the plane of maximum E.M.F. without sparking. 
 
CHAPTER XXVI. 
 
 SPARKING AT COMMUTATORS. ARMATURE 
 DEFECTS. 
 
 WE may now proceed to the consideration of further sources of 
 sparking at the brushes on a commutator, differing materially 
 from those hitherto discussed. These, though still inherent 
 to the machine, are yet scarcely to be included within the 
 general scope and meaning of the term " armature reactions." 
 For any errors under the latter heading, and trouble there- 
 from, are to be understood more especially as questions of cal- 
 culation and design having reference to the machine as a 
 whole ; whereas, the sources of sparking we now propose to 
 inquire into are more of the nature of purely mechanical 
 defects, and are such as will occur in the armature only. Yet, 
 even of these also, though many may arise in the workshop, 
 others may have their origin in a defective design. 
 
 Now the faulty field, it will have been noted, whether too 
 weak, or asymmetrical, or both, did yet not cause any oscillation, 
 or lively swinging to and fro, of the normally fixed planes. We 
 found the planes bent or deflected through the axis of rotation, 
 or unduly inclined. But these effects were permanent and 
 stationary ; and the planes once fixed had no further normal 
 movement. This, however, will not apply to the cases we are 
 now about to investigate. For a fault of a local and mechanical 
 
 82 
 
260 
 
 SPARKING AT COMMUTATORS. 
 
 
ARMATURE DEFECTS. 261 
 
 nature in a revolving armature will be ever changing in its 
 effects within the normally stationary field. Such a fault, thus, 
 at some instants may have no effect, and at other instants be 
 upsetting the whole circuit inside and outside the machine. 
 A result will be that oscillations of the normally fixed planes, 
 or constantly changing deflection at the axis, will be ever 
 recurring. Movements of this kind, moreover, may be steady 
 and continuous, as the armature rotates ; or they may be sharp 
 and intermittent. 
 
 Setting aside the giving way of insulation for this would 
 be equivalent to a short circuit, or an " earth," discussed 
 in a previous Chapter (XXII.) we find these faults are 
 divisible under two main heads, viz., asymmetrical winding, 
 and defective joints. 
 
 To commence with the former, asymmetry, or a want of 
 symmetry, may be produced in various ways. Edison's and 
 Siemens' earlier methods of winding the former illustrated 
 Figs. 14 to 16 are examples in which symmetry was want- 
 ing, owing to the sections differing from one another in total 
 all-round length. Where there are more than one layer of con- 
 ductors along the core, either inside or outside in the case of a 
 Gramme, conductors that should lie next the core may be in 
 the outer layer, and vice versd. Thus the total area enclosed 
 by some of the coils will differ from that enclosed by others ; 
 and so they will generate more E.M.F. if larger, and less if 
 smaller. A long coil or section will also offer more resistance 
 than a short one. But the upshot may be that one-half of the 
 winding would be generating a different output to the other 
 half. Although the necessity for symmetry is well under- 
 stood by manufacturers, it will, nevertheless, be of educational 
 interest briefly to consider the results of its absence. 
 
 Figs. 153, 154 and 155 herewith, represent an armature, 
 which may be either drum or Gramme wound, but of which one 
 half, a, generates less E.M.F. and current than the other half, 6. 
 
262 SPARKING AT COMMUTATORS. 
 
 As regards the drum, the two halves of the winding (as shown 
 Fig. 1) being intermixed, the sketches must be taken as dia- 
 grammatic. Now in Fig. 153, where the weak and the strong 
 half are both bisected by a straight line between the brush points 
 as shown, the self-excited poles of the armature will also come 
 on this line, as indicated by the letters n and s. Assuming a 
 normally symmetrical field, the horns E and H, being equally 
 opposed by the armature poles, will be equally reduced, and 
 the R R plane may be represented by the line through n and s 
 continued, and the distortion of the field will be represented 
 by the inclination of the line yyivxx. 
 
 But with a quarter revolution we find a condition of things 
 as illustrated Fig. 154. The ampere turns of the half a being 
 weaker than those of b, the polarity of the latter is the more 
 powerful, it still being understood that this refers to the self- 
 excited polarity of the armature only. A one-sided polarity 
 thus results, as indicated. The horn E, not being so directly 
 opposed by n, is not sufficiently reduced, so that the short- 
 circuited coil or coils passing it develop too much current, and 
 back- sparking takes place. On the other hand, the horn H is 
 opposed by s abnormally, and is, therefore, reduced too greatly, 
 so that the short-circuited coils cannot generate enough current. 
 Hence forward-sparking ensues. In other words, the R R plane 
 is deflected for the moment into the position R' R'. Further, 
 we may note that as n is abnormally near S, and s abnormally 
 removed from N, N and S both drop, and y y is thus deflected 
 that is, the field is distorted asymmetrically. Fig. 155 is a 
 half revolution ahead of Fig. 154, and will be observed to show 
 reverse results. It will thus be seen that as the armature 
 rotates, the R R plane oscillates with each revolution between 
 the two deflected positions R' R' and R" R", with consequent 
 sparking ; and it may be further observed that the rule that 
 the self-excited polarity of the armature must follow the 
 brushes exactly is here departed from, inasmuch as that rule 
 
ARMATURE DEFECTS. 263 
 
 depended on the armature being wound with perfect symmetry. 
 The mean self-excited polarity will, of course, follow the brushes ; 
 and the oscillations will thus take place about this mean. 
 
 The above, however, constitutes what may be termed a mean, 
 or resultant, effect of asymmetry in armature winding. Re- 
 garded more in detail, it is to be observed that any particular 
 coils that may be larger or smaller than a normal size,- will 
 generate respectively too much or too little E.M.F. when short- 
 circuited by the brush, and so will contain too much or too 
 little short-circuit current. Consequently, in the former case, 
 there will be back- sparking, and in the latter case forward- 
 sparking (see Chapter XX.), the latter possibly being what we 
 have termed "shunt sparking." 
 
 But a point to be noted is, that the asymmetry may be con- 
 fined entirely to the end-winding, each coil being practically 
 symmetrical within the field, and generating the same E.M.F. ; 
 and this view of the subject may be regarded as probably of 
 more importance than the foregoing. For an asymmetrical end- 
 winding, by preventing all the coils or sections from being of 
 the same total length, will cause their resistances corres- 
 pondingly to vary. Hence, the E.M.F. not changing, the 
 short coils or sections, when short-circuited, will have generated 
 within them a larger current than the long coils or sections 
 it being borne in mind that with the very small E.M.F., a very 
 slight variation of resistance will considerably affect the quan- 
 tity of short-circuit current. So there will again be alternate 
 back- and forward-sparking at the brushes. 
 
 Our attention, however, may now be directed to that other 
 source of sparking of a mechanical nature to which we have 
 alluded. We refer to defective joints in the armature winding. 
 This trouble is, of course, especially peculiar to heavy windings 
 of which the coils are constructed of stiff bar or strip, requiring 
 to be riveted, soldered, or otherwise jointed together, as de- 
 scribed in our earlier chapters. A joint improperly effected 
 
264 
 
 SPAKKING AT COMMUTATORS. 
 
 through the parts not being thoroughly clean and bright, or 
 the solder not fulfilling its purpose, or from any other cause, 
 will offer resistance to the flow of current. 
 
 For convenience of reference, such faults due to defective 
 joints may be divided under four heads. These are repre- 
 sented by the letters w, x, y t and 2, in the annexed diagrams, 
 Figs. 156 and 157, of which it will be observed that the former 
 shows one complete heavy bar drum coil, and the latter a wire 
 Gramme coil, with a wire drum coil dotted, while in both the 
 commutator segment is added. A fault w, as will be noted, 
 affects the flow of current into or out of one particular segment 
 
 U c 
 
 jU 
 
 FIG. 156. Diagram of Drum Armature Bar Coil, indicating possible 
 Faulty Joints. 
 
 only. With regard to fault a;, inasmuch as the bar a b apper- 
 tains to two adjacent coils (see Fig. 1), this fault will affect the 
 flow of current between the segment and one of these coils, of 
 which the commencement is shown at c. The arrows indicate 
 the current flow, assuming the segment to be under the 
 positive brush. It is thus to be discerned that a fault x may 
 impede the whole flow into the segment of one half of the 
 armature current, such as we have denoted by c in the 
 diagrams in Chapter XX., and yet permit a free flow of 
 current from the half of the armature similarly denoted by 
 the letter d\ while, in any case, even when that particular 
 
ARMATURE DEFECTS. 
 
 265 
 
 segment is not under the brush, such a fault will impede the 
 flow of current from coil to coil within the winding. Faults y y 
 retard the flow of current within the winding, thus having the 
 latter effect of fault x. One such fault x or ^, however, it is 
 to be noted, can only affect one half of the winding and current 
 at a time, and thus, as the armature rotates, will affect the flow 
 on either side alternately. A fault z combines with all the 
 above the additional effect that, when in series with the brush, 
 like faults w or x, it may check the flow of current to or from 
 both halves of the armature, that is, the whole flow of the main 
 current. 
 
 FIG. 157. Diagram of Gramme or Drum Armature Wire Coil, indicating 
 possible Faulty Joint. 
 
 Now any such retardation of current as above mentioned is 
 to be regarded in two aspects. First, the effects in relation to 
 the main current ; and secondly, the results in relation to the 
 short-circuit currents, as in the coils b and / in the diagrams in 
 Chapters XVII. to XX. The latter results we shall dis- 
 cover to be of the greater consequence, as a matter of fact, in 
 reference to sparking. For as regards the main current, it is 
 to be noted that any of these faults taking effect will be in 
 series with the outer circuit with all its resistance. Thus the 
 actual resistance of a fault may be so small, that when added 
 to the total resistance of the inner and outer circuit, its 
 
SPAKKING AT COMMUTATOES. 
 
 addition thereto may be nearly or quite inappreciable as 
 regards any results produced. Yet the effect on the short- 
 circuit currents may be considerable. In so far, however, as 
 the main current may be affected, that is, reduced in quantity, 
 this would cause"one side of the armature to have fewer ampere 
 turns than the [other side, and so produce a distorted polarity 
 as explained in connection with Figs. 153 to 155, with con- 
 sequent sparking. 
 
 It is undoubtedly, however, the effects on the short-circuit 
 currents that are of the greater importance. For a fault too 
 weak to materially reduce the main current may, nevertheless, 
 almost eliminate the short-circuit current. Thus, suppose the 
 total resistance] of a circuit, inside and outside, to be 3 ohms, 
 while the resistance of one single armature coil is only '0006 
 ohm, it is evident that a faulty joint in such a coil offering, say, 
 006 ohm resistance, would only increase the total resistance to 
 3'006 : but^it would increase the total resistance of the par- 
 ticular coil to *0066, or eleven times greater than normal. 
 Hence, assuming a normal main current of 44 amperes, when 
 the brush would be at the normal non-sparking position, each 
 time this coil was short-circuited taking the position lettered 
 b, that is, in the diagrams in Chapters XVII. to XX. it 
 would be carrying only the eleventh part of the short-circuit 
 current necessary for a sparkless recession of its segment from 
 the brush, or only 4 amperes instead of 44, thus leaving a current 
 of 40 amperes to be broken, with consequent forward sparking. 
 The necessity, therefore, for extreme care in making the joints 
 in a coil so that all the coils shall have a similar total resist- 
 ance or, better still, the advisability of avoiding joints as 
 much as possible is thus the more obvious if sparkless running 
 is to be attained. 
 
 An additional argument is, of course, to be found here for 
 the use of carbon brushes. For these, as was explained in 
 Chapter XXI., are but little affected by the short-circuit 
 
VARIOUS CAUSES. 267 
 
 currents; and, consequently, a defective joint that might cause 
 a metal brush to. spark in the manner above indicated would 
 produce no such effect with a carbon brush. 
 
 Some further causes of sparking which may now be appended 
 are of a more general nature. Thus any source of vibration 
 in running a machine, such as a clumsy joint in the Jbelt, com- 
 bined with too light a shaft, or insufficiently rigid bolting 
 down, may cause a slight jumping of the brushes, with spark- 
 ing as a consequence. A further tightening of the brushes 
 will increase the wear of both them and the commutator seg- 
 ments. The segments, moreover, may wear unevenly, one or 
 more of them perhaps becoming burnt by periodical sparking 
 (such as by a bad joint above discussed), which will then cause 
 the brushes to jump, and so to spark yet more. It is especially 
 important that all the brush gear should be strongly con- 
 structed, so as to hold the brushes firmly, and not to spring. 
 With a field symmetrical as is usual, sparking may be caused 
 by the brush points not being diametrically opposite each 
 other, so that they cannot both be in the non-sparking plane 
 at once, and one brush may spark and not the other. Trouble 
 may also arise from the commutator not being kept clean 
 and free from oil and grit. Vibration, such as may lead to 
 sparking, may ensue from the armature being out of balance. 
 But a further source of trouble in this direction has been 
 found in the use of cast metal segments, instead of rolled 
 metal. This has reference, however, more especially to tram- 
 car motors, wherewith, on account of the jolting, it is necessary 
 to keep the brush springs somewhat tight. The frictional wear 
 is thus great; and it would appear that, owing to the cast 
 segments not being so homogeneous, or so uniform in their 
 hardness and quality, as rolled segments, they do not wear so 
 evenly. Hence, with the cast segments, at least under the cir- 
 cumstances indicated, the commutator much sooner gets out of 
 truth, causing the brushes to jump, with consequent sparking. 
 
CHAPTER XXYIL 
 
 THE TAPER OF COMMUTATOR SEGMENTS. 
 
 A CLOSING chapter may here be added, having reference once 
 more to the construction of commutators, more especially to 
 the formation of the segments to a correct taper, so that 
 accuracy of fit may be ensured. 
 
 When the segments or bars for a commutator are cast, 
 instead of being rolled, they require to be machined or filed on 
 their sides before being put together. To secure accuracy in 
 this operation, it is necessary that the parts be worked up to a 
 gauge. In this manner the requisite exactness in both thick- 
 ness and taper may be obtained. When, on the other hand, 
 drawn copper is used, it is equally necessary to specify the 
 exact taper in ordering, and generally to send a gauge, sample, 
 or template to the manufacturer. 
 
 A very useful and simple gauge or tool has been designed 
 for this purpose by Mr. A. P. Trotter, and was adopted in 1886 
 by Messrs. Goolden and Trotter. This instrument is illustrated 
 on a reduced scale, in Figs. 158 and 159 herewith. There are 
 two ways of stating the taper for a commutator segment or 
 strip. One is to give the angle between the sloping sides, and 
 the other is to describe the slope of one side in relation to the 
 other side as a gradient of " one in so many/' As a circle is 
 
270 THE TAPER OF COMMUTATOR SEGMENTS. 
 
 divided into 360 degrees, the angle for a segment is, of course, 
 arrived at by simply dividing the 360 by the number of 
 segments, or "parts." Thus the angle of a strip for a 
 " 40-part " commutator will be 360/40 = 9degs. But it must 
 be noted that this assumes the parts put together tvithout 
 the insulation. This point requires attention. It has to be 
 borne in mind that the mica or other insulation between the 
 segments does not taper, but is of even thickness. Fig. 160 
 shows how all the strips of a commutator are made to fit in the 
 first instance, minus the insulation ; while Fig. 161 shows the 
 same parts put together with the insulation. This latter at 
 once increases the diameter ; and it should be noted further, 
 in this 'latter case, that the sides of the sections, if continued 
 
 FIGS. 158 and 159. 
 
 inwards, do not meet at the centre. Here, in fact, is demon- 
 strated how great an effect on the ultimate diameter the insu- 
 lation may have, especially when there are a large number of 
 segments, and a high voltage requiring the insulation to be 
 thick. In designing a commutator, it may be observed that 
 
 r> nxd 
 
 >=R -^' 
 
 where r is the inside radius of the commutator when put 
 together without the mica, R the inside radius of the complete 
 commutator, n the number of segments, and d the thickness of 
 the mica or other insulation. 
 
 The principle of the gauge under discussion is bhown in the 
 diagram Fig. 162. The hatched portion represents the section 
 
THE TAPER OF COMMUTATOR SEGMENTS. 
 
 271 
 
 of a commutator segment. In trigonometrical language, the 
 line b c represents the " cotangent " of the angle cay, other- 
 wise denoted by the letter 0, which is the angle of the strip. 
 The radius a b is unity. It will thus be observed that the 
 taper of the section may be expressed in terms of the gradient 
 of the line a c, as compared with the horizontal line x y^ 
 obviously to be determined by the number of times a b will 
 14 go into " b c. Assuming a 40-part commutator, when will 
 be 9, and making use of a table of natural sines, we find the 
 numerical value of the cotangent b c for the angle = 9, to be 
 
 FIG. 160. 
 
 FIG. 161. 
 
 6-3138. The line a b being unity, the gradient of a c will 
 thus be 1 in 6 '3 13 8, which may be given as the taper of the 
 segment. 
 
 It so happens, that for any number of segments in a com- 
 mutator, from about 40 up to 160 or 180, if the number 
 of segments be divided by the corresponding numerical co- 
 tangent of the angle of one of them, a constant, 6 -3, is arrived 
 at when not exceeding one place in the decimals. This is, 
 therefore, useful, since by simply dividing the number of 
 segments by this constant, we arrive at the taper in terms of a 
 gradient. 
 
272 
 
 THE TAPER OF COMMUTATOR SEGMENTS. 
 
 Referring again to Fig. 162, it will be observed that the 
 triangles a b c and a c y are similar ; and cy = ab, and a y 
 = b c. It is also obvious that if we shorten the parallelogram 
 abcy\>j moving c y nearer to a b, we shall open the angle 6 ; 
 and vice versd. Conversely, if the angle be opened b c will 
 become shorter, and if 6 be closed b c will lengthen. Thus for 
 a large number of segments, and consequently a small angle 
 6, b c and a y will be much longer than with a small number 
 of segments and consequent large angle. Now a b being unity, 
 so also is c y. The distance of c y from a b that is, the length 
 of the cotangent be may be calculated for every feasible 
 number of segments in a commutator in the manner stated 
 
 162. 
 
 above that is, by dividing the number 6f segments in 
 each case by the constant 6-3. This being done, the 
 various points where c y will stand on a y, in accordance with 
 the number of segments and consequent size of angle 0, may 
 be marked on a y> and the corresponding number of segments 
 figured on also. A scale will thus be formed on a y. Sub- 
 sequently, in order to fix a c at the proper slope for any number 
 of segments, it will be merely necessary to make cy stand 
 perpendicularly on the mark on the scale having that number. 
 It is in this manner that the scale on the gauge under dis- 
 cussion is formed, with a standard unit (corresponding to c y, 
 Fig. 162), of one inch. Hence, to use the tool, a flat standard 
 inch gauge, held perpendicular to the edge of the 'scale, is 
 
THE TAPER OF COMMUTATOR SEGMENTS. 273 
 
 placed between the jaws exactly at the mark denoting the 
 number of segments. This sets the gauge open at the required 
 angle or taper. In giving instructions, the distance of the inner 
 edge of the segment from the centre (a, Fig. 162) may also be 
 given ; though this must be the small radius with the insula- 
 tion absent, as shown and explained in connection with Fig. 160. 
 Thus the taper and thickness of the segment are both stated. 
 The gauge, after having been set, is fixed by the fly-nut and 
 cross arm, as shown in the illustration. 
 
INDEX. 
 
 PAGE 
 
 Accumulation of Electromotive Force between Generators in Series 
 
 135, 176, 180, 181 
 Air-Gaps, Waste Field, due to ......... 236,237,252 
 
 Alioth and Co. (Messrs.), Commutator of "Helvetia" Dynamo 103, 104 
 Alternating Currents and Long Sparks... ... ... ... ... 136 
 
 Ampere's Law ... ............... 12,131 
 
 Andersen, F. V., Armature Winding ... ... ... ... 47-51 
 
 Arc, Voltaic ............ 126-131, 136, 140, 142-146, 164 
 
 - Back E.M.F. in ..................... 144 
 
 - Comparative Consumption of Electrodes ... ... 130, 146 
 
 - in Self-induction Spark ... ... ... ... ... 145 
 
 - in Sparking at Brushes ... ... ... ... ... 164 
 
 Armature, Balancing of ... ... ... ... ... ... 27, 54 
 
 - Coils, Construction of (see Bars) ............ 31-35 
 
 - Coils, Interchangeability ... ... ... .. ... 55 
 
 - Cooling by Ventilation 
 
 23, 26, 29, 35-37, 41, 42, 47, 49, 66-69, 82, 84, 122 
 - Cooling by Oil or Fluids ............... 82 
 
 Current Direction in ............... 6,7,8,10 
 
 - Drum, Advantages of .. 1, 2, 23, 219, 220, 246, 252, 254,264, 265 
 
 - Drum, Theory of .. ............... 1-16 
 
 - Drum, Windings (see Armatures and End Windings). 
 
 - Drum, Electromotive Force (see Electromotive Force), 
 
 Generation of 6, 8-12, 147-149, 158, 159, 180, 184, 223, 241 
 
 Drum, End Winding (see End Windings). 
 
 Heating of ............... 21,22,28-31 
 
 - Poles, Immobility of ... ... ... ... ... ... 14 
 
 Polarisation, Self- ... ..! ...... 12, 219, 227, 232 
 
 - Polarisation, Ultimate... ... ... ... ... 12, 13 
 
 - Reactions (see Reactions) ............ 217-219 
 
 - Reactions in Weak Field ............ 215-225 
 
 - Reactions in Weak Field regarded analytically ... 232-237 
 
 - Reactions in Asymmetrical Field ... ... ... 245-255 
 
 - Self-Excitation, Polarisation, Magnetisation (see Polarisation) 
 
 12-14, 71-75, 158, 215, 219, 227, 232, 248, 262 
 
 - Self -Excitation, Governed by Brushes 158, 159, 217, 262, 263 
 
 - Plane of Maximum E.M.F. of 
 
 151, 156, 159, 160, 203, 222, 223, 225, 241 
 
 Symmetry of Winding, Necessity for ...... 28, 262, 263 
 
 Wire Drum Winding, Advantages of 
 
 23 
 
276 INDEX. 
 
 PAGE 
 Armatures, Drum 
 
 Alteneck, Hefner von ... ... ... ... .. 51,55 
 
 Andersen, F. V 47-51 
 
 Clark (Latimer), Muirhead and Co. 47-51 
 
 Cronipton and Swinburne ... ... ... ... ... 43-47 
 
 Crompton and Kyle 77-80 
 
 Edison 39-42 
 
 Eickemeyer ... ... ... ... ... ... .. 55-62 
 
 Fritsche ... 83, 84 
 
 Hopkinson ... ... ... ... ... ... ... 51-54 
 
 Kapp 63-67 
 
 Kyle 77-80 
 
 Parsons 80-83 
 
 Sayers 237-243 
 
 Siemens 2,6,17-23,31,55 
 
 Swinburne 34, 35, 67-75 
 
 Swinburne and Crompton ... ... ... ... .. 43-47 
 
 "Westminster" 47-51 
 
 White ... 80 
 
 Cross- Connector for ... ... ... ... ... ... 54 
 
 Assimilated Currents, Potential, or E.M.F. of 175-181 
 
 Assimilation of Currents, General Theory 168-174 
 
 of Main Armature Currents, with Short-Circuit Currents 
 
 185-192, 241 
 
 of Magnetic Flows or Inductions ... ... ... ... 255 
 
 Asymmetry in Armature Winding ... ... ... ... 259-263 
 
 in End Winding ... ... ... ... ... ... 263 
 
 of Field 215,245-255 
 
 of Field, Causes of 253 
 
 of Waste Field 253 
 
 Augmentation or Addition of Currents 
 
 167, 168, 186, 188, 190-192, 241 
 Ayrton and Perry's Table of Specific Resistances ... ... ... 95 
 
 Back E.M.F. in Arc ... 144 
 
 Back-Induction 73, 75, 216-218, 223-225, 232, 233, 247 
 
 Back, or Negative, Lead to Brushes 
 
 152, 153, 184-186, 188-190, 192, 223, 224, 232, 239-243 
 
 Back Sparking 187,188,191,242 
 
 Bars, Conductor, on Armature General Considerations 25-30 
 
 Construction of ... ... . 31-35 
 
 Eddy or Foucault Currents in 28-30 
 
 Heating of 21,22,28-31 
 
 Peripheral Creeping of ... ... ... ... ... ...15,16 
 
 Heating by Foucault Currents, Prevention of 31-37 
 
 Bedding of Commutator Segments ... ... ... 114 
 
 Bending or Deflection of Theoretical or Normally Fixed Planes 
 
 204, 215, 251, 254, 259, 262 
 
 Binding Wires round Armature ... ... 16, 26, 48, 61, 82 
 
 Brushes, Burning of 164,188,190 
 
 Carbon (see Carbon Brushes) 195-202,266-267 
 
 -Elementary Consideration of... ... ... 161 
 
 Governing Armature Self-Polarisation 158, 159, 217, 262, 263 
 
 Cead, Apparent and True 157 
 
INDEX. 277 
 
 PAGE 
 Brushes, Lead, Back or Negative 
 
 152, 153, 184-186, 188-190, 192, 223, 224, 232, 239-243 
 
 Lead, Forward or Positive 
 
 152, 186-188, 190-192, 217, 223, 224, 224, 232, 233 
 Sparking at '(see Sparking). 
 
 Cable, Copper Wire, for Armature Conductor Bars ... ...31, 32 
 
 Carbon Brushes 126,195-202,266,267 
 
 Current Density with ... ... ... ... ... 202 
 
 Heating of Forward Edge 201,202 
 
 Resistance of Contact Cause of Non-Sparking . . 195-201 
 
 Short- Circuit Currents Eliminated 196,201 
 
 Carbons of Arc Lamp, Consumption of . . ... ... ... 130 
 
 Cast Segments for Commutators, Unequal Wear of 267 
 
 Copper and Silver ... ... ... ... ... ... 90 
 
 Gun-Metal 87,90,91,115 
 
 Phosphor-Bronze ... ... ... ... ... 87, 90 
 
 Cessation of Force, Plane of ... 151, 152, 157-160, 166, 193, 204, 221 
 
 Choking of Magnetic Flow or Induction 247,250 
 
 Chord Winding, Swinburne's ... ... ... ... ... ... 67-75 
 
 Circuit, for Current, as Compared with Discharge ... ... ... 129 
 
 Circuits, Short-, Danger of 22 
 
 in Exterior Circuit, as Cause of Sparking ... ... 204-213 
 
 at Commutator ... ... ... ... ... ... 90 
 
 in Armature ... ... ... ... ... ... ... 261 
 
 in Armature Coils under Brush 
 
 161, 163, 166-181, 183-194, 196-201 
 
 in Armature Coils, Currents of, as cause of Non-Sparking 166 
 
 ditto, and Assimilation ... ... ... ... 167-181 
 
 ditto, with Misplaced Brushes 183-194 
 
 ditto, Sole Cause of Sparking with Forward Lead 
 
 186-188, 190-192, 194 
 
 ditto, Cause of Heavy Sparking with Negative Lead 
 
 184-186, 188-190, 192, 224 
 
 ditto, Elimination with Carbon Brushes ... 196-201, 266, 267 
 ditto, Cause of Non-Sparking with Sayers' Winding 239-243 
 
 ditto, Affected by Asymmetrical Winding 262,263 
 
 ditto, Affected by Bad Joint in Winding, causing Sparking, 265, 266 
 
 ditto, with Motor 224,225 
 
 Clark (Latimer), Muirhead and Co.'s Armature Winding ... ...47-51 
 
 Commutators 107-115 
 
 Coils of Armature Winding (see Bars). 
 
 Collector or Commutator ... ... ... ... ... ... ... 85 
 
 Commutator, Burning of Segments 87,89,186,192 
 
 Cones (see Cones) ... 93, 100, 101, 106, 107, 109, 110, 112 
 
 Considerations in Design 86-91,104-122 
 
 Connections with Armature Winding ...61,62, 117-122, 149 
 Commutators 
 
 Clark (Latimer), Muirhead and Co. 107-115 
 
 Crompton 103-106 
 
 Elwell-Parker 89, 122 
 
 Fritsche 83,84 
 
 Gulcher 107-110 
 
 " Helvetia" 103-105,109-111 
 
 Hopkinson 111-114,122 
 
278 INDEX. 
 
 PAGE 
 Commutators 
 
 Kapp 107-113 
 
 Kennedy 107-111 
 
 Paterson and Cooper ... ... ... . ... 111-115 
 
 Mordey's " Victoria " 103-107 
 
 Metal for Segments and Lugs ... 87,90,117-119,267 
 
 Fitting Together 91 
 
 Gauge for Taper of Segments, Trotter's 269-273 
 
 Insulation 86, 88, 89, 91, 93-101, 105-113, 122, 270, 271, 273 
 
 Insulation and Diameter 88,270 
 
 Number of Parts or Segments ... ... ... ... 88 
 
 Size, Length, and Diameter 87-89, 270, 271 
 
 Groove, Easton's Patent 105,106 
 
 Coils (Sayers) 238,239,241 
 
 -End-Play 89 
 
 -Driving 114,115 
 
 Lubrication ... ... ... ... ... ... ... 90 
 
 . Potential Difference between Segments ... 9-12, 93, 148, 241 
 
 - Plane of Maximum E.M.F. of, 151, 156, 157, 160, 181, 222, 223, 225 
 
 Commutation, Plane of ... 8, 60, 72-75, 149-153, 158-162, 192-194 
 
 in Reference to E.M.F. ... 158 
 
 Governed by Brushes 152,153,192-194 
 
 Compound Winding 211,212 
 
 Cones 93,94,98-101,106,107,109,110,112 
 
 Angle of 106,107,109 
 
 Made of Vulcanised Fibre 93, 94, 98, 99, 101, 106 
 
 Built of Mica 93,99,100,101,106 
 
 Contacts, Consumption of ... ... ... ... ... ... 146 
 
 Resistance of, with Carbon Brushes .. 195-202 
 
 Continuous Current, Commutators for ... ... ... ... .. 86 
 
 Current Compared with Static Electricity ... ... ... 129 
 
 Current Transformer, Elwell-Parker 89,122 
 
 Disruptive Discharge of Sparks 126, 127, 136, 142 
 
 Copper for Commutator Segments ... ... ... ... 90,119 
 
 for Lugs 117,118 
 
 Inactive 4-6,25,26 
 
 Counteraction of Cross-Induction, Fischer-Hinnen's Method... ... 258 
 
 Counteraction, Magnetic, or Demagnetisation, of Horns 
 
 75, 219-221, 225, 234, 235, 251 
 
 Crompton and Swinburne, Armature Drum Winding 43-47 
 
 and Kyle, Armature Drum Winding 77-80 
 
 Commutator 103-105 
 
 Cross-Induction 12, 71-75, 216-218, 232-235, 247, 248, 252, 253, 256 
 
 Cause of Field Distortion 249 
 
 with Double-Ended Magnets 256 
 
 Prevention of, by Fischer-Hinnen's Method ... ... 258 
 
 Current, Cause of Deflagration ... ... ... ... ... ... 145 
 
 Cause of Heat 21-23, 25, 28-30, 31-37, 144, 201, 202 
 
 - Cause of Heat in Arc 130, 146, 188, 190 
 
 Continuous, Transforner ... ... ... ... 89, 122 
 
 - Counter to E.M.F 152,153,160,193 
 
 -Density with Carbon Brushes 201,202 
 
 Direction and Generation of, in Drum Winding ... ... 6-12 
 
 -of Magnetism 227,230,236,237,249,256 
 
 Primary and Secondary ... ... ... ... 132-137 
 
 Smelting by ... ... ^ 144 
 
INDEX. 279 
 
 PAGE. 
 
 Currents, Assimilation of 168-174,185-194,241 
 
 Assimilated and E.M.F 175-181 
 
 Augmentation of 167, 168, 186, 188, 190-192, 241 
 
 Foucault, or Eddy 28-37 
 
 - Opposition of ... 167-169, 172, 173, 18a-192 
 
 Opposition of, Energy Lost by 181, 242, 243 
 
 Defective Joints in Armature Winding... 261,263,267 
 
 Defects in Armature Winding as Causes of Sparking ... 259-267 
 
 Deflagration in Spark Due to Current 145,207 
 
 Deflection of Centre Plane of Distortion of Field 251,254 
 
 - of Normally Fixed Planes 204, 215, 251, 252, 254, 259-262 
 Density of Current with Carbon Brushes ... ... ... 201,202 
 
 Difference of Potential between Armature-Winding Sections 11, 19, 22, 241 
 
 of Potential between Commutator Segments 9-12, 93, 148, 241 
 
 Disruptive Electric Discharge 126-128, 136, 142, 143, 145 
 
 Smallness of Time Required for 145 
 
 Distortion of Field 13, 74, 158, 160, 230, 231, 235 
 
 of Field with Double-Ended Magnets 255-258 
 
 of Field, Plane of, Deflected 251-254 
 
 of Field, Caused by Cross-Induction 249 
 
 of Magnetic Flow 230,231 
 
 of Self- Polarisation of Armature 260-263 
 
 Diversion of one Current by another 173,192,224,240,241 
 
 Double- Ended Magnets 245,246,255-258 
 
 Driving Commutators ... ... ... ... ... ... 114, 115 
 
 Conductors on Armature ... 15, 16, 26, 49, 52, 56, 66, 69 
 
 Drum- Winding (see Armature, Armatures, Bars, and End Windings). 
 
 Theory of 1-16 
 
 as compared with Gramme Winding 
 
 1, 2, 219-221, 246, 252, 254, 264, 265 
 
 Generation of E.M.F. and Current in ... ... ... 6-12 
 
 Dynamo, Compound Winding and Sparking 205,211-213 
 
 Separately-Excited and Sparking 205,206 
 
 Series Winding 205,209-211 
 
 Shunt Winding 205,207,208 
 
 as compared with Motor 73, 74, 224, 225 
 
 Earth, Inefficient as Conductor 205 
 
 " Earths " and Sparking 204,205,261 
 
 Ebonite, Specific Resistance of 95 
 
 Edison 2,39-42,261 
 
 Eddy, or Foucault Currents 28-30 
 
 Prevention of 31-37 
 
 Efficiency and Sparking ... ... ... ... ... ... ... 225 
 
 Eickemeyer's Winding 55-62 
 
 Electromotive Force, Generation in Armature 
 
 6, 8-12, 147-149, 158, 159, 180, 184, 223, 241 
 
 Generation in Induction Coil 141 
 
 Accumulation from Generators in Series ... 135, 176, 180, 181 
 
 Commutation of ... ... ... ... ... 158, 160 
 
 of Assimilated Currents 175-181,241 
 
 External, Dependent on External Resistance 
 
 178-181, 205-213, 241 
 Flow of Current counter to _ 152,153,160,193 
 
280 INDEX. 
 
 PAGE 
 Electromotive Force in Secondary Coil ... ... ... ... ... 135 
 
 of Induction Spark 135,136 
 
 of Self-induction Spark 127,140,142 
 
 with Disruptive Sparks or Discharges . . . 127, 136, 142, 143 
 
 cause " Shock " 143,144 
 
 used Internally in Armature 180,181,212 
 
 and Voltaic Arc 130,136,144 
 
 Back, in Arc ... ... ... ... ... ... ... 144 
 
 High Tension, in Crystal Palace Experiments ... 136,143 
 Plane of Maximum, through Armature 
 
 151, 156, 159, 160, 203, 222, 223, 225, 241 
 
 Plane of Maximum, through Commutator 
 
 151, 156, 157, 160, 181, 203, 225 
 
 Electrodes, Consumption of 130, 146, 164 
 
 Elwell-Parker Motor Generator 89,122 
 
 End Windings, Heavy, for Drum Armatures (see also Armature ; Armatures) 
 Circular and Radial Connectors : Edison ... ... ..39-41 
 
 Disc: Edison 41-42 
 
 Volutes, Double Interior : Alteneck ; Hopkinson ; Eickemeyer 51-62 
 
 Volutes, Double Exterior: Crompton and Kyle 77-80 
 
 Volute, Single Interior and Crank : Crompton and Swinburne ; 
 
 Latimer Clark, Muirhead and Co. ... ... ... ...43-51 
 
 Helical, Interior : Kapp ... ... ... ... ... ...63-67 
 
 Helical, Exterior : Parsons; White; Fritsche 80-84 
 
 Connector, Double Volute, Jointless 54 
 
 Energy Lost in Commutating Current 162, 163, 181, 242, 243 
 
 Esson, W. B., on Armature Reactions ... . . ... ... ... 258 
 
 Ewing, Prof. J. A., on Magnetic Induction .. ... ... ... 222 
 
 Excitation, Self, of Armature (see Polarisation) 
 
 12-14, 71-75, 158, 215, 217-219, 227, 232, 248, 262 
 
 External Circuit, Faults in, Causes of Sparking 203,213 
 
 External Voltage of Generator Dependent on External Resistance 
 
 178-181, 205-213, 241 
 
 Fans for Ventilating Armature ... ... ... ... 36, 37, 49 
 
 Fibre, Vulcanised ... 93-99, 101, 106 
 
 -for Cones 93,94,98,99,101,106 
 
 Field, Normal Direction and Distribution of Lines of Force ... ... 6 
 
 Asymmetrical and Armature reactions ... ... 215, 245-255 
 
 Causes of Asymmetry of ... ... ... .. .. ... 253 
 
 Distortion of 13,74,158,160,230,231,235 
 
 Distortion of, caused by Cross-induction . . ... ... 249 
 
 Distortion of, Deflected 251-254 
 
 Distortion of, with Double-ended Magnets 255-258 
 
 Dynamo, Compared with Motor 73, 74, 224, 225 
 
 Self -created by Armature ... ... 73,74,224,239 
 
 Strength of, Relative to Armature Self -excitation 
 
 13, 157, 158, 204-213 215-225 
 
 Waste 225,236,237,252,253,257 
 
 Waste, Asymmetry of ... ... ... ... ... ... 253 
 
 Weak, and Armature Reaction 215-225,247,248 
 
 Weak, Sparking versus Efficiency ... ... ... ... 225 
 
 Weak, Causes of 247,248 
 
 Weak, and Component Armature Reactions ... ... 232-237 
 
 Fischer-Hmnen, Mr. J., Counteraction of Cross-Induction 258 
 
INDEX. 281 
 
 PAGE 
 
 Flash of Lightning 128,145 
 
 at Opening of Switch, Compound Nature of ... 145, 146 
 
 Flashing at Commutators, Compound Nature of 147,164 
 
 "Shunt" ... .165,174,263 
 
 Brilliancy due to Current 145,207 
 
 Flats on Commutator ... ... ... ... ... ... ... 89 
 
 Fleming, Prof. J. A., on Flashing 145 
 
 Flow of Magnetic Induction, Similarity to a Current 227-230, 236, 249, 256 
 
 Internal in Closed Circuit 78,227-229,237 
 
 Confluence and Opposition of Flow 225-237,245-258 
 
 Density 234 
 
 Distortion or Diversion of ... 230, 231, 249 r 250, 255, 258 
 
 Measurement of Strength of ... ... ... ... 233 
 
 Throttling or Choking of 247, 250-258 
 
 Forbes, Prof. G., on Double-Ended Magnets 258 
 
 Force, Lines of 4-9, 75, 148, 158, 222-224, 230, 236, 239 ; 252 
 
 Plane of Cessation of 152, 157-160, 166, 193, 204, 221 
 
 Forward Induction 217,218,223,224,232-243,247 
 
 Induction without Sparking, Sayers' Winding ... 237-243 
 
 or Positive Lead to Brushes 
 
 152, 186-188, 190-192, 217, 223, 224, 232, 233 
 
 Sparking .. 186,189,190,223,224,242,262 
 
 Sparking, Elimination by Sayers' Winding ... ... 237-243 
 
 Foucault, or Eddy Currents 28-37 
 
 Free Magnetism (see Magnetism) ... 225, 229, 236, 237, 252, 253, 257 
 
 Fritsche's Armature Winding ... ... ... ... ... 83,84 
 
 Generation of E.M.F. in Armature 
 
 6, 8-12, 147-149, 158, 159, 180, 184, 223, 241 
 
 in Induction Coil ... ... ... ... .. ... 141 
 
 Gramme Winding ... ... ... ... ... ... Introductory 
 
 as Compared with Drum Winding 
 
 1, 2, 192-221, 246, 252, 254, 264, 265 
 Grooving of Commutator by Brushes ... ... ... ... ... 89 
 
 Giilcher Commutator 107-110 
 
 Gun-metal for Commutators 87,90,91,115 
 
 Heat due to Excess of Current in Conductor ... ... ... 21, 22 
 
 due to Foucault Currents ... ... ... ... ... 28-31 
 
 due to Foucault Currents prevented by Construction . . . 31-37 
 
 prevented by Ventilation 
 
 23, 26, 29 ; 35-37, 41, 42, 47, 49, 66-69, 82, 84, 122 
 
 prevented by Liquids 82 
 
 due to Resistance 21,25,143.144 
 
 in Sparks ... 144 
 
 in Flash at Commutator 207 
 
 in Voltaic Arc . 144 
 
 of Forward Edge of Carbon Brush ... 201,202 
 
 Heavy Drum Winding 25-84 
 
 Helical End Winding : Kapp's 63-67 
 
 -Parsons' 80-83 
 
 " Helvetia " Commutator : Alioth and Co 103,104 
 
 High Tension for Disruptive Sparks ... 127, 128, 135, 136, 140-144 
 
 Experiments at the Crystal Palace 136, 143 
 
 Hooper's Vulcanised Indiarubber, Specific Resistance of ... ... 95 
 
INDEX. 
 
 PAGE 
 Hopkinson's Commutator ... ... ... .. ...111,113,114 
 
 Commutator Lug ... ... ... ... 122 
 
 End Winding 51-54 
 
 Method for the Prevention of Heating 31,36 
 
 Horns, Counteraction or Demagnetisation of 
 
 75, 219-221, 225, 234, 235, 251. 254 
 
 Comparative Effects on, of Gramme and Drum Armatures 219, 220' 
 
 Effect on, of Armature Reactions, with Double- Ended 
 
 Magnets 256-258 
 
 Reversal of 75, 21& 
 
 Hysteresis 208, 210, 222, 
 
 Incandescent Lamps in Parallel, Uniform Voltage for .. ... 211 
 
 Induction Coil 131,132 
 
 -Spark 126,127,135 
 
 Self (see Self-induction). 
 
 Main, of Field (see also Lines of Force) 78, 160, 230-237, 245-258 
 
 - Cross 12, 71-75, 216-218, 232-236, 247-258. 
 
 Cross, with Double-Ended Magnets 256 
 
 - Back ... 73, 75, 216-219, 223, 232, 233, 247-258 
 
 -Forward ... 217,218,223,224,232,237-243,247-258 
 
 Forward, with Dynamo, Inutility of... ... ... ... 224 
 
 Forward, with Dynamo and Sayer's Armature .. 237-243- 
 
 - Component 217-219, 223, 227, 232-237 
 
 Resultant 217-219,231 
 
 Oblique 257,258 
 
 Limited Similarity to a Current ... 227-230, 236, 249, 256 
 
 - Internal Flow in Closed Circuit ... 78,227-229,237,253 
 Throttling in Necks of Pole-pieces 247, 250-258 
 
 Confluence and Opposition of Flow 225-237,245-258 
 
 Saturation by 248,250- 
 
 Augmentation of ... ... ... ... ... ... 250 
 
 Assimilation of ... ... .. ... ... ... ... 255 
 
 Distortion or Diversion of ... 230, 231, 249, 250, 255, 258 
 
 Non-elimination of ... ... ... ... .. ... 236 
 
 Need of Sectional Area for 253,255 
 
 -Density of 234 
 
 Measurement of ... ... ... ... ... . . 233 
 
 Waste Field 225, 236, 237, 252, 253, 257 
 
 Inertia, Electric 141 
 
 Insulation in Armature, Generally Considered 
 
 18, 19, 35, 41, 42, 53, 58, 61, 64, 65, 69, 82 
 
 Destruction of, by Burning ... ... ... ... 22, 37 
 
 in Commutator, Generally Considered 
 
 86, 88, 91, 93-101, 105-113, 122 
 
 in Commutator, Affecting Diameter 8, 270, 271, 273- 
 
 in Commutator Cones (see Cones) 
 
 93, 94, 98-101, 106, 107, 109, 110, 112 
 
 Specific Resistances ... ... 95 
 
 -Air 35,59-61,109,110,113- 
 
 Asbestos 93,99 
 
 - Cotton Tape 35 
 
 Ebonite 95, 98 
 
 Fibre, Vulcanised 93-99,101,106 
 
 -Glass 95,98 
 
INDEX. 283 
 
 PAGE 
 Insulation, Indiarubber ... ... ... ... ._ 95, 98 
 
 Manilla Paper 35,48 
 
 Mica 89,91,93-95,98-101.106,114 
 
 Oil 90 
 
 Paraffin 95-99 
 
 Porcelain ... ... ... ... ... . . ... 98 
 
 -Shellac 35,48,91,95,96 
 
 -Silk Tape 35 
 
 Tape, Silk, or Cotton ... 35 
 
 Varnish 31, 35, 47, 48, 91, 96, 99 
 
 - Vulcanised Fibre 93-99,101,106 
 
 Vulcanite 98 
 
 - Willesden Paper ... 35,48 
 
 -Wood 64,96-98 
 
 Internal Flow of Induction in Closed Magnetic Circuit 78, 227-229, 237 
 Iron Commutator Segments ... ... ... ... ... ... 90 
 
 Joint in Belt, Cause of Sparking ... ... ... ... ... 267 
 
 Joints, Defective, in Armature-Winding, as Causes of Sparking 261-267 
 
 Effect on Short- Circuit Currents 266 
 
 Effects mitigated with Carbon Brushes ... ... ... 267 
 
 Kapp, Commutators of 107,108,111,112 
 
 Armature Winding of ... ... ... ... ... ...63-67 
 
 Kennedy, Commutator of ... ... ... ... ... 107,108 
 
 Knife-edges, for Balancing Armature ... ... ... ... ... 27 
 
 Lamps in Parallel = 211 
 
 Lead to Brushes ... ... ... ... ... ... ... 13,14 
 
 Apparent and True ... ... ... ... ... ... 157 
 
 Back or Negative 152, 153, 184-186, 188-190, 192, 2?3, 224, 232, 
 
 239-243 
 
 Forward or Positive 152, 186-188, 190-192, 217, 223, 224, 232, 233 
 
 Controlling Armature Self -Polarisation 158, 159, 217, 262, 263 
 
 Negative and Short-Circuit Currents ... ... ... 224 
 
 - Positive, with Motor 73,224,225 
 
 Negative, with Dynamo, Armature Creates its Own Field 224 
 -Negative, ditto Sayer's Method 237-243 
 
 Unequal, Sparking at One Brush 248-255 
 
 with Double-Ended Magnets 258 
 
 Leakage to Earth ... 204,205,261 
 
 Lenz'sLaw 153,134,141,194 
 
 Lightning Flash 128,145 
 
 Lines of Force 4-9, 75, 148, 158, 222-224, 230, 236, 239, 252 
 
 Liquid for Cooling Armature ... ... ... ... ... ... 82 
 
 Lubrication on Commutator ... ... ... ... ... ... 90 
 
 Lugs of Commutators, Types of 117-122 
 
 Copper Strip 117-119 
 
 Cast on Segment 119-122 
 
 Bar Parallel with Axis 121, 122 
 
 Magnetic Circuit ... ... ... ... ... ... ... ... 78 
 
 - Flow in Circuit : Similarity to a Current 227-230, 236, 249, 256 
 
 - Flows in Opposition or Confluence in a Circuit 225-237, 245-258 
 
 Flow : Density of Lines of Magnetism 233, 234 
 
284 INDEX. 
 
 PAGE 
 
 Magnetic Flow : Distortion or Diversion of 230, 231, 249, 250, 255, 258 
 
 - Flow in Necks of Pole-Pieces ... 235, 246, 247, 250, 253-258 
 
 - Flow, Throttling of 250,253-258 
 
 Induction (see Induction). 
 
 Permeability ... 132 
 
 Magnetisation, Cross ... 12,71-75,216-218,232,247,248,252,253 
 
 Self, of Armature 
 
 12-14, 71-75, 158, 215, 217, 219, 227, 232, 248, 262 
 Magnetism, Free, Forming Poles : Limited Similarity to Static Change 
 
 of Electricity 229 
 
 Free, forming Consequent Poles ... ... ... ... 237 
 
 - Freej forming Waste Field . . . 225, 236, 237, 252, 253, 257 
 
 - Necessary to Induction Sparks and Effects ... 131, 134, 141, 142 
 
 Non- Elimination of ... ... ... ... ... ... 236 
 
 Magnets, Double-ended ,. 246,255-258 
 
 Manilla Paper 35,48 
 
 Maximum E.M.F., Plane of, in Armature 
 
 151, 156, 159, 160, 222, 223, 225, 241 
 
 Ditto, through Commutator 151, 156, 157, 160, 181, 222, 223, 225 
 
 Ditto, ditto, and Non-Sparking Plane with Weak Field .. 225 
 
 Mica 89,91,93-95,98-101,106,114 
 
 Momentum, Absence of, in Electricity ... ... ... .. ... 141 
 
 Mordey, W. M., " Victoria " Commutator 103-107 
 
 Motor, Armature Reactions with ... ... ... ... 73,224 
 
 Sparking with ... ... ... ... ... .. 225 
 
 - Generator, Elwell-Parker 89,122 
 
 Miiller, Experiments on Fibre ... ... ... ... ... ... 96 
 
 Neutral Conductor, Common to Two Circuits ... ... ... 167-180 
 
 or Non-Sparking Plane 151, 153, 155-157, 160, 166, 172, 252, 255 
 
 or Non -Sparking Plane, Permanent Deflection of 252, 254, 259 
 
 - or Non-Sparking Plane, Oscillation of 183, 203, 213, 259-267 
 Non Elimination of Magnetism ... ... ... . ... ... 236 
 
 Non-Sparking with Carbon Brushes 126, 195-202, 266, 267 
 
 -Points 151,153,155-157,160,166,172,252,255 
 
 Oblique Induction 257,258 
 
 Ohm's Law 171,192,196,240 
 
 Oil on Commutator, Cause of Sparking 90, 267 
 
 Oil for Cooling Armature ... ... ... ... ... ... ... 82 
 
 Opposition of Electric Currents in a Conductor ... .. 167-194 
 
 - of Magnetic Inductions 227-237,245-258 
 
 Oscillation of Planes, Cause of Sparking ... 183,203-215,259-267 
 Oxidation of Contacts 146 
 
 Paper Insulation, Manilla... ... ... ... ... .. 35,48 
 
 Willesden 35,48 
 
 Paraffin 95-99 
 
 Specific Resistance of ... ... ... ... ' ... ... 95 
 
 Parallel, Lamps in, Uniformity of E.M.F. for 211 
 
 Parsons' End Winding 80-83 
 
 Paterson and Cooper's Commutators 111,113 
 
 Permeability, Magnetic 132 
 
 Perry and Ayrton, Table of Specific Resistances 95 
 
 Phosphor Bronze for Commutators 87,90 
 
INDEX. 285 
 
 PAGE 
 
 Plane of Cessation of Force ... 151,152,157-160,166,193,204,221 
 
 Cessation of Force affected by Brushes 159 
 
 Commutation 8, 60, 72-75, 149-153, 158-162, 192-194 
 
 Commutation controlled by Brushes ... 152, 153, 192-194- 
 Maximum E.M.F. through Armature 
 
 151, 156, 159, 160, 203, 222, 223, 225, 241 
 
 Maximum E.M.F. through Commutator 
 
 151, 156, 157, 160, 181, 203, 225 
 
 Nori-sparking ... 151, 153, 155-157, 160, 166, 172, 252, 255 
 
 Requisite Reversal 151, 154, 166, 181, 203, 212, 213, 223, 252, 262 
 
 Planes, Deflection of 204,215,251,252,254,259-262 
 
 Orders of 151 
 
 Oscillation of 183,203-215,259-267 
 
 Undue Inclination of 215,255 
 
 Two, of Commutation .. ... ... ... 75 
 
 Polarisation of Armature, Self- 
 
 12-14, 71-75, 158, 215, 217-219, 227, 232, 248, 262 
 
 Controlled by Brushes 158,159,217,262,263 
 
 Components 217-219, 223, 227, 232-237 
 
 Resultant 217-219 
 
 in Proportion to Field 13, 157, 204-213, 215 
 
 - Oscillation of 262,263 
 
 Ultimate 12, 13, 74, 75, 158, 219, 251, 254, 262 
 
 of Drum as Compared with Gramme 219,220 
 
 Immobility of 14,15 
 
 and Torque 13-15,26,27 
 
 of Coils 15, 16, 26, 27, 162, 163 
 
 Pole-pieces, Flow of Magnetic Induction in 232-237, 245-258 
 
 Throttling of Induction in Necks of 247, 250-258 
 
 Poles of Bar Electromagnet ... 131,229 
 
 Consequent 237 
 
 Potential, Accumulation and Generation of (see Electromotive Force). 
 
 Difference between Armature-Winding Sections 1 1 , 19, 22, 241 
 
 Difference between Commutator Segments.. 9-12, 93, 148, 241 
 Primary Coils 132, 140' 
 
 Reactions, Armature, Resultant and Components ... ... 217-219 
 
 in Weak Field 215-225 
 
 Regarded as Components 217-219, 223, 227, 232-237 
 
 in Asymmetrical Field 245-255 
 
 with Double-ended Magnets 255-258 
 
 Back Induction 73, 75, 216-218, 223-225, 232-236, 247 
 
 withMotor 73,224,225 
 
 Cross Induction ... 12, 71-75, 216-218, 232-236, 247-258 
 
 Cross Induction with Double-ended Magnets ... ... 256 
 
 Forward Induction ... 217, 218, 223, 224, 232, 237-243, 247 
 
 Forward Induction, Utilisation of 224, 237-243 
 
 Oblique Induction 257, 258 
 
 Requisite Reversal, Plane of, 151, 154, 166, 181, 203, 212, 213, 223, 252, 262 
 
 Deflection of 252,254,259-262 
 
 Oscillation of 259-262 
 
 Resistance, Cause of Heat 21, 25, 143, 144 
 
 of Voltaic Arc ... 144r 
 
 Paths of Various, and Ohm's Law 171, 192, 196 240 
 
286 INDEX. 
 
 Resistance of Armature Winding, and Effects 
 
 20-23, 26, 28, 37, 41, 42, 54, 164, 261, 263, 265, 266 
 and the Short-circuit Currents 168-181, 196-201, 265, 266 
 
 Exterior to Generator Necessary for Outer Voltage 
 
 178-180, 205-213, 241 
 
 of Contact with Carbon Brushes 195-202 
 
 Magnetic 78,254 
 
 Resistances, Specific, Ayrton and Perry's Table of ... ... ... 95 
 
 Miiller's Experiments with Fibre ... 96-98 
 
 Resultant Currents 184 
 
 Polarity or Induction 2 17-21 9, 231 
 
 Saturation, Magnetic ... ... ... ... ... ... 248, 250 
 
 Sayers, W. B., Winding of 237-243 
 
 Secondary Coil and Current 132-135,140 
 
 Sections of Armature Winding 18,22,39,46,52,53,55,57,59 
 
 of Armature Winding Potential, Difference between 
 
 11, 19, 22, 241 
 
 or Segments of Commutators, Potential Difference between 
 
 9-12, 93, 148, 241 
 
 Metals for 90,117-119,267 
 
 Gauge for 269-273 
 
 Self-induction, Theory of 139-142,165 
 
 in Armature Winding 162, 164, 165, 188, 192-194, 200, 201, 210 
 
 of Field Magnet Windings 208-210 
 
 Sparking 
 
 126, 127, 130, 131, 137, 139-142, 145, 164, 165, 188, 190, 194 
 
 Shellac 35,48,91,95,96 
 
 Shock Due to E.M.F 143, 144 
 
 Short-Circuit (see Circuit). 
 
 Siemens, Wire- Wound Armature 2, 6, 17-23, 55, 261 
 
 Stranded Armature Conductor Bar ... ... ... ... 31 
 
 High Tension Experiments at Crystal Palace ... 136, 143 
 
 Silk Tape Insulation 35 
 
 Smelting with Arc 144 
 
 Sparking at Commutators ... ... ... ... ... 125-267 
 
 Electric. Elementary Theory of 125-146 
 
 Electric, Orders of 126 
 
 Electric, Statical Disruptive 126-128,145 
 
 Electric, Small Time for 145 
 
 Electric, Induction Disruptive ...126, 127, 131-137, 140, 143 
 Electric, Self-induction Disruptive 
 
 126, 127, 130, 131, 139-142, 145, 164, 165, 188, 190, 194, 208 
 
 Electric, Self -Induction Disruptive Explained ... 139-142 
 Electric, Voltaic Arc 126-131, 136, 140, 142-146, 164 
 
 Electric, Voltaic Arc Explained 130 
 
 - Electric, Continuous Disruptive Discharge ... 126, 127, 142-146 
 
 - Electric, Disruptive, Rupturing Effects of 128,143 
 
 Electric, Disruptive Discharge, Compared with Voltaic Arc 
 
 142, 143 
 
 Electric, Disruptive Discharges and Arcs, High Tension, at 
 
 Crystal Palace 136,143 
 
 - Electric, Disruptive, Effect on Electrodes 128 
 
 Electric, Disruptive Effect of .Lightning 128 
 
 Electric, Arc, Effect on Electrodes 130, 146 
 
INDEX. 287 
 
 PAGE 
 
 Sparking, Electric, Compound Nature of Flash at Switch, and Effects 
 
 at Brushes on Commutator : ... ... ... 147-267 
 
 Elementary Theory of.. 161-166 
 
 Likened to Flash at Switch 163-164 
 
 - Non-sparking Points 151,153,155-157,166,172,252 
 
 due to Misplacement of Brushes ... ... ... 183-194 
 
 due to Weakness of Field 215-225 
 
 due to Asymmetry of Field 245-255 
 
 due to Asymmetry of Armature Winding ... ... 259-263 
 
 due to Defective Joints in Armature Winding ... 261-267 
 
 due to Earths or Short Circuits in Outer Circuit ... 203-213 
 
 due to Failure of Insulation ... ... ... .. ... 261 
 
 due to Various Causes... ... ... ... ... ... 267 
 
 Non-, with Carbon Brushes 195-202,266,267 
 
 Non-, with Sayers' Winding 237-243 
 
 at One Brush only 248-255 
 
 Shunt 165,174,263 
 
 -Back 187,188,191,194,211,242,262,263 
 
 Forward 
 
 185, 186, 189, 190, 206-208, 211, 223, 224, 242, 262, 263, 266 
 
 Forward, Eliminated by Sayers' Winding 237-243 
 
 -at Motors 224,225 
 
 Specific Resistances, Ayrton and Perry's Table of ... ... ... 95 
 
 Static Charge of Electricity, Limited Similarity of Free Magnet- 
 ism to ... .. 229 
 
 Electricity, Disruptive Discharge of 126-1 28, 145 
 
 Stranded Cable for Armature Conductor Bar ... ... ... ... 31 
 
 Swinburne, James, Chord Winding of ... ... ... ... 67-75 
 
 Device for Obviating Foucault Currents ... ... 34,35 
 
 High Tension Experiments at Crystal Palace 136 
 
 Symmetry of Armature Winding ... ... ... ... 27,28 
 
 Tape, Copper, Armature Conductois Bars of ... ... ... 32,47 
 
 Insulation, Cotton or Silk 35,48,64 
 
 Temperature (see Heat). 
 
 Tension (see Electromotive Force). 
 
 Theory, Elementary, of Drum Winding... ... ... ... ... 1-16 
 
 of Electric Sparking 125-146 
 
 of Self-induction 139-142,165 
 
 of Spark at Brush 161-166 
 
 Thompson, Silvanus P., on the Arc 130,144 
 
 on Sparking at Commutators... ... ... 149 
 
 Three- wire System, and Short-circuted Coils under Brushes... ... 167 
 
 Throttling of Induction ... 247, 250-258 
 
 Time for Lightning Flash and Disruptive Sparks ... ... ... 145 
 
 Torque 1, 13-15 
 
 on Armature Winding 15, 26, 27 
 
 Transformer, Continuous Current, El well-Parker 89, 122 
 
 Trotter, A. P., Gauge for Commutator Segments 269-273 
 
 Uniform External Voltage by Compound Winding *...- . ... 211-213 
 Uppenborn's Electrical Calendar j^^&jE^fS 1 ^^ 95 
 
288 INDEX. 
 
 PAGE 
 
 Varnish as Insulation 31,35,47,48,91,96,99 
 
 Vaseline on Commutators ... ... ... ... ... ... 90 
 
 Ventilation 23,26,29,35-37,41,42,47,49,66-69,82,84,122 
 
 with Fans 36,37,49 
 
 Vibration as Cause of Sparking ... ... ... ... 267 
 
 " Victoria " Commutator, Mordey's 103-107 
 
 Volatisation in Arc 130,144,146 
 
 Voltage (see Electromotive Force). 
 
 External to Generator, Dependent on External Resistance 
 
 178-181, 205-213, 241 
 
 External, Uniform, by Compound Winding... ... 211-213 
 
 Voltaic Arc (see Arc, Voltaic). 
 
 Vulcanised Fibre 93-99,101,106 
 
 Vulcanite 98 
 
 Wake and Sanders, Mica Exhibit 100 
 
 Waste Field 225, 236, 237, 252, 253, 257 
 
 Weak Field 215-225, 232-237, 247, 248 
 
 " Westminster " Armature Winding .. 47-51 
 
 " Westminster " Commutators 107-110,111-115 
 
 White, Inventor of Helical End Winding . t 80 
 
 Wiggins and Co., Mica Exhibit 100 
 
 Willesden Paper, Insulation 35,48 
 
 Winding (see Armature, Armatures, Bars and End Windings). 
 
 Wire Drum Winding, Siemens 2,6,17-23,55,261 
 
 Siemens', Advantages ... ... .. ... ... ... 23 
 
 Siemens' Cable Conductors ... ... 31 
 
 Eickemeyer's 55-62 
 
 Wires, Binding, on Armature 16,26,48,61,82 
 
 Wood, Walnut, Paraffined, as Insulation 64, 96-98 
 
INDEX TO DIAGRAMS. 
 
 FIG. 
 
 Ampere's Law, Illustrated 96 
 
 Andersen, F. V., Armature, Winding of ... 23-28 
 
 ^ Commutators of 71,77 
 
 Armature Coil, Plane of, Cutting between Two Commutator Segments 103 
 
 Coil, Plane of, Cutting through One Commutator Segment 104 
 
 Coils, Armature and Field, Diagrammatically Represented 102 
 
 Planes, Positive Lead 105,108 
 
 Planes, Negative Lead 106 
 
 Self -Excited Polarity of, with Gramme Winding 134 
 
 Self -Excited Polarity of, with Drum Winding 133 
 
 Armatures, Drum (see also End Windings) : 
 
 Andersen (also La timer Clark, Muirhead and Co.) ... ... 23-28 
 
 Crompton and Kyle ... ... ... ... 55-58 
 
 Cronipton and Swinburne ... ... ... ... 17-22 
 
 Edison ...14-16 
 
 Eickemeyer 36-42 
 
 Fritsche ... ... ... ... ... ... 63 
 
 Hopkinson (also Mather and Platt) 29-32 
 
 Kapp 43-46 
 
 Parsons 59-62 
 
 Siemens ... ... ... ... ... ... ... ... 6, 7 
 
 Swinburne 47-52 
 
 " Westminster " (see Andersen) 
 
 Back-Induction 54, 129, 130, 141, 145 
 
 Back-Sparking 117,119,121 
 
 Brushes (on Commutator), Elementary Theory of ... ... ... 109 
 
 Elementary Theory of Sparking at ... ... 110 
 
 Elementary Theory of Short- Circuits and Short-Circuit 
 
 Currents through ... ... ... ... 114 
 
 Misplacement of, and Sparking due to 117-121 
 
 Sparking at One Brush Only 148-150 
 
 Carbon, Non-Sparking with 122-124 
 
 hord Winding, Swinburne's 47-52 
 
 Coil, Elementary, and Ampere's Law ... ... ... 96 
 
 Elementary, of Dynamo, and Lena's Law... ... ... 99,100 
 
 Induction, at " Make " 97 
 
 Induction, at " Break " 98 
 
 of Heavy Drum Bar Winding, indicating possible Faulty Joints 156 
 
 of Gramme or Drum Wire Winding, indicating possible Faulty 
 
 Joint ... : ... . ... 157 
 
 u 
 
290 INDEX TO DIAGRAMS. 
 
 FIG. 
 
 Coil of Armature Winding, Plane of, cutting between Two Com- 
 mutator Segments ... ... ... ... ... ... 103 
 
 of Armature Winding, Plane of, cutting through One 
 
 Commutator Segment ... . . ... ... ... ... 104 
 
 Commutation, Plane of (CnP) with Positive Lead 105, 108, 117, 119, 121 
 
 Plane of (CnP) with Positive Lead in Weak Field 129 
 
 - Plane of, with Negative Lead 106,117,118,120 
 
 - Plane of, Oscillation of (slight) 153-155 
 
 Two Planes of, with Swinburne's Chord Winding ... ... 52 
 
 Commutator Segments, Trotter's Gauge for Taper of 158, 159 
 
 Segments, Theory of ditto. 162 
 
 Lugs 79-95 
 
 Effect of Insulation on Diameter of ... ... .. 160, 161 
 
 Planes (M'M', N P) 107, 108, 111, 117-121 
 
 Armature Coils, and Field 102 
 
 Commutators : 
 
 GreneraTForm... ... .' 64,65 
 
 Crompton ... ... ... ... ... ... 66, 67 
 
 Gulcher ... ".'. ... 72 
 
 " Helvetia" (of Alioth and Co.) .. ... 69 
 
 Hopkinson (also Mather and Platt) ... c \. ... " 78,95 
 
 Kapp, Old Forms ... ... ... " 73,74 
 
 Kapp, New Form ... '.'.'. ... ... " ... 75 
 
 Kennedy "... .. ... ... ... ... ... ... 70 
 
 Mordey's " Victoria " ... ... 68 
 
 "Patersoii and Cooper ... ... ... ... ... ... 76 
 
 ''Westminster" (of Latimer Clark, Muirhead and Co. and 
 
 Andersen) . ... 71, 77 
 
 Types of Lugs 79-95 
 
 Components, Resolution of Armature Self -Polarisation into ... 130-132 
 Compound- Wound Dynamo and Faults in Outer Circuit '.. ... 128 
 
 .Conductors, Inactive Copper, of Armature ... ... ... ... 2-5 
 
 Formed to Obviate Foucault Currents 9-13 
 
 Connector for End Winding, Join tless ... ... ... ... 33-35 
 
 Crompton (R. E.) and Co., Commutators of ... ... "... 66,67 
 
 and Kyle, Armature End Winding of ... ... 55-58 
 
 1 and Swinburne, Armature End Winding of ... 17-22 
 
 Cross-Induction, or Magnetisation 53, 129-132, 142-145, 151 
 
 Eliminated (Nearly) 152 
 
 Counteraction of, Fischer-Hinhen's Method ... (page 258) 
 
 Currents, Augmentation of ... ... '.*, . "... ... ... 112 
 
 Assimilation of ... .. ... v .;. ... ... ... 113 
 
 Assimilated, Potentials of ... 115,116 
 
 Short-Circuit, under Brush, Theory of -... ... ... 114 
 
 Short- Circuit, with Misplaced Brushes ... . 118-121 
 Short-Circuit, with Sayers' Winding ... ... ... 146 
 
 Short-Circuit, Counteraction of, at Non-Sparking Point 111, 146 
 
 Short- Circuit, Elimination of, with Carbon Brushes ' 122-124 
 
 Defective Joints in Heavy Drum Winding 156 
 
 Joint in Drum or Gramme Wire Winding ... ... ... 157 
 
 Deflection of Planes with Gramme Winding (see Planes) ... 148-150 
 Double Volute Join tless End Connector for Drum Winding ... ...33-35 
 
 : Volute End Windings (see End Windings) 
 
 Drum Armature, Self -Excited Polarity of ... ... ... ... 133 
 
 Winding, Theory of 1 
 
INDEX TO DIAGRAMS. 291 
 
 FIG. 
 
 Dynamo, Separately Excited and Faults in Outer Circuit ;.. ... 125 
 Shunt Wound, and Faults in Outer Circuit... ... ... 126 
 
 Series Wound, and Faults in Outer Circuit .. ... ... 127 
 
 Compound Wound, and Faults in Outer Circuit ... ... 128 
 
 with Double-Ended Magnets ... 147 
 
 - with Double -Ended Magnets, Fields of 151,152 
 
 . Elementary Coil of, Illustrating Lenz's Law .. 99, 100 
 
 Compactness of, in Reference to Armature End Winding .. 57 
 
 " Earths," as Causes of Sparking " ... 125-128 
 
 Edison's Circular Bar End Winding ... ... ... ... ... 14 
 
 Disc End Winding , 16-16 
 
 Eickemeyer's Wire Drum Winding ... ' ... ... e ".. ...36-42 
 
 End Windings for Drum Armatures : 
 
 Wire Wrapping : Siemens ... ... ... ... ... ... 6,7 
 
 Radial and Circular Connectors : Edison ... ... ... ' ... 14 
 
 Discs: Edison 15,16 
 
 Chord Winding : Swinburne ... ... .. ..47-52 
 
 Helical Connectors, Interior : Kapp ... ... ... ...43-46 
 
 Connectors, Exterior : Parsons ... ... ... ...59-62 
 
 Connectors, Exterior : Fritsche ... ... ... ... 63 
 
 Volute, Single Interior, and Crank : Cromptou and Swinburne ...17-22 
 
 Single Interior and Crank : Andersen 23-28 
 
 Volutes, Double, Interior : Al teneck ; Hopkinson 29-32 
 
 Double, Interior : Eickemeyer ... 36-42 
 
 Double, Interior : Jointless Connector ... ... ...33-35 
 
 Double Exterior : Crompton and Kyle 55-58 
 
 Faults in Armature Winding as Causes of Sparking ... ... 153-157 
 
 in Outer Circuit as Causes of Sparking 125-128 
 
 Field, Armature Coils, and Commutator ... ... ... ... 102 
 
 Weak, and Armature Reactions in 129,144,145 
 
 Asymmetrical, and Armature Reactions in ... ' ... 148-150" 
 
 with Double-Ended Magnets, and Armature Reactions in 151, 152 
 
 Distortion of ..: 105, 1C6, 108, 117, 129, 142-145 
 
 Oscillating or Deflecting, Distortion of 154, 155 
 
 Permanently Deflected, Distortion of 148-150 
 
 Forward, or Positive, Lead to Brushes : Sparking at ... 117, 119, 12f 
 
 or Positive, Lead to Brushes : Planes in relation thereto 
 
 (see Planes) 105,107,108 
 
 Induction ... ' 132, 140, 144 
 
 Sparking 117, 118, UO 
 
 Sparking Eliminated with Sayers' Winding 146 
 
 Sparking at one Brush only ... ... ... ... 148-150 
 
 Foucault Currents, Prevention of : 
 
 Crompton's Method 9-12 
 
 Hopkinson's Method ... ... . . .' ... ... ... ... 8 
 
 Swinburne's Method .. ..'." ... ... .. ... ... 13 
 
 Fritsche's Armature Winding ... ... ... ... ... ... 63 
 
 Gramme Winding ... ... ... ... ... ... Introductory 
 
 Armature, Self-Excited Polarity of 134 
 
 Deflection of Planes with, in Asymmetrical Field ... 148-150 
 
 Heating of Armature Conductor Bars, Methods for Preventing : 
 
 Crompton's ... ... ... ... ... ... ... ... 9-12 
 
 Hopkinson's ... ... ... ... ... ... .. ... 8 
 
 Swinburne's ... ... ... ... ... ... ... ... 13 
 
 " Helvetia" Commutator of Messrs. Alioth and Co. 69 
 
INDEX TO DIAGRAMS. 
 
 FIG. 
 
 Hopkinson, Drs. J. and E., Armature End Winding of ... ...29-32 
 
 Method for Prevention of Heating ... ... ... ... 8 
 
 Commutator of . ... ... ... ... ... ... 78 
 
 Commutator, Connection of, to Winding ... ... ... 95 
 
 Induction Coil at " Make " 97 
 
 -Coil at "Break" 98 
 
 Self-, Diagram Illustrating ... 101 
 
 Back- " 54,129,130,141,145 
 
 -Forward- 132,140,144 
 
 Forward-, Utilised with Say ers' Winding ... .. ... 146 
 
 -Cross-, 53,129,130-132,142-145 
 
 Cross-, Throttling of 148,149 
 
 Cross- and Main-, Throttling of 150 
 
 Cross-, with Double-Ended Magnets ... ... ... 151 
 
 Cross- Eliminated (nearly) with Double-Ended Magnets ... 152 
 
 Cross-, Counteraction of, by Fischer-Hinnen's Method (page 258) 
 
 Oblique-, with Double-Ended Magnets ... ... ... 152 
 
 Insulation, Special Adaptations of, in Commutators ... ... ...66-77 
 
 Effect of, on Diameter of Commutator ... ... 160, 161 
 
 Joints, Defective, in Armature Coils, as Causes of Sparking . . . 156, 157 
 Jointless Double Volute End Connector for Drum Winding ... ...33-35 
 
 Kapp, Gisbert, Drum Winding of 43-46 
 
 Commutator of : Old Forms... ... ... ... ...73,74 
 
 Commutator of : New Form ... ... ... ... ... 75 
 
 Lead (to Brushes), Positive, and Armature Planes ... ... 105, 108 
 
 Negative, and Armature Planes ... ... ... ... 106 
 
 Positive, and Commutator Planes ... ... ... 107,108 
 
 Positive Brushes in Line with Field ; and all Planes . . . 108 
 
 Positive, Correct, with no Sparking 107, 108, 111 
 
 Misplaced Brushes, and Sparking ... ... ... 117-121 
 
 Positive, Excess of, and Back-Sparking ... 117, 119, 121 
 
 Positive, and Forward-Sparking in Weak Field .. ... 129 
 
 Negative, and Forward-Sparking 117, 118, 120 
 
 Negative, and Non-Sparking with Sayers' Winding ... 146 
 
 Unequal, Sparking at One Brush Only ... ... 148-150 
 
 Positive, and Back-Induction 129, 130, 145 
 
 Negative, and Forward-Induction ... ... ... 132,144 
 
 Minimised by Throttling Cross-Induction, with Double- 
 
 Ended Magnets 152 
 
 Minimised or Unnecessary, with Carbon Brushes . . . 122-124 
 
 Lenz's Law, and Elementary Dynamo Coil ... ... ... 99-100 
 
 Lug's, Commutator, Types of ... ... v ... ... ... 79-95 
 
 Magnetic Flow (see also Induction) ... ... ... 135-143 
 
 Flow, Distortion of 142,143 
 
 Magnetisation (see also Induction). 
 
 Self-, of Drum Armature 1.133 
 
 Self-, of Gramme Armature ... ... ... Introductory, 134 
 
 Self-, ef Armature, Asymmetrical 153-155 
 
 Self-, of Armature as Resultant and Components ... 130-132 
 
 with Induction Coils 96-98,101 
 
 Magnets, Double-Ended, for Dynamo Field ... ... .. ... 147 
 
 Double-Ended, and Cross-Induction ... 151 
 
 Double-Ended, and Oblique-Induction ... ... ... 152 
 
 Main-Induction, Throttling of 150 
 
 Misplacement of Brushes (see Lead) Sparking Effects ... ... 117-121 
 
INDEX TO DIAGRAMS. 
 
 FIG. 
 
 Mordey, W. M., " Victoria " Commutator of 68 
 
 Negative Lead and Armature Planes ... 106 
 
 Lead and Forward Sparking 117,118,120 
 
 Lead and Non-Sparking with Sayers' Winding ... ... 146 
 
 Lead and Forward-Induction 132,144 
 
 Neutral, or Non-Sparking Plane (N P), or Points 107, 108, 111, 117-121 
 
 or Non-Sparking Plane, Undue Inclination of (with RE) 129 
 
 or Non-Sparking Plane, Oscillation of (with R R) ... 153-155 
 
 or Non- Sparking Plane, Permanent Deflection of 
 
 (withRR) 148-150 
 
 or Non-Sparking Plane with Negative Lead, and Savers' 
 
 Winding 146- 
 
 Non-Sparking with Carbon Brushes ... ... ... ... 122-124- 
 
 Oblique Induction ... ... ... . ... ... 152 
 
 Ohm's Law, Operation of, Illustrated 112-316, 118-124, 146 
 
 Oscillation of Planes, as Cause of Sparking ... -_; 153-155 
 
 Paterson and Cooper, Commutator of . ... ... 76 
 
 Parsons, Hon. C. A., Drum Winding of 59-62 
 
 Plane of Armature Coil Cutting Between Two Commutator Segments 103 
 
 of Armature Coil Cutting Through One Commutator Segment 104 
 
 Plane of Cessation of Force (C F) with Positive Lead... 105, 108, 129 
 
 of Cessation of Force (C F) with Negative Lead 106 
 
 Plane of Commutation (Cn P), with Positive Lead 105, 108, 117, 119, 121 
 
 with Positive Lead in Weak Field 129 
 
 with Negative Lead ... ... 106,117,118,120 
 
 Double, with Swinburne's Chord Winding ... ... .. 52 
 
 Oscillation of (Slight) 153-155 
 
 Plane of Maximum E.M.F. through Armature (MM) 105, 106, 108, 129, 150 
 
 through Commutator (M' M') 107, 108, 117 
 
 Plane of Non-Sparking, or Neutral Plane (N P) 107, 108, 111, 117-121 
 
 Oscillation of (with R R) 153-155 
 
 Permanent Deflection of (with R R) 148-150 
 
 Undue Inclination of (with R R) 129 
 
 Plane of Requisite Reversal (R R) 105, 106, 108, 111- 
 
 Oscillation of 153-155- 
 
 Permanent Deflection of 148-150 
 
 Undue Inclination of ... ... ... ... ... ... 129 
 
 Planes, Critical : 
 
 Armature Planes ' 105,106,108 
 
 Commutator Planes 107,108 
 
 Oscillation of 153-155 
 
 Permanent Deflection of 148-150 
 
 Undue Inclination of ... ... ... ... ... ... 129 
 
 Polarity, Self -Excited, of Drum Armature ... 133 
 
 Self-Excited, of Gramme Armature 134 
 
 Potentials of Assimilated Currents .'. ... ... ... 115,116 
 
 Reactions, Armature : 
 
 Elementary Illustrations ... ... ... ... ... ...53, 54 
 
 Self-Excited Polarity of Drum Armature 133 
 
 Self -Excited Polarity of Gramme Armature 134 
 
 As Resultant and Components 130-132 
 
 In Weak Field, Shown Graphically 129 
 
 In Weak Field, Shown Analytically 144,145 
 
 In Asymmetrical Field . 148-150 
 
5294 INDEX TO DIAGRAMS. 
 
 FIG. 
 Reactions, Armature : 
 
 With Asymmetrical Winding 153-155 
 
 With Double-Ended Magnets ... ... 151,152 
 
 Resultant Induction of Armature ... ... ... ... 130-132 
 
 Sayers, W. B., Armature Winding of ... 146 
 
 Self -Excitation, Polarisation, of Armature (see Magnetisation, 
 
 Armature Reactions, and Induction) ... ... ... 133,134 
 
 Self-induction, Diagram Illustrating ... ... ... ... .. 101 
 
 Short- Circuits in Outer Circuit, as Causes of Sparking ... 125-128 
 Siemens' Wire Drum Winding ... ... ... ... ... ... 6,7 
 
 Sparking, Induction, at " Make " ... ... ... ... ... 97 
 
 Induction, at " Break " ... ...- 98 
 
 Self-Induction 101 
 
 at Brush on Commutator, Elementary Illustration of ... 110 
 
 with Brushes Misplaced 117-121 
 
 Forward-, at Positive Brush 117,118 
 
 Back-, at Positive Brush 117,119 
 
 Forward-, at Negative Brush 117,120 
 
 Back-, at Negative Brush 117,121 
 
 due to Faults in Outer Circuit 125-128 
 
 due to Weak Field 129 
 
 due to Asymmetrical Fields ... ... ... ... 148-150 
 
 at One Brush only, due to Throttled Cross-Induction 148-149 
 
 at One Brush only, due to Throttled Cross- and Main- 
 Inductions ... ... ... ... ... ... ... 150 
 
 due to Asymmetry of Armature Self -Polarisation ... 153-155 
 
 due to Defective Joints in Armature Winding . . 156, 157 
 
 Non-, Plane (see Neutral Plane) ... 107, 108, 111, 117-121 
 
 Non-, Position of Brush ... ... ... ... ... Ill 
 
 Non-, with Carbon Brushes 122-124 
 
 NOB-, with Sayers' Winding with Negative Lead ... ... 146 
 
 Swinburne, James, Chord Winding of 47-52 
 
 Throttling of Cross-Induction, and Sparking 148-149 
 
 of Cross and Main-Inductions, and Sparking ... ... 150 
 
 of Cross-Induction with Double-ended Magnets ... ... 152 
 
 Volute, Double, Jointless Cross Connector for Drum End Winding 33-35 
 
 " Westminster " Armature Winding 23-28 
 
 Commutators ... ... ... ... .. ... ...71,77 
 
 Wire Winding, Siemens's... ..: ..: ... ... ... ... 6,7 
 
 Eickemeyer's : 36-42 
 
LONDON : 
 
 PRINTED BY GEORGE TUCKER, 1, 2, AND 3, SALISBURY COURT, 
 FLEET STREET, B.C. 
 
far (Bectiitimts, (Electrical (Engineer*, arid 
 (Electrical 
 
 December, 1894. 
 
 THE Publisher of "THE ELECTRICIAN" ha? compiled the following List of Books 
 on Electrical and Allied Subjects, all of which can be obtained direct from 
 "THE ELECTRICIAN" PRINTING AND PUBLISHING COMPANY, LIMITED, Salisbury Court, 
 Fleet Street, London. 
 
 Any of the Works mentioned in this List will be forwarded to any address in the 
 United Kingdom, on receipt of the quoted price. Five per cent, must be added for 
 postage on foreign orders. 
 
 Copies of any known Work on the Theory or Practice of Electricity, or relating to 
 any particular application of Electricity, or of an Electrical or Engineering character 
 (either British or Foreign), can be ordered through this Company. Inquiries respecting 
 the Latest and Best Books on Electrical Subjects will be answered by post, Free of Charge, 
 and all possible assistance rendered to inquirers. 
 
 Daily, Weekly, or Monthly parcels of Miscellaneous Books and Publications, for 
 London, the Provinces, or Abroad, made up on the best possible terms. 
 
 Cheques and Postal Orders to be made payable to "The Electrician Printing 
 and Publishing Company, Limited," and to be crossed " Coutts and Co." 
 
 COTVTEIVTS. 
 
 Electrician " Series, 
 
 PAGE. I PAGE. 
 
 " The Electrician " Electrical Trades' The Incandescent Lamp and its Manufacture 10 
 
 Directory and Handbook ............ 3 Drum Armatures and Commutators ...... 10 
 
 Electromagnetic Theory ............ 4 Electro-Chemistry ............... 11 
 
 Magnetic Induction in Iron and other The Art of Electrolytic Separation of Metals 11 
 
 Electrical Laboratory Notes and Forms ... 12 
 
 Electnc Lamps and Electric Lighting ...... 5 practjca| Notes f of E|eotpica| 8tu(Jents ^ 
 
 l 6, 7 is 
 
 L . ""* 8 The'Ma^acture'of Electric Light Carbons 13 
 
 ' 
 
 A Digest of the Uw of Electric Lighting, Electric Motive Power ............... 14 
 
 &c., &c ...................... 8 Submarine Cable Laying and Repairing ... 14 
 
 Electrical Engineering Formulae ......... 9 Motive Power and Gearing ............ 14 
 
 The Work of Hertz ............... 9 Fine Art Reproductions, &c ......... 15 
 
 Electrical an5 (general list 
 
 "The Electrician" ... ......... ..................... 32 
 
 The "Engineering Magazine" ........................... 2 
 
 Newspapers and Periodicals ...... . .................... 2 
 
 Electricity and Magnetism ........................... 16 
 
 Electric Lighting and Transmission of Power .................. 20 
 
 Telegraphy and Telephony ........................... 23 
 
 Electro-Chemistry and Electro-Metallurgy ............ - ... ...... 24 
 
 Electrical Instruments, Bells, &c ......................... 25 
 
 Reference Books, &c ............................... 25 
 
 Miscellaneous ................................... 27 
 
 "THE ELECTRICIAN" PRINTING & PUBLISHING CO,, LIMITED. 
 
 5527 I, 2, and 3, Salisbury Court, Fleet Street, London, E.G. 
 
2 "The Electrician" Printing and Publishing Co., Ltd., 
 
 PAPERS AND PERIODICALS. 
 
 "The Electrician" Printing and Publishing Company (Limited) forward, post free, 
 on receipt of the annual subscription, any of the following Papers and Periodicals : 
 
 A1VU3FMCAJST. 
 
 AMERICAN JOURNAL OF SCIENCE AND ARTS. (Monthly.) Ann sub., 26s. 6d. 
 
 ELECTRIC POWER. (Monthly.) Ann. sub., 12s. 6d. 
 
 ELECTRICAL AGE. (Weekly.) Ann. sub., 1. Is. 
 
 'ELECTRICAL ENGINEER. (Weekly.) Ann. sub., 1. Is. ; single copies, 8d.; post free, 10d. 
 
 ELECTRICAL ENGINEERING (Monthly.) Ann. sub., 16s. 9d. 
 
 ELECTRICAL INDUSTRIES. (Monthly.) Ann. sub., 19s. 
 
 ELECTRICAL REVIEW. (Weekly.) Ann. sub., 1. Is. 
 
 'ELECTRICAL WORLD. (Weekly.) Ann. sub., 1. 5s. ; single copies, 8d. ; post free, lOd. 
 
 ELECTRICITY. (Weekly.) Ann. sub., 19s. 
 ^ENGINEERING MAGAZINE. 
 
 (Monthly.) Ann. sub., 12s. 6d.; single copies Is.; post free Is. 3d. 
 PHYSICAL REVIEW. (Bi-monthly.) Ann. sub., 13s. 
 PRACTICAL ELECTRICITY. (Semi-monthly.) Ann. sub.. 9s. 6d. 
 RAILWAY REVIEW. (Weekly.) Ann. Sub., 14s. 6d. 
 SCIENTIFIC AMERICAN. (Weekly.) Ann. sub., 16s. 9d. 
 SCIENTIFIC AMERICAN SUPPLEMENT. (Weekly.) Ann. sub., 1. 5s. 
 STREET RAILWAY GAZETTE. (Weekly.) Ann. sub., 1 Is. 
 STREET RAILWAY JOURNAL. (Monthly.) Ann. sub., 1. 5s. 
 STREET RAILWAY REVIEW. (Monthly.) Ann. sub., 1. 
 WESTERN ELECTRICIAN. (Weekly.) Ann. sub., 1. Is. 
 
 BULLETIN DE L'ELECTRICITE ET.DE LA TELEGRAPHIE. (Weekly.) Ann. sub., 12s. 
 
 BULLETIN INTERNATIONAL DE L'ELECTRICITE. (Weekly.) Ann. sub., 8s. 
 
 L'ECLAIKAGE ELECTRIQUE. (Weekly.) Ann. sub., 2. 8s. 
 
 L'ELECTRICIEN. (Weekly.) Ann. sub., 1. 
 
 DIE ELEKTRICITAT. (Semimonthly.) Ann. sub., 9s. 
 
 DER ELEKTROTECHNIKER. (Semi-monthly.) Ann sub., 12s. 
 
 ELEKTROTECHNISCHE ZEITSCHRIFT. (Weekly.) Ann. sub., 1. 5s.; single 
 
 copies, 6d. ; post free, 8d. 
 
 ELEKTROTECHNISCHER ANZEIGER. (Semi- weekly.) Ann. sub., 10s. 
 ELEKTROTECHNISCHER RUNDSCHAU. (Bi-weekly.) Ann. sub., 12s. 
 ELEKTROTECHNISCHES ECHO. (Weekly.) Ann. sub., 16s. 
 L'ELETTRICISTA. (Monthly.) Ann. sub., 12s. 
 L'ELETTRICITA. (Weekly.) Ann. sub., 12s. 
 L'ENERGIE ELECTRIQUE. (Bi-weekly.) Ann. sub., 17s. 6d. 
 L'INDUSTRIE. (Weekly). Ann, sub., 1. 
 
 L'INDUSTRIE, ELECTRIQUE. (Semi-monthly.) Ann. sub., 1. Is. 
 JOURNAL TELEGRAPHIQUE. (Monthly.) Ann. sub., 5s. 
 REVUE PRATIQUE DE L'ELECTRICITE. (Bi-weekly.) Ann. sub., 12s. 
 LA SEMAINE ELECTRIQUE. (Weekly.) Ann. sub., 12s. 
 
 Single Copies of these Journals may "be seen, and those marked * are on sale, at 1> 2 and 3, Salisbury 
 _ Court, but specimens cannot be supplied. __ 
 
 Annual Subscription, 12s. 6d. post free (U.K.); European Continent, <<?., 15s. 
 (Specimen Back Number sent on receipt of 6d. to defray postage. ) 
 
 "THE ENGINEERING MAGAZINE," 
 
 An American Monthly Periodical having an extensive sale in the United States, and containing popularly-written 
 critical articles on Electrical, Engineering and Social Subjects of European interest by leading writers. 
 The Illustrations in this Publication are of the very choicest description. 
 EUROPEAN AGENT: 
 
 CEO. TUCKER, Publisher, "THE ELECTRICIAN" PRINTING & PUBLISHING CO,, Limited. 
 
 "The Engineering Magazine" can be obtained of all Booksellers, and at the Railway Bookstalls throughout 
 
 the Kinqdom. 
 
 Volumes commence in April and October, bound half calf, 10s. 6d. 
 THE ENGINEERING MAGAZINE "TECHNICAL INDEX." 
 
 The Technical Index given in THE ENGINEERING MAGAZINE each month of Articles published in current 
 issues of the technical journals of Great Britain and the United States is becoming more and more appreciated 
 by Engineers and Electricians all over the world. The Index deals with the whole of the best technical 
 literature of these countries ; and the various articles being classified under convenient general headings and 
 consecutively numbered, Subscribers to the Magazine are enabled to easily ascertain anything that has been 
 recently published dealing with subjects of interest to them, as well as the names of the authors. 
 
 Copies of any of the articles enumerated in the Index are suppli 
 of the Magazine, at 15 cents =8d., 30 cents = Is. 4d., or 45 cents = 2s. 
 
 1, 2, and 3, Salisbury Court Fleet Street, London, E.G. 
 
 
"The Electrician" Printing and Publishing Co., Ltd.,"j r) 
 
 PUBLISHED IN JANUARY OF EACH YEAE. ROYAL Suo. 
 Subscription Price, 5s. ; post free 5s. 9d., abroad 6s. 4d. After date of Publication, 7s. 6d. ; 
 
 post free 8s. 3d., abroad 8s. lOd. 
 - THE 
 
 ELECTRICAL TRADES' DIRECTORY 
 
 AND HANDBOOK. 
 
 ONLY RELIABLE 
 
 ELECTRICAL AND KINDRED TRADES. 
 
 (THE BIGr BLUE BOOK.) 
 
 This Directory and Handbook is well known throughout the Electrical, Engi- 
 neering, and Allied Trades in all parts of the World, and no expense is spared to 
 make the work really reliable. New names and corrections to existing entries in 
 the DIRECTORIAL portion of the book are received up to the middle of January, no 
 charge being made for inserting names, addresses, and occupations, and extra inser- 
 tions are made in the Classified Trades' Section at very low rates. In addition to 
 about 600 pages of purely Directorial matter, the HANDBOOK portion of the book 
 contains a mass of interesting information, a great part of which cannot be obtained 
 elsewhere, and in the Biographical Division appear more than 250 sketches of" 
 well-known men in the Electrical World, with many Portraits. 
 
 Following are some of the Permanent Features of this indispensable 
 Reference Book, which has been aptly designated " the combined Kelly and 
 Whittaker of the Electrical Trades " : 
 
 A DIRECTORY OF THE PROFESSIONS AND TRADES CONNECTED WITH ELECTRICITY AND ITS APPLICATIONS : 
 BRITISH DIVISION. \ 
 
 AMERICAN ^wSlSti. | A1 Phabetically Arranged and Classified into Trades and Professions. 
 
 CONTINENTAL DIVISION. 
 
 Summary of Electrical Events for the preceding year. Obituary Notices for the preceding year. 
 PATEN1S, DESIGNS, AND TRADE MARK3.-(Speciatty written.) 
 
 DIGEST OF THE LAW OF ELECTRIC LIGHTING and of ELECTJKIC TRACTION. (Special.) 
 Board of Trade Regulations Concerning the Supply of Electricity. 
 
 London County Council Regulations as to Theatre Lighting. [( Special.) 
 
 Rules and Regulations for the Prevention of Fire Risks arising from Electric Lighting : British aiii Foreign- 
 Lloyd's Suggestions for the Use of Electric Light on Board Vessels. 
 
 Installation Regulations of the London and Provincial Electric Supply Companies. (Special.) 
 Electric Lighting and Electric Traction Notices for the year. 
 
 Provisional Orders and Licenses granted by the Board of Trade in the preceding year. 
 New Companies Registered and Companies Wound Up in the preceding year. 
 
 CENTRAL LIGHTING STATIONS OF THE UNITED KINGDOM. -(Spedxllf/'Coinpiled Folding Sheei.\. 
 Electric Railways and Tramways in the United Kingdom. 
 
 Colonial and Foreign Import Duties on Electrical Machinery and Apparatus. (Specially prepared). 
 Wire Gauge Tables. Resistance and Weight Tables. Comparative Price of Electricity and Gas. 
 Notes on Illumination. Table and Curve of Hours of Lighting. Post Office Regulations. 
 TELEGRAPHIC COMMUNICATION between Great Britain. Colonies, and foreign Countries. 
 TELEGRAPH TARIFFS from the United Kingdom to all parts of the World. 
 The Submarine Cables of the World. The Land Lines of the World. The World's Cable Fleet. 
 Telephone Companies' Statistics. British Government Departments, and Chief Officials. 
 FOREIGN GOVERNMENTS, AND CHIEF OFFICIALS OF ELECTRICAL DlPARTMEXT3.-(pefe&) 
 Local Authorities and Chief Officers in London and Provinces.- (Special.) 
 
 Post ( >ffice Telegraphs : Chief Officials. Institution of Electrical Engineers : Officers, Rules, &c. 
 Northern Society of Electrical Engineers : Officials, Arc. [Officials, Ac, 
 
 Socie'te Internationale des Electriciens, Elektrotechnischer Vereio, American Institute of Electrical Engineeis : 
 London Chamber of Commerce, Electrical and Allied Trades Section : Objects, &c. 
 List of Universities, Colleges, Schools, &c. (with Professors). 
 
 ELECTRICAL LIMITED LIABILITY COMPANIES, FINANCIAL PARTICULARS CONCERNING. (Special.) 
 Principal British Electrical Patents expiring in the near future. [fession. with many Portrait" 
 
 The BIOGRAPHICAL SECTION : More than 250 sketches of the careers of leading men in the Electrical pro- 
 lt will be seen af once from the above Abstract that no similar Publication so efecfirely caters for the 
 
 L'U'c t r ica I Profess w / z . 
 
 A Digest of the Contents and other particulars of the last issue will be sent free on application! 
 
 Full particulars of the Tariff for Advertisements, for Entries under Trade Head- 
 ings in the Classified Trades List, &c., sent free in English, French or German. 
 
 1, 2, and 3, Salisbury Court, Fleet Street, London, E.G. 
 
4 "The Electrician" Printing and Publishing Co., Ltd., 
 
 ' 
 
 ELECTRICIAN SERIES. 
 
 466 pages, price 12s. 6d., post free, 13s. 
 
 ELECTROMAGNETIC THEORY. 
 
 VOL. I. 
 
 BY OLIVER HEAVISIDE. 
 
 THE 
 
 This work was originally meant to be a continuation of the series " Electromagnetic Induction 
 and its Propagation," published in The Electrician in 1885-6-7, but left unfinished. Owing, how- 
 ever, to the necessity of much introductory repetition, this plan was at once found to be impractic- 
 able, and was, by request, greatly modified. The result is something approaching a connected 
 treatise on electrical theory, though without the strict formality usually associated with a treatise. 
 The following are some of the leading points in this volume. The first chapter is introductory. The 
 second consists of an outline scheme of the fundamentals of electromagnetic theory from the Faraday- 
 Maxwell point of view, with some small modifications and extensions upoii Maxwell's equations. 
 The third chapter is devoted to vector algebra and analysis, in the form used by me in my former 
 papers. The fourth chapter is devoted to the theory of plane electromagnetic waves, and, being mainly 
 descriptive, may perhaps be read with profit by many who are unable to tackle the mathematical 
 theory comprehensively. It may be also useful to have results of mathematical reasoning expanded 
 into ordinary language for the benefit of mathematicians themselves, who are sometimes too apt to 
 work out results without a sufficient statement of their meaning and effect. But it is only intro- 
 ductory to plane waves. I have included in the present volume the application of the theory (in 
 duplex form) to straight wires, and also an account of the effects of self-induction and leakage, 
 which are of some significance in present practice as well as in possible future developments. There 
 have been some very queer views prorriulgated officially in this country concerning the speed of the 
 current, the impotence of self-induction, and other material points concerned. 
 
 Vol. II. is in preparation. 
 
 SECOND ISSUE. 370 pages, 150 illustrations. Price 10s. 6d-, post free. 
 
 MAGNETIC INDUCTION IN IRON AND OTHER METALS. 
 
 BY J. A. EWING, M.A., B.Sc., 
 
 Professor of Mechanism and Applied Mechanics in the University of Cambridge. 
 
 SYNOPSIS OF CONTENTS. 
 
 After an introductory chapter, which attempts to explain the fundamental ideas and the 
 terminology, an account is given of the methods which are usually employed -to measure the 
 magnetic quality of metals. Examples are then quoted, showing the results of such measurements 
 for various specimens of iron, steel, nickel and cobalt. A chapter on Magnetic Hysteresis follow?, 
 and then the distinctive features of induction by very weak and by very strong magnetic forces 
 are separately described, with further description of experimental methods, and with additional 
 numerical results. The influence of Temperature and the influence of Stress are next discussed. 
 The conception of the Magnetic Circuit is then explained, and some account is given of experi- 
 ments which are best elucidated by making use of this essentially modern method of treatment. 
 The book concludes with a chapter on the Molecular Theory of Magnetic Induction, and the 
 opportunity is taken to refer to a number of miscellaneous experimental facts on which the 
 molecular theory has an evident bearing. 
 
 "Prof. Ewing is such a great and well-known authority on Magnetic Induction that it is only necessary to 
 call attention to this second issue of his valuable work to insure its being in the hands of all who require 
 information upon this important subject." Philosophical Magazine. 
 
 " Besides being practically the only book dealing with the subject of magnetic induction, this work considers 
 
 the whole subject very exhaustively, both from theoretical and practical points of view The large number 
 
 of graphic representations illustrating the various phenomena of magnetic induction are of great value. . . 
 We can recommend the book as being of considerable importance." Electrical Review. 
 
 " In this book Prof. Ewing has placed before his readers an accurate record of the more recent and important 
 work done in magnetism. The book as a whole presents the best resume of this subject extant." 
 
 Electrical Engineer (New York). 
 
 "This is one of the most important books that has appeared, and coming from an author of such an 
 established reputation it will be looked upon as an undoubted authority upon the important subject of which 
 if treats. No man is better fitted for the task of discussing these questions than Prof. Ewing." 
 
 Electrical World (New York). 
 
 A German Edition is also published, price 8s. 6d. 
 
 I, 2, and 3, Salisbury Court, Fleet Street, London, E.G. 
 
"The Electrician" Printing and Publishing Co., Ltd., 
 "THE ELECTRICIAN" SERIES continue,!. 
 
 NOW Ready. Very fully illustrated, handsomely bound, on good paper, price 7s. 6d. 
 
 ELECTRIC LAMPS AND ELECTRIC LIGHTING. 
 
 Being a Course of Four Lectures delivered at the Royal Institution, April-May.. 1894, 
 
 By PROF. J. A. FLEMING, M.A., D.So., F.R.S., M.R.I., 
 
 Professor of Electrical Engineering in University College, London. 
 
 SYNOPSIS OF CONTENTS. 
 
 I. A Retrospect of Twelve Years Factors in the Development of Electric Illumination The Historical 
 Starting Point Davy's Researches on the Electric Arc The Evolution of Incandescent Electric Lighting- 
 Definition of Fundamental Terms Units of Measurement of Current, Pressure, Work, and f'ower Board of 
 Trade Units Conditions of Public Supply under Acts of Parliament The Heai'ng Effect of an Electric Current 
 Joule's Law Experimental Proofs -Radiation from Incandescent Bodies Temperature of Radient Bodies 
 Surface Efficiency and Specific Radient Qualities of Various Materials Peculiar Properties of Carbon Bright- 
 ness of Various Lights Methods of Photometric Comparison Standards of Light Sun, Moon, Electric Arc, 
 Incandescent Lamp as Illuminants Quality and Intensity of Light Luminosity and Candle-power The 
 Physiological Question of Vision. 
 
 II. The Physics of an Incandescent Lamp Characteristic Curves Relation of Candle-power to Current 
 and Pressure Effects of Position on Candle-power Age of Incandescent Lamps Lamp Mortality Causes of 
 it Judicious and Injudicious Arrangement of Lamps Sockets Switches Fuses Decorative Employment of 
 Incandescent Lampi House Wiring Fire Office Rules Good and Bad Work Causes of Destruction of the 
 Carbon Filament -Molecular Physics of the Glow Limp Edison Effect Large Incandescent Lamps Electric- 
 Meters Methods of Testing and Comparing Glow Lamps Advantage ms Utilisation of Current. 
 
 III. The Electric Arc Lamp Method of Production of the Arc Study of the Arc by Projection Laws of 
 the Electric Arc The Convection of Carbon in the Arc The Crater The Distribution of Electric Pressure in 
 the Arc Arc Lamp Mechanism Recent Improvements Distribution of Light from the Arc Luminoua 
 Efficiency of the Arc Comparison with Incandescent Lamps Street and Interior Arc Lighting Proper Distri- 
 bution of Light Arc Light Photometry The Alternating and Continuous Arc The Inverted Arc The Use of 
 the Arc in Metallurgy Electrical Reduction of Metals by the Arc Temperature of Arc Light Crater The 
 Solar Temperature. 
 
 IV. The Production of Current for Electric Lighting Generating Stations Systems of Supply Low 
 Pressure Continuous and High Pressure Alternating Structure of a Dynamo and Transformer Views of 
 Generating Stations at Home and Abroad Underground Conducting Mains Networks of Conductors House 
 Services Long Distance Transmission Electric Lighting of Rome Tivoli- Rome Transmission Utilisation of 
 Water-power Load Diagrams of Stations Supply of Currents for Purposes other than Light. 
 
 PRESS NOTICES. 
 
 "Treats the whole subject with the lucidity for which Prof. Fleming's expositions are remarkable, and in 
 language which has been, as far as possible, divested of technicalities. . . . Those who may contemplate 
 the electric lighting of their houses will find in the book many valuable hints and practical suggestions." Times. 
 
 " Prof. Fleming possesses the rare gift of being able to interest almost any audience in his subject. . . . 
 The work under notice consists of a transcript of four lectures delivered at the Royal Institution, and though 
 much of the charm which the lecturer was able to throw round the subject is now absent, still enough remains 
 to render a perusal of these pages a recreation rather than a study. . . . Besides supplying the non-technical 
 student with such information as he is sally in need of, he has given to the serious student the best account 
 which we have yet seen of th^ physical properties of arc and glow lamps. . . . The volume contains some 
 220 pages of large type, and is well illustrated throughout. It is the only one we have ever seen that we can 
 thoroughly recommend to the non-technical reader. On the other hand, for the electrical student we cannot too 
 forcibly recommend a careful study of the second and third lectures." Daily Chronie'e. 
 
 " We have no hesitation in recommending all who may be interested in the subject to buy a copy of Prof. 
 Fleming's book, which is well worth the price asked for it ; being fresh as regards matter, and abundantly 
 provided with illustrations which are more or less original. ... As we commenced this article by stating, 
 the book is worth buying for a guide to a fair general knowledge of the principles which underlie the industrial 
 uses of electricity. We can also compliment the publishers upon the way in which the volume is turned out." 
 Journal of Gas Lighting. ' 
 
 "When one comes across a reprint of a series of afternoon Royal Institution lectures he expects an elemen- 
 tary book, with its contents strung t-gether to connect a lot of experiments that the author was anxious to show 
 to an audience already jaded with a variety of subjects no one human brain could take in. Dr. Fleming has 
 given us nothing of this sort, but a concise account of incandescent and arc lamps, followed by a short sketch of 
 electric lighting generally. It is by no means a mere popular book, however, and few electricians, no matter 
 how well up in their subject, will find they can afford to pass it by. ... The binding, printing, and general 
 get up is admirable." J.S. in The Electrician. 
 
 " Dr. Fleming is known for his hich abilities in setting forth scientific facts in a popular lecture in clear and 
 most attractive language, and anyone familiar with the elements of physical science will read the book with 
 great pleasure and derive much souiid instruction. . . . The book is very handsomely got up, and would 
 make an exceedingly suitable presentation volume to anyone interested in, and who already knows at least a 
 little of, the subject." Glasjow Herald. 
 
 DIGEST POST FREE ON APPLICATION. 
 
 1, 2 and 3, Salisbury Court, Fleet Street, London, E.G. 
 
C "The Electrician" Printing and Publishing Co., Ltd., 
 "THE ELECTRICIAN" SERIES continued. 
 
 FOURTH ISSUE. 500 paycs, 157 illustrations. Price 7s. 6d., post free 8s. 
 
 THE ALTERNATE CURRENT TRANSFORMER 
 
 IN THEORY AND PRACTICE. 
 
 By J. A. FLEMING, M.A., D.Sc., F.K.S., M.E.I., &c., 
 
 Professor of Electrical Engineering in University College, London. 
 
 YOL. I. THE INDUCTION OF ELECTKIC CURRENTS. 
 
 SYNOPSIS OF CONTENTS, 
 CHAPTER I. Introductory. 
 
 Faraday's Electrical Eesearclies Early Experiments on Current Induction Electro-Dynamic 
 Induction First Induction Coil Electro-Magnetic Induction Lines of Force Methods of 
 Ampere, Arago, and Faraday Physical Nature of Lines of Magnetic Force. 
 
 CHAPTER II Electro-Magnetic Induction. 
 
 Magnetic Force and Magnetic Induction Tubes of Magnetic Induction Rate of Change of 
 Magnetic Induction through a Circuit Inductance Electromotive Force of Induction Electro- 
 Magnetic Momentum Electro-Magnetic Energy Dimensions of the Co-efficient of Self-induction 
 or Inductance Unit of Self -Induction or Inductance Constant and Variable Inductance Curves 
 of Magnetisation Graphical Representation of Variation of Co-efficient of Induction Magnetic 
 Hysteresis. 
 
 CHAPTER III The Theory of Simple Periodic Currents. 
 
 Variable and Steady Flow Current Curves Simple and Complex Harmonic Motion- 
 Fourier's Theorem Simple Periodic and Sine Curves Current Growth in Inductive Circuits 
 Logarithmic Curves Geometrical Illustrations Graphic Representation of Periodic Currents 
 Mean Value of the Power of a Periodic Current Power Curves Experimental Measurement of 
 Periodic Current and Electromotive Force Wattmeter Method of Periodic Power Mutual 
 Induction of Two Circuits of Constant Inductance. 
 
 CHAPTER IV. Mutual and Self-induction. 
 
 Prof. Joseph Henry's Researches in Electro-Magnetism Mutual Induction Elementary 
 Theory of the Induction Coil Comparison of Theory and Experiment Magnetic Screening and 
 the Action of Metallic Masses in Induction Coils Transmission of Rapidly Intermittent, 
 Alternate, or very Brief Currents through Conductors Effects of Saturation and Magnetic 
 Hysteresis Characteristic Curves of the Series and Parallel Transformer Efficiency of Trans - 
 iormers Ferrari's Experiments. 
 
 CHAPTER V. Dynamical Theory of Current Induction. 
 
 Electric Displacement Maxwell's Theory of Molecular Vortices Velocity of Propagation of 
 an Electro-Magnetic Disturbance Electrical Oscillations Function of the Condenser in an 
 Induction Coil Impulsive Discharges Alternative Paths Impulsive Impedance Relation of 
 Impedance to Periodicity Dr. Hertz's Researches on Electrical Oscillatory Induction Resonance 
 Phenomena Interference Phenomena at Various Distances Recent Experiments Poynting's 
 Views on the Propagation of Electro-Magnetic Energy Possible Direction of Future Research. 
 
 PREZSS NOTICES. 
 
 " It would be very difficult to pick out from amongst the electrical literature of the past ten years any work 
 which' mark?, as emphatically as does Dr. Fleming's book, the manner in which the practical problems of the day 
 
 have compelled electrical engineers to advance in their knowledge of theoretical science It is a book 
 
 which the electrical engineer of the present and of the future alike will read he of the present, if he can ; 
 he of the future, because he must." Prof. Silvanus P. Thompson in " The Electrician." 
 
 " The practical importance and interest of the subject treated is so great that there should be little need to 
 urge students and electrical engineers to make themselves acquainted with this book, but I do urge them 
 nevertheless ; and they may think it fortunate that Dr. Fleming has managed to find time to issue so instructive 
 and readable and well-timed a volume." Dr. Oliver J. Lodge in " Nature." 
 
 " Dr. Fleming's book contains an enormous amount of valuable matter .... which cannot be got anywhere 
 lse in the plain and concise way it is given by Dr. Fleming. It is one of those books every electrician should 
 have." Electrical Review. 
 
 "A most important, timely, and valuable book The author has earned the thanks of everyone 
 
 interested in this great branch of electrical investigation and practice. "Electrical World (^e\v York). 
 
 fl If anyone wants this difficult subject treated in the clearest way, he cannot do better than read this book." 
 
 Indugtrieii. 
 
 1, 2, and 3, Salisbury Court, Fleet Street, London, E.G. 
 
The Electrician" Printing and Publishing Co., Ltd., 
 "THE ELECTRICIAN" SERIES continued. 
 
 SECOND ISSUE. More than 600 pages and over 300 illustrations. Price 12s. 6d., post free. 
 
 THE ALTERNATE CURRENT TRANSFORMER 
 
 IN THEORY AND PRACTICE. 
 
 By J. A. FLEMING, M.A., D.Sc., F.R.S., M.R.I., &c., 
 
 Professor of Electrical Engineering in University College, London. 
 
 YOL. II. THE UTILISATION OF INDUCED CURRENTS. 
 
 SYNOPSIS OF CONTENTS. 
 CHAP. L Historical Development of Induction Coil and Transformer. 
 
 The Evolution of the Induction Coil Page's Researches Callan's Induction Apparatus 
 Sturgeon's Induction Coil Bachhoffner's Researches Callan's Further Researches Callan's 
 Great Induction Coil Page's Induction Coil Abbot's Coil Automatic Contact Breakers 
 Ruhmkorff's Coils Poggendorff's Experiments Stohrer's, Hoarder's, Ritchie's Induction 
 Apparatus Grove's Experiments Apps' Large Induction Coils Jablochkoff's Patent Fuller's 
 Transformer Early Pioneers Gaulard and Gibbs Zipernowsky's Transformers Improvements 
 of Rankin Kennedy, Hopkinson, Ferranti, and others The Modern Transformer since 1885. 
 
 CHAP. II. Distribution of Electrical Energy by Transformers. 
 
 Detailed Descriptions of Large Alternate-Current Electric Stations using Transformers in 
 Italy, England, and United States Descriptions of the Systems of Zipernowsky-Deri-Blathy, 
 Westinghouse, Thomson-Houston, Mordey, Lowrie-Hall, Ferranti, and others Plans, Sections, 
 and Details of Central Stations using Transformers Illustrations of Alternators and Transformers 
 in Practical Use in all the chief British, Continental, and American Transformer Stations. 
 
 CHAP. III. Alternate-Current Electric Stations. 
 
 General Design of Alternating-Current Stations, Engines, Dynamos, Boilers Proper Choice 
 of Units Water Power Parallel Working of Alternators Underground Conductors Various 
 Systems Concentric Cables Capacity Effects dependent on Use of Concentric Cables Phenomena 
 of Ferranti Tubular Mains Safety Devices Regulation of Pressure Choice of Frequency 
 Methods of Transformer Distribution Sub-Stations Automatic Switches. 
 
 CHAP. IV. The Construction and Action of Transformers. 
 
 Transformer Indicator Diagrams Ryan's Curves Curves of Current Electromotive Force 
 and Induction- Analysis of Transformer Diagrams Predetermination of Eddy Current and 
 Hysteresis Loss in Iron Cores Calculation and Design of Transformers Practical Predetermina- 
 tion of Constants Practical Construction of Transformers Experimental Tests of Transformers 
 Measurement of Efficiency of Transformers Calometric Dynamometer and Wattmeter Methods 
 Reduction of Results. 
 
 CHAP. V. Further Practical Application of Transformers. 
 
 Electrical Welding and Heating Transformers for producing Large Currents of Low Electro- 
 motive Force Theory of Electric Welding Other Practical Applications Conclusion. 
 
 PRESS NOTICES. 
 
 "In reviewing the first volume of this work we found much to admire and praise, much to raise high 
 expectations for the volume which was to follow. These expectations have by no means been disappointed. 
 The new volume is in many ways of even greater interest than its predecessor." 
 
 Professor Silvanus P. Thompson in "The Electrician." 
 
 "The book is really a valuable addition to technical literature." Industries. 
 
 11 A valuable addition to the somewhat meagre literature on a subject which is sure to grow in importance, 
 and we congratulate Dr. Fleming on his work." The Engineer. 
 
 " Le sujet traite par le Dr. Fleming est un de ceux qui, pour le moment, attirent 1'attention ge"ne"rale ; son 
 ouvrage est certainement un des plus import-ants de la litterature electrique. Tous les problemes relatifs a 
 1'application des courants alternatifs y sont traite"s avec une tres grande competence et de plus avec une clarte" 
 et avec une precision sans egales. Nous ne ppuvons done que recommander vivement cet ouvrage a 1'attention 
 de tous les Slectriciens." La Lumitre Electrique. 
 
 " L'ouvrage de M. Fleming est une oeuvre vraiment pratique qui doit rendre a l'industrie de grands services 
 
 i> j j ~.,*,, ~ u ' e ii e contient." L'industrie Electrique. 
 
 Das Fleming'sche Werk f iillt entschieden eine Liicke in der Literatur aus und kann durchaus empfohlen 
 verden." Elektrotechnische Zeitschrift. 
 
 1, 2 and 3, Salisbury Court, Fleet Street, London, E.G. 
 
8 "The Electrician" Printing and Publishing Co., Ltd., 
 
 "THE ELECTRICIAN" SERIES continued. 
 
 In Tivo Volumes. Price: stout paper, 2s, post free 2s. 3d. each; strong cloth covers, 2s. 6d., 
 post free 2s. 9d. each. Single Primers, 3d., post free 3\d. 
 
 "THE ELECTRICIAN" PRIMERS. 
 
 (FULLY ILLUSTRATED.) 
 
 A Series of Helpful Primers on Electrical Subjects for the use of Colleges, Schools, 
 and other Educational and Training Institutions, and for Young; Men desirous of 
 entering; the Electrical professions. 
 
 TABLE OF 1 CONTENTS. 
 
 Volume I. THEORY. 
 
 Primer 
 
 No. 
 
 1. The Effects of an Electric 
 
 Current. 
 
 2. Conductors and Insulators. 
 
 3. Ohm's Law. 
 
 4. Primary Batteries. 
 
 5. Arrangement of Batteries. 
 
 6. Electrolysis. 
 
 7. Secondary Batteries. 
 
 8. Lines of Force. 
 
 9. Magnets. 
 
 10. Electrical Units. 
 
 11. The Galvanometer. 
 
 12. Electrical Measuring In- 
 
 struments. 
 
 13. The Wheatstone Bridge. 
 
 14. The Electrometer. 
 
 15. The Induction Coil. 
 
 16. Alternating Currents. 
 
 17. The Leyden Jar. 
 
 18. Influence Machines. 
 
 19. Lightning Protectors. 
 
 20. Thermopiles. 
 
 The object of "The Electrician" 
 Primers is to briefly describe in sim- 
 ple and correct language the present 
 state of electrical knowledge. Each 
 Primer is short and complete in itself, 
 and is devoted to the elucidation of 
 some special point or the description 
 of some special application. Theo- 
 retical discussion is as far as possible 
 avoided, the principal facts being 
 stated and made clear by reference 
 
 /Yolumell. PRACTICE. 
 
 Primer 
 
 ;, NO. 
 
 21. The Electric Telegraph. 
 
 22. Automatic and Duplex Te 
 
 graphy. 
 
 23. The Laying and Repair 
 
 Submarine Cables. 
 
 24. Testing Submarine Cables. 
 
 25. The Telephone. 
 
 to the uses to which they have been i, 
 
 26. Dynamos. 
 
 put. Both volumes are suited to > _ _.. . 
 readers having little previous ac- j ZT - Motors, 
 quaintance with the subject. The I 28. Transformers, 
 mattei is brought up to date, and ) . _ 
 
 the illustrations refer to instruments <\ 29 - * ne Axc Lamp, 
 and machinery in actual use at the 
 present time. It is hoped that the 
 Primers will be found of use in 
 Schools, Colleges, and other Educa- 
 tional and Training Establishments, 
 where the want of a somewhat 
 popularly written work on electricity 
 and its industrial applications, pub- 
 lished at a popular price, has long 
 been felt ; while artisans will find the 
 Primers of great service in enabling 
 them to obtain clear notions of the 
 essential principles underlying the 
 apparatus of which they may be 
 called upon to take charge. 
 
 30. The Incandescent Lamp. 
 
 31. Underground Mains. 
 
 32. Electric Meters. 
 
 33. Electric Light Safety 
 
 vices. 
 
 34. Systems of Electric 
 
 bution. 
 
 35. Electric Transmission 
 
 Energy. 
 
 36. Electric Traction. 
 
 j 37. Electro-Deposition. 
 I 38. Electric Welding. 
 
 "The articles are generally so well written, and the subject matter so judiciously condensed, that there 
 but very little to criticise, though much to praise." Electrical Review. 
 
 " The books are well printed, and we can heartily commend them as stepping stones to more advance< 
 works." Electrical Plant. 
 
 " Clearly written, and all that can be desired in the form of enunciation and explanation." Work. 
 The contents of each one of these volumes is of that quality and description which at once constitute a be 
 a welcome addition to the library of the student or of the artisan." Amateur Work 
 
 they are calculated to be exceedingly usefi 
 ery suitable for presents to boys of a mecbanics 
 wl Board Chronicle. 
 
 Issued annually, price 3s., post free. 
 
 A DIGEST OF THE LAW OF ELECTRIC LIGHTING, 
 
 By 
 
 -A.3STID OTHER STJBCTEJOTS. 
 
 (Revised to January in each year.) 
 
 A. C. CURTIS-HAYWARD, B.A., M.I.E.E. 
 
 An abstract of the Electric Lighting Acts, 1882 and 1889, and of the vaiious documents emanating from the 
 Board of Trade dealing with electric lighting. The digest treats first of the manner in which persons desirous of 
 supplying electricity must set to work, and then of their rights and obligatir us after obtaining Parliamentary 
 powers ; and gives in a succinct from information of great value to Local Authorities, Electric Light Contractors, 
 <fec., up to date. The Board of Trade Regulations, the London County Council regulations as to Theatre 
 Lighting, Britibh and Foreign Rules and E emulations for the Prevention of Fire .Risks arising from Electric 
 Lighting, and the Installation Regulations of the Electric Supply Companies of the Metropolis are also given. 
 
 1, 2, and 3, Salisbury Court, Fleet Street, London, E.G. 
 
"The Electrician" Printing and Publishing Co., Ltd., 
 "THE ELECTRICIAN" SERIES continued. 
 
 800 pages, specially lound lookwise to lie open, 7s. 6d. nctt, post free 7s. 9d. (abroad, 8S., post free] ; 
 large paper edition, with wide margins, 12s. 6d. nctt, post free 13s. (abroad, 13s. 6d. post free}. 
 
 A. POCKET-BOOK OF 
 
 ELECTRICAL ENGINEERING FORMULA, &c. 
 
 BY W. GEIPEL AND H. KILGOUR. 
 
 "With the extension of all branches of Electrical Engineering (and particularly the heavier 
 branches), the need of a publication of the Pocket -Book style dealing practically therewith 
 increases ; for while there are many such books referring to Mechanical Engineering, and several 
 dealing almost exclusively with the lighter branches of electrical work, none of these suffice for the 
 purposes of the numerous body of Electrical Engineers engaged in the application of electricity to 
 Lighting, Transmission of Power, Metallurgy, and Chemical Manufacturing. It is to supply this 
 real want that this most comprehensive book has been prepared. 
 
 Compiled to some extent on the lines of other pocket-books, the rules and formulae in general 
 use among Electricians and Electrical Engineers all over the world have been supplemented by 
 brief and, it is hoped, clear descriptions of the various subjects treated, as well as by concis 
 articles and hints on the construction and management of various plant and machinery. 
 
 A list of the subjects treated will be found below, from which it will at once be seen how 
 Indispensable the book will be to those engaged in electrical work of all kinds. 
 
 Xo pains have been spared in compiling the various sections to bring the book thoroughly up 
 to date ; and while much original matter is given, that which is not original has been carefully 
 selected, and, where necessary, corrected. 
 
 Where authorities differ, as far as practicable a mean has been taken, the differing formula Deing 
 quoted for guidance. 
 
 Units. 
 
 Electrical, Mechanical 
 
 and Thermal Properties 
 
 of Materials. 
 Light and Sound. 
 Gravity. 
 Beams. 
 
 Moments of Inertire. 
 Approximate Formula? 
 The Slide Rule. 
 Mathematical Tables. 
 Miscellaneous Tables. 
 
 Ventilation and Heating. 
 
 Buildings & Foundations. 
 
 Water Power. 
 
 Fuel and Combustion. 
 
 Chimneys and Flues. 
 
 Boilers. 
 
 Feed- Water Heaters and 
 
 Economises. 
 Boiler and Steam-Pipe 
 
 Covering. 
 Boiler Testing. 
 Steam, Gas, and other 
 
 Engines. 
 
 Condensers and Pumps. 
 Injector*. 
 Steam Pipes. 
 
 Mechanical Transmitters. 
 Dynamos and Motors. 
 Regulators. 
 Transformers. 
 Systems of Distribution. 
 Conductors. 
 Lightning Protectors. 
 Arc Lamps. 
 Incandescence Lamps. 
 
 Primary Batteries. 
 Secondary Batteries. 
 Electro-Chemistry. 
 Electro-Metallurgy. 
 Electric Welding. 
 Electric Traction. 
 Mining. 
 
 Testing and Measuring In- 
 struments. 
 Fire Office Rules. 
 Patent Office Information. 
 Wiring Tables. 
 
 A Prospectus, with Specimen Pages, &c., sent post free on application. 
 NOW READY, Fully Illustrated, cloth gilt. Price 2s. 6d. nett; post free 2s. 9d. 
 
 THE WORK OF HERTZ 
 
 AND SOME OF HIS SUCCESSORS, 
 
 BEING THE SUBSTANCE OF A LECTURE DELIVERED AT THE ROYAL INSTITUTION 
 
 By Dr. OLIVER J. LODGE, F.R.S. 
 
 Reprinted, with Corrections to Text and Illustrations, and Additions, with STEEL-PLATE PORTRAIT Frontispiece 
 
 " THE untimely end of a young and brilliant career cannot fail to strike a note of sadness and awaken a chord of 
 sympathy in the hearts of his friends and fellow-workers. Of men thus cut down in the early prime of their 
 powers there will occur to us here the names of Fresnel, of Carnot, of Clifford, and now of Hertz. His was a 
 strenuous and favoured youth ; he was surrounded from his birth with all the influences that go to make an 
 accomplished man of science accomplished both on the experimental and on the mathematical side. The 
 front rank of scientific workers is weaker by his death, which occurred on January 1st, 1S94, the thirty-sixth 
 year of his life. Yet did he not go till he had effected an achievement which will hand his name down to 
 posterity as the founder of an epoch in experimental physics." Extract from Dr. Lodge's "Introductory." 
 
 1, 2, and 3, Salisbury Court, Fleet Street, London. E.G. 
 
10 "The Electrician" Printing and Publishing Co,, Ltd., 
 "THE ELECTRICIAN" SERIES continued. 
 
 Fully Illustrated. Price 7s. 6d., post free 
 THE 
 
 INCANDESCENT LAMP AND ITS MANUFACTURE. 
 
 By GILBERT S. RAM. 
 
 EXTRACT FROM THE PREFACE, 
 
 With the expiration of Edison's master-patent for the carbon incandescent electric 
 lamp, the attention of electric light engineers, as well as of all those who use the light, is 
 unce more directed to the consideration of the lamp itself, to the possibility of obtaining 
 better lamps, and to the probable reduction in price which will naturally follow. Owing 
 to the long prevailing monopoly in the sale and manufacture, there has been little 
 inducement for those interested to experiment and to study the problems connected 
 with the incandescent lamp. As a result of this, the literature of the lamp is very 
 scanty, and is entirely confined to the pages of the leading technical journals. While 
 dynamos, alternators, transformers, arc lamps, and almost every piece of apparatus 
 connected with electrical engineering and lighting, have been written on at length and 
 discussed at meetings of scientific societies, the incandescent electric lamp, which has 
 been the chief cause of the very existence of these machines and apparatus, has been 
 comparatively neglected. With the exception of the valuable series of articles by 
 Mr. Swinburne which appeared in Tlie Electrician some years ago, no comprehensive or 
 detailed account of lamp manufacture has appeared. The manufacture of the incan- 
 descent lamp and the principles underlying it are, consequently, but little known, except 
 to those actually engaged in the work. 
 
 The Author has attempted to impart such information as he has acquired in tht 
 course of a considerable experience in lamp-making, and to give this information wit! 
 as little mathematical embellishment as, under the circumstances, is possible. 
 
 Fully Illustrated. Price 7s. 6d., post free. 
 
 DRUM ARMATURES AND COMMUTATOR! 
 
 By F. MARTEN WEYMOUTH. 
 
 A complete treatise of the theory and the different modes of construction of Drum Windi 
 and also of close-coiled continuous-current Commutators, together with a full resume of som 
 the principal points of consideration that are involved in their design. 
 
 The first chapter relates entirely to the theory of the Drum Winding. An explanation, il 
 hoped with sufficient fulness, is there given of the generation of electromotive force and cui 
 within that form of winding ; and questions of magnetism connected therewith are also conside 
 Chapters II. to IX. are devoted to the description of various methods in which Drum Winding ha 
 been carried out in practice, with special reference to what is termed the "end-winding. 
 Chapters X. to XIII. touch upon the mechanical construction of Commutators ; and Chaptei 
 XIV. to XXVI. deal with Commutator Sparking, with which has become necessarily involved th 
 whole subject of what are known as " armature reactions." The book closes with a chapter on tb 
 Taper of Commutator Segments. 
 
 The various subjects are treated without, or almost without, the use of mathematics. But th 
 author ventures to think that, in this case, the absence of mathematics is far from unjustifiable. 
 
 1, 2, and 3, Salisbury Court, Fleet Street, London, E. 
 
 - 
 
"The Electrician" Printing and Publishing Co., Ltd., 11 
 "THE ELECTRICIAN" SERIES continued. 
 
 Over 300 pages, 106 illustrations. Price 10s. 6d., post free. 
 
 The ART of ELECTROLYTIC SEPARATION of METALS. 
 
 (THEOEETICAL AND PRACTICAL.) 
 By GEORGE GORE, LL.D., F.R.S. 
 
 THE ONLY BOOK ON THIS IMPORTANT SUBJECT IN ANY LANGUAGE. 
 
 SYNOPSIS OF CONTENTS. 
 HISTORICAL SKETCH. 
 
 Discovery of Voltaic and Magneto-Electricity First Application of Electrolysis to the 
 Refining of Copper List of Electrolytic Refineries. 
 
 THEORETICAL DIVISION. 
 
 Section A. : Chief Electrical Facts and Principles of the Subject. Electric Polarity and 
 Induction, Quantity, Capacity, Potential Electromotive Force Electric Current Conduction 
 and Insulation Electric Conduction Resistance. 
 
 Section B. : Chief Thermal Phenomena. Heat of Conduction Resistance Thermal Units, 
 Symbols, and Formula?. 
 
 Section C. : Chief Chemical Facts and Principles of the Subject. Explanation of Chemical 
 Terms Symbols and Atomic Weights Chemical Formulas and Molecular Weights Relation of 
 Heat to Chemical Action. 
 
 Section D. : Chief Facts of Chemico -Electric or Voltaic Action. Electrical Theory of 
 Chemistry Relation of Chemical Heat to Yolta Motive Force Volta-Electric Relations of 
 Metals in Electrolytes Voltaic Batteries Relative Amounts of Voltaic Current produced by 
 Different Metals. 
 
 Section E.: Chief Facts of Electro-Chemical Action. Definition oi Electrolysis Arrange- 
 ments for Producing Electrolysis Modes of Preparing Solutions Nomenclature Physical 
 Structure of Electro-Deposited Metals Incidental Phenomena attending Electrolysis Deconi- 
 posability of Electrolytes Electro-Chemical Equivalents of Substances Consumption of Electric 
 Energy in Electrolysis. 
 
 Section F. : The Generation of Electric Currents by Dynamo Machines. Definition of a 
 Dynamo and of a Magnetic Field Electro-Magnetic Induction Lines of Magnetic Force. 
 
 PRACTICAL DIVISION. 
 
 Section G. : Establishing and Working an Electrolytic Copper Eefinery. Planning a Refinery 
 Kinds of Dynamos Employed Choice and Care of Dynamo The Depositing Room The Vats 
 The Electrodes The Main Conductors Expenditure of Mechanical Power and Electric 
 Energy Cost of Electrolytic Refining. 
 
 Section H. : Other Applications of Electrolysis in Separating and Refining Metals. Elec- 
 trolytic Refining of Copper by other Methods Extraction of Copper from Minerals and Mineral 
 Waters Electrolytic Refining of Silver Bullion and of Lead Separation of Antimony, of Tin, of 
 Aluminium, of Zinc, of Magnesium, of Sodium and Potassium, of Gold Electrolytic Refining of 
 Nickel Electric Smelting. 
 
 Appendix. Useful Tables and Data. 
 
 Second Edition, price Sis., post free. 
 
 ELECTRO-CHEMISTRY. 
 
 By GEORGE GORE, LL.D., F.R.S. 
 
 This book contains, in systematic order, the chief principles and facts of electro-chemistry, 
 and is intended to supply to the student of electro-plating and electro-metallurgy a scientific basis 
 npon which to build the additional practical knowledge and experience of his trade. A scientific 
 foundation, such as is here given, of the art of electro-metallurgy is indispensable to the electro- 
 depositor who wishes to excel in his calling, and should be studied previously to and simul- 
 taneously with practical working. As the study of electro-chemistry includes a knowledge not 
 only of the conditions under which a given substance is electrolytically separated, but also of the 
 electrolytic effect of a current on individual compounds, both are described, and the series of 
 substances are treated in systematic order. An indispensable book to Electro-Metallurgists. 
 
 1, 2, and 3, Salisbury Court, Fleet Street, London, E.C, 
 
12 "The Electrician" Printing and Publishing Co., Ltd., 
 
 THE ELECTRICIAN" SERIES continued. 
 
 Electrical Laboratory Notes & Forms 
 
 ARRANGED AND PREPARED BY 
 
 . J. A. FI^EMJHVO, M.A., F.R 
 
 Professor of Electrical Engineering in University College, London. 
 These " Laboratory Notes and Forms " have been prepared to assist Teachers, Demonstrators 
 
 and Students in Electrical Laboratories, and to enable the Teacher to economise time. The} 
 consist of a series of (about) Twenty Elementary and (about) Twenty Advanced Exercise* 
 in Practical Electrical Measurements and Testing. For each of these Exercises a four-page Eepori 
 Sheet has been prepared, two pages of which are occupied with a condensed account of the theorj 
 and practical instructions for performing the particular Experiment, the other two pages being 
 ruled up in lettered columns, to be filled in by the Student with the observed and calculatec CH 
 quantities. Where simple diagrams will assist the Student, these have been supplied. Thes 
 Exercises are for the most part based on the methods in use in the Electrical Engineering 
 Laboratories of University College, London ; but they are perfectly general, and can be put into 
 practice in any Electrical Laboratory. 
 
 Each Form is supplied either singly at 4d. nett, or at 3s. 6d. per dozen nett (assorted o: 
 otherwise as required) ; in sets of any three at Is. nett ; or the set of (about) Twenty Elementar 
 (or Advanced) Exercises can be obtained, price 5s. 6d. nett. The complete set of Elementar 
 and Advanced Exercises are price 10s. 6d. nett, or in a handy Portfolio, 12s. nett, or bound ii 
 strong cloth case, price 12s. 6d. nett. 
 
 Spare Tabulated Sheets for Observations, price Id. each nett. 
 
 Strong Portfolios, price Is. each. 
 
 The very best quality foolscap sectional paper (16in. by 13in.) can be supplied, price Is 
 per dozen sheets nett. 
 
 1. The Exploration of Magnetic Fields. 
 
 2. The Magnetic Field of a Circular Current. 
 
 3. The Standardization of a Tangent Galvanometer by the Water Voltameter. 
 
 4. The Measurement of Electrical Resistance by the Divided Wire Bridge. 
 
 5. The Calibration of the Ballistic Galvanometer. 
 
 6. The Determination of Magnetic Field Strength. 
 
 7. Experiments with Standard Magnetic Fields. 
 
 8. The Determination of the Interpolar Field of an Electromagnet with Varying Lengths of Air Gap. 
 
 9. The Determination of Resistance and Temperature Coefficients with the Post Office Pattern of Wheatstone 
 
 10. The Determination of Electromotive Force by the Potentiometer. [Bridg< 
 
 11. The Determination of Current Strength by the Potentiometer. 
 
 12. A Complete Test of a Primary Battery. 
 
 18. The Calibration of a Voltmeter by the Potentiometer. 
 
 14. A Photometric Examination of an Incandescent Lamp. 
 
 15. The Determination of the Absorptive Powers of Semi-Transparent Screens. 
 
 16. The Determination of the Reflective Powers of Various Surfaces. 
 
 17. The Determination of the Electrical Efficiency of an Electromotor by the Cradle Method. 
 
 18. The Determination of the Efficiency of an Electromotor by the Brake Method. 
 
 19. The Efficiency Test of a Combined Motor Generator Plant. 
 
 20. Test of a Gas Engine and Dynamo Plant. 
 
 (Ready shortly.) 
 
 21. The Determination of the Specific Electrical Resistance of a Sample of Wire. 
 
 22. The Measurement of Low Resistances by the Potentiometer. 
 
 23. The Measurement of Armature Resistances. 
 
 24. The Standardization of an Ampere-meter by Copper Deposit. 
 
 25. The Standardization of a Voltmeter by the Potentiometer. 
 
 26. The Standardization of an Ammeter by the Potentiometer. 
 
 27. The Determination of the Magnetic Permeability of a Sample of Iron. 
 
 28. The Standardization of a High Tension Voltmeter. 
 
 29. The Efficiency Test of a Transformer. 
 
 30. The Delineation of the Curves of Current and Electromotive Force of a Transformer. 
 
 31. The Photometric Examination of an Arc Lamp. 
 
 32. The Measurement of Insulation and High Resistance. 
 
 33. The Examination of a Secondary Cell by the Potentiometer. 
 
 34. The Efficiency Test of an Alternator. 
 
 35. The Complete Efficiency Test of a Secondary Battery. 
 
 36. The Calibration of Electric Meters. 
 
 37. The Determination of the Hysteresis Curve of Iron by the Magnetometer. 
 
 38. Ihe Determination of Hysteresis Loss by the Wattmeter. , L. 
 
 39. The Determination of the Capacity of a Concentric Cable. 
 
 40. The Complete Hopkinson Test of a Pair of Dynamos. 
 
 1, 2, and 3, Salisbury Court, Fleet Street, London, E.G. 
 
 ' 
 
 ft 
 
 I 
 
 I 
 
 U 
 
 B 
 
"The Electrician" Printing and Publishing Co., Ltd., 13 
 
 " THE ELECTRICIAN " SERIES continued. 
 
 320 pages, 155 illustrations. Price 6s. 6d., post free. 
 
 PRACTICAL NOTES FOR ELECTRICAL STUDENTS. 
 
 LAWS, UNITS, AND SIMPLE MEASURING INSTRUMENTS. 
 By A. E. KENNELLY and H. D. WILKINSON, M.I.E.E. 
 
 SYNOPSIS OF CONTENTS, 
 
 CHAPTER I. Introductory. 
 
 Early Ideas Electricity produced by Chemical Energy Requirements in a good Cell 
 Chemical Action. 
 
 CHAPTER II. Batteries. 
 
 Daniell, Minotto, Thompson Tray, Leclanche, Fuller, De la Rue and Standard Cells. 
 
 CHAPTER III. Electromotive Force and Potential. 
 
 Connecting Cells in Series Distribution of Potential in a Battery. 
 
 CHAPTER IV. Resistance. 
 
 Relative Resistance of Metals Relation between Length, Diameter, and Weight of Telegraph 
 Conductors Resistances in Series and in Multiple Arc. 
 
 CHAPTER V. Current. 
 
 Effect of " Opening " or "Closing" a Circuit Velocity of Current Retardation Period of 
 Constant Flow The Ampere The Coulomb The Milliampere Ohm's Law. 
 
 CHAPTER VI. Current Indicators. 
 
 Detectors or Indicators Directions for Making Detectors for Telegraph and for Telephone 
 Work Indicators for Large Current. 
 
 CHAPTER VII. Simple Tests -with Indicators. 
 
 Tests for " Continuity," for Fault in Telegraph Apparatus, for Identity of Wires, for Insu- 
 lationOverhead Line Insulators G.P.O. Standard Indicator. 
 
 CHAPTER VIII. Calibration of Current Indicators. 
 
 Calibration by Low-Resistance Cells Calibration Curves Simultaneous Calibration of 
 Instruments of Similar Sensitiveness and of Differing Sensitiveness Use of the "Shunt" 
 Comparison by Tangent Galvanometer. 
 
 CHAPTER IX. Magnetic Fields and their Measurements. 
 
 Permanent Magnetic Fields Electro-Magnetic Fields Magnetic Fields of Coils and Solenoids. 
 TABLE OF NATURAL TANGENTS. 
 
 190 pages, 116 illustrations. Price 3s. 6d., post free. 
 
 THE STEAM-ENGINE INDICATOR & INDICATOR DIAGRAMS 
 
 A PRACTICAL TREATISE ON. 
 
 Edited by W. W. BEAUMONT, M.I.C.E., M.I.M.E., &c. 
 
 This useful book considers the object of an Indicator Diagram, or what it is desired that the 
 Diagram shall show ; describes the construction for the Indicator in its various forms ; describes 
 the apparatus necessary for the attachment of the Indicator to the engine, and how to use the 
 instrument ; gives examples of diagrams from all kinds of engines most in use, comparing these 
 diagrams and showing how far they agree with theoretical diagrams ; and shows the most simple 
 methods of calculating and constructing theoretical curves of expansion, and of comparing the 
 actual with the theoretical performance of steam in the steam engine cylinder. 
 
 Fully illustrated. Price Is. 6d., ^>ost free Is. 9d. 
 
 THE MANUFACTURE OF ELECTRIC LIGHT CARBONS. 
 
 A Practical Guide to the Establishment of a Carbon Manufactory. 
 
 Contains the results of several years' experiments and experience in carbon candle-making, and 
 gives full particulars, with many illustrations, of the whole process. 
 
 1, 2, and 3, Salisbury Court, Fleet Street, London, E.C. 
 
14 "The Electrician" Printing and Publishing Co., Ltd., 
 "THE ELECTRICIAN" SERIES continued. 
 
 NOW READY. Over 400 pages, nearly 250 illustrations. Price 10s. 6d. 
 
 ELECTRIC MOTIVE POWER. 
 
 By ALBION T. SNELL, Assoc.M.lNST.C.E., M.I.E.E. 
 
 The rapid spread of electrical work in collieries, mines, and elsewhere has created a demand for a practica 
 book on the subject of transmission of power. Though much had been written, there was no single work dealhr 
 with the question in a sufficiently comprehensive and yet practical manner to be of real use to the mechanics 
 or mining engineer ; either the treatment was adapted for specialists, or it was fragmentary, and power wor 
 was regarded as subservient to the question of lighting. The Author has felt the want of such a book in dealin 
 with his clients and others, and in " ELECTRIC MOTIVE POWER " has endeavoured to supply it. 
 
 In the introduction the limiting conditions and essentials of a power plant are analysed, and in th 
 subsequent chapters the power plant is treated synthetically. The dynamo, motor, line, and details ar 
 discussed both as to function and design. The various systems of transmitting and distributing power by cor 
 tinuous and alternate currents are fully enlarged upon, and much practical information, gathered from actiu 
 experience, is distributed under the various divisions. The last two chapters deal exhaustively with th 
 applications of electricity to mining work in Great Britain, the Continent, and America, particularly wit 
 reference to collieries and coal-getting, and the results of the extensive experience gained in this field ar 
 embodied. 
 
 In general, the Author's aim has been to give a sound digest of the theory and practice of the electric; 
 transmission of power, which will be of real use to the practical engineer, and to avoid controversial poin 
 which lie in the province of the specialist, and elementary proofs which properly belong to text-books o 
 electricity and magnetism. 
 
 NEW VOLUMES IN PREPARATION. 
 
 SUBMARINE CABLE-LAYING AND REPAIRING 
 
 By H. D. WILKINSON, M.I.E.E., &c., fec. 
 
 This work will describe the procedure on board ship when removing a fault or break in a submerged cal 
 and the mechanical gear used in different vessels for this purpose ; and considers the best and most rece 
 practice as regards the electrical tests in use for the detection and localisation of faults, and the varic 
 difficulties that occur to the beginner. 
 
 MOTIVE POWER AND GEARINt 
 
 FOR E^ECTRICAIL, MACMIISTERY. 
 
 BY E. TEEMLETT CARTER, C.E. 
 
 (COPIOUSLY ILLUSTRATED WITH SCALE DRAWINGS d> NUMEROUS PLATES., 
 
 The purpose of this work is the explanation of the principles and practice of modern mechanical moti 
 power and gearing, especially in their application to electrical machinery. Electrical engineering is as much, 
 matter of engines and gearing as of dynamos and cables ; but the conditions of electric light and power distrilr 
 tion are such that a special study of the mechanical plant is necessary. Just as marine or locomotive stef 
 practice is treated in a special manner in works on the subject ; so the Author has endeavoured to hold 
 view the special requirements of electrical practice, and to produce a work on steam and other motive pow 
 which shall be solely devoted to these requirements. 
 
 "MOTIVE POWER AND GEARING" is adapted equally to the needs of the practical engineer and of t 
 student, and the treatment is such as may be easily understood without special mathematical trainir 
 Besides steam plant, as used in electric power stations, the work treats of gas, oil, and water-power engines, a 
 the chapters on these, as well as the section on Gearing, are written on the lines of the latest practice in electi 
 power stations. The best points in the development of motive power for electrical engineering on the Contine 
 and in the United States have also been considered, and are fully treated, and compared with English practic 
 This work constitutes the only existing treatise on the Economics of Motive Power and Gearing for Electric 
 Machinery. 
 
 1, 2, and 3, Salisbury Court, Fleet Street, London, E.G. 
 
The Electrician" Printing and Publishing Co., Ltd., 15 
 "THE ELECTRICIAN" SERIES continued. 
 
 NOW READY. Price 1. 5s., in Sepia or Black; in very Massive Frame, 2. 2s. 
 A Handsome Plate Eeproduction of ROBERT DUDLEY'S famous Painting of 
 
 THE "GREAT EASTERN," 
 
 (By permission of the Executors of the late Sir JAMES ANDERSON.) 
 
 The tirst of a series of Fine Art Plates which it is proposed to issue from time to time, and executed in tht 
 best style, and by the latest and best process of faithful reproduction. The Plate has a high historic value ; it 
 portrays the great ship at the critical moment when, in endeavouring to pick up the 1865 Atlantic Cable, " the 
 picking-up tackle having given out," her Captain gives orders for her return to England. 
 
 At the foot of the picture is a lithographed extract from the letter of Captain (afterwards Sir James) 
 Anderson to the owners of the ship, and afac simile signature. 
 
 The subject measures 24in. by 17in., and is India mounted on fine etching boards, the mount measuring 
 27in. by 21in. The entire plate measures 36in. by 27in. 
 
 Price 10s. 6d., in very Massive Frame, 1. Is. 
 A Handsome Reproduction of HERKOMER'S famous Painting of 
 
 SIR JOHN FENDER, G.C.M.G.,M.P. 
 
 Mounted India on fine etching boards. 
 :ind the whole plate 2Siin. by 
 
 The subject measures 15in. by llin., the mount 18iin. by 
 
 Price 10s. 6d., in very Massive Frame, 1. Is. 
 A Handsome Plate Reproduction of WILLIAM SYMOXDS' fine Painting of 
 
 SIR JAMES ANDERSON, J.P. 
 
 Mounted India on fine etching boards. The subject measures 15in. by llin., the mount 18$in. by 14$in., 
 and the whole plate 282in by 21|in. 
 
 OTHER REPRODUCTIONS WILL FOLLOW SHORTLY. 
 
 STEEL-PLATE PORTRAITS 
 
 WILLOUGHBY SMITH (out of print), 
 MICHAEL FARADAY (Is. extra), 
 SIR JOHN FENDER, G.C.M.G., M.P., 
 SIR WILLIAM THOMSON, F.R.S. 
 
 (LORD KELVIN), 
 (Past-President of the Institution of Electrical 
 
 Engineers), 
 C. H. B. PATEY, C.B. 
 
 (Late Secretary of Telegraphs, G.P.O.), 
 DR. JOHN HOPKINSON, F.R.S. 
 
 (Past-President of the Institution of Electrical 
 
 Engineers), 
 
 DR. OLIVER J. LODGE, F.R.S., 
 PROF. WILLIAM CROOKES, F.R.S. 
 
 (Past-President of the Institution of Electrical 
 Engineers) 
 
 HERMANN VON HELMHOLTZ, 
 
 PROF. W. E. AYRTON 
 
 (Past-President of the Institution of Electrical 
 Engineers), 
 
 LORD RAYLEIGH, F.R.S., 
 
 CYRUS W. FIELD, 
 
 W. H. PREECE, C.B., F.R.S. 
 
 (Past-President of the Institution of Electrical 
 Engineers). 
 
 WERNER VON SIEMENS, 
 SIR JAMES ANDERSON, J.P., 
 JOSEPH WILSON SWAN, 
 ALEXANDER SIEMENS 
 
 (President (1894) of the Institution of Electrical 
 Engineers), 
 
 HAVE 
 
 AND HEINRICH HERTZ, 
 
 BEEN ISSUED AS SUPPLEMENTS TO 
 
 A FEW COPIES OF MOST OF THESE ADMIRABLY-EXECUTED STEEL ENGRAVINGS CAN BE SUPPLIED, 
 Price ONE SHILLING EACH ; post free on Roller, Is. 2d. India mounts, Is. extra. 
 
 Or framed in neat Black Pillar or Brown Ornamental Frames, price 
 FOUR SHILLINGS AND SIXPENCE; carriage paid (U.K.), FIVE SHILLINGS. 
 
 // with mount, Is. extra. 
 
 1, 2, and 3, Salisbury Court, Fleet Street, London, E.G. 
 
16 " The Electrician " Printing and Publishing Co., Ltd., 
 
 ELECTRICITY AND MAGNETISM. 
 
 AN ELEMENTARY TREATISE ON FOURIER'S SERIES, and Spherical, 
 
 Cylindrical and Ellipsoidal Harmonics, with Applications to Problems on Mathematical Physics. By 
 Prof. Byerly, Harvard University. 12s. 6d. 
 
 MODERN VIEWS OF ELECTRICITY. By Oliver J. Lodge, F.R.S., Professor of 
 
 Physics in University College, Liverpool. Illustrated. 6s. 6d. 
 
 THE ELECTRO-MAGNET AND ELECTROMAGNETIC MECHANISM. By 
 
 Silvanus P. Thompson, D.Sc., F.R.S. 450 pages, 213 illustrations. 15s. 
 
 ELECTRICITY: ITS THEORY, SOURCES AND APPLICATIONS. By 
 
 John T. Sprague. Third Edition. Revised and enlarged. 15s. 
 
 MODERN APPLICATIONS OF ELECTRICITY. By E. Hospitalier. Trans- 
 
 lated by Julius Maier, Ph.D. Second Edition. 28s, 
 
 ELECTRICAL PAPERS. For Advanced Students in Electricity. By Oliver 
 
 Heaviside. 2 vols. 31s. 6d. net. 
 
 A COURSE OF LECTURES ON ELECTRICITY, DELIVERED BEFORE THE 
 
 SOCIETY OF ARTS. By George Forbes, M.A., F.R.S. (L. & E.) With 17 illustrations, crown 8vo. 5s. 
 
 SHORT LECTURES TO ELECTRICAL ARTIZANS. By Dr. J. A. Fleming, 
 
 M.A., F.R.S., &c. Fourth Edition. 4s. 
 
 LECTURES IN ELECTRICITY AT THE ROYAL INSTITUTION, 1875-76 
 
 By John Tynclall, D.C.L. 2s. 6d. 
 
 NOTES OF A COURSE OF SEVEN LECTURES ON ELECTRICAL 
 
 PHENOMENA. By John Tyndall. Is. Cd. 
 
 ELECTRIC WAVES : Being Researches on the Propagation of Electric Action 
 
 with Finite Velocity through Space. By Dr. Heinrich Hertz. Translated by D. E. Jones. 10s. 6d. net. 
 
 LECTURES ON SOME RECENT ADVANCES IN PHYSICAL SCIENCE. 
 
 By Prof. P. G. Tait. Third Edition. 9s. 
 
 ELEMENTARY LESSONS IN ELECTRICITY AND MAGNETISM. By Sil- 
 vanus P. Thompson, Principal and Professor of Physics in the Technical College, Finsbury. With illus- 
 trations. Fcap. 8vo, 4s. 6d. 
 
 PRACTICAL ELECTRICITY. For First Year Students of Electrical Engineering. 
 
 By W. E. Ayrton, F.R.S., Professor of Electrical Engineering in the City and Guilds of London Centra: 
 Institution. Sixth Edition. Illustrated throughout, 7s. 6d. 
 
 A TREATISE ON ELECTRICITY AND MAGNETISM. By J. Clerk Max 
 
 well, M.A., F.R.S. Third Edition. 2 vols., demy 8vo, cloth, 1. 12s. 
 
 RECENT RESEARCHES IN ELECTRICITY AND MAGNETISM. By 
 
 Prof. J. J. Thomson, M.A., F.R.S. 18s. 6d. 
 
 MAGNETISM AND ELECTRICITY, AN ELEMENTARY MANUAL OF 
 
 With Examination Questions and many illustrations. By Prof. Jamieson. Third Edition. Crown Svo, 3s. 6d 
 
 ELECTRICITY AND MAGNETISM. By Prof. Balfour Stewart, F.R.S., anc" 
 
 W. W. Haldane Gee, Demonstrator and Assistant Lecturer in Owens College, Manchester. Crown Svo 
 7s. 6d. ; School Course, 2s. 6d. 
 
 AN ELEMENTARY TREATISE ON ELECTRICITY. By J. Clerk Maxwell 
 
 M.A., F.R.S. Edited by William Garnett, M.A. Demy Svo, cloth, 7s. 6d. 
 
 INVENTIONS, RESEARCHES, AND WRITINGS OF NIKOLA TESLA. 
 
 Edited by T. Commerford Martin. 16s. 6d. 
 
 MAGNETISM AND ELECTRICITY. By A. W. Poyser, M.A. Cloth, 2s. 6d. 
 FIRST BOOK OF ELECTRICITY AND MAGNETISM. By W. Perren May 
 
 cock. 84 illustrations. 2s. 6d. 
 
 MANUAL OF ELECTRICAL SCIENCE. By George J. Burch, B.A. 3s. 
 ELECTROMAGNETIC THEORY see page 4. 
 
 1, 2, and 3, Salisbury Court, Fleet Street, London ; E.G. 
 
 
" The Electrician " Printing and Publishing Co., Ltd., 17 
 THE ALTERNATE-CURRENT TRANSFORMER IN THEORY AND 
 
 PEACTICE -m> pa<fc* <: ami 7. 
 
 A.B.C. OF ELECTRICITY. By W. H. Meadowcroft. 2s. 
 
 MAGNETISM AND ELECTRICITY. For the use of students in schools and 
 
 science classes. By H. C. Tarn. F.S.Sc. With numerous diagrams. Cloth, 2s. 
 
 ELEMENTS OF STATIC ELECTRICITY. By P. Atkinson, Ph.D. 7s. 6d. 
 MAGNETISM AND ELECTRICITY. By R. Wallace Stewart. 160 Illustrations 
 
 5s. 6d. 
 
 MAGNETISM AND ELECTRICITY. By Edward Aveling, D.Sc. With 
 
 numerous woodcuts. 6s. 
 
 ELECTRICITY TREATED EXPERIMENTALLY, for the use of schools and 
 
 students. By Linnaeus Cumming, M.A. 4s. 6d. 
 
 ELECTRICITY AND MAGNETISM. By S. R. Bottone. 3s. 6d. 
 
 BOOK E. ARITHMETICAL PHYSICS. Part HA. MAGNETISM AND ELEC- 
 
 TRICITY, ELEMENTARY ANT) ADVANCED. With Supplement on Lines of Force. By C. J. Wood- 
 ward, B.Sc. 2s. 
 
 BOOKF. ARITHMETICAL PHYSICS. Part HB. MAGNETISM AND ELEC 
 
 TRICITY, DEGREE AND HONOURS STAGES. By C. J. Woodward, B.Sc. New Edition. 3s. 6d. 
 
 MAGNETIC INDUCTION IN IRON AND OTHER METALS see page 4. 
 
 ELEMENTS OF EXPERIMENTAL AND NATURAL PHILOSOPHY. By 
 
 Jabez Hogg. More than 400 woodcuts. 5s. 
 
 MAGNETISM AND ELECTRICITY. By W. Jerome Harrison and Charles 
 
 A. White. 2s. 
 
 MAGNETISM AND ELECTRICITY FOR BEGINNERS. By W. G. Baker, 
 
 M.A. Is. 
 
 ELECTRICITY. By Dr. Ferguson : revised and extended by Prof. James Blytb, 
 
 M.A. 3s. 6d. 
 
 MANUAL OF MAGNETISM AND ELECTRICITY. By John Cook, M.A Is. 
 MAGNETISM AND ELECTRICITY. By John Angell. Is. 6d. 
 MAGNETISM AND ELECTRICITY. By F. Guthrie, B.A., Ph.D. 3s. 6d, 
 ELECTRICITY AND MAGNETISM. By Fleeming Jenkin. 3s. 6d. 
 ELECTRICITY FOR PUBLIC SCHOOLS AND COLLEGES. By W. Lardcnj 
 
 M.A. 6s. 
 
 ELEMENTARY TREATISE ON PHYSICS, Experimental and Applied. Trans- 
 lated from Ganot's " Elements de Physique," by E. Atkinson, Ph.D. Twelfth Edition. 15s. 
 
 PAPERS ON ELECTRO-STATICS AND MAGNETISM. By Lord Kelvin. 
 
 Second Edition. 18s. 
 
 PHYSICS. Advanced Course. By G. F. Barker. 21s. 
 
 A TEXT-BOOK OF THE PRINCIPLES OF PHYSICS. By Alfred Daniel 
 
 Second Edition. 21s. 
 
 ELECTRICITY^ AND MAGNETISM. A Popular Treatise. By Ame'de'e 
 
 Guillemin. Translated by Silvanus P. Thompson. 31s. 6d. 
 
 QUESTIONS AND EXAMPLES IN EXPERIMENTAL PHYSICS, SOUND, 
 
 LIGHT, HEAT, ELECTRICITY, AND MAGNETISM. By B. Loewy. 2s. 
 
 LESSONS IN ELEMENTARY" PHYSICS. By Prof. Balfour Stewart. 4s. 6d, 
 ELECTRICITY AND MAGNETISM. By L. Cumming, M.A. 2s, 6& 
 MAGNETISM AND ELECTRICITY. By J. Spencer, B.Sc. Is. 6cf. 
 STUDENT'S TEXT-BOOK OF ELECTRICITY. By Henry M. Noad, Ph.D., 
 
 F.R.S. New Edition, with introduction and additional chapters by W. H. Preece, C.B., F.R.S. 12s 6d* 
 
 HANDBOOK OF ELECTRICITY, MAGNETISM, AND ACOUSTICS. By 
 
 Dr. Lardner. Edited by Geo. Carey Foster, B.A. 5s. 
 
 PRACTICAL NOTES FOR ELECTRICAL STUDENTS see page 13. 
 1, 2 and 3, Salisbury Court, Fleet Street, London, E.G. 
 
18 "The Electrician" Printing and Publishing Co., Ltd., 
 
 ELECTRICAL ENGINEERING FORMULAE see page 9. 
 INTEODUCTION TO THE THEORY OF ELECTRICITY. By Linnaeus 
 
 dimming, M.B. 8s. 6d. 
 
 PHYSICAL TREATISE ON ELECTRICITY AND MAGNETISM. By J. E. 
 
 H. Gordon. 2 2s. 
 
 THE PRACTICAL MEASUREMENT OF ELECTRICAL RESISTANCE. By W 
 
 A. Price. 14s. 
 
 MATHEMATICAL THEORY OF ELECTRICITY AND MAGNETISM. Vol. 
 
 I., Electrostatics; Vol. II., Magnetism and Electrodynamics. By H. W. "Watson, D Sc., and S H. 
 Burbury, M.A. 10s. 6d. each. 
 
 MATHEMATICAL THEORY OF ELECTRICITY AND MAGNETISM, AN 
 
 INTRODUCTION TO. By W. T. A. Emtage, M.A. 7s. 6d, 
 
 ELECTRICITY : A Sketch for General Readers. By E. M. Caillard. 7s. 6d. 
 ELEMENTS OF DYNAMIC ELECTRICITY AND MAGNETISM. By 
 
 Philip Atkinson, A.M., Ph.D. 10s. 6d. 
 
 ALTERNATING CURRENTS OF ELECTRICITY. By T. H. Blakesley. 
 
 Third Edition. 5s. 
 
 .ALTERNATING CURRENTS OF ELECTRICITY : Their Generation, Measure- 
 ment, Distribution, and Application. By Qisbert Kapp. 4s. 6d. 
 
 POLYPHASED ALTERNATING CURRENTS. By E. Hospitalier. 3s. 6d. 
 'EXPERIMENTS WITH ALTERNATING CURRENTS OF HIGH POTENTIAL 
 
 AND HIGH FREQUENCY. By Nikola Tesla. 5s. 
 
 ALTERNATING CURRENTS. An Analytical and Graphical Treatment for 
 
 Students and Engineers. By Dr. F. Bedall and Dr. A. C. Crehore. Second Edition. 11s. 
 
 THE ARITHMETIC OF ELECTRICAL MEASUREMENTS. By W. R. P. 
 
 Hobbs, Head Master of the Torpedo School, H.M.S. " Vernon.' New Edition, Is. 
 
 THE ELECTRIC CURRENT : How Produced and How Used. By R. Mullineux 
 
 Walmsley, D.Sc. 10s. 6d. 
 
 "'THE ELECTRICIAN" PRIMERS see page 8. 
 
 JTHEORY AND PRACTICE OF ABSOLUTE MEASUREMENTS IN ELEC- 
 
 TRICITY AND MAGNETISM. By Andrew Gray, M.A., F.R.S.E., Professor of Physics in the University 
 <College of North Wales. In 2 vols., crown 8vo. Vol. L, 12s. 6d. Vol. II., in 2 parts, 25s. 
 
 ABSOLUTE MEASUREMENTS IN ELECTRICITY AND MAGNETISM. 
 
 By Prof. Andrew Gray. Second Edition. 5s. 6d. 
 
 PRACTICAL ELECTRICAL MEASUREMENT. By James Swinburne. 
 
 Cloth, 4s. 6d. 
 
 EXERCISES IN ELECTRICAL AND MAGNETIC MEASUREMENTS, with 
 
 Answers. By R. E. Day. 3s. 6d. 
 
 ARITHMETIC OF MAGNETISM AND ELECTRICITY. By. Robert Gunn. 
 /AN INTRODUCTION TO PHYSICAL MEASUREMENTS. By Dr. F. 
 
 Kohlrausch. Translated by T. H. Waller and H. R. Procter. 15s. 
 
 PHYSICAL ARITHMETIC. By Alexander Macfarlane. 7s. 6d. 
 
 REPORTS OF THE COMMITTEE ON ELECTRICAL STANDARDS, 
 
 Appointed by the British Association, with a Report to the Royal Society on Units of Electrical 
 Resistance. 9s. 
 
 'ELECTRICAL MEASUREMENT AND THE GALVANOMETER, ITS CON- 
 STRUCTION AND USES. By T. D. Lockwood. Second Edition. 6s. 
 
 THERMO-ELECTRICITY, Theoretically and Practically Considered. By Arthur 
 
 Rust. 2s. 
 
 ELECTRICITY IN THEORY AND PRACTICE. By Lieut, Bradly A. Fiske. 
 
 10s. 6d. 
 
 ELECTRICITY FOR ENGINEERS. By Charles Desmond. Revised Edition. 
 
 10s. 6d. 
 
 POTENTIAL : Its Application to the Explanation of Electrical Phenomena Popu- 
 larly Treated. By Dr. Tumlirz. Translated by D. Robertson, M.A. 3s. 6d. 
 
 1, 2 and 3 Salisbury Court, Fleet Street, London, E.G. 
 
"The Electrician" Printing and Publishing Co., Ltd., 19 
 
 KIRCHOFF'S LAWS AND THEIR APPLICATION. By E. C. Rimington. 
 
 Cloth, Is. 6d. 
 
 ELECTRICAL INFLUENCE MACHINES : containing a full account of their 
 
 Historical Development, their Modern Forms, and their Practical Construction. By John Gray, B.Sc. 
 89 illustrations. 4s. 6d. 
 
 ELECTRICITY IN THE SERVICE OF MAN. A Popular and Practical Treatise 
 
 on the Applications of Electricity in Modern Life. Revised by R. Mullineux Walmsley, ,D.Sc. (Lond.), 
 F.R.S.E. Medium 8vo., with nearly 850 illustrations. 10s. Cd. 
 
 ELECTRICITY IN MODERN LIFE. By G. W. de Tunzelmann. With 88 illus- 
 trations, 3s. 6d. 
 DOMESTIC ELECTRICITY FOR AMATEURS. Translated from the French of 
 
 E. Hospitalier, with additions By C. J. VVharton, Assoc. Soc. Tel. Engineers. Numerous illustrations, 
 demy Svo, cloth, 6s. 
 
 STANDARD METHODS IN PHYSICS & ELECTRICITY CRITICISED, 
 
 AND A TEST FOR ELECTRIC METERS PROPOSED. By H. A. Xaber. Demy Svo., cloth gilt, 
 5s., post free. 
 
 PRACTICAL ELECTRICS : A Universal Handy Book on Every -Day Electric 
 
 Matters. Third Edition. 3s. 6d. 
 
 ELECTRICITY : A Hundred Years Ago and To-Day. By Prof. E. J. Houston. 
 
 4s. 6d. 
 
 ELECTRICITY IN DAILY LIFE. A Popular Account of its Application to 
 
 Every-day Uses. 125 illustrations. 9s. 
 
 ELECTRICITY IN MINING. By Silvanus P. Thompson, D.Sc., F.R.S. 2s. 
 ELECTRICITY FOR SCHOOLS. By J. E. H. Gordon. 5s. 
 A CENTURY OF ELECTRICITY. By T. C. Mendenhall. 4s. 6d. 
 ELECTRICAL ENGINEERING LEAFLETS (HOUSTON AND KENNELLY'S). 
 
 In three Grades Elementary, Intermediate, and Advanced of 35 Leaflets each. Price of single Leaflet, 
 6d., post free. Subscription price for any one Grade, 12s. Cd. ; for any two Grades, 23s. ; for all three 
 Grades, 33s. 6d. Further particulars on application. 
 
 A HANDBOOK FOR OPERATORS IN MEDICAL ELECTRICITY AND 
 
 Massage. By H. Newman Lawrence, M.I.E.E. Is. 
 
 A PRACTICAL TREATISE ON THE MEDICAL AND SURGICAL USES 
 
 OF ELECTRICITY. By Drs. Beard and Rockwell. 200 illustrations. Royal Svo, 28s. 
 
 PRACTICAL APPLICATION OF ELECTRICITY IN MEDICINE AND 
 
 SURGERY. By Drs. Liebig and Rohe". Royal Svo, 400 pages, profusely illustrated, 11s. 6d. 
 
 ELECTRICITY IN THE DISEASES OF WOMEN. By G. Betton Massey, M.D. 
 
 Second Edition. 12mo, 8s. 6d. 
 
 THE WORK OF HERTZ. see page 9. 
 
 A TEXT-BOOK OF ELECTRICITY IN MEDICINE AND SURGERY. By 
 
 George Vivian Poore, M.D. Crown Svo, 8s. 6d. 
 
 A MANUAL OF ELECTRO-THERAPEUTICS. By W. Erb, M.D., translated 
 
 by A. de Watteville, M.D., &c. Demy Svo, 18s. 
 
 ELECTRICITY AND ITS MANNER OF WORKING IN THE TREATMENT 
 
 OF DISEASE. A Thesis for the M.D. Cantab Degree, 1884. By the late William E. Steavenson, M.D. 
 M.R.C.P., Casualty Physician and Electrician to St. Bartholomew's Hospital. To which is appended' an 
 Inaugural Medical Dissertation on Electricity for the M.D., Edin. Degree, written in Latin by Robert 
 Steavenson, M.D. in 1778, with a Translation by the Rev. F. R. Steavenson, M.A. 4s. 6d. 
 
 INTERNATIONAL SYSTEM OF ELECTRO-THERAPEUTICS. For Students 
 
 General Practitioners, and Specialists. 32s. 
 
 TRAITE LMENTAIRE D'^LECTRICITE. By j. joubert. with 321 
 
 illustrations. 6s 
 
 SUR LA PROPAGATION DU COURANT ^LECTRIQUE. By A. Bandsept. Is. 
 ^LECTRICITE INDUSTRIELLE. By D. Monnier. With 33 8 illustrations. 16s. 
 LEgONS SUR LE'LECTRICITE. By Eric Gerard. 2 vols., fully illustrated. 19s. 
 FORMCJLAIRE DE L'ELECTRICITE. By E. Hospitalier. 10th year: 1892. 
 
 4s. 6d. 
 
 TRAIT6 PRATIQUE DE L'ELECTRICITE. By Felix Lucas. 13s. 6d. 
 1, 2 and 3, Salisbury Court, Fleet Street, London, E.G. 
 
20 " The Electrician " Printing and Publishing Co., Ltd., 
 
 ELECTRIC LIGHTING TRANSMISSION OF POWER 
 
 THE ALTERNATE-CURRENT TRANSFORMER IN THEORY AND 
 
 PRACTICE -see pages 6 and 7. 
 
 DYNAMO-ELECTRICITY : Its Generation, Application, Transmission, Storage 
 
 and Measurement. By G. B. Prescott. 545 illustrations. 1. Is. 
 
 "THE ELECTRICIAN" PRIMERS see page 8. 
 
 DYNAMO MACHINERY AND ALLIED SUBJECTS (ORIGINAL PAPERS 
 
 ON). By Dr. John Hopkinson. 5s. 
 
 HOW TO BUILD DYNAMO-ELECTRIC MACHINERY. By Edward Trevert. 
 
 10s. 6d. 
 
 PRACTICAL NOTES FOR ELECTRICAL STUDENTS-^ page IS. 
 CONTINUOUS-CURRENT DYNAMOS AND MOTORS. An Elementary 
 
 Treatise for Students. By Frank P. Cox. 9s. 
 
 THE INCANDESCENT LAMP AND ITS MANUFACTURE-*^ w * 10. 
 THE DYNAMO : Its Theory, Design and Manufacture. By C. C. Hawkins, 
 
 A.M.Inst.C.E., and F. Wall's. 10s. 6d. 
 
 ARMATURE WINDINGS OF ELECTRIC MACHINES. By H. F. Pai shall 
 
 and H. M. Hobart. In preparation. 
 
 DRUM ARMATURES AND COMMUTATORS-^ page 10. 
 PRACTICAL MANAGEMENT OF DYNAMOS AND MOTORS. By F. B. 
 
 Crocker and S. S. Wheeler. Second Edition. 4s. 6d. 
 
 DYNAMO ATTENDANTS AND THEIR DYNAMOS. By Alfred H. Gibbings. Is. 
 
 ELECTRICAL ENGINEERING FORMULAE see page 9. 
 
 DYNAMO AND MOTOR BUILDING FOR AMATEURS. By C. D. 
 
 Parkhurst. 4s. 6d. 
 
 A DIGEST OF THE LAW OF ELECTRIC LIGHTING, &c. seepage 8. 
 
 ELECTRO-MAGNETISM AND THE CONSTRUCTION OF DYNAMOS. 
 
 Vol. I. By Prof. Dugald C. Jackson. 9s. 6d. 
 
 THE MANUFACTURE OF ELECTRIC LIGHT CARBONS see page 13. 
 PRACTICAL DIRECTIONS FOR WINDING MAGNETS FOR DYNAMOS. 
 
 By Carl Hering. 3s, 6d. 
 
 ELEMENTS OF CONSTRUCTION FOR ELECTRO-MAGNETS. By Count du 
 
 Moncel. Translated by C. J. Wharton. 4s. 6d. 
 
 THE STEAM ENGINE INDICATOR, AND INDICATOR DIAGRAMS 
 
 seepage 13. 
 
 ELEMENTARY PRINCIPLES OF ELECTRIC LIGHTING. By Alan A. Camp- 
 
 bell Swinton. Enlarged and revised. Crown 8vo, cloth, Is. 6d. 
 
 ELECTRIC LIGHTING AND POWER DISTRIBUTION: An Elementary 
 
 Manual for Students. By W. Perren Maycock. In three parts. 2s. 6d. each. 
 
 DYNAMO CONSTRUCTION : A Practical Handbook for Engineer Constructors 
 
 and Electricians in Charge. By John W. Urquhart. With numerous Diagrams and Woodcuts. 7s. Gd. 
 
 A GUIDE TO ELECTRIC LIGHTING. By S. R. Bottone. Second Edition. Is. 
 CONTINENTAL ELECTRIC LIGHT CENTRAL STATIONS. By KilHngworth 
 
 Hedges. 15s. 
 
 DYNAMO-ELECTRIC MACHINERY : A Text-Book for Students of Electro- 
 
 Technology. By Silvanus P. Thompson, B.A., D.Sc., M.I.E.E., F.R.S. Just published. Fourth 
 Edition, revised and enlarged. Cloth, 864 pages, 29 folding plates 498 illustrations in text. 24s. post 
 free. 
 
 THEORETICAL ELEMENTS OF ELECTRO -DYNAMIC MACHINERY. 
 
 Vol. I. By A. Kennelly, F.R.A.S. 4s. ed. 
 
 ELECTRIC LIGHTING FROM CENTRAL STATIONS. By Prof. George 
 
 Forbes. Is. 
 
 THE DYNAMO-TENDER'S HANDBOOK. By F. B. Badt. 4s. 6d. 
 1, 2 and 3, Salisbury Court, Fleet Street, London, E.G. 
 
"The Electrician" Printing and Publishing Co., Ltd., 21 
 PRINCIPLES OF DYNAMO-ELECTRIC MACHINES. By Carl Hering. 
 
 Practical Directions for Designing and Constructing Dynamos. With an Appendix containing several 
 Articles on Allied Subjects, and a Table of Equivalents of Units of Measurement. Cloth. 279 pages, 
 59 illustrations, 10s. 6d. 
 
 CENTRAL STATION MANAGEMENT AND FINANCE. By Horatio A. 
 
 Foster. 7s. 6d. 
 
 CENTRAL STATION BOOK-KEEPING. By Horatio A. Foster. 7s. 6d. 
 NOTES ON DESIGN OF SMALL DYNAMOS. By George Halliday. 2s. 6d. 
 ELECTRICITY IN OUR HOMES AND WORKSHOPS. By S. F. Walker. 
 
 Second Edition. Cloth, 5s. 
 
 ELECTRIC LIGHT INSTALLATIONS. By Sir David Salomons, Bart. In 
 
 Three Volumes. Vol. I., 5s. : The Management of Accumulators. Vol. II., 7s. 6d.: Apparatus, Engines, 
 Dynamos and Motors, Instruments, Governors, Switches and Switch Boards, Fuses, Cut-Outs, Connec- 
 tors and Minor Apparatus, Ate Lamps, Practical Applications. Vol. III., 5s. : Application. 
 
 HOUSE LIGHTING BY ELECTRICITY. By Angelo Fahie. Paper, Is.; cloth, 2s. 
 
 DOMESTIC ELECTRIC LIGHTING. By E. C. de Segundo. Is. 
 
 A HANDBOOK OF ELECTRICAL TESTING. By H. R. Kempe, M.I.E.E. 
 
 Fifth Edition, revised and enlarged. 18s. 
 
 ARTIFICIAL LIGHTING IN RELATION TO HEALTH. By R. E. B. 
 
 Crompton. 6d. 
 
 ELECTRIC LIGHTING SPECIFICATIONS, for the Use of Engineers and 
 
 Architects. By E. A. Merrill. Cs. 
 
 ELECTRICAL ENGINEERING: For Electric Light Artizans and Students. 
 
 By W. Slingo and A. Brooker. 640 pages, 307 illustrations. Cloth, gilt, ICs. 6d. 
 
 PRECAUTIONS TO BE ADOPTED ON INTRODUCING THE ELECTRIC 
 
 LIGHT. By Killingworth Hedges. 2i. 6d. 
 
 ECONOMIC VALUE OF ELECTRIC LIGHT AND POWER. By A. R. 
 
 Foote. 4s. 
 
 ELECTRIC LIGHT : Its Production and Use. By John W. Urquhart. Fifth 
 
 Edition, carefully revised, with large additions. 7s. 6d. 
 
 RULES AND REGULATIONS RECOMMENDED FOR THE PREVENTION 
 
 OF FIRE RISKS FROM ELECTRIC LIGHTING, issued by the Society of Telegraph-Engineers. 8vo. 
 sewed, 6d. 
 
 THE PHCENIX FIRE OFFICE RULES FOR ELECTRIC LIGHT INSTAL- 
 LATIONS AND ELECTRICAL POWER INSTALLATIONS. By Musgrave Heaphy, C.E. Eighteenth 
 Edition, 8vo, sewed, 6d. 
 
 ELECTRIC SHIP LIGHTING : For the use of Ship Owners and Builders, 
 
 Engineers, &c. By John W. Urquhart. 7s. 6d. 
 
 ELECTRIC LIGHTING FOR MARINE ENGINEERS. By S. F. Walker. 5s. 
 ELECTRIC LIGHT FITTING : A Handbook for Electrical Engineers. By John 
 
 W. Urquhart. Fully Illustrated. 5s. 
 
 COMPREHENSIVE INTERNATIONAL WIRE TABLE. By W. S. 
 
 Boult. Full particulars of 469 Conductors (4 gauges), Single Wires and Cables, in English, American, 
 and Continental Units. Price 6s. 9d., post free. 
 
 COLLIERY LIGHTING BY ELECTRICITY. By S. F. Walker. 2s. 6d. 
 
 A PRACTICAL TREATISE ON THE INCANDESCENT LAMP. By J. H. 
 
 Randell. 2s. 6d. 
 
 ELECTRIC LAMPS AND ELECTRIC LIGHTING seepage 5. 
 PRACTICAL ELECTRIC LIGHT FITTING. By F. C. Allsop. 5s. 
 INCANDESCENT ELECTRIC LIGHTING. By L. H. Latimer. With additions 
 
 by C. J. Field and J. W. Howell. 2s. 
 
 TREATISE ON INDUSTRIAL PHOTOMETRY, with Special Application to 
 
 Electric Lighting. By Dr. A. Palaz. Translated by G. W. and M. R. Patterson. 12s. 6d. 
 
 WIRING SLIDE RULE (Trotter's Patent). By which can be found at once . 
 
 1. Size of Cable; 2. Length of Cable ; 3. Current Cable will carry; 4. Current Density; 5. Maximum 
 Current; 6, Resistance in Ohms; 7. Sectional Area in Square inchep. Full printed instructions are 
 supplied with each rule. For the pocket. Trice 2*. ed.; post free, 2s. 7d. 
 
 1, 2 and 3, Salisbury Court, Fleet Street, London, E.G. 
 
22 "The Electrician" Printing and Publishing Co., Ltd., 
 WOODHOUSE AND RAWSON WIRING TABLES. Price Paper, Is., 
 
 post free; mounted and glazed. 1?. 6d., post free; in neat cloth case for pocket, 2s. 6d., post free; 
 mounted and glazed and bound in cloth for pocket, 2s. Gd., post free. Printed directions how to use 
 the Tables are issued with each copy. 
 
 THE DISTRIBUTION OF ELECTRICITY. By Prof. George Forbes, M.A., 
 
 F.K.S.E. Is. 
 
 DEVELOPMENTS OF ELECTRICAL DISTRIBUTION. By Prof. George 
 
 Forbes. F.R.S. Is. 
 
 MAY'S POPULAR INSTRUCTOR FOR THE MANAGEMENT OF 
 
 ELECTRIC LIGHTING PLANT. An indispensable Handbook for persons in charge of Electric Lighting 
 plants, more particularly those who have had little or no technical training. Pocket size, price 2s. 6d. ; 
 post free, 2s. 8d. 
 
 PRACTICAL ELECTRIC LIGHTING. By A. Bromley Holmes. Fourth 
 
 Edition. Cloth, 3s. 6d. 
 
 STANDAED TABLES FOR ELECTRIC WIREMEN. By Charles M. Davis. 5s. 
 MAY'S TABLE OF ELECTRIC CONDUCTORS. Showing the relations 
 
 between (1) The sectional area, diameter of conductors, loss of potential, strength of current, and length 
 of conductors ; (2) The economies of incandescent lamps, their candle-power, potential, and strength of 
 current; (3) The sectional area, diameter of conductors, and strength of current per square inch. For 
 office use, printed on cardboard, with metal edges and suspender, price 2s. ; post free, 2s. 2d. ; for the 
 pocket, mounted on linen, in strong case, 2s. 6d.; post free, 2s. 8d. 
 
 ELECTRIC LIGHT ARITHMETIC. By R. E. Day. 2s. 
 
 UNIVERSAL WIRING COMPUTER. By Carl Bering. 5s. 
 
 MAY'S BELTING TABLE. Showing the relations between (1) The number 
 
 of revolutions and diameter of pulleys and velocity of belts ; (2) The horse-power, velocity, and square 
 section of belts ; (3) The thickness and width of belts ; (4) The square section of belts at different strains 
 per square inch. For office use, printed on cardboard, with metal edges and suspender, price 2s.; post 
 free, 2s. 2d. ; for the pocket, mounted on linen, in strong case, 2s. Cd.; post free, '2s. Sd. 
 
 THE GALVANOMETER AND ITS USES. A Manual for Electricians and 
 
 Students. By C. H. Haskins. Second edition. Illustrated. 18mo.,8s. 6d. 
 
 TRANSFORMERS: Their Theory, Construction, and Application Simplified. 
 
 By Caryl D. Haskyns. 4s. ed. 
 
 INCANDESCENT WIRING HANDBOOK. With Tables. By F. B. Badt. 
 HISTORY OF THE TRANSFORMER. By J. Uppenbcrn. Translated from the 
 
 German. 3s. 
 
 ELECTRIC TRANSMISSION OF ENERGY, and its Transformation, Sub- 
 division and Distribution. By Gisbert Kapp. Fourth Edition. 7s. 6d. 
 
 ELECTRIC LIGHT CABLES AND THE DISTRIBUTION OF ELEC- 
 TRICITY. By Stuart A. Russell, Assoc.M.Inst.C.E. 107 illustrations 7s. Gd. 
 
 ELECTRICITY AS A MOTIVE POWER. By Count du Moncel and Frank 
 
 Geraldy. Translated by C. J. Wharton. 7s. 6d. 
 
 ELECTRO-MOTORS : How Made and How Used. By S. R. Bottone. Second 
 
 Edition. 3s. 
 
 ELECTRIC MOTIVE POWER- see page 14. 
 
 THE ELECTRIC LIGHT POPULARLY EXPLAINED. By A. Bromley 
 
 Hoimes. Sixth Edition, Is. 
 
 HOW TO MAKE A DYNAMO. By Alfred Crofts. Fouith Edition. 2s. 
 
 MOTIVE POWER AND GEARING seepage 14. 
 
 THE DYNAMO : HOW MADE AND HOW USED. By S. R. Bottone. With 
 
 39 illustrations Eighth Edition. 2s. 6d. 
 
 HOW TO MANAGE A DYNAMO. By S. R. Bottone. Is. 
 ELECTRIC TRANSMISSION HANDBOCK. By F. B. Badt. 4s. 6d. 
 AMERICAN ELECTRIC STREET RAILWAYS : Their Construction and 
 
 Equipment. By Killingworth Hedges. 12s. 6d. 
 
 ELECTRIC RAILWAYS, RECENT PROGRESS IN. By Carl Hering. 5s. 
 SECONDARY BATTERIES AND THE ELECTRICAL STORAGE OF ENERGY, 
 
 By Dr. Oliver Lodge. Is. 
 
 1, 2 and 3, Salisbury Court, Fleet Street, London, E.C, 
 
"The Electrician" Printing and Publishing Co,, Ltd., 23 
 
 ELECTRIC TRANSFORMATION OF POWER. By Philip Atkinson. 7s. 6d. 
 ELECTRO-MOTORS : The Means and Apparatus employed in the Transmission 
 
 of Electrical Energy and its Conversion into Motive Power. By John W. Crquliait. 7s. 6d. 
 
 THE CHEMISTRY OF THE SECOND AEY BATTERIES OF PLANTS AND 
 
 FAURE. By J. H. Gladstone- and A. Tribe. 2s. 6d. 
 
 THE STORAGE OF ELECTRICAL ENERGY, and Researches in the Effects 
 
 c-eated by Currents Combining Quantity with High Tension. By G. Plants. Translated from the 
 French by Paul Bedford Ehvell. With 89 illustrations, 8vo, cloth, 12s. 
 
 THE ELECTRIC MOTOR AND ITS APPLICATIONS. By T. C. Martin 
 
 and J. Wetzler. Greatly enlarged edition. Quarto, 300 pages, 275 illustrations. 12s. 6d., post free 13s. 6d. 
 
 ELECTRIC BATTERIES, ELEMENTARY TREATISE ON. By Alfred Niaudet, 
 
 Translated by L. M. Fishback. -Sixth Edition, 12s. 6d. 
 
 DOMESTIC ELECTRICITY FOR AMATEURS. By E. Hospitalier. Translated 
 
 by C. J. Wharton. 6s. 
 
 THE ELECTRIC RAILWAY IN THEORY AND PRACTICE. By Oscar 
 
 T. Crosby and Louis Bell, Ph.D. Fully illustrated. 11s. 
 
 HOW TO WIRE BUILDINGS. By Augustus Noll. Second Edition. 6s. 
 PRACTICAL ELECTRICAL NOTES AND DEFINITIONS, for the use o 
 
 Engineering Students and Practical Men. By W. Perren Maycock. Second Edition. Es. 
 
 PRIMARY BATTERIES. By H. S. Carhart, A.M. 67 illustrations. 6s. 
 ELECTRICAL ENGINEERING LEAFLETS (HOUSTON AND KENNELLY'S). 
 
 In three Grades Elementary, Intermediate, and Advanced of 35 Leaflets each. Price of single Leaflets, 
 6d., post free. Subscription price for any one grade, 12s. 6d. ; for any two Grades, 23s. ; for all three 
 Grades, 33s. Gd. Further particulars on application. 
 
 LECONS SUR L'ELECTRICITE. By Eric Gerard. 19s. 
 
 L'ECLAIRAGE A PARIS. By H. Marechal. 16s. 6d. 
 
 LES ACCUMULATEURS ELECTRIQUES. By A. Bandsept. 103 pages. 2s. 
 
 INDUCTEURS DYNAMO-ELECTRIQUES ET PYRO-^LECTRIQUES. By 
 
 A. Bandsept. 6d. 
 
 TRAITE DES PILES ELECTRIQUES ; Piles hydro-electriques Accumula- 
 
 teurs Piles thermo-e'lectriques et pyro-electriques. By Donato Tommasi, Docteur-es-Sciences. 8vo., 
 670 pages, with 160 illustrations. 10s. 
 
 TRAITE tiLJEMENTAIRE DE LA PILE ^LECTRIQUE. By Alfred Niaudet, 
 
 Third Edition. Revised by Hippolyte Fontaine, and followed by a Notice ou Accumulators by E. Hos- 
 pitalier. Illustrated, Svo, 6s. 6d. 
 
 ECLAIRAGE A L'ELECTPJCIT E. By Hippolyte Fontaine. Third Edition, 
 
 entirely re-written, with 326 illustrations, Svo, 13s. 
 
 DIE DYNAMOELEKTRISCHE MASCHINE : Eine Physikalische Beschreibung 
 
 fur den Technischen Gebrauch. By Dr. 0. Frolich. Ss. 
 
 UNTERSUCHUNGEN UEBER DIE AUSBREITUNG DER ELEKTPJSCHEN 
 
 KRAFT. By Prof. Dr. H. Hertz. Price 6s. 
 
 TELEGRAPHY AND TELEPHONY. 
 
 PRACTICAL NOTES FOR ELECTRICAL STUDENTS seepage 13. 
 HANDBOOK OF PRACTICAL TELEGRAPHY. By R. S. Culley, M.Inst.C.E, 
 
 Eighth Edition. 17 plates and 135 woodcuts, Svo, 16s. 
 
 ELECTRICITY AND THE ELECTRIC TELEGRAPH. With numerous illus- 
 trations. By George B. Prescott Eighth Edition, 2 vols., Svo, cloth, 1. 10s. 
 
 ELECTRICAL ENGINEERING FORMULAE seepage 9. 
 
 THE ELECTEIC TELEGRAPH : Its History and Progress, With Descriptions 
 
 of some of the Apparatus. By R. Sabine, C.E., F.S.A. Limp cloth, 3s. 
 
 1, 2 and 3, Salisbury Court, Fleet Street, London, E.G. 
 
24 "The Electrician" Printing and Publishing Co., Ltd., 
 
 SUBMARINE CABLE LAYING AND REPAIRING see page 14. 
 
 RISE AND EXTENSION OF SUBMARINE TELEGRAPHY. By Willoughby 
 
 Smith. 21s. 
 
 COMMERCIAL AND RAILWAY TELEGRAPHY, MODERN SERVICE OF, 
 
 Designed for Students and Operators. Compiled and prepared by J. P. Abernethy. Seventh edition, 
 carefully revised, 427 pages, many illstrations, 8s. 6d. 
 
 MODERN PRACTICE OF THE ELECTRIC TELEGRAPH. By Frank L. Pope. 
 
 Fourteenth Edition, with numerous wood engravings, Svo, cloth, 12s. 6d. 
 
 HANDBOOK OF THE ELECTROMAGNETIC TELEGRAPH. By A. E. 
 
 Loring. 2s. 
 
 "THE ELECTRICIAN" PRIMERS-^ page 8. 
 
 TELEGRAPHIC CONNECTIONS : Embracing Recent Methods in Quadruplex 
 
 Telegraphy. By Charles Thorn and W. H. Jones. 7s. 6d. 
 
 PRACTICAL GUIDE TO THE TESTING OF INSULATED WIRES AND 
 
 CABLES. By Herbert Laws Webb. 4s. 6d. 
 
 INSTRUCTIONS FOR TESTING TELEGRAPH LINES, and the Technical 
 
 At rangements in Offices. By Louis Schwendler. 2 vols. Demy Svo, 21s. 
 
 TELEGRAPHY. By W. H. Preece, C.B., F.R.S., M.I.C.E., and Sir J. 
 
 Sivewright, K.C.M.G., M.A. Ninth Edition, revised and enlarged. 6s. 
 
 TABLES TO FIND THE WORKING SPEED OF CABLES, comprising also 
 
 Data as to Diameter, Capacity, and Copper Resistance of all Cores. By A. Dearlove. 2s. 
 
 ON A SURF-BOUND COAST: Cable-laying in the African Tropics. By 
 
 Arthur P. Crouch, B.A. 7s. 6d.; New Edition, 5s. 
 
 TELEPHONES: THEIR CONSTRUCTION AND FITTING. By F. C. Allsop. 5s. 
 A HANDBOOK OF ELECTRICAL TESTING. By H. R. Kempe. Fifth 
 
 edition, revised and enlarged. 18s. 
 
 TELEPHONE LINES AND THEIR PROPERTIES. By Prof . W. J. Hopkins. 6s. 
 THE TELEPHONE HANDBOOK AND PRACTICAL GUIDE TO TELE- 
 
 PHONIC EXCHANGE. By Joseph Poole. 3s. 6d. 
 
 THE TELEPHONING OF GREAT CITIES, AND AN ELECTRICAL PARCEL 
 
 EXCHANGE SYSTEM. By A. E. Bennett. Is. 
 
 MANUAL OF THE TELEPHONE. By W. H. Preece and A. J. Stubbs. 
 
 Over 500 pages and 334 illustrations. 15s. 
 
 THE TELEPHONE, THE MICROPHONE, AND THE PHONOGRAPH. By 
 
 Count du Moncel. Third Edition, 5s. 
 
 THE MAGNETO HAND TELEPHONE. By Norman Hughes. 3s. 6d. 
 PHILIPP REIS, INVENTOR OF THE TELEPHONE. A Biographical Sketch. 
 
 By Silvanus P. Thompson, B.A., D.Sc. With portrait and wood engravings. 8vo, cloth, 7s. 6d. 
 
 ELECTROMAGNETIC THEORY seepage 4. 
 
 ELECTRICAL ENGINEERING LEAFLETS (HOUSTON AND KENNELLY'S). 
 
 In three Grades Elementary, Intermediate, and Advanced of 35 Leaflets each. Price of single Leaflets, 
 6d., post free. Subscription price for any one Grade, 12s. 6d. ; for any two Grades, 23s. ; for all three 
 Grades, 33s. 6d. Further particulars on application. 
 
 TRAITE DE TELEGRAPHIE ELECTRIQUE. By H. Thomas. With 702 
 
 illustrations. 1. 
 
 TRAITE DE TELEGRAPHIE SOUS-MARINE. By E. Wiinschendorff, 
 
 Inge"nieur des Te"le"graphes, Directeur de Telegraphic Militaire. With 469 illustrations, Svo, 33s. 
 
 LA TEL^PHONIE. By E. Pierard. 7s. 
 
 ELECTRO-CHEMISTRY & ELECTRO-METALLURGY. 
 
 "THE ELECTRICIAN" PRIMERS seepage 8. 
 
 ELECTRO-PLATING : A Practical Handbook on the Deposition of Copper, 
 
 Silver, Nickel, Gold, Aluminium, Brass, Platinum, &c., &c., By John W. Urquhart. Third Edition, 5s. 
 
 ELECTRO-METALLURGY : Practically treated. By Alexander Watt, F.R.S.A. 
 
 Ninth Edition. 12mo, cloth boards, 4s. 
 
 THE ART OF ELECTRO-METALLURGY, INCLUDING ALL KNOWN PRO- 
 
 CESSES OF ELECTRO-DEPOSITION. By G. Gore, LL.D., F.R.S. With 56 illustrations, fcp. 8vo, 6s. 
 
 1, 2, and 3, Salisbury Court, Fleet Street, London, E.G. 
 
"The Electrician" Printing and Publishing Co., Ltd., 25 
 
 THE ART OF ELECTROLYTIC SEPARATION OF METALS- see page 11. 
 A TKEATISE ON ELECTRO-METALLURGY. By Walter G. M'Millan. 10s. 6d. 
 ELECTRO-DEPOSITION : A Practical Treatise on the Electrolysis of Gold, 
 
 Silver, Copper, Mickel, and other Metals and Alloys. By Alexander Watt, F.R S.S.A. With numerous 
 illustrations. Third Edition. Crown 8vo, cloth, 9s. 
 
 ELECTRO-PLATER'S HANDBOOK. By G. E. Bonney. Second Edition, 3s. 
 
 ELECTRO-CHEMISTRY seepage 11. 
 
 ELECTRO-DEPOSITION OF METALS. By G. Langbein 25s. 
 
 THE USES OF ELECTROLYSIS IN SURGERY. By W. E. Steavenson 
 
 M.D..M.R.C.P. Cloth ,5s. 
 
 TRAITfi THEORIQUE ET PRATIQUE D'ELECTROCHIMIE. By Donate 
 
 Tommasi, Docteur-es-Sciences. Large Svo, 1,200 pages. 36?. 
 
 ELECTRICAL ENGINEERING FORMULAE see page 9. 
 
 ELECTRICAL ENGINEERING LEAFLETS (HOUSTON AND KENNELLY'S). 
 
 In three Grades Elemen'ary, Intermediate, and Advanced - of 35 Leaflets each. Price of single Leaflets, 
 6d., post free. Subscription price for any one Grade, 12s. 6d. ; for any two Grades, 23s. ; for all three 
 Grade?, 33s. Cd. Further particulars on application. 
 
 ELECTRICAL INSTRUMENTS, BELLS, <&c. 
 
 "THE ELECTRICIAN" PRIMERS see page 8. 
 
 ELECTRICAL INSTRUMENT MAKING FOR AMATEURS. A Practical 
 
 Handbook. By S. R. Bottone. Fifth Edition. 3s. 
 
 INDUCTION COILS AND COIL MAKING. By F. C. Allsop. 3?. 6d. 
 INDUCTION COILS : A Practical Manual for Amateur Coil Makers. By G. E. 
 
 Bonney. 3s. 
 
 ELECTRICAL ENGINEERING FORMULAE see page 9. 
 DYNAMOMETERS AND THE MEASUREMENT OF POWER. By Prof. J. J. 
 
 Flather. 8s. 6d. 
 
 PRACTICAL ELECTRIC BELL FITTING. By F. C. Allsop. 3s. 6d. 
 THE BELL-HANGER S HANDBOOK. By F. B. Badt. 4s. 6d. 
 ELECTRIC BELL CONSTRUCTION. By F. C. Allsop. 3s. 6d. 
 ELECTRIC BELLS, AND ALL ABOUT THEM. By S. R. Bottone. Fourth 
 
 Edition. 3s. 
 
 ELECTRICAL APPARATUS FOR AMATEURS. By Various Authors. Is. 
 
 REFERENCE BOOKS, Ac. 
 
 "THE ELECTRICIAN" ELECTRICAL TRADES' DIRECTORY AND 
 
 HANDBOOK see ^>age 3. 
 
 WILLING'S BRITISH AND IRISH PRESS GUIDE. A concise and compre- 
 hensive Index to the Press of the United Kingdom, with Lists of the principal Colonial and Foreign 
 Journals. Is. 
 
 HAZELL'S ANNUAL. A Cyclopaedic Record of Men and Topics of the Day. 3s. 6d. 
 STATESMAN'S YEAR BOOK : A Statistical and Historical Annual of the 
 
 States of the World. Revised after Official Returns. 10s. 6d. 
 
 WHITAKER'S ALMANACK. 2s. 6d. 
 
 SELL'S DIRECTORY OF REGISTERED TELEGRAPHIC ADDRESSES: 
 
 From Official Lists supplied by the authority of the Postmaster General. Containing upwards of 40.000 
 firms. 21s. 
 
 KELLY'S POST OFFICE AND OTHER DIRECTORIES. (Particulars and 
 
 lowest prices on application.) 
 
 1, 2 and 3, Salisbury Court, Fleet Street, London, E.G. 
 
26 " The Electrician " Printing and Publishing Co., Ltd., 
 
 ELECTRICAL ENGINEERING FORMULAE see page 9. 
 
 A POCKET-BOOK OF ELECTRICAL RULES AND TABLES, for the use 
 
 of Electricians and Engineers. By John Munro. C.E., and Andrew Jamieson, C.E., F.R.S.E. Tenth 
 Edition. Revised and enlarged. Pocket size, leather, 8s. 6d., post free. 
 
 THE ELECTRICAL ENGINEER'S POCKET-BOOK OF MODERN RULES, 
 
 FORMULA, TABLES, AXD DATA. By H. R. Kemte, M.I.E.E. 250 pages, numerous illustrations. 
 Price 5s., post free. 
 
 USEFUL RULES AND TABLES FOR ENGINEERS AND OTHERS. By 
 
 Prof. Rankine. With Appendix by Andrew Jamieson, C.E. ]0s. 6d. 
 
 A MANUAL OF RULES, TABLES AND DATA FOR MECHANICAL 
 
 ENGINEERS. By D. K. Clark. Cloth, 16s. ; half-bound, 20s. 
 
 MATHEMATICAL TABLES. By the Rev. Prof. Galbraith, M.A., and the 
 
 Rev. Prof. Haughton, D.C.L., F.R.S. 3s. Cd. 
 
 FACTORY ACCOUNTS : Their Principles and Practice. By E. Garcke and J. 
 
 M. Fc'lls. Fifth Edition. 6s. 
 
 LAXTON'S PRICE BOOK. Published Annually. 4s. 
 
 LOCKWOOD'S BUILDERS' PRICE BOOK. Edited by Francis T. W. Miller. 4s. 
 
 ENGINEER'S YEAR BOOK of Formula, Rules, Table?, Data, and Memoranda. 
 
 By H. R. Kempe. 8s. 
 
 GRIFFIN'S ELECTRICAL ENGINEER'S PRICE-BOOK. Edited by H. J. 
 
 Dowsing. 8s. 6d. 
 
 HOUSTON'S DICTIONARY OF ELECTRICAL WORDS, TERMS, 
 
 AND PHRASES. New and greatly enlarged Edition. 560 pages, and nearly COO illustrations. Price 
 21s., post free. 
 
 LAW RELATING TO ELECTRIC LIGHTING. By G. S. Bower and W. Webb. 
 
 Entirely re-written. 8vo, 12s. 6d. 
 
 DICTIONARY OF CHEMISTRY (Watts'). Revised and entirely rewritten by 
 
 M. M. Pattison Muir, M.A., F.R.S.E., and H. Forster Morley, M.A., D.Sc., assisted by eminent contribu 
 tors. In 4 vols. Vols. I. and II., 42s. each ; Vol. III., 50s ; Vol. IV., C3s. 
 
 A DICTIONARY OF APPLIED CHEMISTRY. By T. E. Thorpe, assisted 
 
 by Eminent Contributors. Vols. I. and II., 2. 2s. each ; Vol. III., 3. 3s. 
 
 THE GRAPHIC ATLAS AND GAZETTEER OF THE WORLD. 128 Map 
 
 Plates and Gazetteer of 280 pages. Crown 4to, cloth, 12s. 6d ; half-morocco, 15s. ; full morocco, 21s. 
 
 DICTIONARY OF METRIC AND OTHER USEFUL MEASURES. By 
 
 Latimer Clark, F.R.S., &c. Crown Svo, 6s. 
 
 BLACKIE'S MODERN CYCLOPAEDIA OF UNIVERSAL INFORMATION. 
 
 Edited by Charles Annandale, M.A., LL.D. Eight vols., cloth, 6s. each ; half -morocco, 8s. 6d. each. 
 
 SPONS' DICTIONARY OF ENGINEERING. Three volumes. 5. 5s. 
 CASSELL'S STOREHOUSE OF GENERAL INFORMATION. In volumes. 5s. 
 CASSELL'S CONCISE CYCLOPAEDIA. 7s. 6d. 
 "TECHNICAL INDEX" OF THE " ENGINEERING MAGAZINE."- 
 
 (Particulars on application.') 
 
 PRACTICAL DICTIONARY OF MECHANICS. By Edward H. Knight. 
 
 Three volumes. With 15,000 illustrations. 3. 3s. Supplementary volume, 1. 1. 
 
 HAYDN'S DICTIONARY OF DATES. Twentieth edition ; brought down to 
 
 the Autumn of 1892. Cloth, 18s. ; half-calf, 24s. 
 
 DIRECTORY OF DIRECTORS : A list of Directors of Public Companies, with 
 
 the concerns with which they are connected. 12s. 6d. 
 
 STOCK EXCHANGE YEAR-BOOK : A Digest of Information relating to the 
 
 Origin, History, and Present Position of the Public Securities and Joint Stock Companies known to the 
 Markets of the United Kingdom. If s. 
 
 UNIVERSAL ELECTRICAL DIRECTORY. 4s. 
 
 TABLES AND FORMULAE FOR PLUMBERS, ARCHITECTS, SANITARY 
 
 ENGINEERS, &c. By J. Wright Clarke. With Electrical Tables and Memoranda, by W. Hibbert. 
 Second Edition. Waistcoat pocket size. Is. 6d. 
 
 I, 2 and 3, Salisbury Court, Fleet Street, London, E.G. 
 
"The Electrician" Printing and Publishing Co., Ltd., 27 
 
 DIGEST OF THE LAW OF ELECTRIC LIGHTING, '&c.see pa.je 8. 
 
 ELECTRICAL TABLES AND FORMULAE FOR THE USE OF TELE- 
 GRAPH IXSPECTOES AND OPERATORS. Compiled by Latimer Clark and Robert Sabine. With wood 
 engravings, crown 8vo, cloth, 12s. 6d. 
 
 TABLES AND FORMULAE FOR ELECTRIC STREET RAILWAY EN- 
 GINEERS. Compiled by E. A. Merril. 4s. 6d. 
 
 LIVES OF THE ELECTRICIANS. First Series : Profs. TYNftALL, WHEAT- 
 
 STOXE AND MORSE. By William T. Jeans. Crown Svo. cloth, 6s. 
 
 CASSELL'S NEW POPULAR EDUCATOR. Eight volumes. 5s. each. 
 CASSELL'S TECHNICAL EDUCATOR. Four volumes. 5?. each. 
 BARLOW'S TABLES OF SQUARES, CUBES, SQUARE ROOTS, CUBE 
 
 ROOTS. Reciprocals of all Integer Numbers up to 10,000. Post Svo, cloth, 6s. 
 
 CHAMBERS' ENCYCLOPEDIA. Ten volumes. Cloth, 5; half morocco 
 
 or half calf, 7. 10s. ; and in better bindings up to 15. 
 
 ABC FIVE-FIGURE LOGARITHMS. By C. J. Woodward, B,Sc., Cloth, 2s. 6d. 
 FOUR-FIGURE MATHEMATICAL TABLES. By J. T. Bottomley. 2s. 6d. 
 ILLUSTRATIONS OF .THE C.G.S. SYSTEM OF UNITS. With Tables of 
 
 Physical Constants. By Prof. J. D. Everett. 5s. 
 
 LOGARITHMIC AND TRIGONOMETRICAL TABLES FOR APPROXI 
 
 MATE CALCULATION, &c. By J. T. Bottomley, M. A. Is. 
 
 MATHEMATICAL TABLES. By James Pryde, F.E.I.S. 4s. 6d. 
 
 SPONS' ENGINEERS' DIARY. Published annually. 3s. 6d. 
 
 SPONS' ENGINEERS' PRICE-BOOK. 7s. 6d. 
 
 ELECTRICAL TABLES AND MEMORANDA. By Silvanus P. Thompson, 
 
 D.Sc., and Eustace Thomas. Waistcoat-pocket size, Is. ; post free. Is. Id. 
 
 A POCKET-BOOK OF USEFUL FORMULAE AND MEMORANDA FOR CIVIL 
 
 AXD MECHANICAL ENGINEERS. By Sir G. L. Molesworth, K.C.I.E., and R, B. Molesworth, M.A, 
 Twenty-third Edition. 6s. 
 
 THE MECHANICAL ENGINEER'S POCKET-BOOK. By D. K. Clark. 9s. 
 TABLES, MEMORANDA, AND CALCULATED RESULTS, for Mechanics, 
 
 Engineers, Builders, Contractors, &c. By Francis Smith. Fifth Edition, revised and enlarged. 
 Waistcoat pocket size, Is. 6d. 
 
 ENGINEER AND MACHINISTS' POCKET-BOOK. By Charles H. Haswell, 
 
 ISs. 
 
 REID'S PATENT INDEXED READY RECKONER. With fore-edge Index, 
 
 Third Edition. 23. 6d. 
 
 MISCELLANEOUS. 
 
 WIRE : ITS MANUFACTURE AND USES. By J. Bucknall Smith. 7s. 6d. 
 MODERN PHYSICS, CONCEPTS AND THEORIES OF. By J. B. Stallo. 
 
 Third Edition. 5s. 
 
 NATURAL PHILOSOPHY. By the Rev. Prof. Galbrsith, M.A., and the Rev, 
 
 Prof. Haughton, D.C.L., F.R.S. 3s. 6d. 
 
 PROTECTION OF BUILDINGS FROM LIGHTNING. By Dr. Oliver J. 
 
 Lodge, F.R.S. Is. 
 
 ELEMENTARY PHYSICS, EXAMPLES IN. With Examination Papers. By 
 
 W. Gallatly, M.A. 4s. 
 
 NATURAL PHILOSOPHY: An Elementary Treatise. By Prof. A. Privat 
 
 Deschanel. Translated and added to by Prof. J. D. Everett, D.C.L., F.R.S. Twelfth Edition. Cloth 188. 
 Also in four parts. I. MECHANICS, HYDROSTATICS, &c. ; II. HEAT ; III. ELECTRICITY AND MAGNETISM ; 
 IV. SOUND AND LIGHT. 4s. 6d. each. 
 
 I, 2 and 3, Salisbury Court, Fleet Street, London, E.G. 
 
28 "The Electrician" Printing and Publishing Co., Ltd., 
 
 ELEMENTAEY TEXT-BOOK OF PHYSICS. By Prof. Everett, D.C.L., 
 
 F.R.S. Seventh Edition, 3s. 6d. 
 
 ELECTRICAL EXPERIMENTS. By G. E. Bonney. 2s. 6d. 
 
 THE PHONOGRAPH : and How to Construct It. By W. Gillett. 5s. 
 
 ELECTRICAL ENGINEERING AS A PROFESSION, AND HOW TO 
 
 ENTEJR, IT. By A. D. Southam. 4s. 6d. 
 
 AID BOOK TO ENGINEERING ENTERPRISE. By Ewiug Matheson, M.I.C.E. 
 
 New edition. 850 pages. 1. Is. 
 
 ENGINEERING CONSTRUCTION IN IRON, STEEL, AND TIMBER. By 
 
 W. H. Warren. 16s. net, 
 
 PRACTICAL PHYSICS. By R. T. Glazebrook and W. N. Shaw. 6s. 
 MANUAL OF PHYSICS. By William Peddie, F.R.S.E. 7s. 6d. 
 EXAMPLES IN PHYSICS. By D. E. Jones, B.Sc. Fcap. Svo, 3s. 6d. 
 EXPERIMENTAL SCIENCE. By G. M. Hopkins. 680 engravings. Cloth, 15s. 
 LIGHTNING CONDUCTORS : THEIR HISTORY, NATURE, AND MODE 
 
 OF APPLICATION. By Richard Anderson, F.C.S., F.G.S. Third Edition, 12s. 6d. 
 
 EXPERIMENTAL CHEMISTRY. By Stockhardt, Edited by C. AY. Heaton, 
 
 F.C.S. 5s. 
 
 AIDS TO CHEMISTRY. By C. E. Armand Semple, B.A., M.B. In four parts. 
 
 Parts I. to III., 2s. and 2s. 6d. ; Part IV., Is. and Is. 6d. 
 
 A SHORT MANUAL OF ANALYTICAL CHEMISTRY FOR LABORATORY 
 
 USE. By John Muter, Ph.D., M.A., F.C.S. 6s. 6d. 
 
 LIGHTNING CONDUCTORS AND LIGHTNING GUARDS. By Prof. 
 
 Oliver J. Lodge, D.Sc., &c. With numerous illustrations. 15s. 
 
 PRACTICAL CHEMISTRY, INCLUDING ANALYSIS. By J. E. Bowman and 
 
 Prof. C. L. Bloxam. Eighth Edition. With 90 engravings, fcap. Svo, 5s. 6d. 
 
 CHEMICAL THEORY FOR BEGINNERS. By L. Dobbin, Ph.D., and J. 
 
 Walker, Ph.D., D.Sc. 2s. 6d. 
 
 OUTLINES OF GENERAL CHEMISTRY. By Prof. W. Ostwald. Translated 
 
 by Dr. J. Walker. 10s. net. 
 
 LESSONS IN ELEMENTARY CHEMISTRY. By Sir H. E. Roscoe, M.P., 
 
 F.E.S. Sixth Edition, 4s 6d. 
 
 A TREATISE ON CHEMISTRY. By Sir H. E. Roscoe, M.P., F.R.S., and C. 
 
 Schorlemmer. Vols. I. and II. : Inorganic Chemistry. Vol. L, The Non-Metallic Elements, 21s. 
 Vol. II., Metals, in two parts, 18s. each. Vol. III. : Organic Chemistry. Parts 1, 2, 4, and 6, 21s. each ; 
 Parts 3 and 5, 18s. each. 
 
 THE PRINCIPLES OF CHEMISTRY. By Prof. D. Mendeleeff. Translated by 
 
 George Kamensky. Two Volumes. 36s. 
 
 PRACTICAL CHEMISTRY AND QUALITATIVE ANALYSIS. By Prof. 
 
 Frank Clowes, D.Sc. 7s. 6d. 
 
 INORGANIC CHEMISTRY. By Prof. Edward Frankland, D.C.L., &c., and 
 
 Prof. F. K. Japp, M.A., &c. 24s. 
 
 A SYSTEM OF INORGANIC CHEMISTRY. By Prof. William Ramsey, 
 
 Ph.D., F.R.S. 15s. 
 
 ELEMENTARY SYSTEMATIC CHEMISTRY. By Prof. William Ramsey, 
 
 Ph.D., F.R.S. 4s. 6d. 
 
 PRACTICAL CHEMISTRY, EXERCISES IN. By A. G. Yernon Harcourt, 
 
 M. A., and H. G. Madan, M.A. Fourth Edition, 10s. 6d. 
 
 CHEMISTRY. By Edward Aveling, D.Sc. 6s. 
 
 METAL TURNING. By a Foreman Pattern-Maker. 4s. 
 
 ALUMINIUM : Its History, Occurrence, Properties, Metallurgy, and Applications, 
 
 including its Alloys. By J. W. Pvichards. Second Edition. 511 pages, 1. 5s. 
 
 1, 2 and 3, Salisbury Court, Fleet Street, London, E.G. 
 
"The Electrician" Printing and Publishing Co., Ltd., 29 
 
 PEACTICAL TREATISE ON THE STEAM-ENGINE. By Arthur Rigg. 1. 5s. 
 THERMODYNAMICS OF THE STEAM ENGINE AND OTHER HEAT 
 
 ENGINES. By Prof. C. H. Peabody. 21s. 
 
 AN ELEMENTARY TREATISE ON STEAM. By Prof. John Perry. 4s. 6<L 
 STEAM ENGINES AND OTHER HEAT ENGINES. By Prof. J. A. Ewing, 
 
 M.A., F.R.S. 15s. 
 
 STEAM ENGINE, THE. By George C. V. Holmes. 6s. 
 
 STEAM. By Prof. William Ripper. 2s. 6d. 
 
 STEAM AND THE STEAM-ENGINE. By Andrew Jamieson, Mlnst.C.E. 
 
 Text-Book, 8s. 6d. ; Elementary Manual, 3s. 6d. 
 
 A HANDBOOK OF THE STEAM ENGINE. By John Bourne, C.E. 9s. 
 RECENT IMPROVEMENTS IN THE STEAM ENGINE. By John Bourne, 
 
 6s. 
 
 THE STEAM ENGINE: A Treatise on Steam-Engines and Boilers. By 
 
 D. K. Clark. Two vols. 50s. net. 
 
 THE STEAM ENGINE, THE THEORY AND ACTION OF. By W. H 
 
 Northcott, C.E. 3s. 6d. 
 
 THE STEAM ENGINE. By the Rev. Prof. Galbraith, M.A., and the Rev 
 
 Prof. Haughton, D.C.L., F.R.S. 3s. 6d. 
 
 TEXT-BOOK ON THE STEAM ENGINE. By Prof. T. M. Goodeve, M.A, 
 
 Eleventh Edition, 6s. 
 
 ON GAS ENGINES. By Prof. T. M. Goodeve, M.A. 2s. 6d. 
 
 HEAT ENGINES OTHER THAN STEAM. By H. Graham Harris. Is. 6d. 
 
 THE STEAM ENGINE INDICATOR, AND INDICATOR DIAGRAMS 
 
 see page 13. 
 
 USES OF PETROLEUM IN PRIME MOVERS. By Prof. W. Robinson. Is. 
 
 GAS ENGINES : their Theory and Management. By William Macgregor. 8s. 6d. 
 
 GAS ENGINE, THE. By Dugald Clerk. 7s. 6d. 
 
 THEORY OF THE GAS ENGINE. By Dugald Clerk. 2s. 
 
 GAS AND PETROLEUM ENGINES. By Prof. William Robinson. 14s. 
 
 A PRACTICAL TREATISE ON THE CONSTRUCTION OF HORIZONTAL 
 
 AND VERTICAL WATER, WHEELS. By William C alien. Second Edition. 4s. 
 
 THE TURBINE MANUAL AND MILLWRIGHT'S HANDBOOK. By Charles 
 
 Louis Hett. 2s. 
 
 HEAT AS A FORM OF ENERGY. By Prof. R. H. Thurston. 5s. 
 
 HEAT. By Prof, P. G. Tait. 6s. 
 
 CONVERSION OF HEAT INTO WORK. By Dr. William Anderson, F.R.S., 
 
 Third Edition. 6s. Cheaper Edition. 2s. 6d. 
 
 THE THEORY OF HEAT. By J. Clerk Maxwell, M.A. Edited by Lord 
 
 Hayleigh. 4s. 6d. 
 
 AN ELEMENTARY TREATISE ON HEAT. By William Garnett, M.A., D.C.L. 
 
 Fifth Edition, 6s. 
 
 HEAT A MODE OF MOTION. By John Tyndall, D.C.L. 12s. 
 
 A COLLECTION OF EXAMPLES ON HEAT AND ELECTRICITY. By H, 
 
 H. Turner. 2s. 6d. 
 
 PRINCIPLES OF MECHANICS. By Prof. T. M. Goodeve. 6s. 
 ELEMENTARY DYNAMICS. By E. J. Gross, M.A. 5s. 6d. 
 
 li 2 and 3, Salisbury Court, Fleet Street, London, E.G. 
 
80 "The Electrician" Printing and Publishing Co., Ltd., 
 
 A TREATISE ON ELEMENTARY DYNAMICS. By William Garnett, M.A., 
 
 D.C.L. Fifth Edition, Cs. 
 
 "THE ELECTRICIAN'' PRIME EBsee page 8. 
 
 A TREATISE ON DYNAMICS. By W. H. Besant, D.Sc., F.R.S. Second 
 
 Edition, 7s. 6d, 
 
 MODERN MECHANISM, exhibiting the latest Progress in Machines, Motors and 
 
 the Transmission of Power. Edited by Park Benjamin, LL.B., Ph.D. 30s. 
 
 ELEMENTARY MECHANICS. By Oliver J. Lodge, D.Sc. Lond. Revised 
 
 Edition, 208 pages. Cloth, 3s. 
 
 APPLICATIONS OF DYNAMICS TO PHYSICS AND CHEMISTRY. By 
 
 Prof. J. J. Thomson. 7s. 6d. 
 
 TEXT-BOOK OF MECHANICAL ENGINEERING. By W. J. Lineham. 
 THE CONSERVATION OF ENERGY. By Balfour Stewart. Seventh 
 
 Edition. 5s. 
 
 AN ELEMENTARY TEXT-BOOK OF APPLIED MECHANICS. By 
 
 David Allan Low. 2s. 
 
 MECHANICS. By the Rev. Prof. Galbraith, M.A., and the Rev. Prof. Hatighton, 
 
 D.C.L., F.R.S. 3s. 6<1. 
 
 PRACTICAL MECHANICS AND MACHINE DESIGN, NUMERICAL EX- 
 AMPLES IN". By R. G. Elaine. 2s. Cd. 
 
 THERMODYNAMICS, LESSONS ON. By R. E. Baynes, M.A. 7s. Gd. 
 
 MECHANICS. By Edward Aveling, D.Sc. 6s. 
 
 DYNAMICS OF A PARTICLE, TREATISE UPON. By Prof. P. G. Tait and 
 
 W. J. Steele. 12s. 
 
 MOTIVE POWER AND GEARING seepage 14. 
 
 MODERN SHAFTING AND GEARING. By M. P. Bale. 2s. 6d. 
 
 THE TRANSMISSION OF ENERGY. By Prof. 0. Reynolds, M.A., F.R.S. Is. 
 
 HELICAL GEARS. By a Foreman Pattern-Maker. 7s. 6d. 
 
 APPLIED MECHANICS. By Prof. J. W. Cotterill. Third Edition. 18s. 
 
 LESSONS IN APPLIED MECHANICS. By Prof. J. H. Cotterill and J. H, 
 
 Slade. 5s. 6d. 
 
 HYDRAULIC MOTORS : Turbines and Pressure Engines. By G. R. Bodmer, 
 
 Assoc.M.Inst.C.E. With 179 illustrations. 14s. 
 
 THE MECHANICS OF MACHINERY. By Prof. A. B. W. Kennedy. 12s. 6d 
 MACHINE DRAWING AND DESIGN. By Prof. William Ripper. In one 
 
 Vol., 25s. ; or in five parts, 4s. 6d. each. 
 
 AN INTRODUCTION TO MACHINE DRAWING AND DESIGN. By David 
 
 Allan Low (Whitworth Scholar). With 65 illustrations and diagrams, crown Svo, 2s. 
 
 MANUAL OF MACHINE DRAWING AND DESIGN. By David Allan Low 
 
 and Alfred Wni. Bevis. 7s. Gd. 
 
 MACHINE DRAWING. By G. Halliday. First Course. Part II. ELECTRICAL 
 
 ENGINEERING, 2s. 6d. Second Course, Cs. 
 
 THE MECHANIC'S WORKSHOP HANDYBOOK : A Practical Manual on 
 
 Mechanical Manipulation. By Paul N". Hasluck, A.I.M.E. Crown Svo, cloth, 2s. 
 
 PRACTICAL MECHANICS. By Prof. Perry, M.E. 3s. 6d. 
 
 ELEMENTS OF MACHINE DESIGN. By Prof. W. C. Unwin. Part L, 
 
 General Principles, 6s. Part II., Chiefly on Engine Details, 4s. 6d. 
 
 DRAWING FOR MACHINISTS AND ENGINEERS. (Cassell's Technical 
 
 Manuals.) 4s. 6d. 
 
 ELECTRICITY: WHAT TO DO WITH IT, AND HOW TO DO IT- 
 
 In the Press. 
 
 I 2 and 3 y Salisbury Court, Fleet Street, London, E.C, 
 
"The Electrician" Printing and Publishing Co., Ltd., 31 
 
 ENGINEER'S SKETCH BOOK of Mechanical Movements, Devices, Appliances. 
 
 Contrivances, and Details employed in the Design and Construction of Machinery for every purpose. 
 Nearly two thousand illustrations. 7s. 6d. 
 
 SCIENTIFIC AND TECHNICAL PAPERS of WERNER YON SIEMENS. 
 
 Translated from the German by E. F. Batnber. Vol. I., Scientific Papers and Addresses. Vol. II., 
 Applied Science. 14s. each. 
 
 PERSONAL RECOLLECTIONS OF WERNER YON SIEMENS. Translated 
 
 by W. C. Coupland. 15s. 
 
 POPULAR LECTURES AND ADDRESSES ON YARIOUS SUBJECTS IN 
 
 PHYSICAL SCIENCE. By Sir William Thomson (Lord Kelvin). In three vols. Vols. I. and III. ready, 
 7s. 6d. each. 
 
 JOURNAL OF THE INSTITUTION OF ELECTRICAL ENGINEERS 
 
 (formerly the Society of Telegraph-Engineers and Electricians). Issued at irregular intervals. 
 
 THOMAS A. EDISON AND SAMUEL F. B. MORSE. By Dr. Denslow and 
 
 J. Marsh Parker. Is. 
 
 CHATS ON INYENTION. By John Martin. Is. 
 
 LECTURES ON SOME RECENT ADVANCES IN PHYSICAL SCIENCE. By 
 
 Prof. P. G. Tait. 6s. 
 
 ROMANCE OF ELECTRICITY. By John Munro. Crown Svo, 5s. 
 POPULAR LECTURES ON SCIENTIFIC SUBJECTS. By Prof. H. von 
 
 Helmholtz. Two Volumes. 7s. 6d. each. 
 
 WATER OR HYDRAULIC MOTORS. By Philip R. Bjorling. 208 illustrations. 9s. 
 SCIENCE LECTURES AT SOUTH KENSINGTON. Containing Lectures by 
 
 Celebrated Scientists. Two Volumes. 6s. each. 
 
 ELECTRICAL ENGINEERING FORMULAE see page 9. 
 
 PIONEERS OF SCIENCE. By Dr. 0. J. Lodge. 7s. 6d, 
 
 PROF. CLERK MAXWELL, A LIFE OF. By Prof. L. Campbell, M.A., and 
 
 W. Garnett, M.A. Second Edition. 7s. 6d. 
 
 GLEANINGS IN SCIENCE : A Series of Popular Lectures on Scientific Subjects. 
 
 By Rev. G. Molloy. 7s. 6d. 
 
 ACHIEVEMENTS IN ENGINEERING DURING THE LAST HALF 
 
 CENTURY. By Prof. Vernon Harcourt. 5s. 
 
 SPINNING TOPS. By Prof. J. Perry. 2s. 6d. 
 
 SOAP BUBBLES, and the Forces which Mould Them. By C. Y. Boys. 2s. 6d. 
 
 FORTY LESSONS IN ENGINEERING WORKSHOP PRACTICE. (Poly- 
 
 technic Series.) Is. 6d. 
 
 SCIENCE FOR ALL. Edited by Dr. Robert Brown, M.A., F.L.S. Five 
 
 volumes. 9s. each. 
 
 THEORY OF LIGHT. By T. Preston, los. net. 
 
 THEORY OF SOUND. By Lord Rayleigh. Yols. I. and IL, 12s. 6d. each; 
 
 Vol. III. in preparation. 
 
 THE MATERIALS OF CIYIL ENGINEERING. By Prof. R. H. Thurston. 
 
 Part I. Xon-Metallic Materials, 10s. 6d. ; Part II. Iron and Steel, ISs. : Part III. K"ou-Ferrous Metals 
 and Alloys (Second Edition), 12s. 6cl. 
 
 THE SLIDE RULE : A Practical Manual. By C. N. Pickworth. 2s. 
 HISTOIRE D'UN INYENTEUR: Expose des Decouvertes et des Travaux de 
 
 M. Gustave Trouve". By Georges Barral. With portrait of M. Trouve" and 280 illustrations. 
 
 M^CANIQUE MOL^CULAIRE. By A. Bandsept. Is. 6d. 
 
 MANUEL D'ELECTROLOGIE MEDIC ALE. By Gustave Trouve". 7s. 
 
 PRODUCTION ET UTILISATION RATIONNELLES DE LA CHALEUR 
 
 INTENSIVE. By A. Bandsept. 2s. 6d. 
 
 I, 2 and 3, Salisbury Count, Fleet Street, London, E.G. 
 
THE OLDEST AND BEST ENGLISH ELECTRICAL JOURNAL. 
 
 EVERY FEW AY, PRICE jd., POST FREE 
 
 THE ELECTRICIAN. 
 
 "THE ELECTRICIAN " was originally established in 18C1, in a form similar "to ihe present publication, anq 
 after occupying a leading position amongst scientific journals for a period of about three years, was discontinued, 
 not, however, before it had become apparent that such a publication occupied an essential place amongst the 
 scientific and technical journals. 
 
 From 1864 to 1S7S wes a period during which the absence of " THE ELECTRICIAN" was much fell, nc 
 weeKly paper devoted to the interests of the electrical profession being then in existence ; and, as has fre- 
 quently been the cas j , it was left for gentlemen taking a keen interest in the applications of electricity to tele 
 graphic purposes, whijh we-e at that time the principal electrical industries, to resuscitate " THE ELECTRICIAN," 
 and once again establish it as the recognised leading organ of electrical science and industry. It will thus be 
 seen that " THE ELECTRICIAN " is the oldest electrical journal published. Since its second advent, ard especially 
 since the end of the year 18S7, " THE ELECTRICIAN " has made rapid progress, and has continued to record in a 
 full and thorough manner all the great discoveries and experiments in electrical science, and to illustrate how 
 these could be commerc ally applied and profitably worked. In the volumes of " THE ELECTRICIAN " will be 
 found a most valuable collection of original papers and articles by nearly every leading electrical writer of note, 
 including : 
 
 Dr. W. Spottiswoode, 
 Mr. J. T. Sprague, 
 
 Prof. "W. G. Adams, 
 Mr. G. L. Addenbrooke, 
 Sir James Anderson, 
 Mr. L. B. Atkinson, 
 Prof. W. E. Ayr ton, 
 Prof. W. F. Barrett, 
 Mr. Shelford Bidwell, 
 Mr. T. H Blakesley, 
 Mr. C. Vernon Boys, 
 Major P. Oardew, 
 Mr. E. T. Carter, 
 Mr. R E. Crompton, 
 Prof. W. Crookes, 
 Prof. J. Dewar, 
 Mr. W. B. Esson, 
 Mr. Sydney Evershed, 
 Prof. J. A. Ewirg, 
 Mr. C. A. Faure, 
 Mr. Desmond G.FitzGerald 
 Dr. J. A. Fleming, 
 
 Prof. George Forbes, 
 Prof. W. Garnett, 
 Mr. W. Geipel, 
 Dr. Gladstone, 
 Prof. R. T. Glazebrook, 
 Dr. George Gore. 
 Mr. Oliver Heaviside, 
 Prof, von Helmholtz, 
 Dr. II. Hertz, 
 Mr. Paget Higgs, 
 Dr. Edward Hopkinson, 
 Dr. John Hopkinson, 
 Prof. E. J. Houston, 
 Prof. D. E. Hughes, 
 Mr. Gilbert Kapp, 
 Prof. A. B. W. Kennedy, 
 Mr. A. E. Kennelly, 
 Dr. Oliver Lodge, 
 Prof. J. Clerk Maxwell, 
 Prof. Mengarini, 
 
 Mr. W. M. Mordey, 
 Dr Alex. Muirhead, 
 Irof. John Perry, 
 Mr. W. H. Preece, 
 Mr. J. S. Ra worth, 
 Lord Kayleigh, 
 Prof. Rowland, 
 Prof. Rucker, 
 Mr. Warren de la Rue, 
 Sir David Salomons, 
 Dr. Paul Schoop, 
 Mr. Louis Schwendler, 
 Mr. G. F. C. Searle, 
 Mr. J. 8. Sellon, 
 Mr. Alex. Siemens, 
 Dr. Werner Siemens, 
 Mr. M. Holroyd Smith, 
 Mr. Willoughby Smith, 
 Mr. Albion T. Snell, 
 Mr. W. H. Snell, 
 
 Prof. Balfour Stewart, 
 Dr. W. E. Sumpner, 
 Mr. James Swinburne, 
 Mr. Nikola Tesla, 
 Prof. Silvanus Thompson, 
 Mr. Elihu Thomson, 
 Prof. J. J. Thomson, 
 Sir Win. Thomson (Lord 
 
 Kelvin), 
 
 Mr. H. Tomlinson, 
 Mr. A. P. Trotter, 
 Mr. G.W. de Tunzelmann, 
 Prof. John Tyndall, 
 Mr. F. C. Webb, 
 Mr. F. M. Weymouth, 
 Mr. H. D. Wilkinson, 
 Dr. Zetzsche, 
 
 &c., &c., &c. 
 
 And all papers read before the principal Electrical Institutions throughout the world by men eminent in the 
 Electrical Profession have been given, together with authenticated reports of the discussions thereupon. 
 
 In addition to the above, ''THE ELECTRICIAN " forms a complete record of all the important legal investi- 
 gations which have occupied the attention of the Courts of Justice for the past 14 years, and it is customary for 
 "THE ELECTRICIAN " to occupy a prominent position in the Courts as an authority upon all questions aiiectin^ 
 the Electrical Profession. In this connection it is only necessary to point to the actions which have arisen from 
 time to time upon the Edison and Swan Patents, the Compound Winding Patents, the High and Low Tension 
 Systems, the Accumulator Patents, the Telephone Patents, etc., etc., in which " THE ELECTRICIAN " has figured 
 as a reliable authority, and has been put in evidence and accepted by the parties concerned. 
 
 A regular feature in " THE ELECTRICIAN " has always been the verbatim reports of meetings of Electrical 
 Companies and Corporations, and while " THE ELECTRICIAN " has never trenched upon the grounds legitimately 
 occupied by the financial journals, the information of a financial character given from week to week in its 
 columns is full, reliable, and absolutely unbiassed. It has no interest whatever in any financial schemes, and is 
 devoted entirely to the interests of the profession it was established to serve. " THE ELECTRICIAN " gives full 
 reports of Meetings, &c., held on Thursdays, so that subscribers interested obtain their copies by Friday morn- 
 ing's post. 
 
 The original articles appearing in " THE ELECTRICIAN " are written by gentlemen having no interest what- 
 ever in particular electrical systems, and with but one object, and that the advancement of electrical knowledge 
 and electro-technology generally. Many of these original series of articles have since been revised by their 
 authors, and published in book form. These form the nucleus of the well-known * Electrician " Scries, of 
 which further particulars will be found herewith. 
 
 Finally, " THE ELECTRICIAN " has been of incalculable service to technical education, and has done much 
 to make the general study of electricity the reality it has undoubtedly become. No aspirant to honour and 
 renown in the electrical profession can hope to keep abreast of the never-ceasing stream of discoveries of new 
 applications of electrical science to every day commercial pursuits who does not diligently peruse the columns of 
 " THE ELECTRICIAN," which is pre-eminently the leading electrical journal. 
 
 TERMS OF SUBSCRIPTION: 
 
 YEAR. HALF-YEAR. 
 
 Great Britain .................... CO 1 6 ...... dEO 1O O 
 
 Conn tries in the Postal Union .. 1 4 O ...... O 13 6 
 
 Outside the Postal Union ...... 1 SO ...... O14 6 
 
 QUARTER. 
 
 *0 5 O 
 066 
 O 7 6 
 
 The large and influential circulation of " THE ELECTRICIAN," which far exceeds that of any 
 of its British contemporaries, together with its moderate tariff rates, combine to make the journal 
 
 THE YERY BEST MEDIUM FOR ADVERTISING. 
 
 "THE ELECTRICIAN" PRINTING <& PUBLISHING COMPANY,! Ltd., 
 
 SALISBURY COURT, FLEET LTREET, LONDON, "ENGLAND.