v^ 
 
 P^\A»## 
 
BOUGHT WITH THE INCOME 
 PROM THE 
 
 SAGE ENDOWMENT FUND 
 
 THE GIFT OF 
 
 Henril IB. Sage 
 
 1S9X 
 
 A.J.lpA'^ X0jf/mx 
 
arV18670 
 Electricity in its a 
 
 :ity 
 
 i 
 
 Cornell University Library 
 
 pplication to telegr 
 
 3 1924 031 307 295 
 
Cornell University 
 Library 
 
 The original of tliis book is in 
 tine Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/details/cu31924031307295 
 
ELECTRICITY 
 
 IN ITS 
 
 APPLICATION TO TELEGRAPHY. 
 
ELECTRICITY 
 
 IN ITS 
 
 APPLICATION TO TELEGRAPHY. 
 
 COVERING THE 
 
 SYLLABUS OF THE NEW TECHNICAL EXAMINATION. 
 (Adopted hy the Post Office Telegraph Department.) 
 
 BY 
 
 T. E. HERBERT. 
 
 HONOtTHSMAN AND MEDALLIST IN TELEGRAPHY AND IN TBLEPHONT. 
 LBOTUHBE, MUNICIPAL TECHNICAL SCHOOL, MANOHESTBE. 
 
 FIFTH EDITION. 
 
 WITH FOBTT-EIGHT ILZUSTBATIONS. 
 
 LONDON : 
 F. H. GoLDSPiNK & Co., 4, BuEGON Street, Carter Lane, 
 Doctors' Commons, B.C. 
 
 1897. 
 
[entered at stationers' hall. I 
 
PREFACE TO FIRST EDITION. 
 
 The immediate outcome of the issue of the Syllabus of the Depart- 
 mental Examination, which candidates for superior appointments were 
 required to pass before they could take up such appointments, was 
 a demand for a more popular exposition of the first principles of 
 Electricity, and a more detailed application of these principles to 
 Technical Telegraphy than had hitherto been available, even in the 
 best text-books on the subject. 
 
 The student who had not laboriously followed a course of study in 
 the science which found its practical application in Telegraphy was 
 placed at some disadvantage in commencing to read a technical text- 
 book in which the fundamental facts of the science were taken for 
 granted ; and, so far as could be ascertained, no text-book had been 
 published which filled the gap between the purely scientific text-book 
 dealing with general laws and principles, on the one hand, and, on the 
 other, a technical text-book dealing with the details of an industry in 
 which these laws were applied, and in which the fundamental scientific 
 facts were assumed. 
 
 The writer of the present volume was, therefore, requested by the 
 Editor of the Telegraph Chronicle to undertake a series of papers, in 
 which the steps between the scientific text-book and the technical text- 
 book should be dearly defined and traced in such a way that the initial 
 difficulties in studying the latter might be disposed of, and which at the 
 same time should ta^e into their purview the whole of the ground 
 covered by the Departmental Syllabus. 
 
 The papers were undertaken at short notice, and little time was 
 allowed for elaboration. Yet in spite of the crude form in which the 
 facts were expressed the papers proved to have fulfilled their design, 
 and to have supplied a long-felt want in so complete a manner that, in 
 response to a widely expressed desire that somepermanentform might 
 be given to the articles which appeared in the Telegraph Chroniele,the 
 present volume has been published. 
 
 The technical phraseology in which the subject of Telegraphy is most 
 frequently treated has been abandoned, and tne writer has endeavoured 
 to express the simple facts relating to such tangled questions as 
 Double Current Duplex Working (Differential and Bridge methods), 
 the principle and application of the Wheatstone Bridge to practical 
 testing, the Quadruplex, the Multiplex, and all the important systems 
 of working at present in vogue, which to the tyro seem to be bristling 
 with insurmountable difficulties, in the simplest possible language : and 
 to show that when these facts are related simply as they stand, the 
 study of Technical Telegraphy is not the dry, uninteresting, and 
 difficult subject which it is most frequently supposed to be, but one in 
 which all the parts are coherent and logical, and are capable of 
 interesting treatment. 
 
 In its present form the book differs very considerably in many of 
 its essential features from the papers which appeared in the Telegraph 
 
Vl. 
 
 Chronicle, Where it has been possible to present the facts in a bimpler 
 light this has been done ; and in every possible way the subject has 
 been amplified, so that it covers not merely the Department's Syllabus, 
 but the majority of the ground which students require to traverse in 
 order to enter for the City and Guilds Bxamination. 
 
 The book opens with a consideration of the " first principles " which 
 underlie the subsequent chapters, and an endeavour has been made to 
 present them in a form in which their connection with practical 
 Telegraphy may afterwards be readily comprehended. The chapter on 
 Batteries deals with the chemical phenomena, as well as the electrical, 
 in order that the subject may be thoroughly grasped in all its bearings. 
 Then follows a practical application of Ohm's Law to Electrical 
 Measurement, with an extension of the law to the best grouping of cells. 
 
 General systems of working are next considered, and after the various 
 theories have been handled the connections of the apparatus employed, 
 current required, and causes of the commoner faults are given. 
 
 Test-box Arrangements and a simple explanation of the theory of 
 the new sj^stem of Morning testing are then dealt with, and in turn 
 attention is given to the Wheatstone Bridge, the Universal Battery 
 System, and Multiplex Working. 
 
 The chapter on Repeaters has been thoroughly revised and enlarged 
 by Mr. H. Wilson, of Nevin, to whom the writer begs here to acknow- 
 ledge his indebtedness. In the following chapter the Bridge System of 
 Duplex Working has been treated ; and in order to facilitate explana- 
 tion various parts of the circuit have been drawn in dotted lines, so 
 that the path of the current at different points may be clearly indicated. 
 
 Battery testing is also consideredj and in order that the ground 
 covered by the book shall be as extensive as possible a chapter has been 
 written on the making uj) of special circuits, illustrated by actual 
 examples of the way in which exigencies have been met, and the routes 
 of wires utilised in order to provide ample facilities for the special 
 circumstances which arise from time to time at various places. 
 
 The final chapter consists of a series of miscellaneous notes, which it 
 is hoped will prove particularly useful to students who desire to pass 
 City- and Guilds Examinations. Various points, which it was somewhat 
 difficult to find a place for amongst the remainder of the book, are 
 herein dealt with m a brief manner. 
 
 A long list of questions is appended, about half of which were asked 
 at Departmental Examinations, the majority of the remaining 
 questions coming from City and Guilds papers. 
 
 In every way the treatment of the subject has been extended : and 
 the papers are issued in their present form in the hope that if in their 
 first and cruder form they found such cordial acceptation, and proved 
 themselves of use in any degree whatever, they may now, in proportion 
 to the care which has been exercised in their revision and extension 
 prove themselves additionally useful to all who have been emboldened 
 to undertake the study of Technical Telegraphy. 
 • In conclusion, the writer begs to here pface upon record his sense of 
 mdebtedness to Mr. G. W. Bannister, of Manchester, for much valuable 
 assistance rendered durmg the preparation of both the original mnei-s 
 and the present book. f ^'CI^s 
 
THIRD AND FIFTH EDITIONS. 
 
 Owing to the rapid ezhaustion ,of the first and second editions of 
 this wont, a third edition becsame necessary, and in it the writer has 
 availed himself of many of the suggestions made by readers for its 
 improvement. 
 
 Much matter has been added and the whole brought up to date. 
 The revision of the E.M.F, of the cells in use by the Department has 
 rendered some alterations in the method of taking the constant of the 
 tangent galvanometer and this and the standardisation of the morning 
 test have now been embodied in the present volume. 
 
 The chapter on battery testing has been entirely re-written for the 
 latest forms of instrument, and the writer begs to ezm'ess his thanks to 
 Mr. A, Eden, one of the Technical Officers of the Post Office Tele- 
 graphs, who very kindly undertook the revision of the chapter, 
 
 A chapter has now been devoted to the concentrator as the system 
 is being widely adopted. 
 
 Finally, a chapter on construction has been added in order that the 
 book may also cover the tehole of the syllabus of the City and Guilds 
 Examinations. 
 
 The questions have also been re-arranged and many new ones have 
 been added, chiefly upon the subjects of the new chapters. 
 
 The chapter on chemistry has been re-written by Mr. A. H, Atkins, 
 B.Sc, of the Grammar School, Birmingham, to whom the writer begs 
 to acknowledge his indebtedness. 
 
 T. E. H. 
 
 Engineers'- Department, 
 
 Postal Telegraphs, Manchester. 
 
LoKDON : 
 
 F. H. Q-OLDSPiNK & Co., Printers, 4, Burgon Street, Oartbk Lane, 
 
 Doctors' Commons, E.G. 
 
CONTENTS. 
 
 PAGE 
 
 APPENDIX.— Syllabus op Depaktmbntal Examinations 12 
 CHAPTER I.— Introductory 13 
 
 Lodestone— Magnets— Attraction and repulsion— Compass— Molecular 
 theory of magnetism— Lines of force— Magnetic induction— Voltaic 
 cell- Electrto current and its efCeots— Deflection of Compass needle- 
 Conductors and non-oonductora— Ohm's law. 
 
 CHAPTER II.— Batteries 23 
 
 Theory of simple voltaic cell— Local action- Polarisation — Chemistry — 
 Daniell cell — Positive and negative elements — Exciting and de- 
 polarising fluids— Leolanche cell— Bichromate cell— Dry cells. 
 
 CHAPTER III.— Electrical Measurements 33 
 
 Comparison of currents —Tangent galvanometer — Constant — Com- 
 parison of Ejil.F.'s— Measurement of internal resistance of cells. 
 
 CHAPTER IV.— Arrangement op Cells 41 
 
 Besistance of wires of varying lengths and diameters— Large and small 
 cells- Joining up cells in series and multiple arc— E.M.F. and 
 internal resistance— Characteristic features of various cells. 
 
 CHAPTER v.— The Direct Sounder 45 
 
 Electro-magnet— Up and down oflces- Connections— Current required 
 —Form of battery used— Earth return.. 
 
 CHAPTER VI.— The Single Needle 48 
 
 Induced dial — Lightning protectors — Commutator — Connections- 
 Current required— Form of hattery used. 
 
 CHAPTER Vn.— The P.O. Standard Relay 52 
 
 Theory of relay— Actual arrangements— Neutral rela.ys— Kesistanee in 
 series and parallel— Current required— Siemen's relay. 
 
 CHAPTER VIII.— The P.O. Double Plate Sounder ... 55 
 
 Commutator — Local circuit — Up. down, and intermediate offices — Con- 
 nections and path of current— Faults— Adjustmeni— Current re- 
 quired-Form of hattery used. 
 
 CHAPTER IX.— The Single Current Sounder 59 
 
 Use of relay— Connections— Faults— Adjustment-Current required- 
 Form of hattery used. 
 
CTTAPTEB X.— The Condensek 61 
 
 Capacity— Di-electrio— Microfarad— Post Office condensers. 
 
 CHAPTER XI.— The Double Ctjeeent System 65 
 
 Capacity of line — Eetardation— Theory of double current system- 
 Polarised and non-polarised relays — Double current key connections 
 — Connections of one station — Testing accuracy of connections — 
 Faults— Alteration of connections to work single current — Current 
 required— Form of battery used. 
 
 CHAPTER Xn— The Differential Duplex— Single 
 
 CuKBENr 69 
 
 Theory of duplex working— Method of opposition— Method of combina^ 
 tion— Connections— Duplex and single switch — ^Double split — ^Path 
 of current— Balancing— Current reiuired- Form of battCTy used. 
 
 CHAPTER XHL— The Double Cukkent Duplex 75 
 
 Necessary conditions— Path of current, both keys at rest and depressed 
 — One key depressed— Compensation circuit — ^Balancing capacity of 
 line— Full connections — Effect of heavy loss on line — Current 
 required- Form of battery used. 
 
 CHAPTER XIV.— The Quadkuplex 83 
 
 Theory of system — Connections — ^Adjustments — ^Decrement working — 
 Adjustments— Diplex— Current required- Form of battery used. 
 
 CHAPTER XV.— The Wheatstone Automatic 87 
 
 Theory of system— Transmitter— Connections — Eeceiver —Joining a 
 transmitterto akey-'worked circuit — Speed of working — Capacity — 
 Belf-indnction of receiver— Shunted condenser— Wheatstone duplex 
 —Current required— Form of battery nsed. 
 
 CHAPTER XVI.— Test Box Arrangements 95 
 
 Test-box connections— Through wlres-Intermediatestations— Looping, 
 earthing, crossing, and disconnecting wires — ^Metallic circuit — 
 Forking wires- Battery box. 
 
 CHAPTER XVEL— The Morning Tests 98 
 
 Theory of test— Sent, perfect, and received current— Joint resistance of 
 wires in parallel— Division of cnrrent between condncting paths — 
 Numerical example of test— Insulation resistance— Conditions for 
 accural— Testing battery— Practical arrangements— Standardisa- 
 tion — Connections — Earthed wires— Galvanometer shunts. 
 
 CHAPTER XVm.— Localisation Tests 108 
 
 Test- box galvanometer- Principle of fault localisation— Localising earth, 
 disconnection, and contact faults— Intermittent faults -Covered 
 faults— Stay faults. 
 
 CHAPTER XIX.— The Wheatstone Bridge 115 
 
 Thwry and numerical proof of the principle of the Wheatstone 
 bridge— Practical use of —Best conditions— Actual arrangements- 
 Plug switch— Conductivity test— Loop test. 
 
 CHAPTER XX.— The Universal Battery System 
 
 Theory of system-Practical arrangements— Number of cells required 
 to furnish a given cnrrent— Single needles and double plate sounders 
 - Smgle current sounders— Double aorrent sounders— Form of 
 battery used. 
 
 120 
 
XI. 
 
 CHAPTEK XXI.— The Delaney Multiplex System ... 125 
 
 Theory of system— Form of relay— Self-induotion of relay— Live aud 
 dead segments — Capacity or line— Betardation— Synchronising 
 arrangements- Phonic wheel— Eeed- Corrections— Adjustments. 
 
 CHAPTER XXn.— Eepeatbks 132 
 
 Working speed of lines— Theory of repeater— Automatic switch— Leak 
 circuit— Sparking - Sounder shunts- Keys for speaking— Faults- 
 Connections— Duplex repeater. 
 
 CHAPTER XXIII.— The Bridge Duplex 140 
 
 Theory of system- Douhle current hridge duplex— Path of current, both 
 keys at rest and depressed, and one key depressed — Advantages 
 and disadvantages as compared with differential system. 
 
 CHAPTER XXIV.— Battery Testing 146 
 
 Theory of E.M.F. test— Shunts— Theory of internal resistance— Con- 
 neutiona— Battery testing board— Internal resistance of standard 
 cell — titand^d units. 
 
 CHAPTER XXV.— The Concentrator 156 
 
 Principle of system — Connections— Eing off— Sending "Time"— 
 Equalising resistances- Application to A.B,C. instruments. 
 
 CHAPTER XXVI.— Construction 161 
 
 Choice of routes-Insulations— Arms— Timber— Stays and struts— 
 Wire— StreaaonWire^Weather contact— Undergroundoonstruction. 
 
 CHAPTER XXVII.— Making Up Special Circuits 173 
 
 Principle adopted — Crossing and dividing wires— Making wires good— 
 Eeplaoing faulty lengths— Actual examples— Battery power. 
 
 CHAPTER XXVni.— Miscellaneous Notes 178 
 
 Accumulators- Figure of merit— Winding of apparatus— Moment of a 
 maenet— Heat ■ produced by current— Eesistanoe, weight, and 
 diameter of wires— The Polarised sounder. 
 
 QUESTIONS ... 184 
 
SYLLABUS 
 
 OF THE 
 
 DEPAETMENTAL EXAMINATION. 
 
 I. CroBsing and looping -wires with facility and certainty. 
 
 n. Tracing and localising fanlts in instruments. 
 
 m. Tracing and localising permanent and intermittent earth, 
 contact, and disconnection faults on wires. 
 
 ly. Method of testing the E.M.F. and resistance of batteries, and a 
 general knowledge of the essential features of the various 
 descriptions of batteries. 
 
 V. System of morning testing, both as regards sending and receiving 
 currents, with the necessary calculations in connection with 
 the same. 
 
 yi. Making up special circuits in cases of emergency. 
 
 Vll. Joining up and adjusting single needle, single current and double 
 current Morse, both simplex and duplex, and Wheatstone 
 apparatus. 
 
 VJLU. Fitting a Wheatstone transmitter to an ordinary key-worked 
 circuit. 
 
 IX. A general knowledge of the principles of Quadruples and 
 
 Multiplex working, 
 
 X. Measuring resistances by the Wheatstone Bridge. 
 
 P.O. Circular (7035), June 19th, 1894. 
 
ELECTRICITY 
 
 IN ITS 
 
 APPLICATION TO TELEGRAPHY. 
 
 CHAPTEE I. 
 
 Introductory. 
 Magnetism. 
 
 Before entering upon the subject proper, it is necessary that the 
 fundamental principles of magnetism and electricity shall be clearly 
 apprehended. The present book is an essentially practical work, and, 
 consequently, the pure science of the subject will only be entered into 
 in so far as is necessary for the proper understanding of Telegraphy, 
 which is a practical application of the principles of magnetism and 
 elecirioity. 
 
 The discovery of the peculiar phenomena, which are classed under 
 the heading of magnetism, had its beginning in the finding of a rock 
 to which iron and steel would adhere when brought into contact with 
 it. This stone or rock received the name of lodestone, and is now 
 known to be a compound of the two elements oxygen and iron. Next 
 it was discovered that if steel be rubbed with a piece of lodestone it 
 will in turn become endowed with the power of attracting pieces of 
 iron and steel, that is, it becomes what is now termed a permanent 
 magnet. 
 
 A freely suspended bar of copper would have no tendency to point 
 in one direction rather than another, but a magnet behaves very 
 differently, in that it persists in pointing in a direction which is found 
 to be roughly North and South. Furthermore, the same end of the 
 magnet always points to the North, the other end always pointing to 
 the South, and these extremities are termed the poles of the magnet. 
 The pole which always points to the North is termed the North-seeking, 
 and the other pole is called the South-seeking, or simply North and 
 South poles. If the North pole of a second magnet be approached to 
 the North pole of a suspended magnet, it is seen that the suspended 
 
 B 
 
14 Attraction and R^pulswn. 
 
 magnet tnms away, or is repelled. On pFssenting the South pole to 
 the Xorth pole of the snspendcd magnet the exact opposite takes place, 
 as the suspended magnet is attracted. It is now seen that two S'orth 
 poles repel, and that a South and a North attract one another. By 
 making these experiments with two magnetised knitting needles it is 
 easy to satisfy oneself that similar poles, (.e.g^ two Korth and two 
 Souths) repd, and unlike poles (.e.g., a North and a South) attract one 
 another. To recapitulate experimental facts we have : — 
 
 I. Steel magnetised by being rubbed with a lodeatone. 
 TT. A suspended magnet always points in one direction. 
 in. Like poles rejjel and unlike poles attract. 
 
 The second fact is very suggestive. If the magnet tends to remain 
 in one fixed direction and requires force to displace it, there must be 
 some controlling agency, and it was soon conjectured that the earth 
 must be a magnet whose magnetic poles were, roughly speaking, 
 coincident with the geographical North and South. 
 
 A piece of iron, on being rubbed with a lodestone or with an 
 artificial magnet, exhibits no trace of magnetism when the magnet is 
 removed. However, if the magnet touches the piece of iron, it (the 
 iron) becomes a magnet for the time being, but as soon as the magnet 
 is taken away it once more becomes merely a common piece of iron. A 
 piece of steel, on the other hand, is not so readily made magnetic, as 
 merely touching the steel with a magnet will not impart much of the 
 magnetic property to it. It would then appear that steel resists 
 magnetisation ; but, when magnetised, in spite of the resistance offered, 
 it remains magnetised with little or no tendency to return to the non- 
 magnetic state. Iron, on the contrary, seems to resist to a very small 
 extent, and therefore comes back just as easily to its non-magnetic 
 state. 
 
 The fact that a magnet may be broken into any number of pieces, 
 and yet each will still be a magnet, early suggested that the magnetic 
 properties were molecular. The theory known as the molecular 
 theory of magnetism states that every molecule cf a maffneUe substance 
 is a magnet. "Bj molecule is meant the ultimate particles of which a 
 body is composed. A substance is made up of an infinite number of 
 such particles, and some idea of the state of things may be gatiiered 
 by ima gining the appearance which would be presented by a large 
 number of spherical corks, connected to each other by means of shwt 
 pieces of elastic, immersed in water, some of the corks being tethered 
 to the bottom of the tank. The corks represent the molecules, and 
 the pull of the elastic, the force known as cohesion, which holds them 
 together, 8.e„ the pull whidb every molecule exerts upon every 
 other moleclule. These molecules are exceedingly small, so small 
 
Magnetisation. 15 
 
 indeed, that the head of a pin contains billions. If it were possible 
 to magnify a pea to the size of the world the molecule would 
 be magnified to about the size of a cricket ball. The efEect of the force 
 of cohesion is to keep all the molecules of a body together and make it 
 preserve its shape. Magnetism is, according to the molecular theory 
 due to the turning round of the molecules and placing them so that 
 they all tend to act in the same direction, i.e., so that one molecule 
 does not nullify the efEect of another. 
 
 NNS — NS — NS — NS - NS — NS Q 
 NS — NS — NS — NS - NS — NS D 
 
 Such an arrangement of molecules is represented above, where N 
 indicates the North-seeking and S the South-seeking poles of the 
 molecules composing a magnet. Each pair of small letters represents 
 one of the millions of molecules. The result of this movement is 
 indicated by the large letters N and S, i.e., the North poles all point to 
 the left and the South poles to the right, the result being the formation 
 of a magnet whose North pole is on the left. In steel the molecules 
 appear to be jammed and very difficult to turn round, so that when 
 once turned round, as shown above, the magnetic pull of each molecule 
 on the other is not sufficiently strong to turn them back to the non- 
 magnetic state as represented, thus : — 
 
 / NS — AS - NS — NS — NS — NS \ 
 V SN — SN — SN — SN - SN — SN / 
 
 In this case it will be seen that half the molecules point in one 
 direction and half point in the other ; contequently their combined 
 efEect will be nil. It will also be noticed that each magnet satisfies 
 the attraction of its neighbour. 
 
 The behaviour of iron is very difEerent to that of steel, as its mole- 
 cules are easily turned, but immediately the magnetising force is 
 removed the molecules, acting upon each other, turn back to the non- 
 magnetic state. Magnetisation merely consists then in facing round 
 these molecules ; therefore, when all are faced round the magnet is as 
 strong as it possibly can be, and the iron is then said to be saturated. 
 These theories, supported by innumerable facts, which it is not pro- 
 posed to deal with here, are, however, only of interest to the practical 
 man in so far that they are aids to the remembrance of experimental 
 facts in logical order. 
 
 If a magnet be brought near to a piece of iron it will make that piece 
 of iron into a magnet for the time being. To see why this is so it will 
 perhaps be as well at this juncture to introduce " lines of force" 
 Take a piece of paper and place under it, parallel to each other, two 
 bar magnets having the South pole of the one opposite the North pole 
 
 b2 
 
16 Lines of Force. 
 
 of the other, and spread iron filings over the paper by means of a flour 
 dredger, or other similar contrivance. Having spread a fine layer of 
 filings, tap the table and examine the form of the curves which the 
 filings mark out. They form chains from the North pole of the one to 
 the South pole of the other magnet, at either end. Now reverse one of 
 the magnets so that the North pole of the one is opposite the North of 
 the other, and proceed as before. The filings again form curves, but 
 instead of the lines of the magnet linking with those of the other they 
 turn outwards away from each other. The filings indicate by their 
 arrangement that there is afield of force, and merely serve the purpose 
 of indicating the direction of these hypothetical lines of force and the 
 configuration of the field. 
 
 The peculiar disturbance created in the region of a magnet is known 
 as a magnetic field, and its strength is determined by the amount of 
 attraction or repulsion which it exerts on magnets. Lines of force are 
 hypothetical lines along which an isolated magnetic pole (if such an 
 arrangement could be obtained) would move if perfectly free to do so. 
 A magnetic field is then any space in which lines of force exist. These 
 lines of force in passing through the filings cause each filin g to become 
 a magnet, and in the first instance (unlike poles opposite) the majority 
 of the lines go from the North pole of the one magnet to the South 
 pole of the other, and again from the South to the North at the other 
 end of the magnets. It is noticed that the filings do not collect alone 
 the magnets themselves, the reason of this being that iron and steel 
 conduct the lines of force far better than air, and very few lines stray 
 out of the steel into the surrounding air. 
 
 A line of force does not begin somewhere and end nowhere. Every 
 line is a closed curve, the direction of which is from South to North 
 inside the magnet, and North to South outside ; hence the North pole 
 is sometimes termed a positive and the South a negative pole. In the 
 first-mentioned case the lines start from the inside of the one magnet 
 out into the air, there meeting and joining with the lines proceeding 
 from the second magnet, which are going in the same direction, along 
 this second magnet, out into the air at the other end, and thus by 
 joining the lines from the first magnet, complete the magnetic circuit. 
 In the second case the lines of force tend to cross each other, repulsion 
 takes place, and the lines of each magnet form curves quite distinct 
 from those of the other. The filings serve to indicate the direction of 
 the lines passing through that part of the air where they (the filings) 
 are strewn ; therefore, the central portions of the magnet will be 
 comparatively free from filings, owing to the better conductivity of 
 steel. 
 
 The difference between a magnet and apiece of non-magnetic material 
 is, that the magnet possesses lines of force, whereas the non-magnetic 
 
Voltaic Cells. 17 
 
 material does not ; also, that in the latter case lines passing through it 
 do not render it magnetic, as its molecules are not magnets. If lines 
 pass from right to left through a bar of iron, at the right-hand side 
 of the iron there will be a South, and at the left-hand side a 
 North pole. The statement is identical with that contained in the 
 subsequent paragraph, and if the reader will draw a couple of bars in 
 a straight line with each other — one representing a magnet and the 
 other a piece of iron — he will, by filling in the directions of the 
 lines of force due to the magnet, observe the direction in which they 
 pass through the iron bar, and thus deduce the induced polarity of 
 that iron bar. 
 
 Induction takes places along the lines of force, and the induced 
 pole is opposite in character to the inducing pole ; or, to give an example, 
 a North pole approached to the right-hand side of an iron bar causes 
 that end to become a South pole, the result being attraction. We thus 
 sse that the attraction between a magnet and a piece of iron is the 
 result of previous induction. A line of force passes just as easily 
 through brass, wood, glass, etc., as through air ; hence, the placing of a 
 pieca of copper or other non-magnetic substance in the way of a 
 magnet will not alter in any way the effect it is producing. From what ■ 
 has been already said on the subject of lines of magnetic force, it will be 
 seen that the molecules of a magnetic substance arrange themselves 
 along the lines, and that to magnetise a piece of steel the magnet 
 should be stroked along in one direction only. The end at which the 
 stroke ceases will be opposite in polarity to the pole with which the 
 steel is rubbed.* 
 
 Electricity. 
 
 The arrangement known as a voltaic cell consists essentially of two 
 dissimilar metals immersed in a liquid which wUl form a chemical 
 compound with one of them, e.g., a piece of copper and a piece of zinc 
 immersed in dilute sulphuric acid. For the present this primitive form 
 of cell will serve the purpose of illustrating one method of producing 
 the effect known as an electric current. A battery is merely a 
 collection of cells properly joined up. 
 
 If the zinc end of the battery be joined to one end of a copper wire, 
 and the copper end to the other extremity of the wire, it (the wire) 
 is no longer a common piece of copper wire as it has become endowed 
 with strange properties, the immediate cause of which is that it 
 conveys an electric current. It will also be found, if the battery be 
 
 *For further iDformation on the subject of Magnetism see the "Electrician 
 Primers" or Sylvauus Thompson's "Electricity and Magnetism," chapter ii. 
 
18 Effects of Current. 
 
 sufficiently large, that the wire grows sensibly warm, but this rise in 
 temperature would depend on the method of its exhibition. A S^^'^ 
 of tepid water would possess more heat than a red-hot nail, yet the 
 effects of the two on the flesh would be startlingly difEerent. So it is 
 with the heat generated by an electric current, the best method of 
 exhibiting it being to make the wire as small and thin as possible. 
 The result of motion is friction, and friction always generates heat, 
 hence it was thought that the effects classed under the heading of 
 voltaic electricity were due to the passage of something through or 
 along the wire. Whether such is the case or not is of little consequence 
 to the practical man. If the analogy that electricity behaves like water 
 is helpful, by all means adopt it, but care should be taken to always 
 remember that it is an analogy, and that a perfect analogy is an 
 identity, therefore there must be points where the analogy breaks 
 down. The discussion as to what electricity may really be is far 
 beyond the scope of this book, and therefore will not be entered into. 
 
 The next effect of a wire conveying an electric current which we 
 shall consider is the magnetic effect. Bound the wires there are lines 
 of force," which fact may be proved by placing the wire in a heap of 
 iron filings. The lines of force cause each filing to become a maenet, 
 attract its neighbour, and so on all round the wire, thus forming 
 chains round its circumference, the result being that the filings still 
 retain their positions when the wire is lifted up. The wire does not 
 in any way attract the filings, its lines of force merely cause the 
 filings to arrange themselves round it, thus causing them to present 
 the appearance of adhering to the copper. Unhappily this experiment 
 can only be reproduced by using currents of about forty-five amperes, 
 or about two thousand times as much as is usually employed on 
 telegraphic circuits. It is said that by using very fine filings the effect 
 may be obtained with about ten amperes. 
 
 A compass needle or a suspended magnet will always point to the 
 North, force being required to displace it from that position. In 
 order to illustrate the meaning of this, join the table and the Xorth 
 pole of the compass needle with a piece of stretched elastic, and treat 
 the South pole similarly, the elastic being fixed in such a manner as to 
 form a straight line with the needle and the other elastic string. To 
 twist the needle round we have to overcome the pull of the elastics. 
 Let us now consider the best methods of applying force to accomplish 
 this object. Clearly the best way is to pull the two ends of the needle 
 in different directions, and to pull at right angles to the positions in 
 which the needle is lying. The pull of the elastic represents the effect 
 of the earth's lines of force. A magnet placed at right angles to a 
 compass will tend to send lines of force straight across the needle 
 attracting one pole and repelling the other, or if the earth's field were 
 
Magnetic Field. 19 
 
 not present would place the magnet in a line with it in just the same 
 way that the earth, which may be regarded as a large magnet, holds a 
 compass North and South. 
 
 We might illustrate the pull of the magnet on the needle by 
 two or more pieces of stretched elastic fastened to the ends of 
 the needle, but fastened to the table at right angles to the elastics 
 representing the earth's pull, and also fixed on opposite sides of 
 the needle. Here we have then two tendencies, the needle being 
 urged by two forces at right angles to each other. The position which 
 the needle takes up will then be dependent on the relative magnitude 
 of these forces. If the reader will make a rough sketch of these 
 arrangements and label each part of his sketch in accordance with 
 the foregoing, its meaning and significance will be at once perceived. 
 To deflect a compass needle from its North and South positions, 
 lines of force tending to pass across it, or in other words an opposing 
 magnetic field is requisite. 
 
 It has already been stated that round a wire conveying a current 
 (whatever a current may be) there exist lines of force in concentric 
 circles at right angles to its leugch. Bound the wire, then, there is a 
 magnetic field, and if we set the magnetic field which exists round our 
 wire to oppose the effect of the earth's field on our compass, we shall 
 obtain an alteration in the position of the needle. The difEerence 
 between a wire conveying a current and one not doing so, is the 
 absence of the magnetic field in the latter case, and we have thus a 
 method of exhibiting the passage of an electric current along a wire. 
 The way to place our wire so ihat these lines oppose those due to the 
 earth is to fix the wire above or below in the same straight line with 
 the needle. The wire's lines pass round the wire at right angles to the 
 length, thus tending to cut the earth's lines at right angles, and would 
 in this way produce the desired defiection. 
 
 There is, however, a slight reservation which must be made in 
 explaining the action of magnetic fields on magnets. A compass needle 
 will set itself along magnetic lines of force, but in the cases mentioned 
 it is said that the lines from the current tetid to cross those from the 
 earth. The word "tend "is printed in italics because lines of force 
 can not cross each other, for did they do so we should have two 
 different strengths of magnetic field at one point, which is manifestly 
 impossible. The field due to the earth is altered in its distribution 
 by the field due to our current, and the needle sets itself in the direction 
 in which the re-distributed lines now stand, and we have thus a method 
 of indicating the existence of what is termed an electric current. It 
 would be found that the lines of force round the wire would change 
 in direction on reversing the battery, i.e., by connecting the end of the 
 wire which was joined to the copper pole of the battery to the zinc. 
 
20 Conductors and Noiv-condiictors. 
 
 and the zinc to the end to which the copper was connected. If in the 
 first case the lines pass round in the same direction as the hands 
 of the clock, the north pole of the needle would move to the left. 
 On reversing the battery the lines round the wire would be in the oppo- 
 site direction, therefore the needle would move to the right. We see 
 that there is a reversal in the direction of the lines round the wire, and 
 from this fact we should consider it probable that the direction of 
 passage of the electric current was changed. That the direction of the 
 current is altered may be proved by various experiments mostly coming 
 under the heading of static electricity, or electricity at rest. Current 
 electricity is electricity in motion, and it is with this part of the sub- 
 ject that we have mostly to deal. 
 
 In speaking of directions it is best to arrive at some understanding so as 
 to mak e the expression of a direction as simple as possible. For instance) 
 in speakingof directions of rotation, "clockhandwise" is termed positive! 
 and "anti-clockhandwise" negative, an excess of anything is repre- 
 sented by + and a deficiency by — , and motion takes place from jKJsitive 
 to negative. Electricity may be regarded as always trying to supply a 
 deficiency. The copper pole of the battery is called the positive, and 
 the zinc the negative pole, the direction of the current being from the 
 copper along the wire to the zinc. A current going from SotUh to 
 North, over the compass causes the North pole of the needle to turn to 
 the West. This fact may be called to mind by the word " SXOW," 
 which word comprises the first letter of the most important words in 
 the above sentence. Above the wire the lines will be going in an 
 opposite direction to those at the bottom, therefore if the wire convey- 
 ing the curreut be placed under the needle, the direction of the deflection 
 will be reversed. The best way to ascertain the direction in which the 
 needle will turn is to place the outstretched right hand between the 
 wire and the compass with the palm towards the needle, so that 
 were it the wire, the current would pass from wrist to fingers. The 
 outstretched thumb will then indicate the direction in which the 
 North pole will ttum. 
 
 If we join the poles of our cell by a piece of glass we shall find 
 there is no current through it, for it will not affect our magnetic 
 needle. When a current flows through a circuit that circuit is said to 
 be completed. In the present case our circuit is not complete, for 
 there is no current through it. Glass then will not permit the effect 
 known as an electric current to take place along it, therefore glass 
 is said to be a non-couductor, and copper which will permit the pass^e 
 of a current is termed a conductor. All substances may be divided 
 into these two groups, viz., conductors and non-conductors. Of all 
 known substances silver conducts electricity most freely, and is 
 therefore said to be the best conductor of electricity. Copper follows 
 
Ohm's Law. 21 
 
 Tery closely at its heels, and indeed we may rank all metals and some 
 few fluids amongst good conductors. A non-conducting substance is 
 sometimes called an insulator. 
 
 We have seen that the principal effect of a current is the magnetic 
 effect, and it is by this effect that we know of its existence, hence the 
 most important methods of measurement depend upon it. Before we 
 can compare currents we must have standards just in the same way 
 "that we have standards of lengths, weight and time. Electricity, like 
 everything else, will not move without a motive agency. For instance, 
 water will not move unless pressure of some sort is brought to bear 
 upon it. The word which stands in the same relation to electricity as 
 pressure to water is electromotive force. It has already been stated that 
 some conductors will not allow electricity to pass through them as 
 readily as will others — ^that is to say, they offer more resistance to the 
 passage of the current. If we have a pipe of fixed size with a given 
 pressure, a certain quantity of water will pass through it in a given 
 time, an increase in the pressure applied will result in a larger flow 
 of current. Similarly a larger pipe, corresponding to a circuit of lower 
 resistance, will, with the same pressure, permit a larger flow of water 
 than the smaller pipe. This hydrostatic analogy will, perhaps, render 
 ■the connection between electrical pressure or E.M.F. current and 
 resistance more obvious. The following are the electrical units : 
 
 Current The ampere 
 
 Electromotive Force ... „ volt 
 Resistance... „ ohm 
 
 The Ohm. — This is the resistance offered by a column of mercury 
 106°3 centimetres long, having a sectional area of 1 millimetre, at a 
 temperature of 0° C, I.e., at the temperature of melting ice. This is 
 about the resistance offered by 76i yards of gutta-percha covered 
 wire used to join up instruments. The B.A. ohm, which, up till this 
 year was adopted by the Post Office, was only "9866 of the above true 
 or standard ohm. 
 
 The "Volt. — The E .M .F. of a Daniell cell is rather more than one volt. 
 
 The Ampebe. — This is the current which will flow through a circuit 
 having a resistance of 1 ohm when urged by an electromotive force of 
 1 volt. The ampere is too large a unit for telegraphic purposes, and it 
 has therefore been sub-divided into milli-amperes (milU-ampere = one- 
 "thousandth part of an ampere). 
 
 Ohm discovered that the following relations existed between these 
 units : The strength of current through any circuit varies directly as the 
 electromotive force and inversely as the resistance, or, stated in a more 
 definite manner, the number of amperes flowing through a circuit is 
 
22 Ohm's Law. 
 
 equal to the number of volts of electromotive force divided by the 
 number of ohms of resistance in the entire circuit, or : — 
 Current J-g-^-f"'- i.e., =|E = -f, and E = C X R 
 where C, E, and E are respectively the current resistance, and electro- 
 motive force in the circuil 
 For instance, an B.M.F. of 12 volts applied to a circuit of 6 ohms 
 
 resistance will produce a current of g^ -= 2 amperes. If the E.M.F. 
 
 in a circuit be 30 volts and the current 10 milliamperes, its resistance 
 
 must be .-jjjg- ^ 3000 ohms. If the current is 2 amperes and the resistance 
 
 is 10 ohms, then the E.M.F. in the circuit will be 10 X 2 = 20 volts. 
 These quantities should all be expressed in volts, amperes, and ohms. 
 
Electricity in its Application to Telegraphy. 23 
 
 CHAPTER II. 
 
 Batteries. 
 
 An electric current represents a certaia amount of energy. Energy 
 cannot come from nowhere, and it is not in the power of min to create 
 it. All that we can do is to apply the energy with which the world is 
 endowed to desired objects. A battery is a piece of apparatus which 
 in return for chemical energy gives us electrical energy, and the sooner 
 this fact is realised the better we shall be prepared to understand 
 precisely how this object is realised. A cell supplies us with current, 
 i.e., electrical energy, by robbing the chemicals of which the cell is 
 composed of an equal part of their energy ; or, to put the matter more 
 simply, an electric current is the result of a chemical degradation of 
 the constituent parts of the cell producing the current. 
 
 In the simple cell, which has previously been described, the sulphuric 
 acid attacks the zinc, and with it forms a chemical compound known 
 as sulphate of zinc. If, however, we put a cell of this description to 
 practical uses, we shall find that the current grows rapidly weaker, till 
 at last it almost ceases. The reason of this is that hydrogen gas, which 
 in the re-arrangement of the chemical compounds, is liberated at the 
 copper plate, forms a layer of gas over it, thus covering it up. Then 
 again, the impurities in the zinc, such as iror, tin, arsenic, etc., will 
 form little batteries with the zinc, and waste it away without any 
 equivalent of work done in the external circuit. This evil is known as 
 local action. The words " external circuit " are used in contra-distinc- 
 tion to the " internal circuit," or path of the current inside the cell 
 itself. The trouble to which the impurities in the zinc give rise are 
 got rid of by the process known as amalgamation, which consists in 
 coating the zinc with mercury, thus making it behave as though it were 
 che mically pure. 
 
 The first difficulty with regard to the hydrogen, known as 
 polarisation, is far more serious. The hydrogen forms a non-conducting 
 
24 Ghemistry. 
 
 film over the copper plate, which, in addition to reducing the surface 
 of the copper, also behaves towards the copper like another zinc 
 plate, tending to stop the action of the cell in this way, and has to be 
 got rid of before we can have anything approaching a steady current. 
 In order that the chemical actions of the various cells may be appre- 
 hended, a few notes on chemistry may be desirable. 
 
 A lump of sugar may be pounded with a hammer until the particles 
 are very small, and by using a mortar and pestle their size can be still 
 farther rednced. If, however, these particles are examined by means 
 of a microscope, or even by an ordinary lens, it will be found that 
 the particles have still some size, and evidently, could be made yet 
 smaller. If the sugar is dissolved in some warm water the particles 
 wUl become so small as to be invisible, but we know by the taste that 
 the sugar is there. We may dilute it largely and still recognise the 
 sugar, which is evidently distributed equally throughout the liquid. 
 The question now arises as to whether the sugar is capable of 
 unlimited division. Experiments prove that it is not, but that ulti- 
 mately a particle which will no longer split up and yet remain sugar 
 wOl be arrived at. 
 
 This is called a molecule, and is therefore the smallest particle which 
 can exist free. Sugar is known to consist of carbon and the elements of 
 water, i.e., hydrogen and oxygen, and if a little strong sulphuric acid 
 be added to the sugar solution the acid will combine with the water, 
 black carbon will be liberated, and the vessel will become very hot. 
 It may be remarked in passing that this heat is caused by the chemical 
 action, and is always given off during chemical union. Now it is clear 
 that the molecule has separated into still smaller particles of sub- 
 stances quite distinct from sugar. These particles are called atoms, and 
 every molecule is so split up. The carbon produced has not yet been 
 decomposed, and is therefore called an element. It is evident that com- 
 pounds are made up of elements, but it is also as certain that 
 elementary atoms dq not exist alone but in groups or molecules, 
 generally of two atoms, but in one or two cases of one, two, three, four 
 or even six atoms. An atom may therefore be defined as the smallest 
 particle of matter that can enter into combination. An atom of 
 hydrogen is represented by H, but in the free state as a molecule as 
 Hz. An atom of oxygen is represented by O, a molecule by O2, and a 
 molecule of ozone by O3, showing that in this gas three atoms are 
 condensed into the same space as two of ordinary oxygen. 
 
 In order to avoid writing the names of each of the elements composing 
 a compound a sort of chemical shorthand called symbols is used, thus 
 Hydrogen is represented by H ; Sulphur, S ; Copper, Cu ; Carbon, G 
 Oxygen, O ; Manganese, Mn ; Zinc, Zn ; Mercury, Hg ; Chlorine, CI 
 Potassium, K ; Chromium, Cr ; Nitrogen, N ; Lead, Pb ; Sodium, Na, 
 
Chemistry. 25 
 
 A number placed after a symbol indicates that number of atoms are 
 to be considered. For instance, a molecule of sulphuric acid consists 
 of two atoms of hydrogen, one of sulphur, and four of oxygen. In 
 order to write this quickly each element has been given a distinctive 
 symbol, and to indicate the number of atoms in each element in the 
 compound a figure stating the number is placed after each element. 
 Where, however, there is only one atom the figure " 1 " is understood 
 and is not written. Sulphuric acid is therefore HaSOj. In order to 
 express more than one molecule of a substance a large figure is placed 
 before the compound thus — 2NHiCl. This means two molecules of 
 NH4CI, or two atoms of nitrogen, eight of hydrogen, and two of 
 chlorine. 2H2 means two molecules of hydrogen, each of which con- 
 sists of two atoms, i.e., four in all, 4Cr2(S04)s is the symbol for four 
 molecules of chronium sulphate, each of which consists of two atoms 
 of chromium, three of sulphur and twelve of oxygen. 
 
 Chemical compounds are known by the thousand, but there are also 
 groups of elements which have no independent existence, but which 
 can take part in chemical reactions and be transferred from one com- 
 pound to another. These groups are called radicles. An atom of 
 hydrogen H or oxygen are elementary radicles. SOi, NOs, HO are 
 compound radicles. Radicles are also divided into metallic as Na,. 
 Zn, NH,, and acid as CI, SO4, NO3. 
 
 Compounds are formed by the union of these radicles in almost every 
 conceivable way, but the great majority belong to the three following 
 groups : — 
 
 1, Acids which are extremely active chemical compounds and are 
 
 formed by the uni6n of hydrogen, H, with an acid radicle, as : — 
 H2 SO4 Sulphuric Acid, 
 HNOs Nitric Acid, 
 HCl Hydrochloric Acid, 
 
 2, Bases. Hydrates or hydroxides are formed by the union of the 
 
 radicle hydroxyl HO with a metallic radicle, as : — 
 Potassium hydrate KHO, 
 Ammonium hydrate NH4HO, 
 
 3, Salts are formed by the union of a metallic and an acid 
 
 radicle : — 
 
 Zn SO4 Zinc Sulphate, 
 
 NH4 CI Ammonium Chloride, 
 
 The metals are often called positive elements and the non-metals 
 negative, or still better a list of the elements can be made, the highest 
 being the typical negative element oxygen, and at the bottom the 
 
26 Chemistry. 
 
 typical positive element potaasinm. In this list each metal is negative 
 to those below and positive to all above it. Also the further they are 
 apart the more intense the chemical affinity. The radicles may also be 
 classified as + positive and — negative, thus SO< is negative because 
 it is combined with the positive H, and the definition will become : — 
 
 An acid formed of hydrogen and a negative radicle, 
 . A hose formed of hydraxyl (HO) and a positive radicle, and 
 A salt formed of a negative and a positive radicle. 
 
 Now to split up a compound requires energy, and when separated they 
 possess potential energy. When the elements reunite the energy is 
 reproduced, and may exhibit itself in the form of heat, light, electricity, 
 etc. Thus the sun for ages decomposed the carbonic acid gas in the 
 coal forests, and with the carbon the plants were built up. This energy, 
 after lying dormant for millions of years, is now utilised by us in the 
 form of heat, really the original heat poured out by the sun ages ago. 
 
 Thus we see man can neither create nor destroy energy, and it is 
 equally certain that he cannot create or destroy matter. All he can 
 do is to change its distribution, and in so doing change its external 
 appearance. Hence, in any chemical equation, every atom taking part 
 in the reaction must be accounted for, i.e., if yon have ten atoms on 
 one side of yonr equation you must have ten on the other. The quan- 
 tities of each element must be the same, but their distribution may be 
 entirely different. 
 
 If a piece of pure zinc is placed in dilute acid, as for instance, sulphuric 
 acid, no action takes place. If now, a piece of carbon, platinum or 
 copper is also placed in the solution there is still no action. Lastly, 
 if while they are in the liquids the elements are connected outside 
 either by touching them together or joining them by a wire, the zinc 
 immediately dissolves, while hydrogen gas is given ofE from the surface 
 of the copper. At the same time the wire is very dtffierent to what it 
 was in its normal state. It will carry, deflect a magnetic needle as in the 
 telegraph, make soft iron into a magnet, and produce other electrical 
 phenomena according to the strength of the action. The energy of the 
 chemical action exhibits itself as electricity, and a current of electricity 
 is said to flow from the copper to the zinc outside, and from the zinc to 
 «opper inside, thus completing the circuit . 
 
 Several theories have been put forward to account for the current in 
 the wire, but there is nothing certain known of what takes place in the 
 -wire when the connection is made. The copper plate is the positive 
 pole and the zinc the negative, which can be remembered thus : 
 
 CoPper Positive, 
 ZiNc Negative. 
 
Chemistry. 27 
 
 The vessel is called a cell, and a number of cells put together and 
 joined up is called a battery. The action in the cell may be represented 
 thus : — 
 
 Before the poles are connected. After connection. 
 
 Zn + HjSO, = ZnSOi + H2 
 
 One molecule and One molecule of form One molecule and One molecule 
 
 of zinc sulphuric acid of zino sulphate of hydrogen. 
 
 The copper plate is not acted upon by the chemical constituents of 
 the cell, and is therefore omitted from the equation. 
 
 The reaction depicted by the above symbols represents a molecule 
 of zinc acted on by a molecule of sulphuric acid, the result being that 
 the zinc and the two atoms of hydrogen of the sulphuric acid change 
 places, when we have zinc sulphate with hydrogen free instead of 
 hydrogen sulphate (another name for sulphuric acid) and zinc free : — 
 
 The solution of zinc sulphate at once leaves the zinc, and being 
 heavier than water slowly sinks, so that fresh surfaces of zinc are con- 
 tinually being exposed to the action of the acid. The current will 
 cease when all the zinc has been consumed, i.e., reduced to zmc 
 sulphate. But in the case of the copper the hydrogen collects in a 
 thin layer and clings to the surface. This is called polarisation, and of 
 course interferes seriously with the current. The retnedy is to intro- 
 duce another liquid which will unite with the hydrogen and remove it 
 continuously. In the description of the various cells these will be 
 mentioned as they occur. In the instance we have used for our illus- 
 tration, that is the two elements zinc and copper and sulphuric acid, 
 the sulphuric acid is called the exciting fluid, and the liquid used to 
 remove the hydrogen is called the depolarising fluid, and may be a 
 solution of copper sulphate forming sulphuric acid and metallic copper. 
 The latter forms a layer on the original copper element and therefore 
 the action is not impeded. 
 
 According to the most recent investigations, when sulphuric acid 
 dissolves in water part of it at once separates into the radicles H2 and 
 SOi, or ions, as they were first termed by Faraday. These are in 
 equilibrium with themselves, and the copper and zinc, but directly the 
 connection is made and there is a means of escape for the energy : the 
 zinc attracts the SO4 and the hydrogen goes to copper. A fresh portion 
 of HjSOi decomposes, and so the action goes on continuously. 
 
 Similarly when sodium chloride, NaCl, is dissolved in water it will 
 immediately split up into the ions Na and CI. Now, if a current of 
 electricity is sent through this solution the Na is attracted by one pole 
 and is set free, forming NaHO, and the CI is attracted by the other 
 and escapes. Some more NaCl decomposes, and the action goes on till 
 all the salt is decomposed. 
 
28 The Daniell Cell. 
 
 Daniell. — Daniell decided to nse copper snipbate to form the 
 chemical compoond with hydrogen. The Daniell cell consists of an 
 outer jar or vessel in which is placed the zinc plate. Inside this outer 
 Tessel is placed another jar of nnglazed earthenware, which latter is 
 pervious to water, and is therefore called a porous pot, its object being 
 to prevent admixture of the two solutions without stopping the 
 current. Inside the porous pot is a copper plate immersed in a 
 saturated solution of copper sulphate with a few undissolved crystals, 
 to prevent the solution becoming weak as the cell is being worked. 
 We have thus exactly what we want, viz., a conducting liquid which 
 will get rid of the hydrogen and not interfere with the action of the 
 cell. The space which contains the zinc (between the outer jar and 
 the porous pot) is now filled up with soft water, and if the copper be 
 joined to the zinc by a piece of wire a current will circulate through it. 
 In doing this the hydrogen will make its appearance at the copper 
 plate, where it will be immediately seized upon by the copper 
 sulphate, thus forming pure copper (which is deposited on the copper 
 plate) and sulphuric acid, which finds its way into the outer jar. 
 The action of the cell at first, then, is to form zinc oxide and 
 liberate hydrogen at the copper plate, but as has already been 
 pointed out, the copper sulphate combines with it, thus forming 
 pure copper and sulphuric acid. The cell is short-circuited 
 until the solution in the outer jar has become dilute sulphuric 
 acid, or if the cell is wanted immediately a little sulphuric 
 acid is added to the water, and it is from this point that we shall 
 trace its action. The action in the outer jar is, then, the same as that 
 of our simple cell, viz., 
 
 Zn + HjSOi = ZnSO, + H2 
 
 These two atoms of hydrogen appear at the copper plate, meet the 
 copper sulphate and unite with it, thus : — 
 
 H2 + CUSO4 = H2SO4 + Cu 
 
 Two atoma of and One molecnle of form One molecule of and One Molecule 
 
 Hydrogen Copper Sulphate Hydrogen Sulphate of Copper. 
 
 Hydrogen sulphate is merely another name for sulphuric acid an 
 has been employed to render the changes in distribution of the chemical 
 compounds more obvious to the unchemical reader. 
 
 We thus see that in this form of cell the strength of the sulphuric 
 acid solution is always kept up to standard, for as fast as the sulphuric 
 acid is used up in the outer cell it is given back from the inner. 
 
 It has already been stated that the zinc is the negative and the copper 
 the positive terminal of a cell. As the zinc is the element which is 
 degraded and which furnishes the electrical energy we term zinc the 
 positive element . In the external circuit the current flows from copper 
 
The Leclanche Cell. 29 
 
 to zinc, but in order that the circuit ma? be complete (without there 
 can be no current) the current must flow from zinc to copper inside the 
 cell, hence to express these two somewhat conflicting facts we say that 
 the zinc is the positive element, but the negative pole or terminal of the 
 cell. Similarly the copper is the negative element but the positive pole. 
 
 The positive metal or element is always the one which is acted on or 
 consumed by the liquid. The liquid which consumes the zinc, when 
 the circuit is completed, is called the exciting fluid, as opposed to the 
 liquid taking up the hydrogen which is called the depolarising fluid. 
 The hydrogen causes polarisation unless means are employed to get 
 rid of it, or in other words depolarise the cell. 
 
 Leclanche. — In this form of cell zinc is selected for the positive 
 and carbon for the negative element. The carbon, which is placed in 
 the porous pot, has at its top a terminal fixed to it by means of a lead 
 lug, the remainder of the space inside the pot being filled up with pebble 
 manganese (manganese dioxide). The outer jar contains the porous 
 pot and the zinc element which is generally made in the form of a rod. 
 The outer cell is filled up with a solution of salammoniac and the 
 inner with water, or water with a little of the salammoniac solution 
 added to improve its conductivity. The porous pot, which will not 
 ordinarily conduct, must be wet in order that it may permit the passage 
 of the current. When the cell is thus made up there will be no action 
 whatever, as the salammoniac will not attack the zinc, in fact will not 
 have the slightest effect on it when the circuit is not closed. When the 
 circuit is closed the chemical compounds are entirely altered in their 
 distribution. Salammoniac is a compound of ammonium and chlorine, 
 and is represented by NHjCl (one atom each of nitrogen and chlorine 
 together with four atoms of hydrogen). When the circuit is closed we 
 shall find the zinc and the ammonium in the salammoniac change places 
 and thus part of the zinc will be consumed to form zinc chloride (a sub- 
 stance that readily dissolves in the salammoniac solution), ammonia 
 (which, however, immediately combines with water to form ammonium 
 hydrate) and hydrogen set free. This hydrogen appears at the carbon 
 plate, and has to be got rid of in order that it may not give rise to the 
 troubles described in the case of the Daniell cell. The depolariser used 
 in the Leclanche is the manganese dioxide packed round the carbon 
 plate . Immediately the hydrogen is liberated at the carbon plate it (the 
 hydrogen) seizeson the manganese, robbing it of an atom of its oxygen, 
 which combines with the oxygen to form water. The action may be 
 depicted thus : — 
 
 Outer Cell :— 
 
 
 
 
 
 Zn 
 
 + 
 
 2NH4CI 
 
 + 
 
 2mo 
 
 One moloule 
 
 together 
 
 two molecules of 
 
 and 
 
 two molecules 
 
 of zinc 
 
 with 
 
 Salammoniac 
 
 
 of Water 
 
 
 
30 The Bichromate Cell. 
 
 = ZnCh + 2NH4HO + Hj 
 
 form one molecule ol together two molecules of and two atoms of 
 Zinc Chloride with Ammonlnm Hydrogen. 
 
 Hydrate 
 
 Inner Cell : — 
 
 Hj + 2MnO. = MnzO, + H2O 
 
 Two atoms of and two molecules of form one molecule of and one molecule 
 Hydrogen Manganese Manganese at Water. 
 
 Dioxide Sesqnioxide 
 
 In the inner cell we see that the two molecules of manganese 
 dioxide taking part in the reaction have between them given away one 
 atom of oxygen to the hydrogen, thns forming water. On the other 
 hand, the manganese dioxide, instead of being two, has become one 
 molecnle of a lower oxide, namely, Mn^Os, the sesqnioxide of 
 manganese. Ammonimn hydrate is written NH4HO instead of NH5O 
 from chemical considerations into which it is not proposed to enter. 
 
 Fxjlleb's Mekctjry Bichromate . — This cell consists of a porous pot 
 containing a thick zinc rod flattened out at the bottom so as to stand 
 opright. The outer jar holds the porous x>ot and the carbon plate. 
 The porous pot is filled up with dilute sulphuric acid, the outer jar 
 containing the depolarising fluid, which in this case is a solution 
 composed of a mixture of sulphuric and chromic acids. Chromic acid 
 is dear, bnt bichromate of potash and sulphuric acid are cheap, and as 
 chromic add is formed when these two substances are brought together, 
 it is much cheaper to charge our own cell with bichromate and 
 sulphuric acid. As in the Daniell the action of the sulphuric acid on 
 the zinc results in the formation of zinc sulphate which dissolves in the 
 solution, and the liberation of hydrogen which makes its appearance at 
 the carbon plate where the chromic and sulphuric acids seize upon it 
 and with it form water and chromium sulphate. In the outer jar when 
 the sulphuric acid is added to the bichromate of potash the following 
 reaction takes place : — 
 
 K^Cr^O, + H,0 + H^SOi 
 
 One molecule of Bichromate and one molecule and one molecule of 
 of Potash of Water Sulphuric Acid 
 
 = K^SO, + 2H2CrO, 
 
 form one molecule of Potassium and two molecules of 
 Sulphate Chromic Acid. 
 
 The potassium sulphate is a harmless bye-product which dissolves in 
 the solution and gives no further trouble, therefore in the subsequent 
 expressions it will be ignored. 
 Inner cell : — 
 
 3Zn + 3HjS0. = SZnSO, + 3Hj 
 
 Three molecules and three molecules form three molecules and six atoms of 
 of Zinc of Sulphuric Acid of Zinc Sulphate Hydrogen. 
 
The Dry Cell. 31 
 
 This hydrogen makes its appearance at the carbon plate where it 
 meets the depolarising ilnid. 
 
 3H2 + 2HjCr04 + SaSO, 
 
 Sis atoms of and two molecules of and three molecules of 
 Hydrogen Chromic Acid Sulphuric Acid 
 
 = Ck(SO0» + 8H2O 
 
 form one molecule of and eight molecules 
 
 Chromium Sulphate of Water. 
 
 This is the reaction if the cell is working properly, but if the cell is 
 worked too heavily without being replenished, another set of reactions 
 will take place, the result of which will be the formation of crystals of 
 chrome alum upon the carbon plate. This has to be guarded against. 
 There is yet another feature about this cell which should be noted, and 
 that is the method of keeping the zinc well amalgamated. At the bottom 
 of the porous pot is placed two ounces of mercury, which keeps the 
 zinc bright and clean. A zinc freshly raised from the sulphuric acid 
 solution should appear bright like silver ; should this not be the case 
 more mercury ought to be added. The action in the outer jar is somewhat 
 <:omplicated, and it is therefore thought desirable to append a general 
 description of the action without symbols. The hydrogen, on coming 
 into the outer jar, meets the chromic and sulphuric acid, where it robs 
 the chromic acid of some of its oxygen, the remaining parts of the 
 chromic and sulphuric acids then re-combine to form chromium 
 sulphate and water. 
 
 The Gassnbr Dry Cell. — This cell, which made its appearance in 
 1888, was the first successful dry cell. As a matter of fact, these dry 
 oells are all modifications of the Leclanche, and are, needless to say, wet. 
 Were they not so their internal resistance would be infinite, con- 
 sequently they could not generate a current, so that the term dry cell 
 is somewhat of a misnomer. The cell consists of a zinc containing vessel 
 which forms the positive element, and contains the negative element 
 which is a block of carbon. The remaining space is filled up by a jelly 
 consisting of zinc oxide, ammonium chloride, sulphate of lime, 
 chloride of zinc, and water. It was found that these cells could be 
 regenerated by passing a current through them in the opposite direction. 
 This is the case with all dry cells, and they may be regarded as 
 accumulators or secondary cells, A dry cell as distinguished from an 
 ordinary wet cell consists of a sealed cell in which the electrolyte is held 
 in a damp gelatinous condition. It is equally obvious that no gases must 
 be liberated, or the cell will burst, a fate which overtakes a very large 
 number of a certain form of dry cell. 
 
 The E.C.C, Dky Cell. — In this form of cell, manufactured by the 
 Electric Construction Company, a round zinc containing vessel serves 
 as the positive, and a carbon block as the negative element, and is 
 
 C2 
 
32 The Dry Cell. 
 
 likewise a modification of the Leclanche. Round the carbon plate is 
 placed a damp black paste consisting of carbon dust, manganese 
 dioxide, together with the oxides of magnesium, calcium, and iron. 
 Bound this black paste, in the space between it and the zinc, there is a 
 white paste consisting of lime, ammonia, water, and salammoniac. 
 This is the construction of the cell, and it will be seen at once how 
 similar it is to that of the Leclanche. Probably its chemical action ia 
 also similar, but it will be complicated by the presence of the other 
 substances mentioned. 
 
 The Hellensen Dky Cell. — This cell consists of a hollow circular 
 carbon, the interior of which is filled with silicate cotton, and surrounded 
 first with a black paste, then with a white paste, and finally a square 
 zinc containing vessel. The whole cell is then packed with sawdust 
 into a slightly larger carbon box and the top is run in with pitch. 
 The black paste consists chiefly of carbon and manganese dioxide, 
 with the salts of silicon, magnesium, calcium, and iron. The white 
 paste consists of water, lime, ammonia, and the oxides of zinc and 
 magnesium. This cell is an exceedingly efScient one and is much used 
 by the Department as the local battery for working sounders and 
 double plate sounders. 
 
 The Obach Dry Cell consistsof a zinc cylinder of a thickness suitable 
 to the size of the cell, with a connection wire attached to the inside of 
 the cylinder. This latter contains a carbon rod, surrounded by a 
 depolarizer, composed of a mixture of powdered peroxide of man- 
 ganese and carbon intimately mixed and pressed into close contact 
 with the carbon rod. A brass terminal is fixed into the top of the 
 carbon in the following manner : — A verticle hole is provided in the 
 rod into which the pin of the terminal is inserted, and a molten alloy, 
 composed of bismuth, lead and tin, is poured around the pin. As this 
 aUoy expands slightly in cooling, it establishes a perfect contact 
 between the pin and the carbon. An exciting paste composed of 
 plaster of Paris and flour mixed with a solution of ammonium chloride 
 surrounds the depolariser, A space is provided in the upper part of 
 the cell to receive the gases and water which are formed during the 
 working of the cell, and this space is filled with sawdust as aa 
 absorbent for the moisture. The cell is sealed at the top with a layer 
 of bitumen, small holes being provided for the escape of the gases. 
 The cell is contained in a strong cardboard box impregnated with 
 paraffin wax, which makes it damp proof, and thus insulates the cell. 
 The cells are remarkable for their large output. 
 
 These dry cells all possess one very serious disadvantage, and that is 
 that once they become dry they are useless, and it is no uncommon thing 
 to see dry cells sufEering from this trouble being soaked in a bucket of 
 water. 
 
Electricity in its Application to Telegraphy. 33 
 
 CHAPTER III. 
 
 Electkical Measukements. 
 
 The magnetic field round a wire is strictly proportional to the 
 strength of current flowing in that wire, and in order to estimate the 
 strength of that magnetic field we must have a standard magnetic 
 field with which to compare it. If we compare the relative strengths 
 of the magnetic fields due to the respective currents, we shall then 
 know the relative strengths of the two currents. In order to ascertain 
 the relative strengths of the two magnetic fields we compare them with 
 the earth's field, which does not sensibly vary. 
 
 B 
 
 Figure 1 represents a single loop of wire traversed by an electric 
 current, the dotted circles with the arrows indicating the position and 
 direction of the lines of force. Let us place the coil in such a 
 position that the lines of force due to the earth pass from A to B 
 i.e., place it North and South. A compass needle points North and 
 South, and anything placed in a straight line with the position which, 
 
34 Tangent Galvanometer. 
 
 the needle would occupy is said to be in the magnetic meridian. The 
 position of the loop of wire may then be described by saying that it 
 lies in the magnetic meridian. A freely suspended magnetic needle 
 hung in the centre of the loop, will point to the North, i.e., will place 
 itself in a line with A and B. If now a battery be joined up, the 
 magnetic field due to the current will tend to place the needle in a line 
 with this magnetic field, i.e., the needle will tend to point East and 
 West. Here we have, then, twosmagnetio fields opposing one another. 
 If these magnetic fields are equal in strength the needle will take up a 
 position midway between the position North and South and the posi- 
 tion East and West, i.e., 45° from the North and ^oith, or normal 
 position. This position is, therefore, known as a deflection of 45°. 
 
 The earth's field has always its one particular value, hence, if the 
 current be reduced, the magnetic field due to it will also be decreased, 
 consequently the needle will be pulled more in the direction of the 
 earth's field (N. and S.) than in the direction of the current's field (B 
 and W.) ; but if increased it will be pulled more in the direction of 
 the current's field than in that of the earth, and it will then be seen 
 that the needle takes up various positions dependent upon the value 
 of the current strength in the loop. 
 
 Up to a deflection of 45° the current's field has more leverage on the 
 needle than the earth's field, because the current at 0° is pulling at 
 right angles to the needle. At 45° the current and the earth are each 
 pulling at an angle of 45° to the needlie, consequently they have an 
 equal leverage. From 45° to 90° the earth has by far the greater pull. 
 This will mean that up to 45° a small increment in current will produce 
 a greater increment in deflection than if the deflection was over 45°. 
 These relations may be expressed by saying that the current varies as 
 the trigonometrical tangent of the angle of deflection, in which case 
 twice the deflection will not mean twice the current. A table of 
 tangents must be used thus: With " A " current the needle deflects 
 30° and with " B " current 80°. What are the relative strengths of the 
 two currents? A table of tangents shows that the tangent of 30° 
 is expressed by the number -5774, and tan 80° by 5-6713; then "A" 
 current is to " B" as -5774 is to 5-6713, and not as 30 to 80. This is 
 briefly the theory of the tangent galvanometer, but in practise it is 
 very greatly simplified. 
 
 The scale of the galvanometer is not marked in equal divisions of 
 degrees whose tangents have to be hunted up, but in divisions which 
 are proportional to the tangents of the angles of deflection, hence 
 every current is proportional to the number of divisions over which 
 the needle passes. The zero positions of the needle are necessarily at 
 A and B, but as this part of the scale would be unreadable owing to 
 the shadow of the coil a light pointer is attached at right angles to 
 
Galvanometer Constant. 
 
 35 
 
 the needle, so that the zero position of the pointer is moved round to 
 East and West. The pointer is a fine aluminium wire which has the 
 advantage of being both light and non-magnetic. A piece of looking- 
 glass is let in underneath the scale, and when the pointer and its image 
 on the mirror coincide the observer's eye is directly above the pointer, 
 and hence it is not possible to read a deflection incorrectly. The 
 needle itself must be short in order that the tangent law may be 
 satisfied. Beversal of current entails reversal of magnetic field, and 
 hence of deflection. We have now all the essential features of a 
 simple tangent galvanometer. Such an instrument with but one turn 
 of wire would not, however, serve our purpose, as the currents with 
 which we deal would not be strong enough to produce readable deflec- 
 tions. The needle must be about one-twelfth of the diameter of the 
 ring, hence it is' not expedient to reduce this diameter beyond ten or 
 
 FiGUKE 2. 
 
 "t^K 
 
 twelve inches. The only way to increase the sensitiveness of the 
 instrument is to wrap our wire round and round many times, as 
 indicated in Figure 2. 
 
 The arrows indicate the direction of the magnetic field. It will be 
 seen also that near the centre of the coil the filings are clustered 
 together very thickly, thus indicating that the strength of the magnetic 
 field at this point is great. The number of lines of force will be 
 augmented with every turn of wire, i.e., a current of one milliampere 
 flowing through 1000 turns will produce the same magnetic effect as a 
 
36 Adjustment of Standard Cell. 
 
 current of one ampere flowing through one turn, and will therefore 
 give the same deflection. We thus see that the sensitivnees of the 
 galvanometer can be increased to any desired degree. 
 
 The tangent galvanometer used by the Post Office is of the skew 
 scale pattern invented by Mr. Eden. The arrangement consists of an 
 instrument set exactly as described above and then rotated through 
 60°. This gives a very much wider range as a sufficiently strong 
 current would pull the needle through to the other side of the ooU, and 
 then still further turn it to the extreme position. The divisions are 
 large at the centre of this scale, which is obviously a very great advantage. 
 
 The Daniell is used as the standard cell, and when made of 
 
 chemically pure materials its E.M.F. is 10947 B.A. volts. The 
 
 resistance of the tangent galvanometer is 320 B.A. ohms, and to this 
 
 is added another 750 B.A. ohms by means of a plug on the base of 
 
 the instrument. If the internal resistance of the cell is 24'7 B.A. ohms 
 
 r0947 I 
 
 the current through the galvanometer is 750-1-330 + 34-7 ^^ ioM 
 
 ampere equals 1 m.a. Such a current on a jtroperly adjusted instru- 
 ment produces a deflection of 80 divisions, hence each division 
 
 represents ^ ot a milliampere. The internal resistance of the 
 
 standard is usually only about 7 ohms, and a deflection of about 81^ 
 divisions therefore represents the current (1"016 m.a.) flowing through 
 the instrument in order that 80 divisions may accurately represent 
 one milliampere. The current has to traverse the internal resistance 
 of the cell, and this resistance must always be added to the 
 external resistance when calculating the current through a circuit. 
 
 On a brass stalk, fixed on to the ring of the instrument, is a small 
 filidJTig bar magnet, the purpose of which is to increase or reduce the 
 sensitiveness of the galvanometer as may be desired. The magnet 
 should always be placed in a straight line with the stalk and the zero 
 of the scale. If this is not done the pointer may not be at zero. If 
 the needle is not at zero when the magnet is removed the position of 
 the galvanometer should be changed untU it is. The magnet, in prac- 
 tice, is usually employed to increase the sensitiveness of the 
 galvanometer by producing a magnetic field at the centre of the coil 
 which opposes the earth's field, thus giving the current a weaker field 
 with which to compare its own. The North pole of the magnet faces 
 the North in this case, and it is obvious that the nearer the magnet 
 approaches the more is the earth's field reduced, and the more sensitive 
 does the galvanometer become, but by adjustment of this distance the 
 galvanometer is reduced to standard sensitiveness, viz„ when 80 
 divisions represent 1 m.a,, and this operation is termed " taking its 
 constant." 
 
Measurement of E.M.F. 
 
 37 
 
 The B.A. ohm has now been abandoned in favour of the true or 
 standard ohm, which has been found to be 10135 B.A. ohms. The 
 volt has also raised to 1"0135 B.A. volts, and the unit of current has 
 
 been left unchanged as i-^l^^'^^ = 1 ampere. For instance, the resist- 
 ance of the tangent galvanometer in standard ohms is 315$ ohms, and 
 the resistance coil is 740 ohms, hence the total resistance of the instru- 
 ment is 1055f standard ohms. If the resistance of the standard 
 cell is 24i standard ohms, the current through the circuit will be 
 
 1 m.a. : 
 
 r08 standard volts 
 
 =1 m.a. The E.M.P. of the standard cell 
 
 1080 standard ohma 
 
 is 1'08 standard volts, In order that the constant may be accurately 
 taken the resistance of the standard cell should be measured as 
 described in Chapter XXIV. and the deflection which must be produced 
 in order that 80 divisions may be equivalent to 1 m.a. is indicated by 
 the table below : 
 
 Table Showing Standard Cell Deflections Eequired with 
 Different amounts of Resistance to Ensore that 80 
 Divisions Equal One Milliamperb, 
 
 Total Res. of 
 circuit in stan- 
 dard ohms 
 
 Ditto ordinary 
 ohms 
 
 Deflection re- 
 quired on 
 outer scale. 
 
 1061 
 
 1064 
 
 1067 
 
 1070 
 
 1073 
 
 1076 
 
 1080 
 
 1075-4 . 
 
 1078-4 
 
 1081-5 
 
 1084-5 
 
 1087-6 
 
 1090-6 
 
 1094-7 
 
 81i 
 
 8U 
 
 81 
 
 801 
 
 80i 
 
 80J 
 
 80 
 
 1083 
 1097-7 
 
 79i 
 
 Measurement of Electromotive Force. 
 
 It may be mentioned that the Department are now introducing 
 sealed standard cells. They consist of an Obach dry cell joined up 
 with a resistance of about 1100 ohms to two terminals upon the outside 
 of a box, the top of which is fastened down and secured by a seal. 
 Upon the outside of the box is marked the number of divisions corre- 
 sponding to 1 m.a., but the 750(o plug should not in this case be removed. 
 
 Ohm's law states that -r ^ C where R is the total resistance in cir- 
 cuit. If the total resistance in the circuit has always one value we 
 may say that the E.M.F. is proportional to the current passing through 
 the circuit, i.e., the greater the E.M.F. applied to the circuit the corre- 
 spondingly greater will be the current produced. Consider a circuit 
 composed of a cell and galvanometer. The current through the 
 circuit is equal to the E.M.F. divided by the sum of the galvanometer 
 resistance and the resistance which the cell ofEers to the passage of the 
 
38 Equal Deflection Method. 
 
 current through it (called the cell's internal resistance). If the 
 resistance of a circuit is always the same, then the current which will 
 flow through it, due to given E.M.F.'s, will be directly proportional ta 
 those E.M.F.'s. 
 
 The internal resistances of cells usually range from "1 ohm to 20 
 ohms. If, now, we make the resistance of the galvanometer very 
 large we shall render the difference in internal resistance between 
 various cells negligible, consequently the B.M.F. of the cells will be 
 directly proportional to the respective currents which they will 
 send through the galvanometer, i.e., proportional to the deflections 
 produced. Let us suppose the resistance of our galvanometer 
 is 320 ohms. Two cells are given you, one has an E.M.F. of 
 one volt with one ohm internal resistance, the other an E.M.F, 
 of two volts with four ohms internal resistance, The first is 
 joined to the high resistance galvanometer. The current through 
 11 2 2 
 
 it is g2o J. 1 = "32I ^^PS's- I"! tlis second case = g^ , ^ ^ -^ 
 
 ampere. The current in the latter case is practically twice the former, 
 but if greater accuracy be desired, the 750 ohms may be inserted. Thus, 
 to compare E.M.F.'s we must employ a high resistance galvanometer, 
 or failing that, must put a few hundred ohms in our circuit in order to 
 prevent errors due to the internal resistances of the cells. In the case 
 cited, if we had used a galvanometer of negligible resistance the 
 currents would have been one ampere in the first case and half-an- 
 ampere in the second, whence we might have argued that the £.M,F, 
 in the first case was twice as much as in the second, instead of vice versa. 
 The currents are measured by the number of divisions passed over by 
 the pointer, and in future deflections will be instanced instead of 
 currents. 
 
 With the 750 plug out the Daniell gives 80 divisions, which 
 corresponds to 1'08 standard volts, therefore the E.M.F. of any cell in 
 
 standard volts is equal to-gj- multiplied by the deflection produced. The 
 
 resistance coil is marked 750, which is the value of the resistance in 
 B.A. ohms, and it is thought desirable to still call it the 750 ohm coil 
 on this account, although it is only 740 in standard ohms. It should 
 be noted that the resistance of the circuit should not be altered when 
 comparing E.M.F.'s, for instance, if the 750 ohms is in circuit when the 
 deflection with the standard cell is obtained, it must also be in circuit 
 when the deflection with the cell to be compared is observed. This 
 does not apply to the measurement of currents. 
 
 The foregoing method is in practice only applicable to the case of 
 single cells. In order to keep the current through a circuit at a 
 constant value while the E.M.F. is raised the resistance of that circuit 
 
Internal Resistance. 39 
 
 must be augmented in proportion. Join up a rheostat, a galvanometer, 
 and standard cell. The cell sends a current through the galvanometer 
 and rheostat. Adjust the resistance in the rheostat till a convenient 
 deflection of 30° to 50° is obtained. The resistance of the galvanometer 
 and rheostat together should be high, say 300 or 400 ohms, in order 
 that the internal resistance of the cells may not introduce errors. 
 With this apparatus it is easy to measure the E.M.F. of a 20-cell 
 battery. All that has to be done is to remove the standard cell, and in 
 its place insert the battery. The deflection on the galvanometer will 
 be very much greater, probably right to the end of the scale. Now 
 insert resistance in the rheostat till the same deflection is obtained 
 as that which was obtained from the standard cell. The current- 
 through the galvanometer is then the same as before, therefore the 
 E.M.F.'s of the cells are in proportion to the resistances in circuit in 
 each case. The advantage of this method, known as the equal 
 deflection method, is that any form of galvanometer may be used. A 
 numerical example is appended. 
 
 The standard cell, with a 60-ohm galvanometer and 440 ohms in 
 the rheostat in circuit, gives a deflection of 40°. A second battery with 
 4940 in rheostat under the same conditions gives the same deflectioni^ 
 Then I'OS volts is to the E.M.F. of the second battery as 60 + 440 is 
 
 to 4940 + 60; therefore its E.M.F. = I'OS x ™ = 108 volts. In 
 
 this case if the E.M.F. of the standard cell is given in standard volts- 
 the E.M.F. of the second cell will also be in standard volts, and it is 
 of no consequence whether the resistances are given in standard or 
 B.A. ohms. 
 
 The method is made use of in the Post Office battery testing 
 instrument (g.u.). 
 
 Another useful method is that due to Wiedemann, Put the 
 standard cell and the cell to be compared in series with the galvano- 
 meter copper to zinc and read the deflectioui Now reverse one of the 
 cells so that they tend to send the currents in opposite directions 
 through the galvanometer. If the cells have an equal E.M.F., the 
 second deflection will be nothing. If not, the E.M.F. of the greater ig 
 equal to that of the lesser multiplied hy the difference between the larger 
 and smaller deflections, divided by their sum, 
 
 -E. = E.^ 
 where dz is the larger deflection. 
 
 Meastjkement op Internal Resistance, — The half deflection 
 method is too well known to require more than brief treatment. A 
 cell, rheostat and galvanometer are placed in circuit, and resistance is 
 added by means of the rheostat till a convenient deflection is obtained^ 
 The resistance is then changed to a higher value until the deflection is 
 
40 Thomson's Method. 
 
 halved. If the current, due to a steady E.M.F., through a circuit is 
 halved, then the resistance of that circuit must have been doubled. 
 First of all we have the resistance of the battery and galvanometer 
 and the small resistance in the rheostat in the circuit. In order to 
 reduce the current to half its former value a resistance equal to the 
 fium of the internal resistance of the battery galvanometer, small 
 xesistance in the rheostat must have been added, hence the resistance 
 of the cell is equal to the amount by which the small resistance in the 
 rheostat is increased, minus the galvanometer resistance, added to the 
 email resistance in the rheostat. A convenient rule is that the 
 resistance of the cell is obtained by doubling the smaller resistance, adding 
 ike resistance of the galvanometer and deducting the sum from the larger. 
 
 With a galvanometer of 6 ohms resistance, and 10 ohms in the 
 rheostat, a deflection of thirty divisions is obtained. On raising the 
 resistance in the rheostat at 30 ohms, the deflection fell to half its 
 former value, viz., to 15 divisions. The resistance of the cell is 
 ttheref ore : — 
 
 30— [(2 X 10) + e] =4 ohms. 
 
 Thomson's method is, however, by far the better. Two rheostats 
 :and a galvanometer of known resistance are required. First, connect 
 up one rheostat, with no resistance inserted, in series with the cells 
 •(whose resistance it is desired to measure) and the galvanometer. 
 Now connect the terminals of the battery by means of the second 
 arheostat, and put in resistance until a suitable deflection is obtained. 
 If the resistance in this rheostat be nothing, then all the current will 
 :go through it and no deflection will be produced on the galvanometer. 
 It is, however, better to put a high resistance in at first so as not to 
 cause undue polarisation. Having obtained a suitable deflection 
 remove the rheostat which joins the terminals of the battery- Now 
 insert resistance in the remaining rheostat till the same deflection as 
 before is obtained, when the resistance of the battery will be equal to 
 the resistance in the first rheostat multiplied by that in the second rheostat 
 ■and divided by the galvanometer resistance. It is obvious that in all 
 these comparative tests all the quantities must be stated in either 
 B.A. or in standard units, and not in a mixture of both. During the 
 remainder of this work B.A. and standard units will not be dis- 
 tinguished, it being understood that to convert B.A. volts and ohms 
 into standard volts and ohms it is only necessary to multiply by "9866. 
 The resistance or apparatus, wire, etc., now in use are all stated in 
 the old or B.A. ohm, unless specially marked, " standard ohms." The 
 B.A. is approximately 1 J°/o less than the standard ohm. 
 
 Yet another method is to place the cells so that their E.M.F.'s 
 oppose each other, and measure them in the ordinary way by means 
 of the Wheatstone bridge, described in Chapter XIX. 
 
Electricity in its Application to Telegraphy. 41 
 
 OHAPTER IV. 
 
 Akrangbmbnt as Cells. 
 
 The E.M.F. of a cell is totally independent of its size, so that a cell 
 made out of a percussion cap and a grain of zinc would have just the 
 same E.M.F. as a cell as large as a Lancashire boiler, composed of the 
 same materials. There is, however, another important factor which 
 cannot be overlooked, and that is the internal resistance of the cell. 
 With a given pressure, it would not be possible to get as much water 
 through a small pipe as through a large one. This applies equally to- 
 conductors and electric currents. The current will not pass along 
 a very fine wire as readily as through a thick one, or with a given 
 E.M.F. it is impossible to get as much current through the thin 
 as through the thicker wire. It was found by experiment that th& 
 resistance of a wire depends upon four factors, viz. : — 
 
 1. Its length. 
 
 2. Its sectional area. 
 
 3. Its composition. 
 
 4. Its temperature. 
 
 If two wires are of the same size and coQiposition, but their lengths- 
 different, their resistances will be directly proportional to their lengths. 
 If they are of the same length, but different in gauge, then their 
 respective resistances will be inversely proportional to their sectional 
 areas, which is again proportional to the square of the diameter. 
 
 Numerical example. — If the resistance of five miles of wire "072. 
 inches in diameter is 50 ohms, what is the resistance of one mile of 
 similar wire '036 inches in diameter ? 
 
 Resistance of 5 miles "072 diameter wire = 50 ohms. 
 
 „ 1 ,, " ^ 2 " 
 
 1 -036 „ ='||5 X 10 =40 ohms. 
 
 It will be seen that '036 is half of '072. Therefore the resistance of 
 the former is not twice but the " square of two " times the latter, viz.,. 
 40 ohms. 
 
42 Joining up Cells. 
 
 The resistance of all metals rises when the temperature is raised, and 
 this variation, within the limits of our climatic conditions, may be 
 assumed to be proportional to the temperature without serious error. 
 The resistance of copper increases about '21 per cent, for each degree 
 Fahrenheit or about '38 per cent, per degree Centigrade. 
 
 It is evident, from a consideration of the above principles, that the 
 larger the cell the less resistance will it oSer to the passage of the 
 current, thus the greater the current obtainable from it. The larger 
 the plates the thicker is the liquid between them compared with the 
 -distance between, the former corresponding to sectional area in the case 
 ■of the wire, 
 
 In joining up several cells to work together we must see that the 
 -cells all tend to send the current in the same direction, and this can 
 ■only be done by joining the copper of one cell to the zinc of the next, 
 and so on. It will be seen that in such a case each cell adds its pressure 
 to that of the others, but in so doing it also adds that most 
 ■objectional quantity, its internal resistance. If, now, a large current 
 is wanted, say half-an-ampere, it would not be any use joining cells in 
 this way, because even if the external circuit in which the current is 
 required is of negligible resistance the cells would not supply the 
 current, as their own internal resistance would have to be overcome, 
 hence a thousand cells would not give any more current than one cell. 
 In this case the internal resistance of the cells must be decreased. 
 This can be done in two ways, viz., by making the cells very much 
 larger or by joining them up in parallel. 
 
 If we had two pumps, each having the same pressure and 
 delivering the same quantity of water, we should expect that if 
 they were both turned to send the water in the same direction 
 through one big pipe there would then be twice the quantity 
 of water going through the pipe. It is precisely the same with our 
 ■cells. Supposing we had a ten-cell Daniell resistance 5'4 ohms 
 per cell short circuited through a wire of negligible resistance. 
 
 .mi_ i. 1-08 X 10 _ _ ^„ ,. 
 
 The current =' 5-4 ^ 10 — "2 ampere. If the two copper terminals 
 
 and the two zinc terminals of two such batteries be connected, and the 
 ■circuit completed from copper to zinc as before there would be exactly 
 the same E.M.F. as from one battery, but there would be twice the 
 amount of current, or in other words, it will be precisely similar to a 
 ten-cell battery, each of whose cells has a resistance of half 5' 4 ohms. 
 Cells joined up in order to reduce the internal resistance are said to be' 
 joined in parallel or where both series and parallel arrangements are 
 used together they are said to be joined in multiple arc. In any case 
 of arrangements of this character count all the cells in series. Then 
 the total E.M.F. will be the E.M.F. of one cell multiplied by the 
 
Uses of Cells. 43 
 
 number of cells in series, and the total internal resistance will be the 
 internal resistance of one set divided by the number of similar sets. 
 
 For instance, if there are three batteries of 20-cell Daniell, each 
 having an internal resistance of 5 ohms per cell, joined in parallel, what 
 is the total E.M.F. and internal resistance of the combination. 
 
 Total E.M.F. == number in series x 1-08 = 20 x 1'08 = 21-6 volts. 
 
 Total resistance of one set ^ 20 x 5 = 100 ohms. Three sets in 
 
 100 
 
 parallel have a total resistance of ^^^ = 331 ohms. 
 
 When a battery is worked its E.M.F. and internal resistance alter to 
 a certain extent. The greater the current and the longer it is sustained 
 the greater are the chemical changes. Even when comparatively small 
 telegraphic currents are being supplied the value of the E.M.F. and 
 resistance change very considerably if the current is sustained for a 
 long time. The worst condition under which cells are allowed to 
 remain are as follows in B.A. units : — 
 
 minimum 
 Daniells, '96 volts = 10 \ fall 
 
 Bichromates, 182 „ =15% „ 
 Leclanches. 1-2 „ =25% „ 
 
 Internal resistance must come between 
 Daniell, large 5 ohms and 2*5 ohms 
 
 „ small 8 „ „ „ 
 
 Bichromate 1 quart, 4 ohms and 2'5 ohms 
 
 „ '3 quarts, 1'5 ohms and '5 ohms 
 
 Leclanche, ordinary, not more than 4 ohms. 
 The Leclanche polarises very quickly if the current is kept on for 
 any length of time because the hydrogen arrives at the carbon plate 
 <juicker than the manganese can get rid of it. On the other hand, 
 there is absolutely no local action, so that the cell is used where the 
 -working is not heavy, as for instance, on " receiving news wires," where 
 only BD's, etc. have to be sent. Again, at a local office where the 
 work is not heavy they are eminently adapted owing to the small 
 amount of attention they require. 
 
 The Daniell is by far the most constant battery in use as polarisation 
 is entirely eliminated, but owing to its low E.M.F. and internal 
 xesistance it is not adapted for supplying heavy currents. The small 
 size is used chiefly on simplex sounder circuits. The large size is used 
 for duplex circuits and for local batteries. Owing to the local action 
 and the attention required it is not used for the purposes mentioned 
 in the case of the Leclanche. 
 
 The Bichromate gives by far the heaviest currents, bat owing to its 
 local action and skilled attention required it is quite unsuitable for 
 .giving small currents intermittently. For working Quadruplex, 
 
44 Uses of Cells. 
 
 Wheatstone sending, Duplex, and Multiplex circuits, and for universal 
 working it is the only form of primary cell which could be employed. 
 
 Dry cells answer all the purposes for which the Leclanche is employed, 
 but their internal resistance are much less. It is of course eminently 
 adapted for transport, and is called into requisition in the case of most 
 of the special arrangements. A ten ohm resistance coil is put in the 
 box with every ten cells in order that too much current may not be 
 obtained from them if they are accidently short-circuited. 
 
 It is also used as a local battery and is placed in the hood behind the 
 instrument. 
 
Electricity in its Applicatioyi to Telegraphy. 45 
 
 CHAPTER V. 
 
 The Direct Sounder. 
 
 It has already been pointed out that by wrapping a wire round 
 many times its magnetic effect is greatly enhanced, i.e., there are a 
 greater number of lines of force passing through it. Iron conducts 
 these lines far better than air, hence if we place iron in a coil of wire 
 (called a solenoidj the number of lines passing through it will be 
 increased according to the quality of the iron core and armature. In 
 the case of the sounder it is necessary to pull down the armature with 
 some force, hence the iron is added. In order to have a magnetic 
 circuit (which is precisely similar to an electric circuit) the lines must 
 pass down one core, along the connecting piece of iron, up the other, 
 out into the air, along the armature, through the air back to the first 
 core. For this to be the case the arrangement of the poles must be 
 
 n — s 
 
 8 N. 
 
 n 
 
 -S 
 
 A hand rule was given on page 20 enabling us to predict the 
 direction in which the N pole of a needle would move when a wire 
 conveying a current was placed over it. The. same rule applies for 
 finding the N end of a solenoid. Place the right hand on the top of 
 the solenoid so that the current flows in the direction of wrist to 
 fingers, the outstretched thumb then points to the N pole. As the 
 other limb of the electro-magnet has to be opposite in polarity to the 
 first the direction of the current round must be reversed. This may be 
 done by winding the wire on in the opposite direction, or by winding it 
 in the same direction and connecting the end of the one to the end of the 
 other, or beginning to beginning . The reason that the armature is pulled 
 down is that poles of opposite polarity are induced in the armature, 
 as shown by the letters in italics. The easier and truer way to think of 
 the phenomenon is to remember that lines of force always try to 
 
 D 
 
46 The Direct Sounder. 
 
 shorten themselves by exerting a pull along their length, and in this 
 case they are shortened by the nearer approach of the armature. 
 
 In a great Department like the Telegraphs uniformity must be 
 obverved in connections, otherwise an immense amount of time will 
 be wasted in the case of a fault occurring. The station nearest to 
 London is the up-station. The up-station joins the positive pole of the 
 battery to line in order to efEect a mark, or as the practician crudely 
 puts it, " sends copper to line to mark." Stations inserted between the 
 up and down are called the intermediate stations. 
 
 The connections of the sounder and direct writer are very simple. The 
 line is brought to the centre of the single current key. The positive 
 terminal of the battery is connected to the front stop of the key and the 
 negative terminal to earth. The back stop is joined to the sounder, whose 
 other terminal is connected with the negative pole of the battery and to 
 earth. These are the connections for an up-station, but if the set is 
 required for a down station, the middle stop should be connected to 
 earth and the sounder and battery to line. A galvanometer (see 
 Chapter VI.) is always employed, and this is joined to the middle 
 contact so that a current will flow through it, through the key and 
 sounder to earth. The current flows through the earth in order to 
 complete the circuit. It will be seen that the down station sends 
 copper to earth instead of to line. The value of these arrangements 
 will become apparent later, when the discussion of the test box 
 arrangements is entered into. The galvanometer is usually of a 
 simple form, as its only object is to indicate the passage of a current 
 and not to estimate its strength. About 50 milliamperes are required 
 to work this instrument. The resistance of the instrument was 
 formerly 40 ohms but has now been reduced to 20 ohms. 
 
 The connections of the direct writer are precisely similar, and the 
 method of joining up will be clearly apprehended from the above. Z 
 and are joined to the negative and positive poles of the battery 
 respectively. At an up-station U, the up-line terminal is earthed, and 
 D connected to line. At a down-station U is connected to line and D 
 earthed. Fifteen to twenty milliamperes should be allowed to work 
 the instrument. Small Daniells are used for both direct sounders and 
 direct writers. 
 
 In order that the circuit may be complete we must have two wires or 
 lather two separate conducting paths between the two offices. The 
 earth will act as the second conducting path and obviate the necessity 
 for a second wire. It, moreover, offers no resistance. The question as 
 to the currents going out, say from London to Birmingham, crossing 
 each other is not a real difficulty, as it is obvious that the batteries 
 joining positive to earth and those joining negative to earth will 
 combine together in series, the excess of current being transmitted by 
 
The Direct Wt'iter. 47 
 
 the earth as an ordinary conductor. For instance, if three 5-cell Daniells 
 are joined up to an earth plate, two with positive and one with 
 negative to earth. The two dissimilar poles join together in series 
 sending a current through the two lines in series, no current passing to 
 «arth in respect of these two batteries. The third battery sends a 
 current to earth along the line. The batteries satisfy each other as 
 nearly as possible, the remaining cells producing the current which 
 flows through earth. The two batteries combining together have 
 double the E.M.F., but the two lines have double the resistance, thus 
 the current strength is not affected. 
 
 d2 
 
48 Electricity in its Application to Telegraphy. 
 
 CHAPTEE VI. 
 
 The SiNfiLE Needle. 
 
 It has already been pointed out that reversal of current entails 
 reversal of the direction of the consequent magnetic field, and hence 
 reversal of the direction in which a magnetic needle will turn. 
 Provided we are supplied with a means of reversing the current with 
 sufficient facility, we have then only to consider the form of the coils 
 and needle in the construction of the single needle instrument. The 
 coils are wrapped as closely round the needle as possible in order to 
 concentrate the magnetic field. Such an arrangement would not serve 
 in the case of a galvanometer for direct measurement, as the calculations 
 connected with it would be too complicated and inaccurate. In order 
 that the needle may not be reversed in polarity by lightning or by 
 the continuous passage of the current, it is rendered magnetic by 
 induction from a couple of large magnets, which it would be impossible 
 to demagnetise or reverse by currents through the coils. The two 
 forms of induced dials are Varley's and Spagnoletti's, the latter being 
 the later and more efficient. Illustrations of the actual shapes may be 
 found in Preece and Sivewright's Telegraphy. All the Post Office 
 galvanometers are made so that the needle will move in the direction 
 in which the current is passing, and this remark applies in the case of 
 the single needle. If the current passes from the left to the right-hand 
 terminal, then the deflection will be to the right. Each coil is 
 detachable from the frame, one wire being connected to the brass 
 frame and the other left free in each case. The current flows in at 
 the protruding wire, through the first coil out on to and across the 
 frame, thence on to the frame of the second coil and out. All the 
 galvanometers now supplied have induced needles precisely similar to 
 the single needle coils. 
 
 The induced dial was invented to get rid of the trouble due to 
 lightning. Lightning protectors are employed on every circuit, for if 
 protection were not afforded hundreds of coils would be fused and 
 therefore destroyed on the occasion of every thunderstorm. ' The 
 principle of the lightning protector is very simple, and though there 
 
Lightning Protectors. 49 
 
 are many forms they are all identical in principle. When dealing 
 with lightning we confront quite a new aspect of our subject. The 
 'Currents are of an exceedingly sudden character, hence protection 
 which will suffice for slow heavy currents will be useless in the case 
 of lightning. If an explosive such as guncotton is laid on a stone 
 in the open air and is simply ignited with a match nothing beyond 
 a "fizz" will take place, but if it be fired with enormous rapidity 
 by means of a detonator the air above it, possessing inertia, will 
 not move instantaneously with sufficient rapidity, i.e., it will offer an 
 ■enormous amount of inertia resistance, and the result will be that the 
 stone will be blown to pieces. Lightning behaves in an analogous 
 manner to the guncotton. The electro-magnetic inertia resistance 
 offered by a coil of wire to a lightning discharge is enormous and if we 
 give it a chance to rebound and avoid traversing the coil by cracking 
 through a small air gap it will naturally do so at a path of less resist- 
 ance. This means that a current cannot reach a finite value in an 
 infinitely short time, and later the reason of this will be dealt with. 
 
 In the case of the needle, the two wires coming from the dial are 
 often wrapped closely together, so that the current will break down 
 the insulation between them and jump across. The reel form provides 
 also an escape to earth. The P.O. protector consists of two plates 
 fiaparated by a thin disc of mica, in which there are three holes. The 
 top plate is connected to line and the bottom one to earth. The 
 ordinary telegraphic current will not jump across to the other plate 
 '(which is connected to earth), but the lightning discharge will do so 
 and thus save the instrument. The vacuum protector consists of a 
 email bulb, from which the air has been previously exhausted, in which 
 two points are approached within a short distance, one being connected 
 iio earth and the other to line and instrument. At intermediate offices 
 a special protector with three plates separated by two discs is employed. 
 The centre plate is connected to earth, and the up and down lines 
 ■connected to the other plates. 
 
 But to return to the single needle. The commutator has six 
 terminals. The Z and G terminals are connected to the zinc and 
 •copper poles of the battery. The two terminals X and Y are joined 
 by a brass strap. A is the up line terminal and B is connected to the 
 Tight hand terminal of the coils, whose left is the down line terminal. 
 If a positive current enters at A it passes to the middle terminal / of 
 the right hand key, along its length through its back stop to Y, along 
 the connector to X, through the left back stop to e, and thence to B 
 through the needle coils to the down line or earth. It will be seen 
 that if either of the back stops are dirty, or if X and Y are not con- 
 nected the wire will appear to the distant station to be disconnected. 
 This is the receiving circuit. 
 
50 
 
 Gonnections- 
 
 When the right-hand key is depressed the top contact, which is 
 joined to the positive ipole of the -battery, is connected to the middle 
 of the key, which is in turn connected to A. At the same time the 
 front screw is depressed, and connects the negative pole to B. If the 
 line and earth of the instrument be connected a current will flow 
 through the coils,Scausing a deflection to the right. 
 
 If now the right ikey be released and the left depressed, the positive 
 pole of the battery is connected through the left top contact to the 
 
 ^^^^^W 
 
 Figure 3. 
 
 middle of the key to B. The zinc is connected by means of the front 
 stop to A. Now the current will flow from B through the needle coils 
 back to A, causing a deflection to the left. 
 
 The down line is the line to the down station, and the down line 
 terminal is joined to the down line. At a down station there is na 
 down line, therefore the down line terminal is joined to earth. At 
 the up station the up line terminal is earthed and the down line 
 terminal is joined to the down line or line to the down station. If 
 the station be up, then A, the up line terminal, is connected to earth ;, 
 but if a down station, to line. 
 
Faults . 51 
 
 If either of the front stops do not make contact when the back top 
 stops are doing so, or vice versa, no current will be sent out. To test 
 our instrument it is only necessary to connect A and the left terminal 
 of the dial and work the keys. If when calling it is noticed that 
 the needle frequently does not follow the keys, the cause will be that 
 the distant office is trying to stop you. If now, when the keys are 
 allowed to rise, there is no reply, the fault is either in your own 
 receiving circuit or in the distant keys. If, however, by putting on 
 a galvanometer it is found that the distant station is sending out 
 currents, then the fault is in your own keys. If on the other hand 
 currents are not sent out the fault is either in their keys or battery. If 
 the disconnection is in the coils or' line then the needle will not move 
 for either station, and the only remedy is to go to the test box and see 
 whether the line or instrument is faulty, In case of a fault at once 
 examine the contacts and much time will be saved, 
 
 It will be seen that in the case of a down station the up-line terminal 
 is connected to the line to the up-station and the down-line terminal 
 to earth. The up station has the up line terminal earthed and the 
 down line terminal connected to line. The only differences then in 
 connections between the up and down stations is in the reversal of 
 line and earth. Depression of the right key joins the positive pole of 
 the battery to A, the up line terminal. At rest A is joined to B 
 through X and Y. 
 
 The connections of the commutator are somewhat different when 
 the circuit is to be worked from a universal battery as, were they of 
 the form shown in Figure 3, it would be possible to short circuit 
 the battery by depressing both keys. In order to avoid this the 
 connections have been altered to those shown in Figure 6. In all 
 instruments which are now sent out these connections are altered, but 
 it is as well to examine them before using the commutator for 
 universal working. (See Chapter XX.) 
 
 The resistance of the single needle is 200 ohms, and from 15 to 20 
 milliamperes should be allowed for working it, Leclanche No. 3 cells 
 are usually employed on such circuits. The lightning protector at 
 very small offices is placed inside the case below the dial. 
 
52 Electricity in its Application to Telegraphy. 
 
 CHAPTER VII. 
 
 The P.O. Standard Relay. 
 
 The standard relay essentially consists of a fixed electro and a 
 moveable permanent magnet. This permanent magnet is so arranged 
 that it will be deflected to the right when a positive current enters the 
 relay from the up side. 
 
 §) ® (^ 
 
 Figure 4. 
 The arra,ngement is represented conventionally in Figure 4, from 
 which it will be seen that there are two electro-magnets, each pole of 
 which has a protruding soft iron polepieoe. Between these polepieces 
 the moveable permanent magnets are fixed. The soft iron tongues are 
 fixed at right angles to the bras^ connecting rod and to the flat surface 
 of the inducing magnet. If these tongues were pieces of magnetised 
 steel the working currents would very soon demagnetise and reverse, 
 or at all events weaken them, thus reducing the sensitiveness of the 
 
The Standard Relay. 
 
 53 
 
 relay. To avoid this trouble the tongues are made of soft iron, rendered 
 magnetic by induction from a powerful horseshoe magnet placed 
 behind them. 
 
 We have now to consider which end of the magnet we should prefer 
 to make North. The earth's line of force in England pass in such a 
 direction that if a poker is held vertically and struck whilst in that 
 position, the top becomes a South and the bottom a North pole. If 
 the top of the magnet be made North the earth's lines wiU tend to 
 reduce its nugnetism, therefore, it will be preferable to make the top 
 of the magnet into a South pole, and this is accordingly done. 
 
 Figure 5. 
 
 It will be seen that the magnet is wound with two distinct wires, one 
 between U and D, and the other between TJ-circle and D-circle. In 
 practice, the covering of one wire is made white and that of the other 
 green. 
 
 The two tongues are connected together by a brass rod, to the top 
 of which (above the soft iron tongues) is fised the contact tongue. If 
 a positive current enters at IT (the np-Une terminal), it will, in passing 
 round the coils, make the right electro-magnet North at the top and 
 South at the bottom, and the left magnet South at the top and North 
 at the bottom. The top tongue is by induction made South, and will, 
 therefore move to the right. At the same time the bottom tongue is 
 made North, and it also will move to the right. It should be noted, 
 however, that the fixed end of the tongues are opposite in polarity to 
 the inducing magnet, but that the moveable ends are of opposite 
 polarity to the fixed. This doubles the force with which the tongue 
 will move to the right, thus rendering the relay twice as sensitive as if 
 
54 Connections. 
 
 there were only one tongue. It is obvious that the coil U-cirole to D- 
 circle behaves in exactly the same manner. 
 
 Each coil has a resistance of 200 ohms, so that if D-oircle be joined 
 to U by a connecting strap, the current will pass round the first coil and 
 then round the second coil in the same direction, i.e., the relay is con- 
 nected up in series. Its resistance will be 400 ohms, as the current has 
 now to traverse 200 ohms twice. To connect the relay in parellel U 
 and U-circle, and D and D-circle must be connected. The current now 
 splits, half going through the U D coil and half through the U-circle 
 D-oircle coil. We have added a second conducting path for the current, 
 and have thus made it twice as easy for the current to circulate through 
 the relay as through one coil, and its resistance will consequently be 
 100 ohms. 
 
 The top four terminals are, as we have seen, the ends of the coils. 
 The bottom three (T, S, and M) are the local terminals. The left is the 
 spacing, the right the marking, and the centre one the tongue. 
 
 If, when no current is passing through the relay, the soft iron 
 tongues are nearer to the left hand than to the right hand electro- 
 magnet, they will, by induction, make the iron of the nearer electro- 
 magnet into a magnet of opposite polarity ; hence attraction will take 
 place, and the tongues will be held to the nearer electro-magnet. 
 
 The relay is so constructed that the two contacts S and M can be 
 moved by means of the regulating screw, so as to approach the tongues 
 nearer to the right or left electro-magnet, as may be desired. The 
 nearer they are moved to the iron of the electro-magnet the greater is 
 the pull exerted between them and that iron, consequently it is easy to 
 make the relay more or less sensitive, and, when necessary, to give it a 
 distinct bias to either marking or spacing. 
 
 The most sensitive condition of the relay is when the tongue will 
 move to either side and will stay at which ever side it is placed. This 
 is known as the neutral position. It will be noticed that the tongues 
 will not take up an intermediate position between M and S and return 
 to it when deflected. In order to accomplish this, two springs are 
 added, by adjustment of which the tongue will always return to its 
 central position. A relay with springs is called an A neutral or a B 
 neutral in the case of those wound with 100 ohms coils. Post 0£5ce 
 standard relays are tested at the Instrument Factory, HoUoway, and 
 must work well with half a milliampere. 
 
 The Siemen's relay is now fast disappearing, and it is, therefore, not 
 proposed to enter' into its construction very fully. SufiBce it to say 
 that its principle is, in the main, similar to the P.O. relay. It has a 
 single electro-magnet with one tongue, which is magnetised by means 
 of a bar magnet whose other end is attached to the soft iron 
 connecting piece of the electro-magnet. 
 
Electricity m its Application to Telegraphy. 55> 
 
 OHAPTER VIII. 
 The P.O. Double Plate Sounder. 
 
 The single needle, even when fitted with sounding pieces, is not at 
 all suitable for large offices where there is a great deal of noise. With 
 the object of producing louder signals the form known as Neale's 
 sounder was introduced, but even this, though a great improvement, 
 was not sufficiently loud. Then, again, it is a very awkward shape to 
 fit on to tables along with sounders, which are much lower than the 
 needle writing desk. The present form of Double Plate Sounder i& 
 the result of evolution from the original Bright's Bells. 
 
 The commutator is the same as in the single needle with the XY 
 strap removed. The tongue of the relay is held in a central position, 
 between the S and M contacts by means of two springs. If a current 
 passes through the relay in the direction U to D the tongue of the 
 relay moves to the right, joining T and M, thus completing the circuit 
 of a battery. This battery is called a local battery in contradistinction 
 to the larger or line battery joined to Z and C of the key. Similarly, 
 the circuit of the local battery is termed the local circuit and is com- 
 pleted through an electro-magnet, which attracts its armature and 
 thus produces a signal. When, however, the current passes through the 
 relay from D to U, the tongue passes over to S, and thus completes the 
 left-hand local circuit. When sending, the instrument at the sending- 
 station is not affected, hence a galvanometer is usually added to the 
 apparatus. The galvanometer used for this purpose is of the f ornk 
 known as " single current." It has two stop pins, and the dial is not 
 graduated. 
 
 The actual connections are illustrated in Figure 6. A, as before, is- 
 the up-line terminal, and B, or rather the left-hand side of the galva- 
 nometer, is the down line terminal. The connections of the set can 
 easily be borne in mind by remembering that the X and Y and the 
 D and TJ terminals of the relay, which face each other, are joined 
 
56 Path of Current. . 
 
 across diagonally. The tongue of the relay is joined to the local 
 battery. It will be seen that when T and S are in contact a dot is 
 iormed, and when T and M are together a dash is the result. 
 
 The instruments are joined up in exactly the same manner at up, 
 intermediate, and down stations, with the exception of the up-line 
 and down-line terminals. In the case of local circuits the H.O. is the 
 up-station, hence, in the circuit known as the MR-Elizabeth Street, 
 MR is up, Elizabeth Street down, and Oheetham, which comes between 
 MR and Elizabeth Street, is intermediate. At MR the up-line 
 terminal is connected to earth, and the down to line. Oheetham, the 
 intermediate station, is down to MR, consequently the up-terminal is 
 connected to the up-line going to MR, and the down-line terminal is 
 joined to the Elizabeth Street line. Oheetham then is up to Elizabeth 
 Street, but down to MR. The up-terminal at Elizabeth Street is 
 connected to line, and the down-terminal to earth. This method of 
 ■connection applies to any and every instrument. 
 
 Now to trace the path of the current. The later form of commutator, 
 which renders it impossible to short-circuit the battery, is shown in 
 Figure 6. The sending circuit is very similar to that of the older form 
 of commutator described in the case of the needle. If the right key 
 be depressed, the positive pole of the battery will be connected to A 
 via the top right-hand contact and centre of right-hand key. The 
 negative pole of the battery will at the same time be joined to B via 
 the front right-hand contact to X, then through the back stop of the 
 left key, along its length to the centre and on to B. Similarly when 
 the left key is depressed is joined to B and Z to A via the back 
 stop of the right hand. It will be seen that depression of both keys 
 <5uts the battery circuit. 
 
 Only one instrument, connected as an up station, is shown in the 
 figure for the sake of space, but it will be of assistance if readers will 
 draw two instruments connected up by a line and earth. If the down, 
 station depresses his right key the current flows from the up-line 
 terminal (which at a down-station is connected to the line) to line, 
 thence to the left side of our galvanometer deflecting it to the right, 
 thence to B to the middle contact of the left key, along its length, 
 through the back stops to X, through the relay in the direction of U to 
 D, thence to Y, through the back stop of the right key to A, and thence 
 to earth and back to the battery. The current has pulled over the 
 tongue of the relay, joining T and M, so completing the local circuit 
 through the right sounder. The depression of the distant station's 
 left key merely reverses the current, thus actuating the left sounder. 
 
 If, now, the back contacts of the keys be dirty, or in any other way 
 ■cause disconnection, the receiving circuit will be cut. The appearance 
 ■of this fault will be that the distant station is stopping you (as seen by 
 
Double Plate Sounder Connectiofis. 
 
 57 
 
 Figure 6. 
 
58 Adjustment. 
 
 -the galvanometer, which is in the line circuit), but that when the keys 
 are released no sisals are received. The sending faults are the same 
 as described in the case of the single needle. 
 
 When the relay is out of adjustment, the first thing to be done is to 
 take off the springs and render the relay neutral by means of the spacing 
 screw, at the same time noticing whether the local circuit is perfect. 
 The spacing screw should then on no account be touched, as this is the 
 most sensitive condition. Having done this, put on the springs and 
 tighten them so that they hold the tongue in the central position with 
 the least possible tension. The instrument will probably work 
 satisfactorily now. 
 
 If the signals are not received on one of the bells it will probably be 
 because its spring is too tight, or it is too far from its magnet, has too 
 much play, or all three. This is soon remedied, but care must be taken 
 not to get the armature too close to the magnet, as in that case sticking 
 •(due to residual magnetism, or magnetism which remains when the 
 -current has ceased) is the result. Frequently one of the relay springs 
 becomes loose, and sending produces a series of A's. Unless you see at 
 a glance what the trouble is, proceed in the first mentioned manner. 
 Dirty local contacts or broken local connections are revealed during 
 adjustment for neutrality. "Weak locals will require the bells to be 
 given very little play with a weak spring in order to make working 
 possible. The S and M contacts being too far apart, or too close 
 will prevent the reception of clear distinct signals, but these are obvious 
 faults. 
 
 A sufficient number of Leclanche cells (No. 3 size) to furnish a 
 current of from 14 to 17 m.a. should be provided for the line battery. 
 A 2-cell Large Daniell is employed for working the local sounders 
 which have a resistance of 20 ohms. The terminal of these coils are, 
 as in the case of the sounder, joined by a resistance of 500 ohms in 
 order to prevent sparking at the contact points of the relay. (See 
 ■Chapter XXII.) 
 
Electricity m its Application to Telegraphy. 59 
 
 CHAPTEE IX. 
 
 The Single Cureent Sounder. 
 
 Only in the case of very short lines is the line current used to directly 
 produce the reading signal. The use of a relay saves battery power, 
 renders the signals much firmer, and obviates the necessity for further 
 regulation. The amount of energy wasted in overcoming the resistance 
 
 Liwe «n 
 
 ^^{Bi 
 
 Figure 7. 
 
 of a line is equal to the product of the resistance of that line and the 
 square of the current flowing through it, so that the smaller the 
 current we can do with the better. The line current then is only used 
 to complete a local circuit when a goodly proportion of the energy of 
 the small local battery is returned in producing a loud signal. 
 
60 Single Current Sounder Connections. 
 
 Joining up the apparatus is now the only matter requiring con- 
 sideration. The positive pole of the battery is always connected to the 
 front stop of the key. The middle stop is connected to the galvanometer 
 whose other terminal forms the down line terminal . The back stop is 
 connected to the receiving circuit, so that the only subject now left 
 for consideration is how to join the relay up to it- U is the up line 
 terminal of the relay, and must therefore be connected to earth in 
 the case of an up station. Here again the apparatus connections are 
 precisely similar at both stations. At an up station, the U side of the 
 relay and the negative pole of the battery are earthed, but at a down 
 station they are both connected to line as shown in Figure 7. 
 
 Now suppose the up station depresses his key. A current leaves 
 through the galvanometer along the line to the U of the down station's 
 relay, through which it passes in the direction of U to D, along the key 
 through the galvanometer to earth, thence back to the negative pole of 
 the battery. The circuit is thus completed, and the tongue of the 
 relay is pulled over to M, and completes the local circuit. The relay 
 is, of course, given a bias towards spacing, so that when the current 
 ceases the tongue returnsto S, thus ending the mark. The local circuit 
 is dotted and requires no explanation. 
 
 The most common fault is one of regulation, and its appearance is 
 that of sticky dashes and missed dots. These two opposite effects 
 usually elicit the remark " send firmer." The cause of the fault is that 
 the tension spring of the sounder is too tight, causing sharp dots to 
 disappear. To counteract this the spacing bias is reduced, thus causing 
 the dashes to run together owing to the bias being insufficient. 
 This appearance may be caused also by the armature being allowed to 
 approach the magnets too nearly, the result of which is heavy dots 
 and sticky dashes. The Telegraphist in charge of the circuit gives the 
 screw three or four turns towards spacing and light dots and 
 sticky dashes again result. Sometimes, when the locals are weak, 
 the armature is approached very near to the magnets, but this is 
 a practice which should not be followed. The armature should 
 
 always clear the magnets by at least ^ inch when down. When 
 
 desired, a local inker may be substituted for the sounder, or both may 
 be worked together. 
 
 A sufficient number of small Daniells to send out a current of from 
 15 to 20 m.a. are provided for the line batteries. The locals for 
 sounders and local inkers are 5-cells large Baniell. 
 
Electricity in its Application to Telegraphy. 61 
 
 CHAPTER X. 
 
 The Condenser. 
 
 A telegraph line has another awkward property besides resistance, 
 and that is capacity. The capacity of india-rubber bladders might 
 conveniently be defined as the quantity of water which it would contain 
 at a standard pressure. Similarly the electrical capacity of a body is 
 the quantity of electricity which it will contain at a standard electrical 
 pressure, viz., one volt. 
 
 In order to appreciate the nature or cause of capacity let us refer 
 back to the diSerence between a conductor and an insulator. The 
 former allows the passage of a continuous current, whereas the latter 
 does not. If a fixed spring is bent it will, when the pressure on it is 
 released, fly back. If this takes place in air it flies backwards and 
 forwards many times before coming to rest. This corresponds to the 
 discharge of a leyden jar through a pair of discharging tongs. If, 
 however, the spring is immersed in treacle it will, when bent, return 
 slowly to its normal position without any oscillation. This is the 
 kind of discharge observed in the case of a telegraph circuit. The 
 treacle produces the same effect on the motion of the spring as does 
 the electrical resistance of the circuit on an electric current. We may 
 then consider that a current applied to an insulator strains it, thus 
 permitting the current to flow into the di-electric to a certain extent, 
 the current ceasing immediately the "fly-back" tendency of the 
 insulator is equal to the forward pressure of the battery producing the 
 strain. The insulator in this connection is termed a di-electric. In 
 the case of an aerial telegraph line the air is the di-electric, and in the 
 case of a cable or underground wire the gutta-percha or whatever the 
 insulation may be between it and the earth. 
 
 We have seen then that if two conductors are in dissimilar electrical 
 conditions, the insulating medium experiences a strain. The effect of 
 capacity is to provide a path of no resistance until the backward 
 strain of the di-electric is equal to the forward pressure of the battery, 
 i.e., until the condenser is charged. When the battery is removed the 
 
 E 
 
62 Theory of the Condenser. 
 
 current rushes out of the condenser in the opposite direction to the 
 charging current. It should, however, be remembered that this 
 current rushes out of a line wire at whichever end of the wire is con- 
 nected to earth. If we are considering the receiving end of the wire, 
 the rush of current is in the same direction as the charging current. 
 
 The current has to charge the wire before any efEect can be produced 
 at the distant end, thus some of the current which should be producing 
 a mark is lost. When the battery is removed the wire discharges part 
 of the current running through the receiving instrument in the same 
 direction as the sending current which has just ceased, and thus the 
 mark is prolonged. The remainder of the current will rush out at the 
 sending end if the switch is not then in such a position as to cut the 
 wire. Should it not be so, then the whole of the electricity stored in 
 the wire rushes through the distant receiving apparatus, thus prolonging 
 the mark. 
 
 The quality of material of which the spring is composed afEeots the 
 distance to which a given pressure will bend it. The same remark 
 applies to the material of which the di-electric is composed in regard 
 to electrical pressure. In comparing di-electrics, air is taken as the 
 standard, and is said to have unit specific inductive capacity. A current 
 changing in direction first of all strains the di-electric in one direction, 
 then relieves the strain and again strains it, ibut 'in the opposite 
 direction. The analogy of the spring will also tend!to render it 
 obvious why a condenser permits the passage of an alternating current. 
 This is the easiest and most correct way to regard capacity, but it may, 
 perhaps, be as well to explain the conventional or electro-static method 
 of regarding capacity, 
 
 Two insulated conductors may be thrown into opposite electrical 
 conditions by the attachment of a battery or frictional machine. The 
 plate connected to the positive pole of the battery assumes a positive, 
 and that connected to the negative pole assumes a negative potential, 
 and when oppositely charged plates are joined by a wire a current flows 
 from the conductor of higher " electrical level " or "potential" to the 
 lower, thus causing the bodies to regain their normal uncharged 
 condition. The quantity of electricity which the bodies will contain 
 depends upon the proximity, size, electrical pressure applied, and, lastly, 
 upon the materials insulating them. In the case of a telegraph line, the 
 wire corresponds to the one and the earth to the other plate of our 
 condenser. The air which insulates the line from the earth corresponds 
 to the di-electric medium of the condenser. 
 
 The three essential parts of a condenser are then two conductors 
 separated by an insulator (called in this connection a di-electrio). 
 The di-electric should be as thin as possible in order that it may 
 readily give way to the strain caused by throwing the conductors 
 
Practical Arrangemsnts. 
 
 63 
 
 on either side of it into different electrical states. The actual thickness 
 employed is determined by its insulating properties. If the two con- 
 ductors are allowed to come into contact the condenser becomes a mere 
 resistance. The larger the surface of di-electric which is strained the 
 greater is the amount of electricity which the arrangement will 
 retain or discharge, i.e., the greater is the capacity. iThe last factor 
 which needs consideration is the material lof the di-electric. Two 
 conductors separated by air, and having a capacity which we will call one, 
 would, if separated by dense flint glass of equal thickness, have a capacity 
 of 7-37, or if by paraffin2-29. Mica (664) is the best substance to employ, 
 
 FlGUKB 8. 
 
 but its cost is prohibitive for any but standard purposes. Paraffined 
 paper is used by the Department, and answers the purpose admirably. 
 The unit of capacity must of course be in terms of volts, amperes, 
 and seconds to fall in with the remainder of the system of electrical 
 units. Accordingly, a condenser of unit (i.e., one farad) capacity is 
 one which, when a cell of one volt E.M.F. is joined to it, will be 
 charged with a unit quantity of electricity, i.e., the quantity which 
 flows through a circuit of one ohm resistance at a pressure of one volt 
 in one second. The farad is far too large for practical purposes, and 
 the microfarad is always used in speaking of condensers. Its prefix 
 " micro " signifies " one-millionth of." All that now remains to be 
 said on this subject relates to the actual form end connections of the 
 apparatus. 
 
 e2 
 
64 Practical Arrangements. 
 
 Some condensers are for convenience divided into two sections so 
 that each section may act, when required, as a separate condenser, thus 
 saving the room which would be taken up by a second box, The 
 vertical strokes in Figure 8 represent ithe sheets of tin-foil, and the 
 vertical spaces between the upper and lower bars the di-electric. To 
 insert capacity a plug connecting the block and A or B is inserted. 
 In this it is the converse of a resistance box where, when a plug is 
 removed, the current instead of passing from block to block through 
 the plug has to traverse a resistance coil. The blocks each bear a 
 number which represents the capacity in microfarads, which will be 
 inserted on putting in a plug between that block and A or B as the 
 case may be. All the ends of the tin-foil are joined together and go 
 to one terminal, the A block forming the other terminal. To get the 
 full seven-and-a-quarter microfarads capacity A and B must be 
 connected together by means of two little screws provided for the 
 purpose. As soon as the paper insulation between the tinfoil leaves 
 deteriorates, and it is possible to send a current through it the 
 condenser should be exchanged. The capacity is of course reduced 
 when the insulation resistance between the terminals falls. 
 
 Condensers are also made in three sections for use on very long lines 
 with submarine cable in order that a more accurate balance may be 
 obtained. 
 
Electricity m its Application to Telegraphy. 65 
 
 CHAPTER XI. 
 
 The Double Current System, 
 
 It has been preTiousIy remarked that a telegraph line has, besides 
 resistance, a certain amount of capacity, and that this capacity delays 
 the current in producing its effect at the distant end when the circuit 
 is completed, and tends to prolong the current on the breaking of the 
 circuit (page 62). The result of this is that sharp dots are entirely 
 lost on long underground wires, where the capacity is great, con- 
 sequently firmer, and, therefore, slower signalling is necessary. This 
 occurs when the single current is employed. A special key which 
 immediately it is allowed to rise sends a current in the opposite 
 direction, is used on all such circuits. 
 
 It is obvious then that a relay such as the standard relay, which will 
 only respond to currents in one direction, must be used. Such a relay 
 is termed a polarised relay, in contradistinction to a non-polarised relay 
 which will work with a current in either direction. Although the 
 relay used for a single current working is always a polarised relay 
 it need not necessarily be so. 
 
 Let us consider what takes place on the closing of the circuit. The 
 current flows into the wire, but has to satisfy the capacity of the wire, 
 by flowing into it (as a condenser) until the wire is charged, before any 
 effect is produced on the distant instrument. When the circuit is 
 broken the wire discharges itself, thus prolonging the signal unduly. 
 
 The double current system has for its object the quickening of the 
 discharge of the line, which is ordinarily comparatively slow. The 
 current in the reverse direction clears the line and enables the relay to 
 be worked in its most sensitive, and, consequently, most rapid position 
 of neutrality, the result being that the speed of working is greatly 
 enhanced. In practice the relay is worked with a slight bias to spacing 
 in order that the tongue may not remain over to M, and by so doing 
 complete the local circuit and thus run down the local battery. If the 
 relay is exactly in the neutral position the sounder probably chatters 
 due to slight contact, etc., and this is most irritating to hear. 
 
 The double current key must provide also for the receiving circuit. 
 When the key is at rest a current is flowing from it to line, so that a 
 
66 
 
 Double Current Key. 
 
 switch must be provided in order %o cut the battery out and join up 
 the relay to line. The double current key now in use has five 
 terminals. 
 
 Z and C (Figure 9) are respectively connected to the zinc and copper 
 of the. battery. When the switch* is at " receive " the middle terminal 
 is connected through the right-hand 'terminal by means of the right 
 switch lever and the left-hand contact. When the switch is turned to 
 "send" the insulated halves of the key are connected to the bottom 
 
 Q o O 
 
 \'' 
 
 of' b 
 
 "3 
 
 
 c o 9 
 6 6 
 
 
 o o o 
 
 \ 
 
 S»*VCF 
 
 MARK 
 
 FiGUBE 9. 
 
 RectwE. 
 
 terminals. -The bottom springs are connected to opposite poles of the 
 battery to the top set, so that depressing the key reverses the 
 connections as shown in Figure 9. 
 
 On a double current set, the galvanometer should be joined up in 
 paraUel, i.e., so that the current splits, half of which circulates through 
 each coil, and this may be done by joining the left top to the right 
 
Double Current Sounder. 
 
 67 
 
 bottom terminal and the right top to the left terminal. By joining the 
 two bottom terminals the two coils will be joined up in series, i.e., the 
 current will have to pass through the coils one after the other. The 
 terminals of the galvanometer (Figure 10) were left disconnected in 
 order to call attention to this point. 
 
 2/A/f To 
 
 on £. 
 
 t.lfJK TO 
 
 Figure 10. 
 
 Connections. — The right terminal of the key is joined up whether 
 the switch is at " send " or at " receive," and it will, therefore, form one 
 of the line terminals. When the switch is at " receive " the centre 
 terminal is connected to line (by means of the right-hand terminal). 
 Evidently the relay must be inserted here with its other terminal 
 
68 Faults, 
 
 joined to the other line. Bat the left terminal of the key must 
 also be similarly connected in order that currents may be sent out, and 
 as it is an up-line terminal the relay's U terminal must be joined to 
 the up-line also. The right terminal of the key is a down-line terminal, 
 and to this the galvanometer is connected with its free terminal to 
 form the down line. 
 
 The correctness of the apparatus connections may be tested by 
 connecting D of the relay to the right terminal of the galvanometer, 
 and working the key. It should, however, be noted that this does not 
 completely prove the connections as D to the centre of the key is not 
 included, 
 
 When the up 'station turns his switch the current flows from C 
 through the key to earth thence to the down earth plate which is 
 connected to the galvanometer, through the galvanometer, through the 
 key to the middle terminal, through the relay in the direction D to 
 U, to line, thence through the up station's galvanometer back to the 
 up battery. This current holds the tongue of the relay to S. Depression 
 of the key reverses the current and causes the completion of the local 
 circuit, i.e., causes a mark. The current which fills up the intervals 
 between the marks is termed the spacing current, in contradistinction 
 to the marking current. 
 
 A common fault is for the key to send one current, in which case 
 the key should be changed if the cause is not a dirty contact at the 
 springs. In the earlier forms of the double current key a disconnection 
 at the switch was not a rarity, but this would be revealed by the test 
 mentioned above. If after having called for some time it is noticed 
 by the galvanometer that the distant station is trying to stop yon, but 
 that when yon turn your switch to receive you get nothing, the fault 
 is in your own receiving circuit. If the connections are right then 
 either the key is faulty or one of the connecting wires is broken 
 inside the gutta percha. 
 
 It is sometimes desired to make a double current set work single 
 current. This is done as follows : — Disconnect the two wires going to 
 the left terminal of the key and join them to the Z wire, leaving Z 
 and the left terminal free. Now connect Z and the centre terminal 
 by means of a piece of wire. If the distant station is up and the 
 double current set has been !used as an up station the necessary 
 alterations (i.e., reversal of line and earth) must also be made at the 
 test-box. 
 
 A sufiBoient number of small Daniells to send a current of from 14 to 
 17 m.a. are required. The local battery consists of a 5-cell large 
 Daniell. It is obvious that a local inker may be substituted for the 
 sounder if desired. 
 
Electricity in its Application to Telegraphy. 69 
 
 CHAPTER XII. 
 
 The Differential Duplex.— Single Current. 
 
 Duplex working is affected by arranging the apparatus so as to act 
 in the ordinary way when only one station is working his key, but 
 immediately both keys are depressed some action is arranged to 
 
 L/NB 
 
 Figure 11, 
 
 intervene which will cause both sounders to be actuated. It is obvious 
 that immediately either station releases his key the- ordinary simplex 
 working ensues. 
 
 Such an arrangement is depicted in Figure 11 and will abundantly 
 explain the principle upon which such circuits are worked. Two P.O. 
 
70 Principle of Biiplex. 
 
 standard relays are sketched with the U and D-oircle terminals joined 
 to the middle contact of a single current key. Consider now that A. 
 depresses his key, A current flows from the battery to U and D-circle. 
 There are two paths open for the current, the one through D-cirole 
 U-circle to line and the other through the rheostat back to the battery. 
 If, now, the rheostat has resistance inserted equal to that of the line 
 and distant receiving circuit together, the current will have to split 
 between two paths of equal resistance, and will therefore split equally 
 between them. The current passing through the rheostat is called the 
 compensation current and proceeds in the direction U to D, causing a 
 pull of the relay to M. The other, or line current, goes in the 
 direction D to U, thus tending to space ; but if the resistance of the 
 two circuits, viz., the line and the compensation circuits, are equal, 
 there will be no effect produced on the sender's apparatus when his 
 key is working as his current passes through the two coils of his relay 
 in opposite directions. This applies equally to B. Now let us see 
 what becomes of A's line current. It flows along the line to B, where 
 it passes through the U-circle D-circle coil of the relay in the 
 direction U to D,'and thence through the key to earth, thus causing a 
 mark. B's current behaves in a similar manner when the circuit is 
 worked simplex. 
 
 Now, supposing that while A has his key depressed B also depresses 
 his. The result is that A and B both connect the positive pole of the 
 battery to line, thus virtually cutting the line circuit, for neither 
 battery can send a current against the E.M.F, or backward pressure of 
 the other. The only path available for the current is for A, through 
 the compensation circuit in the direction U to D, thus causing a mark. 
 Similarly at B, the compensation current registers a mark so long as 
 both keys are depressed. Immediately either key is allowed to rise 
 the conditions of ordinary simplex working are restored when the line 
 current produces the mark. 
 
 There is yet one other case to be mentioned, and that is when one 
 key is in the intermediate position between the front and back stops, 
 i.e., when it is rising to end a mark. The current then passes through 
 the distant relay from U-cirole to D-circle, and U to D through the 
 rheostat to earth and back to- the battery thus causing the desired 
 mark. The current thus traverses a resistance twice as great as when 
 the key is resting on the back stop. The current is then half its former 
 value, but as it traverses both coils the magnetic effect is the same as 
 in the former cases. At the end where the key is depressed more 
 current will pass through the compensation than through the line side, 
 and this might cause a mark at this end. This would, however, merely 
 tend to prolong the signal which the raising of the key at the other end 
 was meant to end. This difficulty could easily be surmounted by 
 
Differential Duplex. 
 
 71 
 
 employing a key which would not break connection with the battery 
 until it also touched the earth stop. The quadruples increment key 
 would answer the purpose. 
 
 In the Post Office system the preponderating current in the 
 compensation side merely tends to space, hence no effect is produced 
 from the momentary upsetting of the balance. This position is^ 
 however, of so momentary a nature that no ill effects are observed 
 from the alteration in the resistance of the circuit. When similar 
 poles of the battery are connected to line the circuit is said to b& 
 worked by the method of opposition. 
 
 L//^£r 
 
 \\ 
 
 §^- 
 
 ,. 
 
 .-'#- 
 
 » ^ 
 
 \^^/ 
 
 
 
 H 
 
 Kl 
 
 
 
 f^l 
 
 1 
 
 V 
 
 
 
 Dovm 
 
 
 Figure 12. 
 
 In the system described above duplex working is effected by causing 
 the line side of thejrelay to be unaffected when both keys are depressed, 
 the line side of the relay ordinarily causing spacing. If now th& 
 compensation circuit be made to cause spacing, and the line side to cause 
 marking, we must so arrange matters that when both keys are depressed, 
 the current passing through the line coils of the relay shall be twice that 
 flowing in the compensation circuit. When this is so duplex working 
 will again be possible. This is the system adopted by the Department 
 and the augmentation of the line current is brought about by connecting 
 opposite poles of the battery to line when both keys are depressed. 
 
72 
 
 Practical Arrangements. 
 
 The result is that the two batteries combine together to send a current 
 through the line side of both relays, and this current is obviously twice 
 as great as if only one battery was engaged in its production. The 
 current, however, is opposed to the compensation current, so that the 
 total effect on the relay is practically the same as in the former case. 
 
 Now let us go yet a step further. To obtain the condition of things 
 mentioned above we should have to reverse the battery and relay 
 <x]nnections at one station. The same object will be served by reversing 
 the line and earth as in the case of an up or down station and this is 
 accordingly done. Figure 12 shows such a set joined up, galvanometers) 
 
 DcE 
 
 Figure 13. 
 
 switches, etc., being omitted for the sake of simplicity. When A only 
 depresses his key the current splits equally through the two coils of 
 the relay, the one current proceeding along the line through the back 
 stop of B's key, through the TJ D coil to earth, thus causing a mark. 
 Similarly, when only B depresses his key, A's relay only is affected, 
 thus we see that the instruments work in the ordinary way when only 
 one key is depressed. When now both keys are down A's current 
 splits as before, half going through the compensation circuit and half 
 through the line circuit, but the current in this circuit is augmented by 
 the line half of B's current. We have now currents flowing in both A 
 and B's compensation circuits, but a current of twice the strength flows 
 -through the line circuit and its preponderance causes a mark at both 
 
Balancing. 73 
 
 stations. ImmediatelT either key is released the ordinary simplex 
 working ensnes. In practice some slight alterations are made. For 
 instance, duplex working is not always desirable or possible, consequently 
 a switch is provided in order that simplex working may be resorted to 
 when required. The full connections are depicted in Figure 13. 
 
 It will be seen that the current splits twice, viz., at the galvanometer 
 and at the relay. Only one station is shown in Figure 13, as the 
 instrument connections are the same at each end. At the up station 
 the np-line terminal is earthed. If now the key is depressed the 
 compensation current flows from the battery to the strapped terminals 
 of the galvanometer through the left-hand coil, through the rheostat 
 and switch to D, thence through the D IT coil and switch back to the 
 battery, thus holding the tongue to S ; the line currents meanwhile 
 flowing through the right coil of the galvanometer to line (in the case 
 of an up station), through the distant station's apparatus to earth, 
 thence back to ithe up station, through earth, through the U-circle, 
 D-circle coil and back to the sender's battery. This nullifies the effect 
 of the compensation current. When both keys are depressed the line 
 currents combine as already described, and it is by this preponderance 
 that the working is effected. 
 
 The object of the double split is merely to simplify the form of the 
 duplex switch. It will also be noticed that a small resistance block is 
 inserted between the back stop of the key and the relay. This resistance 
 is made equal to the internal resistance of the battery, so that the 
 resistance of the circuit is not altered by the depression of the key- 
 The connections of the duplex switch are such that the lever stands 
 between the two terminals which are connected, and at the same time 
 the opposite two terminals are also joined. The figure shows the 
 switch at duplex. 
 
 The path of the current when the down station only is working is to 
 earth through the U-cirele D-circle coil of the up relay in the direction 
 n to D through the small resistance block, through the key and right 
 coil of the galvanometer to line. 
 
 As single current working is only resorted to on short lines, no trouble 
 is experienced as regards capacity, therefore the only point which 
 remains to be dealt with is as to balancing. We know that if more 
 current circulates through the line side of the relay than through the 
 compensation side, the relay will mark on depression of the key. This 
 can only occur when the resistance in the rheostat is greater than that 
 of the line. If the resistance be too little the relay will be unaffected. 
 The galvanometer needle indicates which coLL has the more current. 
 If your own marks are received on moving the large arm of the 
 rheostat to 4,000 this resistance is too much, as more current is flowing 
 to line than through the compensation circuit. Leaving the 400 arm 
 
74 Balancing. 
 
 at decrease the resistance till the deflection is reversed in direction, 
 the proper resistance to be inserted then lies between the last two 
 values. Now adjust the 40 arm similarly till a perfect balance is 
 obtained. If a wire in the compensation circuit is broken, or if the line 
 goes earthy your own marks will be received straight. If the line is 
 disconnected no marks will be registered on the sounder at the sending 
 end, but the galvanometer will move in response to the spacing currents 
 -flowing through the compensation circuit. 
 
 A sufficient number of large Daniells to provide a current of from 
 15 to 20 m.a. should be used for the line battery. This apparatus 
 should not be used for circuits over 25 miles in length, and even under 
 this distance it is frequently more desirable to work double current. 
 The local battery is a 5-cell large Daniell- 
 
Electricity in its Application to Telegraphy. 75 
 
 CHAPTER Xin. 
 
 The Double Cukkent Duplex 
 
 On any but the shortest lines the double current system is adopted, 
 and this is more especially the case in large towns where the sections of 
 underground are of considerable length. The application of the 
 double current system to the working of a circuit produces the 
 electrical equivalent of the " flicking " of a whip lash by throwing it 
 forward and then suddenly jerking it back, thereby producing a very 
 sharp and heavy "flick." The '"flick" is produced by the sudden 
 withdrawal- With the double current duplex three conditions must 
 be satisfied : — 
 
 1. When both keys are at rest (with both switches to send) the 
 tongues of both relays must be held over to spacing. 
 
 2. When both keys are depressed both relays must mark. 
 
 3. When one key is at rest and one depressed the depressed key's 
 relay must space and the other relay must mark 
 
 When both keys are at rest or both depressed the two batteries 
 combine together to produce the same effect on the two relays, the 
 compensation circuit having only half as much current flowing through 
 it as the line circuit, owing to the two batteries acting together. 
 When both keys are at rest the tendency of the compensation circuit 
 is to mark and that of the line circuit to space. When A's key is 
 depressed the current is reversed in his relay, the compensating circuit 
 now tending to space. The two batteries have now similar poles to 
 line, and the result is that the compensation circuits at either end 
 are unopposed, and A's produces spacing and B's marking. Similarly, 
 if B's key is depressed a mark is registered at A. Figures 14 and 15 
 illustrate this, the dotted portions of the circuit indicating the parts 
 in which the line current predominates over the compensation current, 
 or vice vena, the full lines indicating the path of the smaller currents. 
 
 It will be seen that in Figure 14 switches, etc., are excluded for the 
 sake of simplicity. The current splits twice, and the instrument 
 connections are the same at both stations as in the case of the single 
 
76 
 
 Double Current Duplex. 
 
 current sounder. The current from A's battery flows to C of the key 
 and thence to the relay, half going through the U D coil and left-hand 
 coil of the galvanometer back to the negative pole of the battery, and 
 the other half proceeding through the D-circle U-circle coil along the 
 line through the right-hand coil of the galvanometer to the negative 
 pole of B's battery, with which it joins through the D-circle U-circle 
 coil of his relay to earth, thence to the right-hand coil of A's 
 galvanometer, back to the negative pole of A's battery. Half of B's 
 
 p* — - . ....—-_. «._._.. ___- - — — --. 
 
 FlKUBE 14. 
 
 current splits through his compensation circuit. The currents in the 
 compensation circuits are produced by each battery separately, but the 
 current in the line current is produced by both in combination, and is 
 thus twice as strong. Both keys are shown at rest in the figure, but it 
 will be seen at once that if both keys are depressed the current is 
 everywhere changed in direction, thus producing a mark at both 
 stations. In both cases the line current is twice as great as the 
 compensation current. 
 
 In Figure 15, the connections, when the down station has depressed 
 his key. are shown. The two batteries in this case have similar poles to 
 
Path of Current. 
 
 77 
 
 line, and since they are of equal E.M.F. or strength, there is no flow of 
 current along the line. The compensation circuits are now the only 
 paths available, and as there is no current in the line coils of the relay 
 the coils in the compensation circuit are left unopposed. At the 
 up station, where the key is at rest the compensation circuit side of the 
 relay causes a mark, whereas A where the key is depressed the 
 current through that compensation clrcuic has been reversed, and it 
 causes a space. 
 
 LtlMC 
 
 FiGUBB 15. 
 It h as already been mentioned that the necessity for double current 
 working arises from the fact that all wires possess more or less capacity. 
 In the case 'of a short line where the capacity is exceedingly small a 
 single current duplex may be used. In the case of a long line duplex 
 working would be out of the question if only a rheostat were used for 
 
78 Effects of Capacity. 
 
 balancing purposes. The nature of the line must be imitated as closely 
 as possible in all respects, and it has already been remarked that the 
 current has to flow into the wire until the wire becomes fully charged 
 before any effect is produced at the distant end. The current flows 
 into the wire very rapidly, or in other words, there is a heavy rush of 
 current into the wire, so that if a rheostat only were used to balance 
 the circuit this rush of current into the wire would cause a kick on the 
 relay. The wire would act as though it offered little or no resistance 
 during the time that it was being charged as a condenser. This would 
 mean upsetting the balance for this instant, and the heavy rush of 
 current would overpower the steady flow which would immediately 
 commence through the rheostat. On a long line this momentary 
 upsetting of the current balance would be quite sufficient to render 
 duplex working impossible. Again, on the cessation of the current the 
 wire would discharge itself thus producing another kick. These kicks 
 occurring in the middle of a mark would split them up. 
 
 In the normal position, with both switches to send, on a double 
 current duplex circuit the line is fully charged and a current flowing 
 along the line from the down to the up station in the direction 
 D to U, through the line coil of the down relay. When A depresses 
 his key his battery opposes the up-station's battery and the line is 
 left free to discharge itself. The discharge occurs in the opposite 
 direction to the current which caused the charge, i.e., along the line 
 through the TJ-circle D-circle coil of the down relay to the left-hand 
 terminal of the key where it joins with the battery (which is sending 
 a current in the same direction) through the right-hand coil- of the 
 galvanometer to earth. By allowing the key to rise the wire again 
 becomes charged and a rush or current in the opposite direction occurs, 
 causing the needle to kick downwards. 
 
 When a wire is charged the current flows into the di-electric (air or 
 G.P. as the case may be) until its backward pressure stops the current 
 from flowing further. At the other edge of the di-electric, i.e., at the 
 earth, current flows out of the di-electric, When the wire is allowed 
 to discharge the current flows out on to the wire and in from the 
 earth, and in this way the circuit is completed. 
 
 Now let us apply this knowledge to practical telegraphy. If we have 
 a long line balanced only by a rheostat we have seen that when the 
 line discharges a momentary current will flow from U to D through the 
 relay. Normally with one key depressed there is no current in the 
 right-hand coil of the galvanometer. Depression of the key disconnects 
 the battery from the line and the discharge then occurs, causing a kick 
 to marking. The needle of the galvanometer kicks upwards indicating 
 a momentary marking current. This is caused by the passage of the 
 current through the right-hand coil of the galvanometer to earth. 
 
Compensation Circuit. 
 
 79 
 
 An upward kick of the needle on depressing the key indicates that 
 the line possesses capacity or rather that it possesses more than the 
 compensation circuit. If now we can give to our compensation circuit 
 an amount of capacity equal to that of the line we shall avoid all 
 trouble, as the rush of current into the condenser in the compensation 
 circuit will balance the rush of current into the wire or cable. 
 
 It should be borne in mind that even though the capacity of two 
 circuits may be the same, and though they contain the same quantity 
 of electricity, yet their discharges may be very different in character. 
 A condenser discharging through 50 ohms will give double the current 
 that it would produce when discharging through 100 ohms, but in the 
 
 FlQUBE 16. 
 latter case the discharge or current would occupy twice the{time, 
 thus I the quantities of electricity would be the same in both cases. 
 The effect of the compensation coil is always {opposite to that of the 
 line coil, hence the two equal and opposite I effects will balance each 
 other, andithe current balance will not be disturbed. Capacity is given 
 to the compensation circuit by means of condensers joined across the 
 rheostat as shown in JFigure 16, 
 
 A paraffin paper condenser discharges itself more rapidly than a tele- 
 graphTline because in the latter case the charge is distributed overia long 
 length of wire. We must therefore make our condenser I discharge 
 itself more slowly, and for this purpose the jP.O.icondensers'are made 
 in two sections. Resistance inserted in the i>ath ofiithe discharge 
 slows down the rate at which that discharge takes place. 
 
 Dr. Lodge, some time ago, managed to so arrange mattersj that the 
 charge of a leyden jar oscillated backwards.>ndiforwards^t the rate 
 
 f2 
 
80 
 
 Balancing Capacity. 
 
 of a musical note, and consequently the sparks due to the discharges of 
 the jar, instead of giving a sharp snap, pitched a musical note. In the 
 case of a telegraph circuit there is no oscillation, as the discharging 
 resistances inserted are far too high, but this digression has been 
 entered into in order to illustrate the fact that the rate of discharge 
 of a condenser may be modified by altering the resistance of the 
 discharging circuit. 
 
 The rheostat is balanced so far as the resistance of the circuit 
 is concerned, The method of utilising the condenser is shown in 
 Figure 16. 
 
 OOnvn UNE o. E. 
 
 o 
 
 somas/! 
 
 ? ' ! 
 
 a 
 
 #H 
 
 FiGTJBE 17. 
 
 B| is the' rheostat, and across its terminals the condenser is joined. 
 Two variable resistances are inserted between the condensers and the 
 rheostat in order to make the condenser discharge occur at the same 
 time as the line discharge. As has already been remarked, the capacity 
 of the line is distributed over a very long length, and in order to 
 imitate the discharge of the line more closely, two condensers are 
 used, and it is for this purpose that condensers are made in two 
 sections, each of which is a complete and independent condenser. 
 
 B,, called the retardation coils, is usually a low resistance box, 
 by means of which 10 to 1,100 ohms may be inserted. The second 
 
Balancing Capacity. 81 
 
 resistance, B3, called the condenser coils, is the set of resistance coils 
 which is inserted between the two blocks of the condenser. It was 
 formerly a fixed resistance of 2,000 ohms, but is now usually a box of 
 coils, by means of which any resistance from 50 to 4,050 ohms may be 
 inserted. The condenser A discharges itself first, and as its discharge 
 begins to decay B comes into action. When a current flows through 
 the rheostat there is a difference of potential between its terminals, 
 and it is to this difference of potential that the condenser owes its 
 charge, the exact difference of potential in volts being the current in 
 amperes, multiplied by the resistance in ohms in the rheostat (Ohm's 
 law). The quantity of electricity in coulombs in the condenser will 
 be equal to the difference of potential in volts between the ends of the 
 rheostat, multiplied by the capacity of the condenser in farads. The 
 condenser current has two paths, one through the rheostat and the 
 other through the compensation circuit, and it is upon this latter that 
 we rely to counterbalance the line discharge. 
 
 The first thing to be done is to balance the current. If there be no 
 resistance in the compensation circuit the sounder is held down by that 
 circuit, as, being the path of less resistance, more current flows through 
 it and your marks are received reversed on your own sounder ; if too 
 high your own marks are received straight. Adjust the larger arm of 
 the rheostat till it is found what the balancing resistance required is to 
 400 ohms, Put the arm at the lower value and increase the resistance 
 by means of the 40 arm till a good balance is obtained. This method 
 is much quicker than trusting to chance. The resistance should be 
 balanced by seeing if the needle stops at the same point with key 
 raised and key depressed after the capacity kick has passed away. 
 
 The capacity should now be balanced, and this may be done by 
 means of the galvanometer, though scarcely so well as by the passage 
 of working signals, the reason being that the galvanometer is hardly 
 sensitive enough. It has already been shown that an upward kick of the 
 needle on depression of the key indicates that there is more capacity in 
 the line than in the compensation circuit. If the compensation circuit 
 possesses more than the line the direction of the kick will be reversed, 
 i,e., downwards on depressing and upwards on raising the key. If, now, 
 we get an upward kick on depressing the key and another upward kick 
 on raising the key it indicates that the rushes of current into the line 
 and into the condenser, though they may be equal in magnitude, are 
 occurring one after the other, thus each preponderates in turn. The 
 object of the retardation coils is to adjust the time of the discharge. 
 Hence, summarising these facts, we have the following rules for 
 balancing the capacity of a circuit : — 
 
 1. If the needle kicks upwards momentarily on depression of the 
 key, the capacity should be increased. 
 
82 Simplex and Duplex Working. 
 
 2- If the needle kicks dowmjoards momentarily on depreesion of 
 
 the key, the capacity ahonld be decreased. 
 3. If the needle kicks upuxird» momentarily both on raising and 
 depressing the key, the second resistance, Bj is wrong. 
 
 The valne of Bi varies so little with the weather, etc., as seldom to 
 require any adjustment. If, while working, yon suddenly begin to 
 receive your own marks reversed, the line has increased in resistance, 
 and is probably cut, but if the marks are straight, either the wire is to 
 earth or the compensation drcnit is cut. Which is in fault may be 
 ascertained by turning the duplex switch, when if the line does not 
 appear earthy the compensation circuit is faulty. If it is found 
 impossible to balance owing to your not being able to reverse the 
 deflection, see if either split wire is off. The actual connections of 
 the apparatus are depicted in Figure 17. 
 
 The effect of heavy loss on a line is to render it unfit for duplex 
 working. The reason will be at once apprehended when it is 
 remembered that duplex working depends upon the preponderance of 
 the current in the line coil over that in the compensation coil. When 
 a wire is earthy its resistance is greatly decreased, consequently the 
 resistance in the rheostat is decreased. The strength of current 
 flowing through both coils is therefore increased. Now, marking is in 
 one case effected by the current from the distant station joining with 
 your own. If now, three-quarters of the distant station's current 
 flows to earth, only one quarter will appear at your end to join with 
 the heavy current you are sending out. The line side of the relay is 
 then slightly increased, but scarcely in a sufficient degree to render it 
 perfectly certain to overcome the strong puU of the compensation 
 circuit. It is, however, obvious that simplex working might be easy 
 and certain, as it does not depend on a preponderance of current in one 
 coil over that in another, but on the current which would flow through 
 both coils in series, hence the pull on the tongue would be at least 
 twice as strong. 
 
 liarge Daniells are used upon all double current duplex circuits less 
 than 150 miles long, and a sufficient number are allowed to provide a 
 current of from 14 to 17 milliamperes. Above 150 miles Bichromates 
 are used, A 5-cell large Daniell is used for the local. 
 
Electricity m its Application to Telegraphy. 83 
 
 CHAPTER XrV. 
 
 The QuADBtJPLEx, 
 
 Qnadruplex working depends upon the fact that currents differ from 
 each other in direction and in strength. Two relays are used, one an 
 ordinary polarised relay, which is only affected by changes in the 
 direction of the current, the other a non-polarised relay, which is 
 worked only by a strong current without reference to its direction. 
 
 QUADRUPUEX. __jjNe^ 
 
 l/fiSVK 
 
 Figure 18. 
 
 No matter how strong a current is, if it flows through the " A " or the 
 polarised relay in the direction D to U spacing will always result, but 
 with the "B" relay this is not so. The " B" relay has a soft iron 
 tongue which is pulled away from the electro-magnet by means of a 
 spring. This adjustable spring is made fairly strong, in fact, so strong 
 that a current of 10 milliamperes will not overcome it, whereas the 
 "A" relay will work with anything over J-milliampere. 
 
84 Theory of Quadruplex. 
 
 Two keys are used, one reversing the current and the other adding 
 more cells to the battery, thus increasing the cnrrent strength. In 
 order that the " B " marks may be clear and distinct the " B " relay is 
 held over to marking by the spring, the effect of the strong cnrrent 
 being to pnll the tongue away from M. The marks would thus be 
 received reversed on the sounder were it connected to T and M, but 
 instead of this an uprighting sounder is employed. Its armature is 
 held down permanently, but on the passage of a heavy current through 
 the "B" relay the circuit is broken, and the armature rises and 
 completes the soimder circuit, as shown in Figure 18. A weak cnrrent, 
 direction D to U, affects neither relay, but if in the direction U to D 
 the "A" relay marks. If strong in the former direction " B" only 
 marks, but if in the latter both mark. Both relays are differentially 
 wound, so that in the case of a " B " relay, if equal currents flow in 
 the direction TJ to D in one coil and D to U in the other, one current 
 magnetises and the other de-m^netises, hence no effect is produced. 
 
 The " A " key (Figure 18j is an ordinary double current key with the 
 switch removed, but joined up to " send." The " B " or increment key 
 is a modified sii^le current key. It consists of a lever playing between 
 two springs so adjusted that the lever always touches one of them. 
 At one point both are touched, thus short circuiting the battery, and 
 it is for this purpose that the resistances B^ and 8 are employed, 
 sparking at the battery being thus avoided. 
 
 The " A " key is joined up to the smaller battery through the 
 bottom contact of the " B " key. Depressing the " B " key adds to this 
 smaller battery a larger one, thereby increasing the current strength. 
 At one point the " A " key short drcuits the smaller battery through 
 B', which resistance prevents sparking on raising or depreasii^ the key 
 further. This short circuiting is of so momentary a nature as not to 
 affect the marks. 
 
 This arrangement is duplexed, the current splitting at the " B " relay 
 and passing throngh it, the " A " relay, and galvanometer. One coil of 
 the galvanometer is connected to line, and the other to the rheostat, 
 etc., which is joined to earth. A switch C is added, which cuts out the 
 batteries and inserts a resistance equal in value to their internal 
 resistance. The Fignie gives the connections for an up office, but in 
 order that the method of combination and not opposition may be used 
 the two bottom wires on the galvanometer and the battery are reversed 
 at a down office. 
 
 If both " A " keys are at rest, or both are depressed, the line sides of 
 the relay carry a cnrrent twice as strong as the compensation sides, 
 and therefore the requisite space or mark is formed. The " B " relay 
 is not affected, as the eSFecUve cnrrent is only of normal strength owing 
 to the compensation circuit opposition. Similarly if only one "A" 
 
Adjustments. 85 
 
 key is depressed, the compensation circuit causes the mark. Indeed, 
 the quadruplex behaves exactly as the double current duplex, for 
 when both "B" keys are depressed the preponderating current is 
 much stronger than when they are raised. For instance, suppose the 
 "B" current (i.e., the current required to work "B") is 30 m.a., the 
 two batteries combine together and 60 m.a. flows through the line side 
 of both relays, but this 60 is opposed by 30 in the compensation circuit 
 at the distance end, and consequently the efEective current is only 
 30 m.a., which, however suffices to work both B relays. The depression 
 of one B key augments the current at the distant end, thus actuating 
 the B relay there. At the sending end the heavy current affects both 
 relays equally and oppositely, hence no effect is produced. The B 
 battery consists of rather more than twice the number of cells in the 
 A battery, hence the current is three times as strong. 
 
 The most awkward case for consideration is when one A key and the 
 distant B key only are simultaneously depressed. Let us suppose that 
 the current is 10 m.a. When all keys are depressed at each end the 
 batteries combine together and 60 m.a. flow the line and 30 m.a. through 
 each of the compensation circuits. This 30 m.a. upon one coil of each 
 relay causes the B relays to be actuated. If one A key be released 
 there will be no current in the line circuit (assuming the insulation to 
 be perfect) as the batteries oppose each other, and 30 m.a. will flow 
 through both compensation circuits, thus actuating both B relays. At 
 the end where the A key is at rest the current will be in the direction 
 TJ to D, whereas at the depressed key end it will be in the direction D 
 to U thus spacing. If now the station which has the A key depressed 
 releases his B key we shall have one A key and the distant B key 
 depressed. At the depressed B key end we shall have, as before, 30 m.a. 
 in the compensation circuit in the direction U to D, but as the distant 
 battery opposes this we shall only have 20 m.a. in the line coil in the 
 direction D to U, i.e., a net effect of 10 m.a., which is in the direction 
 U to D. This will cause the A relay only to mark. At the depressed 
 A key end we shall have 10 m.a. in the compensation circuit due to 
 the A battery in the direction D to U, and 20 m.a. in the line coil 
 in the direction D to U due to the distant B battery. This is 
 equivalent to 30 m.a. in the direction D to U upon one coil, and hence 
 the balance is preserved. At this end the A relay spaces and the B 
 relay marks. A skeleton diagram, showing only one relay at either end 
 with the batteries joined direct to it without the keys, will render the 
 foregoing perfectly apparent. 
 
 Sometimes the B key is made to decrease the current. In this case 
 the uprighting sounder is removed and the local sounder and battery is 
 connected to the T and M terminals. When all keys are at rest the B 
 relays are held away from M by the strong currents, but immediately 
 
86 The Diplex. 
 
 the cnrrent strength is changed by depression of the B key the spring 
 polls the tongue over, thereby causing a mark. This system is known 
 as decrement working. 
 
 Adjustuents. — ^The circuit is balanced in exactly the same manner 
 as the duplex. But it is better to balance with full power, i.e., the 
 " B " key depressed. The play on the uprighting sounder should be as 
 small as possible. The missing of dots on the " B " side may be due 
 to:— 
 
 1. The spring of the relay being too tight. 
 
 2. The armature of the uprighting sounder haying too much play. 
 
 3. Failure of the sounder local. 
 
 4. Bad adjustment of the sounder and its spring. 
 
 The defect which is causing bad working may be ascertained by 
 -watching and listening to the uprighting sounder. A fault in the 
 distant " A " key, such as the sending only one current, may not affect 
 reading on the "A" side, as the relay will be re-adjusted by the 
 Teceiver, probably without his ascertaining the cause of the sticky or 
 light marks, as the case may be. The " B " side will not be able to read, 
 and we shall have the perfect station reading on "A," but not on " B," 
 -when "A" sounder is working. At the imperfect station they may 
 read perfectly on both sides. 
 
 Bichromate cells are used exclusively on all Quadruplex and Diplex 
 circuits. The current required is 10 to 15 m.a. for the " A," and 20 to 
 30 m.a. for the " B " side, thus the full current is from 30 to 45 m.a. 
 
 The Diplex. — Amodification of the quadruplex known as the Diplex 
 is often employed, and is very useful for cricket-ground work, etc., 
 where all the work is from the CG-. At the CG. there is an " A" and 
 a " B " key, but only one relay, which is of standard pattern. At the 
 other end there is only one key, but "A" and "B " relays. The con- 
 nections are the same as those of the quadruplex, but at the CG-. the 
 " B " relay is omitted, the wires passing through an " A " relay and 
 the galvanometer only. At the other end the "B" key and the 
 second half of the battery are omitted, the "A" key being joined 
 direct to the battery. 
 
Electricity in its Application to Telegraphy. 87 
 
 CHAPTER XV. 
 
 The Wheatstone Automatic. 
 
 The transmitter merely consists of a specialised form of double 
 current key, the object of which is to send out signals mechanically at 
 a greater rate than is possible by hand. The centre set of holes on 
 the perforated slip are for the purpose of guiding the paper only. If 
 the transmitter be run without slip it sends out currents of equal 
 duration, first in the one direction then in the other, thus forming a 
 series of dots. If the two holes in the perforated slip are vertical the 
 transmitter moves as though there were no slip in it, but if the lever 
 which stops the mark or effects the spacing meets an obstruction, in the 
 
 PERFORATED Strp 
 
 Figure 19. 
 
 form of a part of the slip without holes, the mark is continued till the 
 lever is able to rise through the next hole and a dash is formed. 
 
 The arrangement by which this is brought about is shown above. 
 The double current key is replaced by a divided lever so arranged that 
 each lever can touch one pole of the battery. When the top lever U 
 touches Cu the positive pole of the battery is joined to the up line, but 
 when U touches Zu the negative pole is so connected, and the current is 
 
88 
 
 Principle of Transmitter. 
 
 thus reversed in direction. This lever is controlled by the rods S and 
 M. A weight acting npon a train of wheels causes the beam PPi to 
 rock np and down thus, if there is no slip in the transmitter allowing 
 A and A' to rise alternately. When A rises K moves forward, pnshing 
 D over to Zd. ■■ The lever is held in this position by means of the 
 jockey wheel J. When A' rises D is poshed to Cd by K' and the 
 cnrrent is thus reversed and the mark ended. 
 
 It will be seen that the rising of one rod produces an effect and that 
 the rising of the other withdraws it, i.e., the rising of one rod put Z 
 to line, and after it has gone down to the level of the paper Z is still 
 
 FiGUEE 20. 
 to line until the other rod rises and reverses the current. If now this 
 other rod rises immediately afterwards a dot is formed, bnt if instead 
 of doing this it is met by the whole paper it has to wait until the next 
 hole appears, which is a dot later, and instead of two dots with a 
 space between them being formed the original mark is 'continued, thus 
 giving a dash three times as long as a dot. The spring and wheel J 
 are responsible for the continnation of the mark. 
 
 The four contact points of the transmitter have to be very exactly 
 adjusted, or only one current is sent out. Frequently it happens 
 through want of adjustment of the four contact points ithat the 
 transmitter sends out firmer marking than spacing currents, or mce 
 versa, and this fault is termed a marking or spacing bias as the case 
 
Wheatstone Connections. 
 
 89 
 
 may be, and is indicated by a permanent deflection of the galvanometer 
 -when the transmitter is running without slip. To make this adjust- 
 ment the transmitter should be run slowly without slip, and when the 
 deflections on either side of zero, observed on the galvanometer, are 
 equal the speed should be increased. If now no deflection is produced 
 the adjustment is perfect. The bolting of a transmitter is a 
 mechanical fault due generally to a speck of grease on the friction 
 wheels in the gear case, and should be dealt with by the mechanics. 
 
 The next point is the connections of the instrument . The transmitter 
 must cut out the double current key, as, if it did not the marks would 
 be reversed if the key was accidentally depressed and none would be 
 
 FiarKE 21. 
 
 sent if the key switch were to receive. The batteries are led straight 
 to the C and Z terminals of the transmitter (Figure 20). When the 
 transmitter is stopped Z and C are joined straight through to Z and 
 O of the key by means of MKZ and MKC. This is accomplished by 
 means of the three lever switch ABC. is joined to MKC by means 
 of the lever A which is on the left-hand contact when the transmitter is at 
 rest. Similarly Z is joined to MKZ by means of C which is resting on 
 its right contact, D is joined to K by means of B thus joining the 
 down line on to D of the key. In this position it will be seen that the 
 transmitter puts the battery and down line through to the key. 
 
 The path of the received current through -the apparatus (Figure 21) 
 is through the galvanometer, transmitter D to K, K to D of key, 
 through the key and receiver on to the up line. 
 
 The receiver merely consists of a standard relay, the tongue of 
 which can be caused to raise an inked wheel on to the paper which is 
 
90 Speed of Working. 
 
 pnlled along at a uniform speed. The two inked wheels are necessary 
 in order that the ink may not at the higher speeds he thrown on to the 
 paper in too large quantities. The sending circuit of the key is 
 precisely the same as that of the double current sounder. 
 
 The turning of the transmitter switch releases the mechanism and 
 the beam rocks up and down, at the same time the three levers, A, B, 
 and C, are thrown across. C and Z are joined to the left and right two 
 contacts by means of the levers A and C respectively. At the same 
 time D is joined to the lower half of the divided lever. U is per- 
 manently joined to the upper half, but when the transmitter is at rest 
 IT is dis beyond the lever. When the transmitter is started the three 
 switch levers are thrown across, thus joining the battery to the four 
 contact points, and joining the divided lever to the galvanometer and 
 earth. 
 
 To convert a double current set into a complete wheatstone first 
 remove the relay and join the two back wires to the receiver, taking 
 care to cross them, as the U terminal of the receiver is on the left. The 
 local circuit should be j oined to T and M . Disconnect the two battery 
 wires going to Z and C of the key and join them to Z and C of the 
 transmitter. Now join MKZ and MKC to Z and C of the key. Dis- 
 connect the wire going to D of the key and join it (the wire) to D of 
 the transmitter. Now join D of the key to K of the transmitter, 
 Lastly, join IT of the transmitter to U of the key and the set is 
 complete. 
 
 Speed op Wobking. — On aerial lines the speed of working is only 
 limited practically by the working speed of the apparatus employed. 
 If, however, there is a long section of underground such as that 
 between Manchester and Liverpool the speed is limited by the 
 capacity of the circuit, as it (the speed) is proportional to the product 
 of the circuit's capacity and resistance. Hence, in choosing a wire for 
 a race meeting, it is best to obtain one with as little underground as 
 possible. 
 
 There is yet another bad influence at work, and that is the self- 
 induction of the receiver, and it may be as well to here make a few 
 explanatory remarks on the subject of self-induction, as this property 
 of the receiver is a serious obstacle to the production of rapid changes 
 of current strength in the circuit. When a battery is connected to an 
 electro-magnet and a current flows Imes of force are formed, and 
 it is to these lines that we owe all the macfuetic effects of the current. 
 If a conductor is cut through by a line of force an electro-motive 
 force is called into existence in that conductor. This is the principle 
 of dynamos, where conductors are continually dragged through a 
 magnetic field and the currents produced iare collected by means of 
 
Self IndiKtion. 
 
 91 
 
 brashes and commutators. The current starts the lines of force, andf 
 in starting the lines have to cut through the windings of the electro- 
 magnet, and this they do in such a way that the generated Ejyi.F. 
 opposes the original action to which they owe their origin, i.e., is in an. 
 opposite direction to the E.M.F. of the battery. 
 
 The establishment of the magnetic field takes time, and if there are 
 three or four receivers in circuit this time will be greatly increased. 
 We should then have to lower the speed very considerably before each 
 mark could come out clear and distinct without interfering with the 
 subsequent mark. The opposition currents set up in the receiver, if 
 
 FiGTJKE 22. 
 they are allowed to come into prominence, very greatly distort the 
 marks. In order to minimise these effects a shunted condenser is used, 
 as shown in Figure 22. 
 
 This arrangement greatly minimises the effect of self-induction in. 
 the receiver. When the condenser is not charged it acts as a short- 
 circuit across the rheostat, and thus, owing to the decreased resistance 
 of the circuit a very powerful current is provided, owing to the 
 diminution of the resistance of the circuit, sufficiently powerful to 
 produce the normal current in spite of the backward pressure due to 
 the self-induction of the 'receiver. When the condenser is charged its 
 resistance rises to infinity and the decreased current continues to flow 
 throngh the rheostat. The backward pressure from the receiver coils- 
 
92 Shunted Condenser. 
 
 lasts only so long as it takes the lines of force to attain their steady 
 position. It is only during the time that the condenser is being 
 charged that the heavy current is provided, and we therefore see that 
 the arrangement merely provides a heavy current at the commencement 
 of the mark. During the greater portion of the mark the current is 
 flowing steadily through the rheostat without any backward pressure 
 from the receiver coils. 
 
 When the current is stopped the discharge from the condenser 
 opposes that from the receiver, and the result is the almost immediate 
 cessation of the current, provided that the capacity of the condenser, 
 the resistance of the shunt, and the self-induction in the receiver bear 
 a certain relation to each other. The receiver coils should be joined 
 in parallel, as their self-induction is thereby greatly diminished, A 
 circuit which has unit (absolute) self -inductance is defined as one in 
 which one line of force is started when a unit current (absolute, i.e,, 
 ten amperes) is applied to it. The practical standard, the henry, is 100 
 million times the value of the absolute unit. However, this digression 
 emphasises the fact that wherever there is a magnetic effect produced 
 there is also self-induction. 
 
 Beferring again to Figure 22, it will be seen that the condenser 
 becomes charged by means of the potential difference between the ends 
 of the rheostat. When the current is stopped the condenser dis- 
 charges. Two paths are open, the one through the rheostat, and the 
 other through the receiver as shown by the arrows in Figure 22. The 
 current which passes through the receiver is the only one which has 
 any effect, good or bad, on the marks. The arrow over the receiver 
 ehows the direction of the current due to self -induction, and it will be 
 seen that this current and the condenser current are in opposite direc- 
 tions, and that they neutralise each other. The value of the current 
 which passes through the receiver can be altered by altering the 
 rheostat and the condenser. If there be no resistance in the rheostat 
 the condenser cannot become charged. Again, if there be no capacity 
 inserted there can be no effect. 
 
 The resistance inserted in the rheostat has yet another good effect, 
 and that is that it increases the total resistance of the circuit, and con- 
 sequently the B.M.F. due to the self-induction of the receiver pro- 
 duces less current than if the circuit were of lower resistance, and we 
 therefore see that in adjusting a condenser it is preferable to make the 
 resistance high and the capacity low rather than vice versa. 
 
 It will naturally be asked, but why cannot a contrivance of this 
 sort be used to neutralise the capacity of the line ? The reason is not 
 far to seek. In this case the capacity is distributed throughout the 
 whole length of the line, and therefore any corrective measure must 
 also be distributive. Various corrective measures have been proposed 
 
Wfieatstone Duplex. 
 
 93 
 
 in the case of doable wires, but the only feasible remedy is due to 
 Mr. Preece. This measure is distributive, as it proposes to so design 
 the two conductors of a cable that the condenser effect of every inch 
 shall be exactly balanced by the current which one wire induces in 
 the other. This efEect is sometimes termed mutual induction, but 
 since mutual induction is defined as induction between two distinct cir- 
 cuits the term reduction of self-induction would seem more appropriate 
 to this case. If mutual induction is not possible then the effect of the 
 contrivance is merely to reduce the self-induction of the circuit and to 
 
 Boww Line on E 
 
 UP LIME OK E 
 
 0„ . O 
 
 -iHBDH 
 
 m 
 
 \o^ 
 
 PiGUKB 23, 
 
 increase its capacity. Lord Kelvin and Professor Oliver Heaviside 
 hold this view, and consider that the device is a step in the wrong 
 direction. They maintain that the capacity should be reduced by 
 increasing the distance between the conductors, thereby reducing the 
 capacity and increasing the self-induction. The writer is led to believe 
 that actual tests of cables designed on this principle will shortly be 
 made which will definitely settle the point. 
 
 The galvanometer is joined up in parallel and the coils are each 
 shunted by a resistance of 300w, in order to minimise the self- 
 inductance. (Figure 23.) The discharge on breaking the circuit (caused 
 by the lines again cutting the windings) prefers to travel round the 
 
 G 
 
94 Wheatstone Dv/plex. 
 
 shunt rather than through the circuit whose resistance may be between 
 3,000 and 6,000 ohms. 
 
 The adjustment of the current should be made before any capacity 
 is inserted. The normal current will then produce a deflection of 
 about 35°. The capacity should now be adjusted, and then all will be 
 in working order. 
 
 The electrical principles of the Wheatstone Duplex apparatus are 
 precisely similar to the ordinary double current duplex. The only 
 difference is that a second duplex switch is added for the purpose of 
 cutting out your own battery and inserting a resistance equal in value 
 to its internal resistance, in just the same way as in the quadruplex . The 
 transmitter is inserted with the double current key, and the receiver i» 
 in place of the standard relay. All the adjustments are precisely 
 similar to those previously described in the case of the wheatstone 
 simplex and double current duplex. The full connections are given in 
 Figure 23. 
 
 On all Wheatstone circuits a sufficient number of Bichromate cells to 
 produce a current of from 20 to 25 milliamperes are used. 
 
Electricity in its Application to Telegraphy. 95 
 
 CHAPTER XVI. 
 
 Test Box Arrangements. 
 
 In very small offices the wires are led straight to their respective 
 instruments, but in large offices this plan is not feasible. The wires 
 come to a terminal on the test-box, and from there are joined through 
 to the instruments. The up-lines are joined to the upper row of 
 terminals, and the down Imes to the bottom set. At very large offices 
 this arrangement is sometimes departed from, and the wires are put in 
 
 Figure 24. 
 groups and marked according to the routes they traverse. For instance, 
 there are fourteen wires going from Manchester to Liverpool by the 
 underground route, and these wires have a brass label above them 
 " Liverpool UG- " at Manchester. This arrangement is particularly 
 useful in tracing contacts or, indeed, for finding a particular wire quickly. 
 The Test Box enables us to at once determine whether a fault is in 
 the instrument or on the line, and also to make any change in the form 
 of apparatus. For instance, suppose it is found necessary to work 
 Wheatstone on a circuit owing to pressure, and that the form of 
 apparatus normally in use on that circuit is double-current duplex. We 
 
 g2 
 
96 Test Box Alterations. 
 
 look round and see where there is a Wheatstone set on a wire which is 
 being worked at key speed. We then cross the two instruments, i.e., 
 join the busy wire to the Wheatstone set, and put the other one on the 
 double current duplex. 
 
 At the test box we have both wires from the instrument, so that 
 alterations may be made without difficulty. The first row of terminals 
 are joined to the up lines, the second and third to the up and down 
 instrument teiininals respectively, the fourth to earth, the fifth and 
 sixth to the up and down terminals respectively, and the seventh to the 
 down lines. In Figure 24 A is joined through to its instrument in the 
 usual manner. The black centre row of terminals are all connected 
 together and joined to the various earth plates. The line comes to a, 
 which is joined through to the U terminal of the instrument by means 
 of a link, the D terminal of the instrument being earthed in a similar 
 manner. In this case we are down to a. 
 
 Through Wiees. — By this term is meant wires which only pass 
 through the box for testing purposes. For instance, LV — NT passes 
 through the Manchester box, and at MB, it is termed a through wire. 
 An example of such a wire is shown at 6. The up or LV side is joined 
 to h, and the first instrument terminal is connected (at the back of the 
 box) to the down instrument terminal at h, to which the NT or down 
 side of the wire is connected. Sometimes in smaller offices the instru- 
 ment terminal is joined to the down side, but this is not so convenient, 
 Intebmediatb Stations — An example of this is shown at q and v. 
 The up line comes to g, where it is joined through to the instrument 
 whose other terminal is (at the back of the box) joined to down 
 instrument terminal, this being connected to the down line by means 
 oiji, strap.^ To divide a wire means to terminate it at an intermediate 
 point, ^d ^^om; thence ^orfc, each si4e as a separate circuit. An 
 example is shown at t and y, where s and x instruments are used for the 
 purpose. To simply come intermediate on this wire, the second 
 instrument terminal at s should be joined to fs' first instrument 
 terminal, and y and its first should be strapped across ; t should now 
 be joined to the first instrument terminal of 8. 
 
 Looping Wiees. — This merely means join together, .^ and pare 
 shown looped. If you are asked to loop ten wires, join them all 
 together unless specially requested not to do so, e.g., unless you are 
 requested to loop TS 5 and 7 and TS 6 and 8. 
 
 Eaething Wiees.— The wire c is earthed. Eemove the link and 
 join the wire to a black terminal. 
 
 Ceossing Wiees. — To cross a wire means that another wire is to be 
 substituted for it. If LV sends a D8 to MR. " Cross LV— NT, and 3 
 racing," the LV side of LV— NT, is to be changed with the MR— LV 
 wire known as 3 racing. An example is shown at a, where the up side 
 
The Battery Box. 97 
 
 of the through wire is crossed with r. The down terminal should 
 now be joined to u. If the LV— MR length of the LV— NT wire is 
 workable at simplex, z should be joined to r instrument. The wires 
 must, of course, now be crossed also at LV, and an example of this 
 is shown at g and m, where g is put on to the m instrument and m on 
 to the g instrument. 
 
 Disconnecting Wires. — Remove all connections from the wire as 
 shown at s. 
 
 Forming a Metallic Ciecuit.— An example of this is shown at 
 to and X, where x is used in place of the ordinary earth return. This 
 arrangement is only resorted to during the prevalence of earth currents. 
 
 An example of an up wire put to an instrument usually used for 
 a down set is shown at /and n. 
 
 Forking Wires. — D and £fare shown as forked on to £ instrument. 
 A rheostat is inserted with E, whose resistance is much less than that 
 of jD, and suitable resistance is inserted so that the current may 
 be equally divided between the two stations. The down instrument 
 terminal may be earthed or joined to a down line if it is desired to 
 include some other office. If joined to an up station that station 
 must be requested to reverse. To fork a through wire it is merely 
 necessary to join one terminal to the wire it is desired to fork it with. 
 
 H and 1 are shown joined through. It signifies that MR gives up 
 two wires in order to form a wire for some important purpose. For 
 instance, an LV — NT wire might be made up of LV — MRs and MR 
 NTz. J is tapped to one instrument, i.e., there are two instruments in 
 circuits at the one office on 1 circuit, 3 is tapped on the leak principle, 
 but a rheostat might be inserted with advantage if the circuit is 
 anything but a local one, otherwise too much current will be 
 appropriated by the leak. 
 
 The Battery Box. — At large offices the test box is divided into 
 three divisions, viz., the line box, the line battery box, and the locals. 
 The arrangement of the battery box is exceedingly simple. The 
 negative pole of the battery is generally led to the left, and the 
 positive to the right hand terminal. The two terminals immediately 
 below are the instrument terminals from Z to C of the key. 
 Ordinarily these terminals are joined through by means of a link, thus 
 joining Z of the battery to Z of the -key. Now in order to increase 
 the power it is only necessary to disconnect the C link and join the C 
 terminal to Z of the battery which you utilise for the purpose, and 
 then join the C of that battery to the right-hand instrument terminal. 
 To cross a battery, i.e., to work the instrument by a second set of 
 batteries, all that is necessary is to join the instrument terminals to 
 that second set. The local box is precisely similar, and it is thought 
 that the foregoing renders its arrangements sufficiently clear. 
 
98 Electricity in its Application to Telegraphy. 
 
 CHAPTER XVII. 
 
 The Morning Tests, 
 
 In order that the Telegraph Service may be maintained in efficient 
 condition it is necessary that faults may be quickly detected and 
 remedied, if possible, before the heavy traffic for the day sets in. 
 With this object every important wire is tested in the morning at 
 seven o'clock. The system now in use is due to Mr. Eden, and is at 
 once rapid and simple to manipulate. It moreover possesses the great 
 advantage of placing the galvanometer and the testing battery under 
 the control of the testing office, and no doubt can then be entertained 
 as to the accuracy of the constant. 
 
 The principle of the test is to send out a current on one wire, receive 
 it back on a second, and observe the difference between the current 
 sent out and the current received. The current received is less than 
 the current sent out by the amount which passes to earth through the 
 insulators. Now, instead of using two galvanometers, one wound 
 differentially is employed, and the received current is made to pass 
 round the second coil in the opposite direction to the sent out current, 
 which passes through the first coil. The method of demonstrating the 
 theory of the test employed in this chapter will be numerical, as the 
 rigorous proof would render necessary the use of algebra. All the 
 quantities will be stated in B.A, units, as the instruments now in use 
 are all marked in B.A. ohms. 
 
 Figure 25 represents a telegraph line with a total resistance of 20 
 ohms having connected to it a battery whose E.M.F. is "18 volts or 
 180 millivolts (these figures are employed merely to simplify matters, 
 and the internal resistance of the battery is assumed to be practically 
 nothing). In the centre of the line there is a fault whose resistance is 
 40 ohms. 
 
 When no fault exists the current passing through the circuit is 
 
 2Q^ =9 milliamperes. The current which the battery will send 
 
 through the circuit when the insulation is perfect is called the " perfect 
 current!' The existenceof afault diminishes the resistance of the circuit. 
 For instance, the resistance of the above circuit with a 40w fault is 18w 
 instead of 20u when perfect. The reason is that the current flows 
 
Eden's Morning Test. 
 
 99 
 
 from A to 0, where it now has two paths open for it. Part of the 
 current goes through the fault and part through the OB. The joint 
 resistance of the two paths is equal to the product of their respective 
 resistances divided by the sum of their resistances. In the present case 
 
 the joint resistance of 40 and 10 = 
 
 40 
 
 _ip _ 400] _ 
 60 
 
 = *^ = 8«. The 
 
 40 -I- 10 
 
 resistance of the entire circuit is AC + joint resistance of fault and 
 CB =r 10 + 8 =: 18w. With the same E.M.F. the sent out current added 
 to the received current, i.e., the current passing A, has always the same 
 Talue (twice the perfect current), no matter how the insulation 
 resistance of the circuit may vary. Example. — Circuit perfect. Sent 
 
 out current = ^ = 9 m.a. 
 30 
 
 _ A IO*>vf- 
 
 ft |Oe>l>n>^. G 
 
 O 
 
 
 
 
 vfed 
 
 : — 
 
 
 
 ^-^ 
 
 E— 
 
 
 
 % 
 
 
 t 
 
 > 
 
 o 
 
 
 ,*■ 
 
 
 @ 
 
 
 
 
 
 (1 
 
 
 ? 
 
 
 j: 
 
 E 
 
 
 E 
 
 Figure 25, 
 
 Received current the same as none escapes = 9 m.a. Their 
 sum = 18 m.a. 
 
 180 
 
 With 40io fault. Sent out current = -^ := 10 m.a. 
 
 Received current ^ 8 m.a. 
 
 Their sum = 18 m.a., as before, when the circuit resistance was 
 assumed to be perfect. 
 
 The current sent out was 10 m.a. in this case. 10 m.a. reached C (as 
 A to C is assumed to be perfect). At C the current has two paths open 
 to it. The current always divides in inverse proportion to the resist- 
 ance of the respective paths. If the paths have equal resistance the 
 current splits equally, but if one is twice as great as the other, twice as 
 much current wUl pass through the path of lesser resistance. In the 
 present case four times as much will pass along CB as wUl flow 
 through the fault. Hence, if we divide the current received at C (i.e., 
 the sent out currents) into five equal parts, four will go along CB and 
 one wiU flow through the fault. The sent out current = 10 m.a. 
 
 .*, the current received at B =: 4'5ths of 10 =: 8. 
 
 The current through the fault := l'5th of 10 ^ 2. 
 
100 Theory of Test. 
 
 Sent out cmrent + Beceived current always the same. For every 
 milliampere which passes out at our fault J m.a. eostra is sent out, and 
 i m.a. less is received. 
 
 Sent out current with 40<d fault = 10 m.a. 
 
 Beceived „ _„ „ = 8 m.a. 
 
 Current flowing through fault = 2 m.a., i.e., 1 m.a. extra is sent 
 out, 1 less received, and 2 pass through fault. 
 
 Now imaafine that there is a galvanometer at A and at B, both of 
 which have negligible resistance. The deflection on the A galvanometer 
 corresponds to 10 m.a., and that on B to 8 m.a. The current sent out 
 minus the current received is, of course, equal to the current flowing to 
 earth at C. The sent out current minus the received current is also 
 equal to twice as much as the received is diminished by when the 
 insulation falls from the perfect to a lower value. 
 Perfect current received ^ 9 m.a. 
 Received current with 40k» fault := 8 m.a. 
 Amount by which received current is diminished = 1 m.a. 
 
 2 X 1 = 2 
 
 10 
 
 — 8 
 
 = 2 (9-1), ».e., 
 
 Sent out 
 
 Beceived 
 
 Twice amount ty which 
 
 current. 
 
 current. 
 
 received current is diminished. 
 
 One point in a line can only be at one potential at one time. The 
 point C is at a certain potential above the earth. Again, B is connected 
 to earth as is the end of our 40 ohm leak, therefore they are at the 
 same potential (lower than C). The diagram assumes no earth 
 resistance, so that the negative pole of the battery, the end of the 
 fault and B, may be regarded as all joined together. When three- 
 points are all connected together by paths of no resistance they must 
 be at the same potential, just in the same way G can obviously only 
 be at one potential. Now, by KirchofFs* second law the difference of 
 potential between any two points in a conductor is equal to the 
 resistance between the two points multiplied by the current in the 
 conductor. Between the two poles of the battery, i.e., between A and 
 B there is a difference of potential of 180 millivolts. 
 
 Between A and C the difference of potential = lOu x 10 m.a. =^ 
 100 millivolts. 
 
 Between C and B the difference of potential = 10 X 8 nLa. ^ 80 
 millivolts. 
 
 Between A and B 80 + 100 = 180 m.a., which is correct. 
 
 Between G and earth 40U x 2 m.a. = 80 miUivolts. 
 
 This also must be correct, as G is at 80 millivolts above B (assumed 
 to be zero). The points of zero potential are B, negative pole of 
 battery and end of leak. 
 
 ''KirchoffB laws are merely deductions from Ohm's law. 
 
Insulation Resistance. 101 
 
 If on our circuit (Figure 25) we get 10 m.a. sent out and 8 m.a, 
 received, what is the resistance of the fault ? First find the current- 
 through the fault. This is equal to sent — received current, j.e., 10 — 
 8 = 2. 
 
 Potential difference between O and B = Potential difEerencfr 
 between G and £, i.e., lOu x 8 m.a. ^= Besistance of fault x 2 m.a. 
 
 .'. twice the resistance of the fault = 80ui. 
 
 .*. Resistance of fault ^ 40 ohms, which is so. 
 
 This is the working in a particular case, but the results obtained are 
 none the less general. The resistance of the battery was supposed to 
 be nothing, but if its resistance were 20w the only effect would be that 
 A would be the centre of the resistance of the circuit in place of G. 
 
 As a matter of fact, these results would not hold were the fault 
 anywhere but in the centre of the line. In order that this may be so 
 a resistance of 10,000 ohms is put at each end of the* line. The 
 insulation resistance of the wire may then be assumed to be equal to 
 one fault, or leak to earth, whose resistance is the same as the insulation 
 resistance. For instance, if a wire is 10 miles long, and has an average- 
 insulation resistance of 10,000 ohms per mile, its insulation may b& 
 considered equivalent to a fault of 1000 ohms at the centre of the line 
 if the 10,000 ohm coUs are used at each end of the line to throw it 
 (the line) into the centre of the circuit. Further, instead of having 
 two galvanometers, one wound differentially is employed, thus its 
 deflections are proportional to the sent — received currents, which =^ 
 current through fault. 
 
 The total insulation resistance, or the resistance of the equivalent- 
 fault = resistance between G and B (J resistance of circuit) X current 
 in GB (received current) divided by the current through the fault. 
 
 The received current is equal to the perfect current minus half the 
 difference between the sent and received currents, or minus half th& 
 observed deflection. Therefore the insulation on any line is equal to- 
 the perfect current minus half the deflection produced multiplied by 
 half the total circuit resistance and divided by the deflection produced.- 
 
 (perfect cnrrent — j deflection) X j oct. res. 
 
 deflection. 
 
 If the perfect current and deflection are both stated in tangent 
 divisions the insulation resistance will be in ohms. 
 
 One great advantage of this test is that it enables us to test two 
 wires at once. Where a second wire is not available to loop with it the 
 wire is earthed. In ithis case only the sent cnrrent is observed, and 
 the reading or sent out-current is equal to the perfect current + i 
 leakage. Multiplying by 2 gives us twice the perfect current plus the 
 cnrrent finding earth through the insulators. The received cnrrent is^ 
 equal to twice the perfect current minus the sent current. If now w» 
 
102 
 
 Morning Test Connections, 
 
 8ATTCRY TSST-BOX. 
 
 DRY BATTBRY L OAH.BATT. 
 
 Ic I. 40clU Mio* 
 
 LINE 
 T«ST - BO)« 
 
 
 
 EARTH. 
 
 SW\TCH CONNtCTIONS IM POSITIONS 
 a" "3' "4-" 
 
 FiauBE 26. 
 
Morning Test — Practical Arrangemente. 103 
 
 multiply this by half the circuit resistance and divide by the loss we 
 have as before the insulation resistance in ohms. In testing single wires 
 the wire is earthed through 10,000u at the distant end. 
 
 Practical Arrangements. — In order that the galvanometer con- 
 stant may be correct it is necessary that the resistance of the standard 
 cell should be accurately known. The method of taking this test will 
 be explained in Chapter XXIV. Having ascertained the resistance of 
 the standard cell, reference should be made to the table given on 
 Page 38, and a deflection, corresponding to the internal resistance of 
 the cell, should be obtained in order that eighty divisions may accu- 
 rately represent one milliampere. The constant is then taken by 
 turning the switch to position 1. Insert the 750 coil by removing the 
 plug, and see that no shunts are in use. Now join Z of the standard 
 cell to the left and C to the right-hand test box terminal (Figure 26). 
 The path of the current is from C to circle-4, through the 750 coil, 
 through the galvanometer to 2-circle, thence to E to 2 of the switch to 
 3, to one of the galvanometer to G-, thence through the switch and back 
 to the battery. The galvanometer key short-circuits terminals 2 and 4. 
 This is the only case in which the 750(d coil is in circuit. In all other 
 cases it should be short-circuited by the plug. A, B, C, and D are left 
 disconnected in this case. 
 
 Having verified the galvanometer constant it is next necessary to 
 obtain the correct battery power to perform the test. To do this the 
 switch is turned to position 2, T is joined to 40 cells, and C and D are 
 connected by a strap. The current flows from C of the battery to F of 
 switch to 4, through the 10,000w block to 3 of switch to 1, to 4 circle 
 through the short-circuited 750 coil to 2 circle, thence through the 
 10,000u block back to the battery. The total resistance it has encoun- 
 tered (20,190io) is made up of two 10,000w coils, one coil of the 
 galvanometer IGOui and SOu internal resistance. A deflection of 108^ to 
 111^ should be obtained. Each cell added will increase the deflection 
 by 2f divisions. The deflectioa aimed at is 110, i.e., the E.M.F. of the 
 battery should be 55'5 volts. Only one coil of the galvanometer is 
 used, therefore the deflection would be 220 if both coils were used, i.e , 
 the current is 2'75 m.a. 40 divisions on half -coil = 1 m.a. 
 
 .'. Perfect current in tangent divisions on any line = 
 
 E.M.F. in miUivolta x coPBtant 55522'5 x ^- 
 
 Total resistance of circuit. B 
 
 To test a loop turn the switch to position 3. The current flows in 
 the negative direction from D to O, through the 10,000io coil to 2-oircle 
 through the galvanometer to 4-circle to the right test box terminal, 
 along the line to the left test box terminal, through the second lO.OOOui 
 block from 4 to 2 of the switch, through the coil 3 to 1 from G- to F to 
 earth. The current 2 to 4 is the sent current, and 3 to 1 is the received 
 
104 Practical Arrangements. 
 
 current. When the line is perfect no deflection is obtained, as the 
 received current is equal to the sent current, none being lost by 
 leakage. If the deflection obtained with the loop is above the maxi- 
 mum permissible turn the switch to position 1. If the deflection 
 is in the ordinary direction the wire on the right test box terminal is 
 the least perfect. If the deflection is in the other direction the left 
 wire is the offender. If no deflection is obtained the fault is probably 
 close to the looping office, and you cannot be certain which is wrong 
 without an independent test. The principle of this test is that, if the 
 fault be, say, in the middle of one of the wires, the current will nearly 
 all flow through the fault as a path of lesser resistance, thus making 
 the current in the galvanometer coil joined to the faulty wire pre- 
 ponderate. The path is Z through 10,000(<i to a point where it splits, 
 one part going direct to 2, and the other through the switch to 3 to 1 of 
 the galvanometer, through the switch again to the other line terminal- 
 Where a loop is not available the distant office earths the wire through 
 10,160a>, and the test is taken with the switch at position 3. Before any 
 tests are taken all loops should be. tried with a detector and battery 
 to see that the wires are looped. The 10-cell Daniell is provided for 
 measuring the resistance of earthed and disconnected wires in case a 
 bridge is not available. The insulation resistance of a wire disconnected 
 at the distant end maybe obtained by joining up the galvanometer and 
 sending the current through one coil only. Its insulation resistance ia 
 
 equal to Reflection °^^^- To measure currents turn the switch to 
 position 1, The wires to be tested are brought to springs, and switch 
 plugs are inserted which are connected to the line test box terminals. 
 This saves a great deal of time, as the insertion of a plug cuts out the 
 instrument circuit. The wires are led from the street to the springs, 
 which rest on contacts, which latter are connected to the line terminals 
 on the test box, so that the insertion of a plug cuts out the test box 
 side of the circuit. 
 
 The deflection corresponding to 200,000a» for the open and one 
 megohm per mile for the covered work is given, and any deflection 
 above this value indicates that the wire is earthy. If there ia one 
 distinct fault its position may be found by taking readings with it at 
 position 1 and position 4. The distance of the fault in ohms from the 
 testing office is found by multiplying the galvanometer resistance plus 
 the resistance of the loop by the larger deflection, dividing the result 
 by twice the larger minus the smaller, and subtracting 160 from this. 
 
 In order that the condition of lines may be seen at a glance, the 
 insulation resistance per mile should be given. The morning test gives 
 ns the total leakage on a line, and calculations will have to be made in 
 order that the condition of wires maintained by the Department, and 
 
standardisation of Readings. 105 
 
 ■^es maintained by the Bailway Companies for the Department, may 
 be compared. Also, this information is valuable as indicating which 
 lines require overhauling. A line may be in a bad condition but yet give 
 less than the deflection representing 250,000u per mile. In order that 
 these deflections may be directly comparable, they are all brought up 
 to a standard. The deflections are modified so that they represent 
 exactly an equivalent amount of leakage on a line 100 miles long, 
 whose conductor resistance is 500(». In order to demonstrate the 
 principle upon which the tables are calculated, let us refer back to 
 Figure 25. If the resistance of our circuit had been lOu, i.e., 5a< on 
 «ither side of the fault, the battery would send rather over 6 m.a. 
 through the fault. This is due to the diminution of the resistance of 
 the circuit. With a conductor resistance of 20u two milUamperes flows 
 through the fault. Again, with a conductor resistance of 40bi, 1'8 m.a. 
 flows through the fault. If, now, we compared the leakage or currents 
 through the faults in these three cases, we should come to the con- 
 clusion that in the first place the insulation resistance was only a third 
 of that in the first case. Clearly, then, we shall have to take account 
 of the conductor resistance of the circuit. Tables have been prepared 
 by means of which this correction may be made without calculation. 
 Nextly, we have to correct to standard mileage, An insulation 
 resistance of 40(i> on a 20-mile line is not equivalent to an insulation of 
 40«j on a 10-mile line. The insulation in the 'former case is twice as 
 good as in the second case. 
 
 In practice, two tables are used ; the first for converting the deflec- 
 tion observed into the total insulation resistance in ohms, and also for 
 correcting to standard conductor resistance, viz., SOOu. The total 
 insulation resistance in ohms is read ofE at the value opposite the 
 deflection (actually observed) under the conductor resistance of the 
 wire. Nextly, on the same table, under the 500u conductor resistance 
 «olumn, opposite the nearest value to the insulator resistance previously 
 found, the deflection opposite to it is noted. This deflection represents 
 the equivalent leakage in tangent divisions on a wire having a conductor 
 resistance of 500w, but of the same mileage as the wire under test. This 
 deflection must then be corrected to standard mileage, viz., 100 miles. 
 This can be readily calculated, as the insulation resistance per mile is 
 equal to the total insulation resistance multiplied by the mileage. The 
 standardised deflection or equivalent leakage on a standard line is equal 
 to the deflection previously found multiplied by 100 and divided by the 
 actual mileage of the wire under test. Tables for this conversion are 
 also supplied. 
 
 It will thus be seen that we have a leakage upon our lines given in 
 divisions of leakage upon a line of 100 miles in length and 500w 
 conductor resistance were the lines in the same condition as the line 
 
106 Galvanometer Shunts. 
 
 under test. Upon the standard line 154 divisions represents 250,000ii> 
 per mile, and 73 divisions represents l,000,000ii>, or one megohm per 
 mile. Thus instead of putting down the values of the equivalent 
 leakages upon the standard line in divisions we may write down the 
 insulation resistance per mile in megohms. This is accordingly done, 
 and in place of the second table mentioned above giving leakages in 
 divisions one giving the insulation resistance per mile is substituted, 
 and these values are entered upon the morning testing sheets. The 
 working standard deflection represents the condition in which it will be 
 possible to obtain the maximum working speed Wheatstone. 
 
 Galvanometer Shunts. — If we have a galvanometer of 90u resistance, 
 and across its terminals a place resistance of 90(ii, any current which we 
 send through the instrument will split equally between the galvanometer 
 and the resistance coil, i.e., the current will be halved, or the deflection 
 produced will be only half as much as if the galvanometer had no 
 resistance with which it had to share the current, i.e., as if it were 
 tmshunted. If the resistance were 45 twice as much current would got 
 through the shunt as through the galvanometer, i.e., the current may be 
 imagined as divided into three parts, only one of which goes through 
 the galvanometer. Similarly, if the resistance employed be 9<a the 
 current is divided into 10 parts, only one of which afEeots the 
 
 galvanometer. This is called a t^ shunt. The resistance of a shunt is 
 
 equal to (power of shvmt)-! "^Uing 10 the power of the shunt, we find 
 
 that the resistance of the shunt must be j^q^ = -J, i.e., Wit. The 
 
 object of the shunt is to give greater range to the galvanometer. 
 On P.O. galvanometers 80 divisions = 1 m.a. If we wish to measure 
 50 m.a. we must reduce the sensibility of the galvanometer as we 
 
 cannot read 4,000 divisions of our scale. The i shunt will reduce 
 the deflection to 100 divisions. The constant of the instrument is now 
 80 divisions per 40 m.a. Of course terminals 3 and 2-circle must be 
 joined. The joint resistance of the galvanometer and shunt is always 
 
 power of ahunt °* ^^^ resistance of the galvanometer unshunted. For 
 
 instance, the resistance of the instrument with-gi?, shunt is sir of 
 
 (2 X 160) = Iw. Both coils are joined together and the shunt comes 
 between the 1 and 4-oircle. To use a particular shunt insert a plug at 
 the value required. 
 
 The shunts affixed to the P.O. tangent galvanometer are l-20th, 
 l-40th, l-80th, l-160th, l-320th. Their resistances are respectively 
 l-19th, 1.39th, l-79th, 1.159th, and 1.319th of 320a>. The combined 
 
Galvanometer Shunts. 107 
 
 resistance of the galvanometer in each case is IBu, Sw, 4(i>, 2u, and lu^ 
 Depression of the key short-circuits the galvanometer. When a shunt- 
 is used the deflection observed must be multiplied by the power of the' 
 shunt in order to compare the current producing it with a current 
 through the unshunted galvanometer. In other words, the constant of 
 the galvanometer when shunted is equal to the constant unshunted 
 
 divided by the power of the shunt. 171 divisions with ■^. shunt = 
 
 13,680 divisions, i.e., 171 m.a. 
 
 80 
 
 The current is then -stt = 1 m,a. per division. 
 
108 Electricity in its Application to Telegraphy. 
 
 OHAPTER XVIU. 
 
 Localisation Tests. 
 
 In the previous chapter we have seen how faults may be detected by 
 the morning test before the work of the day sets in. In order that 
 faults may be quickly remedied it is essential that the Engineers 
 Department should be given some data as to where the fault is, for 
 instance, it would be absurd to merely remark that a wire was cut 
 between London and Glasgow. When a wire is reported to the test 
 clerk as faulty he immediately proceeds to find out as nearly, as may be 
 the locality of the fault in order that he may advise the Engineer 
 responsible for the good working of the length in which the fault occurs 
 and his lineman. For this purpose wires are led into the test boxes of 
 various offices en route, and in practice faults are localised only between 
 two successive offices. 
 
 In order that these tests may be made rapidly a special galvanometer 
 and switch is used. The ordinary form of test box galvanometer is 
 shown in Figure 27. It will be seen that a somewhat complicated switch 
 consisting of four quadrants of brass surrounded by a ring of the same 
 metal, but insulated from the quadrants, is employed. The ring is 
 connected to earth and four grooves are cut in it so as to admit of a 
 brass plug being placed between it and any of the four quadrants. 
 We may then earth any one of the quadrants by placing a plug between 
 that quadrant and the earth. C and Z are two opposite quadrants 
 connected respectively to the copper and zinc poles of the battery. X 
 and Gr are the two other quadrants and it will at once be seen that we 
 may connect Z to either X or G- by placing a plug in the groove between 
 Z and X or between Z and G. The same remarks apply to C which may 
 be connected to X or G at will. If we connect Z to G then we shall 
 desire to connect C to X and this as we have seen can readily be done, 
 Now in order to reverse the battery, i.e., connect Z to X and C to G, it 
 is merely necessary to put in a plug between Z and X and one between 
 C and G. It will be seen that X is connected to the left hand terminal 
 of the instrument, and that G goes to the centre terminal, wluch in turn 
 
Test Box Galvanometer. 
 
 109 
 
 is joined through the galTanometer to the right hand terminal. If we 
 desire to test a relay in order to ascertain whether the continuity of the 
 coils is perfect, we should join the coils of the relay to the two outer 
 terminals. By inserting plugs as explained above we should have a 
 battery, galvanometer, and the coils of the relay in circuit. If the 
 continuity of the circuit is perfect then a deflection will be observed on 
 the galvanometer. If we connect a wire to the right hand terminal of 
 the galvanometer we may, by plugging G- and R put it to earth through 
 the galvanometer. We may connect the zinc pole of the battery to line, 
 
 FiGrBE 27. 
 
 or as it ia termed, send a zinc current to line by plugging Z and G and 
 earthing the positive or copper pole of the battery by plugging C and B. 
 In order now to send a copper current, i.e., to reverse the battery, Z and 
 B and G and G should be plugged. In order to observe a current on a 
 through wire the right and middle terminals may be used and no plugs 
 must then be inserted. Again the same object could be accomplished 
 by using the two outside terminals and plugging X, Z and Qr. To put 
 on a current remove the Z and G plug and insert it between X and G 
 or remove Z and X plug and plug G and G. 
 
 H 
 
110 Localisation of Faults. 
 
 A. most excellent modiflcation of this arrangement was designed by 
 Mr. Matthews of Manchester. It consists of a modified duplex switch. 
 Three positions are used, left, right, and central. The two outer 
 positions send copper and zinc to line, and the central one connects the 
 galvanometer to earth. Contacts are tested by joining a second battery 
 to the wire in contact and testing the suspected wires by means of the 
 earthed galvanometer, i.e., with the switch in the central position. 
 
 Earth. — ^An earth fault is at once apparent from the increased 
 deflection observable on the galvanometer in connection with the 
 instrument on the circuit. This is caused by the reduction in the 
 resistance of the circuit due to the fact that the current is now given 
 an additional conducting path, and this, as we have seen, increases the 
 conductivity of the circuit or reduces its resistance. It should, in 
 testing for &,ultB, be ascertained beyond doubt that the fault is not in 
 the apparatus or leaders to test box before testing the wire. Having 
 ascertained by sending a current through the receiving apparatus and 
 also by receiving a current at the test box from the apparatus, that the 
 fault is a line fault, we should then proceed to make our test. The 
 principle of the test is to ascertain up to what point the wire is clear 
 of earth. If we join our wire to the right hand terminal of our 
 galvanometer and plug Z and G- and C and E, we shall send a zinc 
 current to line. If the wire is disconnected and the insulation from 
 earth is perfect, it is obvious that no current can flow through the 
 galvanometer, but if on the other hand the wire is to earth the circuit 
 will be completed from the zinc pole of the battery, through the galvan- 
 ometer, along the line to earth, through the fault and back to C of the 
 battery, thus causing deflection. We have the wire disconnected at 
 successive points and apply the test wire. 
 
 If the wire has slipped off the insulator and is touching the bolt, 
 the wire will be dead or full earth, but the fault may not be so pro- 
 nounced as this, in which case it would be termed a slight or partial 
 earth. The three terms usually used are partial earth, earth, and full 
 earth. Partial earth is generally due to the breaking of insulators. All 
 earth faults should be tested with the 10-cell Daniell as described on 
 page 104, and defined in tangent divisions. 
 
 As an example of a localisation test, let us suppose that MB- 
 Huddersfield 3 is to earth. MB-HF passes through Oldham, Rochdale, 
 Todmorden, and Halifax. Ask a middle ofiSce, say Bochdale, to 
 disconnect the wire. When a BQ is received saying that this has been 
 done, join the test wire (a length of wire joined to the right hand 
 terminal of the galvanometer) to line. A battery is thus connected 
 to the line with the galvanometer in circuit. If the wire appears dis. 
 to Rochdale, then clearly the earth fault is not between Manchester 
 and Bochdale, i.e., it is between Bochdale and Huddersfield. Now ask 
 
Disconnections and Contacts. Ill 
 
 Sochdale to join the wire through, and, at the same time request Halifax 
 to disconnect it. If now the wire does not appear disconnected, i.e., if 
 a deflection is produced it is faulty, and since we have proved that it is 
 olear to Rochdale, the fault must be between Bochdale and Halifax. 
 Now have the wire cut at Todmorden, and if the wire proves clear to 
 there, then the fault is between Todmorden and Halifax. The 
 Engineer responsible for this section of line, his local lineman, and the 
 Superintending Engineer of the District, should now be informed that 
 MR-HFais to earth between Todmorden and Halifax. If the wire is 
 on the Railway, the Telegraph Superintendent of that Railway must be 
 advised. These remarks apply to any class of fault. 
 
 Disconnection. — Faults of this character are localised in the same 
 manner as earth faults, viz., by connecting a galvanometer with an 
 earthed battery to line, but the wire is earthed at the various points. 
 The continuity of the wire is proved good up to a certain point, 
 beyond which to the next point it is found that a wire is disconnected, 
 therefore, the fault must be between these points. For instance, let 
 us suppose MR-HUsis reported to us at Manchester to be disconnected 
 by the Divisional OfScer. MR-HUs passes through the test boxes at 
 Rochdale, Todmorden, Halifax and Bradford. 
 
 MR to HX. Wire earthed at HX. — Dis. (no deflection). 
 
 MR to Todmorden „ „ Todmorden. — Dis. (no deflection). 
 
 MR to Rochdale „ „ Rochdale. — Earth (deflection). 
 
 MR-HUs dis Rochdale and Todmorden. 
 
 Contact. — It should be first ascertained that the fault is not in the 
 leaders to the test box. If it is in the leaders, working one instrument 
 will afiect the other when both lines are disconnected from the 
 apparatus at the test box. In a large office the simpler plan is to 
 disconnect both instruments from both lines and earth, and put a 
 current on from the first instrument to the second by joining a 
 terminal of one to the right hand terminal of the galvanometer, and 
 joining a terminal of the other instrument to the left hand terminal of 
 the galvanometer, at the same time plugging X and Z and C and Gr, A 
 contact manifests itself by currents being received on a wire when 
 neither station are sending on the wire. In order to ascertain whether 
 two wires are in contact we should join an earthed battery to one wire, 
 and an earthed galvanometer to the other. If the wires are touching, 
 and in contact, then the current sent on the first wire will be received 
 i.e., the second. This is an important point, as if a battery and galvano- 
 meter are merely joined between the two wires a current will be 
 indicated on the galvanometer if both wires are earthy though they 
 may not be in contact. Some time ago an amusing error was made by 
 a Test Clerk who did not pay due regard to this point. The North side 
 of a through wire was reported in contact with a wire going due South. 
 
 h2 
 
112 Intermittent Faults. 
 
 Both wires were to earth, consequently a deflection was observed as 
 explained. Frequently both contact and earth occur at the same point 
 and this fact is duly chronicled by the efficient Test Clerk. Having 
 ascertained that the two wires are in contact we may localise by 
 having one wire disconnected at successive points. The wires should be 
 joined to the two outside terminals of the galvanometer and C and B, 
 G and E and Z and X plugged. This puts a zinc current on to the 
 first line and earths the second one through the galvanometer. In order 
 to see whether there is also earth remove X and Z and plug Z and G. 
 
 If a wire is merely reported in " contact " we must ascertain with 
 what wire it is in contact, and the usual method of doing this is to 
 connect an earthed galvanometer to the faulty wire and successively 
 touch each wire following the same route with an earthed battery and 
 watch for the wire which when touched causes a deflection of the 
 galvanometer. This is done without disconnecting the circuits, and only 
 results in an RQ or two if the wire touched is working. The better 
 plan is to join the battery to the faulty wire and touch the working 
 wires with the galvanometer. If two wires are in full contact and are 
 not to earth it is obvious that one may be worked by suspending the 
 other wire and having it disconnected at both ends. If this latter 
 precaution were not taken the first wire would be to earth through 
 the apparatus at the end of the second wire. 
 
 As an example of a test for localising a contact, let us suppose that 
 MR-BDi and LV-NT, are reported to Test MR to be in contact. 
 These wires only travel together from MR to BD, therefore, the 
 contact is somewhere between these two points. Both wires pass 
 through Manchester, Rochdale, Todmorden, Halifax, and Bradford. 
 We have then to ascertain whether the fault is between MR and 
 Rochdale, Rochdale and Todmorden, Todmorden and Halifax, or 
 Halifax and Bradford. The test would then be proceeded with as 
 under : — 
 
 MR to Todmorden (both wires dis. at Todmorden) deflection, i.e., 
 contacts occur at these two points and not between Todmorden 
 and Bradford. 
 
 MR to Rochdale (both wires dis. at Rochdale) no deflection, i.e., 
 contact does not occur between these two points, therefore it 
 is between Rochdale and Todmorden. 
 
 If, in this latter case, we had had a deflection, then the fault would 
 have been between MR and Rochdale. If half-a-dozen wires are all 
 reported in contact with each other, take these tests with any two, and 
 having discovered the locality prove that all are in contact at this 
 point. 
 
 Contacts may differ in degree, and the three gradations generally 
 mployed are slight contact, contact, and fall contact. Slight contact is 
 
Covered Faults. 113 
 
 usually due to kite strings — contact and full contact is frequently caused 
 by wire, etc.. thrown across the wire by ignorant or malicious children. 
 
 Intermittent Faults. — ^If faults occur on a wire and disappear, 
 then occur again and again disappear, etc., they are termed intermittent 
 faults. In every case proceed as before and carefully watch for a 
 recurrence of the fault. If the fault recurs often enough to cause 
 delay on the circuit, cross the faulty length with a length of good wire 
 which can be spared. If the faults catuse little or no trouble and are a 
 long time in recurring, they may be localised by crossing successive 
 lengths of the wire as under, but this test is not found very reliable in 
 practice. MR-BDi is to earth intermittently, MR-HUa is in good 
 condition, cross MR-BD, with MR-HUj at RO, and cross the BD and 
 HU instruments at MR, Result : — Faults reappear on MR-BD,, i.e., 
 in crossing the MR-RO length we have not crossed the faulty length, 
 or fault is not MR and RO. Now cross the length RO to Todmorden 
 by requesting the latter office to cross MR-HUs and MR-BDi. 
 Result : — Fault now manifest itself on MR-HUj. The latter cross has 
 removed the fault from MR-BD, to MR-HUj, therefore, we have 
 discovered the faulty length, viz., RO to TM. 
 
 Covered Wire Faults. — These are faults which cause far more 
 trouble to the Engineer's Department than to the Commercial Branch. 
 The symptoms of a covered fault are as follows : — ^In testing the wire 
 in the morning it appears below standard. When placed on the test 
 box galvanometer with a few cells it appears nearly perfect, though 
 appearing better with a copper than with a zinc current. If the power 
 be increased and a zinc current put on, the needle will slowly travel 
 over till a big deflection (showing full earth) is obtained. On reversing 
 the battery the needle will slowly march back till the wire appears 
 nearly perfect. The ezplaaation of this pheaomenon is that a current 
 flowing from the wire through the fault to earth forms au oxide of 
 copper at the fault. This oxide is a non-conductor, and hence the 
 fault is covered up. A zinc current reduces these oxides and keeps the 
 wire 'clean, thus rendering the fault more pronounced. A common 
 mistake is to increase the power on a circuit immediately it appears 
 earthy. A good example of this occurred some time ago. A special 
 wire was earthy in the underground, and the power was increased to 
 the verge of absurdity. This resulted in very nearly full earth, and 
 the wire was stopped, A lineman was sent out to cross this wire with 
 a local wire in the streets. The special wire, meantime, had a rest and 
 recovered somewhat, at all events the local office was able to work 
 through it with their half-dozen Leclanches, and the special wire was 
 in good order. Submarine cables are worked with the least possible 
 power for these reasons, A hexode battery connected to an Atlantic 
 cable would probably cause a hundred or two faults. 
 
114 Stay Faults. 
 
 If a covered fault exists, which is intermittent and causes much 
 trouble, it is as well to develope it by means of a zinc current, but 
 this should only be done in very extreme cases, as other faults may be 
 thereby developed. 
 
 Stay Faults. — These are earth faults caused by the wire almost 
 touching the stay. Water is held by capillary attraction between the 
 wire and stay, and when a current is sent through this drop of water 
 the resistance of this fault is increased. It differs in symptoms from 
 the covered fault, inasmuch as a current in either direction runs it up. 
 
 The crossing and making good of faulty wires is dealt with in 
 Chapter XXV. 
 
Electricity m its Application to Telegraphy. 115 
 
 CHAPTER XIX. 
 
 The Wheatstone Bridge. 
 
 The principle of the Wheatstone Bridge will be proved by means of 
 a numerical example in order to avoid the use of algebra. The bridge 
 consists of three known resistances, by means of which and a 
 galvanometer, the fourth or unknown resistance, is determined. The 
 arrangement is shown in Figure 28. C is the unknown resistance. 
 
 FiGTJBE 28. 
 
 The other three sides of the figure represent the three known 
 resistances. When the point Q is at the same i potential as the point 
 H, there can be no flow of current between them, and therefore no 
 deflection on the galvanometeri When this state of aSairs obtains the 
 
116 Theory of Wheatstone Bridge. 
 
 prodDct of the rraistance of the opposite sides are eqnal, i.e., A x D ^ 
 
 A 
 
 C X B, i.e., C = ^xD. IfA = B, then C = D, or the resistance 
 
 nnplngged in the box. The two sides A and B are termed the 
 proportionals. In the case given above A = 8 ohms, B = 12 ohms, C 
 = 4, D ^ 6. The resistance of the battery is assumed to be '8(o, and its 
 E.M.F. 8 volts. As no current passes throngh OH when a balance is 
 obtained we may ignore it, as in the present case, where 8 x 6 = 12 x 
 4, The resistance of the entire circuit is then the joint resistance of 
 E6F and EHF pins the resistance of the battery, t.«., 
 (8 + 4) X (12 + 6) , .-_ 12x 18 , .B_216 ■ .« _ ,^ , .« _ 
 (8 +4) + (12 + 6)"^ "~12 + 18'*'''~30''" »—'''+ «- 
 8 ohms. 
 
 E ,1-8 volts , 
 
 E=^--8^h5^=l*™P«'^- 
 
 The difference of potential between F and 6 ^ the current in FG- 
 multiplied by its resistance. The total current ^ 1 ampere, but it splits 
 between the two paths in inverse proportion to their resistances, 
 8 ohms and 12 ohms, 2.e., as 12 ; 8, or as 3 : 2 respectively. 
 
 .'. 3-5ths of the total current (one ampere) goes EGF = '6 ampere, 
 and 2-5ths of the total current (one ampere) goes EHF =: '4 ampere. 
 By ohms law C X B = E. 
 
 .•. Difference of potential between F and G = 4 x "6 ^ 2'4 volts. 
 
 Between G and fe = 8 x 6 = 48 volts.- 
 
 Between F and H = 6 x "4 = 2-4 volts. 
 
 Between H and E = 12 x "4 = 4-8 volts. 
 
 From this we see that the potential of G above F = potential of H 
 above F. .'. the relationship is true in this case. By upsetting the 
 balance, it will be found that G and H are at different potentials, and 
 therefore a flow of current will take place between G and H. For 
 instance, let us in the present case suppose that B and D are crossed, 
 thus making B =r 6 ohms and D = 12 ohms. This will not alter the 
 value of the current passing through the circuit as the resistance of EHF 
 has not been altered by the change. The products of the resistance 
 of the opposite sides are no longer equal (8 x 12 does not equal 4x6) 
 therefore, if the relationship indicated is true there will be a difference 
 of potential between G and H. 
 
 Difference of potential between F and G ^ 4 x '6 = 2'4 volts. 
 
 Between G and B = 8 x '6 = 4-8 volts. 
 
 Between F and H = 12 x '4 = 48 volts. 
 
 Between H and E = 6 X '4 = 24 volts. 
 
 We see that we have made no mistake as the difference of potential 
 between B and F via G equals that via. H, viz., 4*8 + 2*4 := 7'2 volts. 
 
Practical Arrangements. 
 
 117 
 
 With respect to F, G is at 2'4 volts, and H with respect to the same 
 point is at 4*8 volts, therefore there is a difference of potential between 
 Q, and H, and a galvanometer connected between these two points will 
 indicate a current through it. 
 
 The practice of the bridge is equally simple. A and B consists of 3 
 coils, viz., 10, 100, 1000 ohms. D consists of coils of various values, so 
 that any resistance from 1 to 11,110 ohms can be at once unplugged. 
 The actual arrangements are shown in Figure 29. BC and CA are 
 the proportionals ; K: puts on the battery, Ki inserts the galvanometer. 
 The battery key should always be depressed before the galvanometer 
 key, in order that the current may have attained a steady value 
 before the galvanometer is joined up. The current does not attain 
 its maximum strength so quickly through a wire possessing self- 
 induction as through the plain resistance coils of the bridge. 
 Besistance must always be inserted in BC and CA, otherwise the 
 galvanometer will be short-circuited. In order that the bridge may 
 be as accurate as possible, the following rules must be observed : — 
 
 (a) The resistance inserted proportionals BC and CA must be as 
 near to the resistance to be measured as possible. 
 
 (5) The right-hand key must be depressed before the left. 
 
 (c) The keys should be depressed for as short a time as possible. 
 
 Suppose we wish to measure the resistance of a line the resistance of 
 which we know to be about 700 ohms. Take out 1,000 in BO and 1,000 
 in CA. Take out the infinity plug, and note the direction of the deflection. 
 The method is to reverse the direction of the deflection, and then we 
 know the resistance we are measuring is between these two values. 
 
 Besistance iinplugged. 
 
 Deflection towEirds 
 
 Unknown resistance is 
 
 Infinity 
 iCOO ohms 
 400 „ 
 600 „ 
 650 „ 
 630 „ 
 640 „ 
 645 „ 
 643 „ 
 642 ., 
 
 Bight 
 
 Left 
 
 Bight 
 
 Left 
 
 Left 
 
 Bight 
 
 Bight 
 
 Nil 
 
 Between and infinity 
 Between and 1000 ohms 
 Between 400 and 1000 „ 
 Between 600 and 1000 „ 
 Between 650 and 600 „ 
 Between 680 and 650 „ 
 Between 640 and 650 „ 
 Between 640 and 645 „ 
 Between 640 and 643 „ 
 642 ohms 
 
 In this case a deflection to the right means the resistance in the box 
 is greater than the unknown resistance, and to the left that it is less. 
 In the case worked out on Page 116, where B ^ 6 and D ^ 8, H was at 
 a higher potential than G, but if we had made the product of C and B 
 exceed that of A and D, then G would have been at a higher potential 
 
118 
 
 Condxictivity Test. 
 
 than H, and hence the coirent through a galvanumeter placed between 
 6 and H would have been in different directions. We thus see that a 
 deflection in one direction indicates that the resistance unplugged in 
 box is too much and in the other direction it is too little. 
 
 In the P.O. form a plug-switch is added which reverses A and B. It 
 consists of four quadrants arranged as in test box galTanometers . Two 
 opposite quadrants are joined to line and earth respectively, and the 
 other two are joined to A and B. The object is that tests may be 
 taken with both poles of the battery, as the effect of the direction of 
 the cnrrent on faults are different. The mean of' the two values 
 obtained may be assumed to be the correct value. 
 
 Conductivity Tests.— U the wire is earthed at the distant end F (or E) 
 should be earthed, and E (or F) should be connected to the wire. If 
 the wires are all joined together at the distant end, E and F should be 
 joined up as shown in Figure 29. If four wires, a, b, c, and d are looped 
 together, each wire must be measured twice thus : — 
 
 a and h, b and c, a and c, c and d. 
 
 O 
 
 ^ fOCO fOC fo C /o foO fOOO J^ 
 
 1 * 
 
 T 
 
 loo /oo 
 
 Z^oo 
 
 /o 
 
 do 
 
 
 RESISTAHCe TO BE TCSTCa 
 
 ) 
 
 hooo AOoo 
 
 'Hllh 
 
 FiGUEE 29. 
 
 By adding to the resistance of a and b that of h and c, we have the 
 sum of twice the resistance of b and that of a and c. If now we 
 subtract the resistance of a and c from this value we have left twice 
 the resistance of h. 
 
 The resistance of a is then obtained as follows : — To the resistance 
 
 obtained for a and b, and that obtained for b and c. From the result 
 
 subtract that obtained for a and c, and finally divide by 2. Knowing 
 
 b we know a, hence c and d may be at once ascertained, 
 
 „ . , , , (Kes. a and 6) + (Bes. 6 and c)— (Bes. a and c) 
 Besistance of o = o 
 
Loop Test. 
 
 119 
 
 Loop Test. — ^It is sometimes very desirable to ascertain the exact 
 locality of a fault, and if a second wire is available for the purpose 
 this may be done very easily. 
 
 The principle of the test may be comprehended from Figure 30. 
 
 3) 
 
 Figure 30. 
 
 The good and the faulty wires are looped at the distant end, e.g.y 
 MK-BLi and MR-BLz looped at BL for MR. The two wires are made 
 to form the third and fourth arms of the bridge (the proportionals- 
 forming the first and second arms). The conductivity resistance of 
 the loop is first measured. The earth fault will not cause any material 
 error as the circuit is metallic. Suppose the resistance of the loop is- 
 800 ohms. The battery is practically joined to F. In this case FG- 
 forms one arm, and FD + the resistance in the box HD forms the' 
 other. The proportionals a and & should each be 1,000 ohms. In- 
 order that a balance may be obtained we must add to FD sufl&cient 
 resistance to make it equal to FGr. When this has been done twice 
 (HD + DF) = total circuit resistance plus the resistance taken out ia 
 the box in order to obtain balance. 
 
 , DF = 
 
 total circuit resistance — resistance unplugged. 
 
 The resistance of DF gives then the distance of the fault from the 
 testing office. By dividing this resistance by the resistance per mile 
 of the wire, obtained from the conductivity test, we shall ascertain 
 approximately the position of the fault in practice. The results of 
 the tests are valueless when the resistance of the fault approaches the 
 insulation resistance between the lines. This test is not used very 
 often in the case of aerial lines. 
 
 Actual Connections. — Earth the battery, and disconnect it from D, 
 Join the good wire to E, and the faulty one to F, 
 
120 Electricity in its Application to Telegraphy. 
 
 CHAPTER XX 
 
 The Universal Battbby System. 
 
 The object of the system is to work the cells as efficiently as possible, 
 i.e., to get from them the greatest possible amount of energy at the 
 minimmn cost. One set of cells is used to work several circuits. Suppose 
 we have ten circuits, each having a resistance of 1,000 ohms, and that our 
 "battery is a perfect battery, having an E .M. F. of 10 volts, and an internal 
 resistance of nothing. The current through one wire would then be 
 
 YWn = 1^ °^-^- ^ ^^ were placed in parallel, i.e., bunched together at 
 
 the sending office their joint resistance would be 100 ohms, consequently 
 
 the current through them would be ttttt ^ 100 m.a. This 100 m.a., 
 
 liowever, has to divide between the ten circuits, each of which will 
 take one-tenth or 10 m.a. as before when only one wire was joined up, 
 and thus we see that with a battery of absolutely no internal resistance 
 the addition of wires of equal resistance in no way affects the 
 current through each, and hence an indefinite number of circuits 
 might be worked simultaneously from one such battery. When only 
 four keys are down the resistance of the circuit is 250 ohms, and the 
 current leaving the battery is 40 m.a. This has to divide between the 
 ionr circuits, hence the current sent ^ut to each station is not 
 diminished when other keys are depressed. We have been considering 
 the battery resistance as nothing, and we see that no diminution of 
 current takes place when all the circuits are worked simultaneously. 
 If the battery has any internal resistance, which it, of course, must have, 
 -the current will undergo a diminution. The extent of this diminution 
 will depend upon the values of the line resistances, and upon the 
 internal resistance of the cells, hence we see that it is necessary to use 
 batteries whose internal resistance is as small as possible. The internal 
 resistance of the cells limits the number of circuits that can be 
 worked off one battery. Accumulators are now being very much 
 Tised on account of their exceeding low resistance, but they are 
 only possible where there is also an electric light plant from which 
 
Theory of System. 121 
 
 the cells can be charged. The circuits connected up to a universal 
 
 battery must be of approximately equal resistance. If they are not 
 
 resistances must be inserted in the path of the battery current to make 
 
 them so. The incoming current does not, of course, pass through these 
 
 resistances. 
 
 It is not possible in practice to obtain circuits of precisely equal 
 
 resistance as special resistance blocks would have to be wound foreaclt 
 
 circuit. The outcome of practical experience is that circuits may be 
 
 grouped together if their resistances do not differ by more than 25°/ot 
 
 i.e., the resistance of the longest circuit must not exceed that of the 
 
 shortest circuit on the group by more than 25%, If a circuit has a 
 
 resistance less than this, resistance blocks must be inserted in the 
 
 battery leads to bring it up to the mean resistance of the circuits ; for 
 
 instance, suppose we have five circuits whose resistances are 1,400, 
 
 1,350, 1,260, 1,020, and 560 ohms respectively, which it is " desired to 
 
 group on to one battery. The highest resistance is 1,400 therefore the 
 
 lowest must not be less than 1,400 — 25°/,, of 1,400 = 1,050 ohms. The 
 
 first three circuits fall between 1,050 and 1,400 ohms, and therefore 
 
 need not be touched. Now, the mean resistance of the set should be ; 
 
 1,400 + 1,050 1 „.,- , 
 H — ■ — = 1,225 ohms, 
 
 therefore, we must bring the other two up to about this value. The 
 actual point to which we bring them depends upon the sizes of resist* 
 ance blocks usually made by the Department. For instance, if 94f 
 ohms were actually required, a 100-ohm block should be used. In the 
 present case we should use a 200-ohm block for the 1,050 ohm circuit, 
 and a 700-ohm block for the other, but this latter circuit should not be 
 grouped with these high resistance circuits if it can be avoided. The 
 objection to grouping circuits of very different lengths together arisea 
 from the fact that in wet weather the insulation of the long circuits 
 will be very much less than that of the shorter circuits, consequently 
 their resistances will be less and they will appropriate more than their 
 fair share of current. 
 
 The standard number of circuits to be worked from one battery is 
 five, but there is no reason why this number should not be increased 
 provided we have suitable batteries. In London 120 circuits are worked 
 from one set of accumulators. It should be noted that the internal 
 resistance of the battery employed must never under any circumstances 
 be more than the combined resistance of the group. The combined 
 resistance of the group should be taken as the resistance of the highest 
 circuit divided by the number of circuits on the group. In calculating 
 battery power for a group, take the total current required when all 
 circuits are working, as the current required to work one instrument 
 multiplied by the number of circuits, and see how many cells will be 
 
122 Universal Battery System— Current Required. 
 
 required to famish this total current through an external resistance 
 equal to the resistance of the highest circuit divided by the number. 
 Take the present case, let us suppose we wish to work these circuits 
 from one battery and that they are single current sounders each 
 requiring 15 m.a. 
 
 Total current required ^ 15 X 5 =; 75 m.a. 
 
 Combined resistance of circuits = -^ ^ 280 ohms. 
 
 The question has now resolved itself into how many cells, and of 
 what form, will be reqmred to send 75 m.a. through 280io. The 
 internal resistance of the cells most be taken into account as the 
 circuit is of such a low resistance. The current through a circuit due 
 to a battery may be ascertained by dividing the E.M.F. of that battery 
 ty the sum of the external resistance of the circuit and the internal 
 resistance of the battery. The number of cells (X) required to furnish 
 a given current may be found by dividing the resistance of the circuit 
 <R) by the difference between the E.M.F. of one cell (e) divided by the 
 current required (C), and the internal resistance of one cell (r), or 
 
 Resistance of circuit. 
 
 Number of cells =^ 
 
 E.M.F. of one cell -r . , -o £ n 
 
 k ^-7 -. — J — Internal Res. of one cell. 
 
 Current required 
 
 N= ^ 
 
 e 
 
 r. 
 
 c 
 
 All quantities must be expressed in amperes, ohms, and volts, or 
 milliamperes, millivolts, and ohms. 
 
 In the present case let us suppose it has been found more convenient 
 to use Bichromates, and the E.M.F. and resistance per cell must be 
 taken as the lowest figrure at which the battery is allowed to remain 
 without being considered faulty, viz., 1°82 volts and 4 ohms. 
 
 280 
 /. Number of cells ^ j^og = 1378, or 14 cells. 
 
 ■075 ~ * 
 
 Since it is not possible to supply three quarters of a cell it is usual 
 to add one cell if any fraction of a cell is required. Thirteen cells will 
 not quite supply 75 m.a., therefore 14 should be used. In the present 
 case large Darnells might be used, but Bichromates are preferable and 
 are now almost exclusively used. 
 
 In the case of single needles or double current sets, two such 
 batteries are required, one to send positive currents and one to send 
 negative currents to line. If two batteries were not used we shonld 
 have the cells short-circuited when one operator depressed the left key, 
 
Connections. 123 
 
 ;at the same time that another operator had the right key down. One 
 operator would earth the negative pole and join the positive to line, 
 ■while the other would earth the positive and put the negative to line ; 
 hence both poles of the battery would bs connected to earth, and 
 neither would get any current. 
 
 Single Needles and Double Plate Soonders.— The left hand 
 front contact screw and the back right hand spring must be removed. 
 Two batteries are employed, one of which sends all the negative 
 currents (having its positive pole earthed), and the other (having its 
 negative pole earthed), sends all the positive currents to line. The 
 usual arrangement is a battery with its centre earthed, but this is merely 
 a matter of convenience. In the case of Up offices when the left key is 
 depressed, positive current flows to to the left top contact spring 
 through the centre of the key to the needle, thence through the line 
 to earth, and back to the battery. Similarly when the right key is 
 depressed a negative current goes to line via the front contact screw. 
 In the case of Down offices the connections must be reversed as the 
 battery would be short-circuited when an up and a down station de- 
 pressed the same key together, as in one case a positive current would 
 be sent to line and to earth in the other. In order to avoid this the 
 battery is reversed at a down station, the negative pole of the battery 
 being connected to G and the positive to Z. In order that the marks 
 may not be reversed, A is earthed and the dial is joined to the line. 
 The connections of the dial are then crossed. At an intermediate 
 station no alteration is necessary. Two equalising resistances are 
 required, one for each half of the split battery, and these are included 
 in either battery lead, i.e., between Z of the key and the battery and 
 between C and the battery. The current from the outside offices does 
 not therefore pass through them. If it did it would necessitate an 
 increase of battery power at each station which would more than 
 'Counterbalance any economy arising from the use of the system at the 
 head office, It will thus be seen that the outside offices are not in any 
 way interfered with. The connections for Double Plate Sounders are 
 analogous. The connections of the commutator must be as Shown in 
 Figure 6, otherwise the accidental depression of both keys will short- 
 circuit the battery and stop all stations. 
 
 Single Current Sounders. — Down and int«rmediate stations 
 cannot be worked aS. the same set of batteries, The equalising 
 resistances are inserted between the front stops of the keys and the 
 batteries where necessary. For an up office the negative pole of the 
 battery is earthed, and the positive pole is connected to the key of every 
 set which it is desired to work from the battery. When no relay is 
 used, as in the direct sounder and direct writer, up and down inter- 
 mediate offices may be worked oS. one battery. 
 
124 Connections. 
 
 Double Cubrent Soitnder. — Here again a split battery is employed. 
 If the ordinary D.C. key is used the negative pole is joined to the left 
 and the positiye to the right-hand terminal of the key. The centre 
 terminal is connected to the relay, the other side of which is earthed. 
 The right-hand bottom terminal of the key goes through the galvano- 
 meter to line. The left terminal is left disconnected. When the switch 
 is tnmed the right terminal is connected to Z, and when the key is 
 depressed to C. When the switch is at receive the cnrrent passes 
 through the key and relay to earth. These are the connections for an 
 up of&ce, but if it is desired to make the office down, the battery and 
 relay must both be reversed. At an intermediate office the connections 
 are similar, but if the up line is longer than the down line an equalising 
 resistance should be inserted as the current forks off. A special form 
 of key consisting of a two-way switch and a small S.G. key on one base 
 is however generally employed with universal batteries. The line is 
 led on to the top lever of the two-switch way, one side of which 
 connects it (the line) through to the relay, and the other way connects 
 the line to the centre of the single current key. This key has four 
 terminals, the top two being Z and 0, the bottom left being U (con- 
 nected to relay) and the right D (connected to the line). 
 
 The greatest care must be exercised in using instruments worked 
 from a universal battery for special work, as reversing line and earth 
 of the instrument at the test box will cause short-circuiting of the 
 battery through the galvanometer when the key is to send and at rest, 
 and through the relay when depressed. Signals would, however, be 
 received perfectly, and this fact would tend to confuse matters were the 
 true cause not snspected. 
 
 The bichromate battery is almost invariably used for universal 
 working owing to its low internal resistance, and to the fact that it is 
 most efficient when heavily worked, which is not the case with Leclanche 
 cells. On the other hand, the Bichromate would be just as inefficient at 
 a small local office as the Leclanche would be for universal working. 
 Duplex and multiplex circuits are frequently worked from the same set 
 of cells with success and with great economy. Accumulators are now 
 being introduced at all the large offices, and are most economical. 
 
Electricity in its Application to Telegraphy. 125 
 
 CHAPTER XXI. 
 
 The Delany Multiplex System. 
 
 The object of this system is to increase the carrying capacity of the 
 lines, and at the same time to avoid the delays consequent upon 
 punching. Its advantage over the quadruplex is that if the pressure is 
 great in one direction all the arms can be worked in the direction of 
 the pressure. The essential principle of the system is to let each 
 operator have, in turn, exclusive use of the line for a short period of 
 time. If the time during which each operator has to wait before he 
 again has the use of the line is made exceedingly short, i.e., less than 
 the time taken to send a dot, his working is not interfered with. Now, 
 in order that each operator may in turn be joined up at regular 
 intervals, the line is brought to a revolving arm carrying a metal 
 brush, which makes contact with a series of segments arranged in a 
 circle and connected to the respective operators. 
 
 In order that operator 1 at A may be in communication with operator 
 1 at B, the trailer at A must be on segment 1 at the same time that the 
 trailer at B is on segment 1, hence we see that the trailers must move 
 synchronously in order that opterator 4 at A may be joined to operator 4 
 at B when their turn comes iEor use of the wire. A special set of 
 apparatus is used for the purpose of keeping the trailers in exact 
 synchronism. It will be noticed that sixteen segments are depicted 
 in Figure 31. The object of this arrangement is to give each operator 
 the use of the wire more frequently, but for a shorter period. Each 
 operator has exclusive use of the wire four times per revolution of the 
 trailer, but for only a quarter as long a time as if there were only four 
 segments. In practice the sub-division is carried out to a much 
 greater extent. There are 144 working segments, which can give each 
 operator the use of the wire twenty-four times in one revolution. 
 When two sets are used the term diode is used, when three triode, 
 -when four tetrode, when five penthode, and when six hexode. 
 
 I 
 
126 The Multiplex System. 
 
 The trailer makes three revolutions per second, thus joining^ 
 each operator to the line for gg| of a second every ^ of a 
 
 second. That is to say, operator 1 is joined to the line for g^r second 
 
 then operators 2, 3, 4, 5, and 6 in turn for the same period, which is, of 
 course, the time the trailer takes to pass over one segment. When the 
 
 trailer again touches operator I's segment -=, has elapsed since the 
 
 last contact. The current forming a " dot " lasts -^ second when 
 
 keying twenty-five words per minute. A dot is the shortest current 
 which is sent, therefore we must so arrange matters that a dot cannot 
 be sent so quickly that the trailer has not time to join up the apparatus 
 at least once. In the present case the margin is ample. If the 
 trailer is touching segment 1 at the precise instant when the key is 
 
 A 6 
 
 *' frC^' "-"^ A' ^•^i'' 
 
 „ <5—^ •^— <> \t- 
 
 Figure 31. 
 
 depressed, we shall have the use of the wire three times during the 
 sending of the dot. If the trailer be in any other position only two 
 
 currents, each lasting ^ second, will reach the distant end. A dash 
 
 a 
 
 lasts three times as long as a dot, viz., for of ''^ ^ second. During the 
 time the key is held down, connecting the sending battery to the 
 segment, the line is joined up eight or nine times ; thus eight or nine 
 currents flow to line and through the distant relay, which is joined to 
 the operator I's segment. The relay at the distant end is wound with 
 a very large number of turns of wire, having a resistance of 1,200 
 ohms. Further, the moving parts are somewhat heavy. The current, 
 which lasts -^ second passes through the relay. The magnitude 
 
 effects of a transient current upon the heavy tongue of the relay 
 depends upon : — 
 
 (a) the number of turns of wire on the relay magnets ; 
 
 (6) the duration of the current ; 
 
 (c) the strength of the current. 
 
The Distributor and Relays. 127 
 
 In order to get the best effects, a must be made very large ; h we 
 cannot interfere with, as this is practically fixed by the length of the 
 dot; c is made as great as possible, firstly, by employing a large 
 battery, and secondly, by nullifying the self-induction of the relay as 
 completely as possible. In order to do this a condenser of 10 micro- 
 farads capacity is placed across the terminals of the relay. This 
 almost entirely neutralises the bad effects. As has already been 
 explained, the effect of self-induction is to set up an E.M.P. which 
 opposes the current at starting. Now, this back E.M.P., acting in 
 the opposite direction to the battery, lowers the effective E.M.P., 
 hence the smaller current will flow. The result of the self-induction is, 
 then, very similar to that which would be produced if a resistance were 
 inserted in the circuit and the current thereby reduced. Self-induction 
 was originally described as an apparent increase in the resistance 
 of a circuit when an B.M.F. was applied suddenly. For instance, if 
 the E.M.F. of the battery be 50 volts and the back E.M.F. 25 volts, the 
 current will be that due to 50 — 25, i.e., 25 volts. As the current is 
 of so short a duration as gs this back B.M.F. is very serious, and 
 the condenser is employed to nullify it. The truth of this may be 
 verified by removing one of the condenser wires, when it will be found 
 that no signals are received however fine the adjustments are made. 
 This fault often occurs in practice, but is quite obvious, as no other 
 arm is affected. It may be well here to mention that the operators' 
 segments are separated by somewhat smaller segments, termed dead 
 segments. These are interposed in order to prevent the trailer joining 
 up two segments at the same time. Contact between two arms is 
 frequently caused by metal dust getting between the live and dead 
 segments, in which case the plate of dead segments should be removed 
 and both live and dead segments dusted with a brush. 
 
 Let us now consider what occurs when the letter " a " is sent from one 
 arm. The switch is first turned, and spacing currents are set to line every 
 ^ second. Now the key is depressed and three marking currents go 
 to line in the time that the key is depressed. Next, when the key rises 
 three spacing currents are sent, and when depressed for the dash, nine 
 marking currents leave. When the key next rises, spacing currents 
 succeed each other at the rate of 72 per second. The moving parts of 
 the apparatus are very heavy, so that the quick succession of heavy 
 impulses keeps the relay to mark or space, as the case may be. The 
 relay is kept nearly neutral but with a slight spacing bias. The heaviness 
 of the moving parts renders the relay sluggish in action, consequently, if 
 we make our impulses strong enough, and succeeding each other at 
 suf&ciently short intervals of time, we shall have our marks reproduced 
 -without sensible distortion. Each operator is provided with a similar 
 
 i2 
 
128 Theory of Multiplex. 
 
 set of appaiatns, and their indiyidnal use of the wire does not 
 interfere with that of any of the others, consequently, if the line is 
 short, and we have perfect synchronism, we shaU have moltiplez 
 transmission. 
 
 Unfortunately we have to deal with another distnrbing influence, 
 and that is that the time taken for a current to arrive at its maximum 
 strength becomes very appreciable. If we connect a battery to a line 
 possessing a large amount of capacity, the current takes a definite time 
 before its effect is observable. The usual, though not scientifically 
 correct method of stating this fact, is to say that it takes a certain 
 time for a current to reach the distant end and that the greater the 
 capacity the greater the time. Suppose now we connect a battery to 
 No. 1 segment. A current flows to line during the time the trailer 
 occupies in passing over it. This current will not be observable at 
 the distant end till a short interval later if the line possesses capacity, 
 consequently we must endeavour to express the amount of retardation 
 (in time) in segments. That is to say, if the retardation amounts to 
 "002 second, none of the current will appear while the trailer is on 
 segment 1. It will all appear, after the battery has been disconnected, 
 on segment 2, In this case we shall have to allot two segments to each 
 operator. The first will have to be joined to the sending part of his 
 key and the second to his relay. When his switch is at receive both 
 segments are connected to the relay as, should the weather be wet 
 the capacity of the line will be less and the current will make its 
 appearance earlier, viz., on the latter half of the sending segment. 
 Where the retardation amounts to '002 second hexode working is 
 possible. This is the time which a current takes to attain its maximum 
 strength on a London-Birmingham wire. Where the retardation is 
 within "004 second, or equal to two segments, four arms only can be 
 worked, as a current sent on the first segment is received on the third 
 segment. Beyond this triode working only is possible. Hence we see 
 that this difficulty of the retardation of the line has been overcome by 
 sending only from the first segment and then bunching all the 
 segments together, by means of the key switch, to receive upon. 
 Single current working is occasionally adopted as in the case of the 
 Manchester-Grimsby wire. 
 
 Next, we have to keep the two motors in synchronism. Firstly, we 
 must obtain a motor to twist the trailer, which moves as regularly as 
 possible. The phonic wheel, invented by Paul La Cour, answers the 
 purpose admirably. The wheel or armature of the motor consists of a 
 toothed wheel of soft iron placed in front of an electro-magnet. A 
 peculiar kind of flywheel is employed, consisting of a circular box 
 with a circular groove filled with mercury. When the speed of rotation 
 of the motor becomes slightly slower, the ring of mercury still rotates, 
 
Synchronising Arrangements. 
 
 129 
 
 thus tending to preserve uniform rotation. If currents are sent 
 through the electro-magnet at regular intervals the motion of the 
 motor is perfectly uniform. The method of obtaining currents 
 which succeed each other at regular intervals is somewhat singular. 
 A rigid bar of mild steel, set into vibration by means of an electro- 
 magnet, interrupts the current and thus furnishes the necessary regular 
 current impulses. 
 
 \ ©--r T P 
 
 <,— 
 
 ii 
 
 II 
 II 
 II 
 
 I- 
 
 
 E 
 
 MAIN BATTERY 
 
 ttHE 
 
 6 
 
 
 MAIN 8/ITTERY 
 
 FlGUKB 32. 
 
 The battery which drives the motor also drives the reed (Figure 
 32, A side). The current from the battery B passes through the motor- 
 magnet, through r, and through the reed contact back to the battery. 
 The reed magnet is connected across the terminals r, (7 ohms), but 
 its circuit includes the up-contact and lever of the uprighting sounder, 
 which is permanently held down by the battery Bi and the relay. The 
 current splits at r„ a small proporton (between l-5th and l-6th) goes 
 
130 Synchronising Arrangements. 
 
 through the reed magnet and lever of the uprighting sounder. Immed- 
 iately the current flows the reed is attracted, so breaking the circuit 
 at B. The reed then flies back and again makes contact. In this way 
 a series of currents succeed each other at regular intervals. It will be 
 seen that the reed is worked by the difference of potential between 
 the ends of ri. The reed magnet controls the rate of vibration of the 
 steel bar. If, now, the lever of the uprighting sounder be held up for 
 a second, the bar will be released from the control of the magnet, and 
 will vibrate backwards and forwards much quicker than before, but, 
 of course, with reduced amplitude. ' This makes and breaks the current 
 more quickly, thus increasing the speed of rotation of the trailer. 
 There are sis sets of segments, which are evenly distributed round 
 the distributor. Three sets consisting of three segments are used for 
 sending corrections, and three sets of five segments of smaller size, 
 which occupy the same space as three ordinary or sending segments, 
 are used for receiving corrections. We must so arrange matters 
 that, if the first motor lags slightly behind the second motor, the 
 lever of the uprighting sounder shall be raised in order to increase 
 the speed of rotation of the motor slightly so as to remedy 
 the defect. If we connect a battery to the first sending segment 
 a current will reach the distant end instantaneously if the line has 
 no capacity. The two trailers are not at the same part of the 
 distributor at the same time. While the trailer at A is on the 
 sending segments the trailer at B is on the receiving segments. 
 On a line of two segments retardation a current sent on the first 
 sending segment will be received on the fourth and fifth receiving 
 segments. If now the correcting relay, with a spacing bias, be 
 connected with the third receiving segment, no current wUl pass 
 through the relay if the two trailers are m synchronism. If now the 
 trailer at B lags behind the trailer at A by one segment, the current 
 which is sent from the first segment passes through the relay at B, thus 
 pulling the tongue of the relay away from the spacing contact as 
 shown at B. This disconnects the reed magnet for an instant, thus 
 increasing the speed of rotation of the trailer for an instant pulling it 
 up into exact synchronism. The current is represented as being sent 
 from the first sending segment, and as being received on the third 
 receiving segment instead of on the fourth. The current sent to 
 line is thus received on the connecting apparatus, and produces the 
 effects shown in Figure 31 (B side). It should, however, be remembered 
 that when the current is received on the third receiving segment at B 
 the trailer at A has meanwhile moved on to the third sending segment. 
 It will thus be seen that the correcting relay is joined one segment 
 behind the position upon which the sendii^ current should be received. 
 A sends three currents and B sends three currents per revolution. 
 
Adjiistments. 131 
 
 "We have thus seen that the trailer is regulated six times per revolution. 
 In the circuit of the first arm two galvanometers are inserted, one in 
 the sending circuit and one in the receiving circuit. In order to see 
 whether the hexode is in steady synchronism the distant ofiB.ce holds 
 down his key, and the needle is carefully watched. If the needle 
 vibrates, i.e., does not give a steady deflection, we may be sure that the 
 corrections are either too heavy or too light, in which case the correct- 
 ing relay and uprighting sounder must be adjusted. When the 
 instruments are out, the corrections, which are sent every time the 
 trailer passes over the sending segment, are received on the arms. If 
 they pass from A to P arm, your trailer is moving too fast. From F 
 to A too slowly. A rheostat is placed in circuit with the reed for the 
 purpose of adjustment. The weaker the current through the magnet, 
 the quicker is the rate of vibration of the reed. A sliding weight is 
 placed on the reed for the purpose of making large adjustments 
 in speed. Sliding towards the fixed end increases the speed. 
 
132 Electricity in its Application to Telegraphy. 
 
 CHAPTER XXTT. 
 
 Befeaters. 
 
 The working speed of any telegraph line is inversely proportional 
 to the product of the resistance and capacity of that line. The total 
 capacity of a line is equal to the capacity of a unit length multiplied by 
 the length of the line. Similarly the total resistance is equal to the 
 resistance per mile multiplied by the length of the line in miles. In 
 order to obtain the total product of the capacity and resistance we shall 
 have to multiply the capacity per mile and the resistance per mile by 
 the square of the length of the line. Hence we see that the working 
 speed of a line varies inversely as the square of the length. Suppose we 
 have an serial line 250 miles long and that the working speed is 120 
 words per minute. If now we divide the line at the centre the working 
 speed to this intermediate point will not be 2 X 120 but ^ X 120, i.e., 
 480 words per minute. The other half of the line will also work at the 
 rate of 480 words per minute. Thus when the signals appear at the 
 intermediate point they are transmitted on to the second line at the 
 same rate. The repeater, owing to mechanical inertia tends to reduce 
 this theoretical speed. For instance, 120 words a minute can be worked 
 on a London-Dublin wire, but with a repeater at Nevin the speed is 450 
 words per minute. We thus see that the insertion of a repeater into a 
 long line results in a great increase in the workng speed . 
 
 The wire terminates at this intermediate point, passing through 
 a relay whose tongue is joined to the other line. In Figure 33 A the up 
 station is sending to the down station. It will be seen that the up line 
 passes through the relay to earth. The down line is connected to T, 
 and a divided battery is joined between S and M. When the tongue of 
 the relay is over to the left, the left-hand half of the battery sends a 
 negative current on to the down line. If the up station depresses hi» 
 key the tongue will move over to M, thus putting a positive current on 
 to the down line, which causes the down line station's relay to mark. 
 
 Figure 33 B shows the down station sending . In this case it will be seen 
 that the battery has been crossed, A positive current sent by thedowa 
 station passes through the relay, causing T to go over to 8, thus joining 
 up the right-hand Jjattery, which sends a positive current on to the up 
 
Principle of Repeater. 
 
 133 
 
 line. This will cause the up station's relay to apace. In order that 
 either the down or up station may send automatic switches which 
 will, when the up station turns his switch, make the connections 
 indicated in Figure 33 A have to be provided. Further, when 
 the up station turns his switch to receive and the down station begins- 
 to signal, the connections indicated at B must be made. The relays 
 employed are of the ordinary type, so that the tongue is always touching 
 
 A 
 
 DOWN L.IME 
 
 
 FlGUKB 33. 
 
 either S or M when at rest. It will thus be seen that if there is no 
 current passing through the relay a current will yet be on the line. 
 The automatic switches are each worked by a neutral relay, which is- 
 joined in series with the line relay. Its S and M points are joined 
 together and connected to a battery. A. local circuit is formed from S 
 and M through the coils of the automatic switch to T with the battery 
 in circuit. Ordinarily T is held between S and M by the springs of 
 
134 Automatic Switch. 
 
 -the relay, but immediately a current passes through the relay T is 
 joined to S or M, in which case the local circuit is closed, and the tongue 
 of the line relay is joined to the other line. 
 
 In adjusting the auto-relay care should be taken to place the least 
 T)0ssible tension on the spiral springs for only by so adjusting can we 
 -expect them to work ef&ciently for any considerable time. If adjusted 
 with little tension there will be two forces tending to restore the 
 iongue to the midway position on the cessation of the main current. 
 One spring pushing into the centre and the other pulling. With a 
 considerable amount of tension on the springs and the tongue — say — 
 on the left hand contact point : the left hand spring will tend to keep 
 it in that position while the right hand spring will tend to pull it over 
 to the right hand contact point, and very uncertain working will be 
 the result. 
 
 The automatic switch consists of an electro-magnet with two 
 armatures, which play between two contact points, as shown in Figure 
 .34. The armatures are shown at rest, in which position they are held 
 by springs. When a current flows through the electro-magnet the 
 armatures are attracted and the levers move towards the magnet, thus 
 joining the three top terminals together and leaving the others dis- 
 connected. The centre terminal is joined to the tongue of the relay, the 
 Tight goes to line, and the left through the receiver to earth. 
 
 Thus we see that a portion of the current sent by the line relay 
 will, instead of going to line, leak through our receiver, and upon this 
 current we rely to ascertain the state of the marks. A resistance coil 
 is inserted in series with the receiver to regulate the amount of current 
 ■through the leak circuit. With a battery of very low internal resistance 
 it makes little difference to the distant station whether the resistance 
 in this circuit be 5,000 ohms or 20,000 ohms, the great utility of the 
 leak coils being to allow just sufficient current through the receiver to 
 -work it properly. In this way any little irregularity in the signals 
 will be shown up, whereas with a stronger current flowing, the defect 
 may be covered up. 
 
 When we reduce the leak resistance from 20,000 to 5,000 ohms we do 
 not, as a rule, appropriate current due to the distant station. We 
 ■simply increase the total current flowing from the battery, and of this 
 increased current the receiver with 5,000 ohms takes a greater share 
 than it did with 20,000 ohms in circuit. 
 
 A condenser of 1 mf. capacity joined in series with a 2,000 ohm 
 Tesistance coil is placed across the leak resistance in order to neutralise 
 the self-induction of the receiver or relay coils, and so reduce the 
 tendency to sparking at the transmitting relay contact points. 
 
 To reduce the sparking at the receiver and leak relay contact points 
 •each sounder is shunted with a resistance of 500 ohms, thus afEording 
 
136 Self-Inductio7i. 
 
 an alternative path for the extra current or currents due to self- 
 induction, instead of sparking at the contact points. 
 
 It will now be well to see how the neutral relay affects the- 
 closing of the auto-switch. Suppose both relays are at rest as- 
 shown in Figure 34, Kow let the down station turn his switch. A 
 current flows through the left galvanometer to the right top terminal 
 to the right bottom terminal of the right automatic switch through 
 the down line relay, and through the down neutral relay to earth. 
 This causes the neutral relay to move over to S. This closes the 
 automatic switch, thus joining T of the down relay to line. Further, 
 the down line relay has been held over to S, thus immediately the- 
 automatic switch is closed by the neutral relay a current flows on to- 
 the up line. Now when the down station depresses his key both 
 relays move over to M. 
 
 The automatic switch is made very sluggish in action so that the 
 armatures will not be released during the time that the neutral 
 relay is moving over to M. Thus we see that a succession of 
 marks will cause the tongues of both relays to vibrate backwards- 
 and forwards, but that the auto connections will be maintained 
 so long as marks are being sent. When the down station ceases 
 signalling, the tongue of the neutral relay will return to its 
 central position, and, perhaps, a fifth of a second later the automatic 
 switch will open. It is made sluggish by placing a shunt across the 
 terminals of the coil and giving it a high co-efficient of self-induction. 
 This will mean that when the current through the coil ceases another 
 current due to self-induction will spring into existence, and as has 
 already been explained, it will be in the same direction as the battery 
 current, thus tending to perpetuate its effects. The shunt provides a 
 good conducting path for these currents which consequently take 
 effect upon the automatic switch, preventing it from releasing its 
 armatures when only a very short break in the current occurs, such as 
 that caused by the tongue of the relay moving from S to M, or vice 
 versa. A shunt of 500i<i is placed across the terminals of sounders and 
 double-plate sounders in order that the E.M.F. due to the breaking of 
 circuit may not cause sparking at the relay points. The E.M.F. of self- 
 induction at the breaking of the circuit joins with the E.M.F. of the 
 battery, and thus a high E.M.F. is the result. The shunt provides a 
 conducting path, and thus no trouble in the shape of sparking results. 
 The auto-shunt is of low resistance, and thus the self-induction current 
 tends to unduly prolong the effect of the battery current, which is 
 what we desire. The sounder shunt is of too high resistance to cause 
 trouble in prolonging the marks. 
 
 Precisely the same effects as described in the case of the down 
 station take place when the up station turns his switch. The 
 
Repeater Faults and their Causes. 137 
 
 current flows through the down automatic switch and up relays 
 'Closing up the auto. It is necessary to provide keys for speaking 
 -to the terminal stations, and these are of the universal typB, 
 They come in the line circuits at the points indicated by a cross in 
 Figure 34. When at receive the lines are joined through as shown, 
 biit when the switch is turned everything beyond the galvanometer is 
 ■cut out. The main split batteries are also employed for this purpose. 
 .Shunted condensers, which are essential, are inserted between the two 
 jrelays in order that their self-induction may be neutralised as 
 explained on Page 91, thus preventing the speed of working from 
 being greatly reduced. 
 
 Preparatory to commencing the daily speed trials — either simplex or 
 •duplex — the contact points should be well burnished, closed up to 
 ishort circuiting point, then opened the least shade and the tongue set 
 in a neutral position, i.e., in such a position that it will remain at which 
 -ever side it may be brought by the finger. Officers in charge of speed 
 trials at terminal offices should remember that the standard speeds are 
 :flxed by the Engineer-in-Ohief's Department under the best possible 
 conditions as regards weather, with batteries in perfect condition, with 
 apparatus specially adjusted, and with experienced technical officers at 
 each end, and any question as to actual speed should therefore be 
 'decided in favour of the wire rather than against it. 
 
 Although it is impossible to enumerate the whole of the faults 
 likely to occur on a repeater board, perhaps the following, with 
 probable causes, may be of benefit to those placed on repeater duties : — 
 
 Auto-switch jumping whilst stations working. 
 
 (1) Too much resistance in the rheostat. 
 
 (2) Contact points of auto-relay too wide apart. 
 
 (3) Contact points of auto-relay dirty. 
 
 (4) Too much tension on spiral spring of auto-relay. 
 
 (5) Auto-battery weak. 
 
 (6) Contact or intermittent earth on line. 
 
 (7) Faulty in sender's key. 
 Auto-switch jumping whilst circuit at rest. 
 
 (1) Too little tension on spiral springs of auto-relay. 
 
 (2) Contact points too close, 
 
 (3) No, or insufficient resistance in rheostat. 
 This fault might also be due to lightning discharges. 
 Auto-switch sticking whilst circuit at rest. 
 
 (1) Faulty adjustments of spiral springs of auto-relay. 
 
 (2) Faulty adjustment of auto-switch armatures. 
 
 When a station reports a current being left on at the repeater a 
 glance at the galvanometers will usually suffice to show whether this 
 
138 
 
 Bepeater Faults and their Causes. 
 
 is so or not. If both galvanometers show a deflection, the current is 
 probably coming into the repeater from one side or the other. To 
 make certain of this put over both switches to " send," and if the 
 armatures of the auto-switches are at once released the fault is not at 
 the repeater. If, however, the levers fail to return to their normal 
 position, the fault is with the repeater and must be at once attended to. 
 
 When the current received is very weak and no means of increasing 
 it, it will probably be found necessary to increase the sensitiveness of 
 the auto-relay by decreasing the tension of the spiral springs. 
 
 Signals running together or failing. 
 
 (1) Line fault. 
 
 (2) Sender's key faulty. 
 
 (3) Misfitting slip. 
 
 (4) Contact points of transmitting relay dirty. 
 
 Figure 35. 
 
 Sending batteries short circuiting. 
 
 (1) Contact points closed up touching the tongue. 
 
 (2) Line to earth very near. 
 
 (3) Lightning protector short circuited. 
 
 (4) Key screwed down. 
 
 This fault would cause the indicator beU to ring. 
 
 When an ofSce advises that they are unable to gain the attention of 
 the repeater clerk, and it is found that no call is being received the 
 key switch should be put to " send," and if the terminal station's 
 current is received on the galvanometer the fault is likely to be in :— 
 
 (1) The right-hand lever of the auto-switch. If not there— 
 
 (2) Examine switch of key, 
 
 (3) Arms of rheostat. 
 
 (4) And all other connections of the incoming current. 
 
Duplex Repeater. 139 
 
 Weak signals in the leak circuit might be due to— 
 
 (1) Too much resistance in leak coils. 
 
 (2) Faulty receiver adjustment, or — 
 
 (3) Earth on line, close up to repeater. This latter fault might 
 
 cause the indicator bell to ring. 
 
 An earth fault might be caused by lightning discharges short circuit- 
 ing the protectors. 
 
 When a fault becomes apparent it is a good rule to be sure of its- 
 nature before altering any of its adjustments. 
 
 The current required is from 20 to 25 m.a. and Bichromate cells are 
 exclusively used as line batteries. The auto-locals consist of 10 cells- 
 Large Daniell. 
 
 The duplex repeater is very similar, but is much more simple, as 
 automatic switches are unnecessary, just in the same way that the- 
 switch on a double current key is unnecessary for duplex working, A 
 quadruplex A key would do just as well. In a duplex repeater we- 
 have to provide for simplex working, and a switch is used to form the- 
 connections of Figure 34. Besistance must be introduced in the- 
 rheostat before turning to duplex, otherwise a fused transmitting 
 relay may result. The duplex connections are shown in skeleton in 
 Figure 35. R represents the compensation circuit, the neutral relays 
 and the automatic switches are all cut out of circuit, and it is thought 
 that the above figure will make matters sufficiently clear in view 
 of the exhaustive explanation which the double current duplex 
 received. 
 
140 Electricity in its Application to Telegraphy. 
 
 CHAPTER XXTTI. 
 
 The Bridge Duplex. 
 
 The principle of this system of working will be readily understood 
 -from the description of the differential duplex. In the differential 
 system we make the sending current split through the two coils of our 
 own receiving apparatus, and if the current splits equally the relay is 
 not affected, but when one path is closed by an opposing E.M.F., the 
 •other coil takes all the current. Again, when the current from our 
 own end joins with that of the distant end, we have tl\e current flowing 
 through one coil twice as strong as that flowing through the other coil, 
 <!onsequently the stronger current produces an effect. "We may then 
 say that the differential system depends upon balance of current. On 
 the other hand, the Bridge system consists in splitting the sending 
 ■current to two points, between which are connected the receiving 
 apparatus. If the potential at these points is the same, there will be 
 no flow of current through the receiving apparatus, but if we stop one 
 ■of the paths beyond this point the other only will be available, and the 
 rreoeiving apparatus will practically be a shunt upon it. 
 
 C E and C F (Figure 36) are two equal resistances, between which 
 we cause our sending current to divide at A, HG and HK 
 •serving a like purpose at B. First, we must ascertain that the 
 arrangement will work simplex. Suppose A depresses his key. 
 The current splits at C to E and F. If now, the resistance in 
 -the rheostat B is equal to that of the line and distant receiving 
 :apparatuB, the potentials at B and F will be equal — vide Chapter 
 XIX. If the potentials ai E and F are equal there will be no flow 
 of current through the relay. The one half of the current flows 
 -through the rheostat back to the battery, whilst the other half 
 proceeds along the line to G where there are two paths open for the 
 current— one through the relay to K, the other through the proportional 
 arm GH, through the key to earth. The current flowing through the 
 Telay now splits at K, part passing through KH and part through R to 
 earth. The portion of the current which passes through the relay in 
 the relay effects the mark as it passes in the direction U to D. As only 
 a, portion of the received current passes through the relay, it is obvious 
 
Theory of Bridge Duplex. 
 
 141 
 
 that we shall require more battery power for bridge working than for 
 differential working. Precisely the same will occur to A if B depresses 
 his key. B's relay will be unaffected, whilst A.'8 will record a mark. 
 
 In the present case, C E and G F have been made equal, but this is not 
 necessary so long as the same proportion is maintained between the 
 other two arms of the bridge, viz., between the line circuit and B. The 
 same remark, of course, applies to B. We have seen that if the product 
 of the resistances of the opposite sides of a Wheatstone Bridge are 
 equal, there will be no flow of current through the galvanometer. This 
 arrangement is merely a slight modification of the Wheatstone Bridge. 
 
 Now, let us suppose that A and B have their keys depressed simul- 
 taneously. A's current flows to C, and there splits, one part going via 
 
 LI NE 
 
 FiGTIKE 36. 
 
 C F and R back to the battery. The other part, on reaching E, is met by 
 B's current flowingthrough H &. The two currents oppose, and the result 
 is that no current flows through the line. At A the current flows from 
 E through the relay in the direction U to D (causing a mark), through 
 the rheostat back to the battery. Thus we see that E and the relay 
 become a shunt across F. Similarly at B, the relay is worked by the 
 part of his own current which flows through H G and the relay. When 
 the E.M .F.'s of the two batteries oppose each other on the line, the efEect 
 is precisely similar to a disconnection of the line. We then see that 
 when one key only is depressed, a mark is registered in the ordinary 
 way, and that when both are down together the part of the current, 
 
 K 
 
142 
 
 Theory of Bridge Duplex. 
 
 •which wonld otherwise traverse the line, make its way through the 
 relay, thus registering a mark at either end. 
 
 Now we might, if we desired, make the currents flowing to E and G- 
 combine with each other by reversing one of the batteries. This 
 would make the current through the line twice as strong as that 
 through the compensation circuits. Again, by making use of both 
 these efEects we may work double current. The system of joining up 
 may be made uniform, and a distinct up and down station obtained. 
 This is what is done in practice. A skeleton diagram of the duplex 
 <K>nnection8 are given in Figures 37 and 38. Both keys are shown at rest 
 in Figure 37. The current flows from the positive pole of the battery 
 at A to C where it splits, part going on to E and part to F, The part 
 
 FiGUKE 37. 
 
 going to E goes to earth at A ; thence to B's earth-plate through the 
 left coil of his galvonometer to his battery, with which it combines 
 and flows on to H . Here, again, the current splits, going part to K and 
 part to O. This latter part flows on to the line through the right-hand 
 coil of A's galvanometer back to the battery. Thus we see that 
 through the dotted part of the circuit there is a current flowing which 
 is the result of the combined batteries. This current flows along the 
 line. Now, to return to station A. It was pointed out that the current 
 splits at C, and that a part passed along C P, indicated above by dots 
 and dashes. Now, at F the current again splits, part goins through the 
 relay in the direction D to U, thus joining the main current. The 
 
Double Current Bridge Duplex. 
 
 143 
 
 result of the current through the relay is a space. A part flows through 
 the rheostat and left coil of the galvanometer back to the battery > 
 Similarly, at station B the current splits at K, part joining the main 
 current via the relay, and part passing through the rheostat back to 
 the battery. It should be noted that the current flowing through the 
 rheostat is due to one station's battery and not to both acting in 
 concert, A space is registered at both stations, and it will be observed 
 that this is due to the relays being worked by the currents shunted 
 from C E and G H respectively. The path of the main current is 
 dotted. 
 
 Figure 38. 
 
 Now let A depress his key. The batteries oppose each other and 
 no current flows via the line or earth. The path of the current is now 
 through the left galvanometer coil, through R, along FC and FEC in 
 parallel, back to the battery, that part of the current which takes the 
 path through the relay (in the direction D to IT) causing a space. 
 Now, let us see what has taken place at B. The current flows to H 
 where it splits, part coming along HK through the right-hand coil of 
 the galvanometer back to the battery. That part of the current which 
 goes via HGK passes through the relay in the direction U to D, thus 
 
 e2 
 
144 Advantages and Disadvantages of System. 
 
 causing a mark. This is merely a case of a divided circuit, the arm 
 HGr, with the relay, forming a shunt upon HK. Thus we see that the 
 depression of a key causes a mark at the distant end. If, now, both 
 keys are deapressed, the batteries will combine together, and we shall 
 have precisely the same relations between the currents in the various parts 
 of the circuit that obtained in Figure 37, but, of course, the current 
 will be reversed in direction everywhere, and we shall thus have both 
 relays registering marks. It will be noticed that the galvanometers 
 are inserted with one coil in the line circuit and one in the compensation 
 circuit. This is done in order that an accurate balance may be obtained. 
 When the balance is perfect, depression of the key at A will not affect 
 A's galvanometer. It will be seen that E and G are the up line 
 terminals. The galvanometer at B has been reversed for the sake of 
 simplicity in drawing and explanation. R should go to the left-hand coil 
 of the galvanometer and the right-hand side to E. The shunt current 
 through the relay has everywhere been indicated by dots and dashes. 
 
 The arms E and C P are usually each SOOOw, and a condenser of 
 suitable capacity is joined across each. The object of these condensers 
 is to regulate the quantity of electricity which shall flow into the 
 cables. Before any effect can be produced the capacity of the cable 
 has to be satisfied. The condenser which is joined to the arm con- 
 nected to line acts as a short circuit upon that arm until the condenser 
 has become charged. This reduction in the resistance of the circuit 
 provides a heavy rush of current during the time that the condenser 
 is being charged. When the condenser has become charged the normal 
 current flows through the 3000w coils. The time taken for a current to 
 rise to its maximum value in any given circuit does not depend upon the 
 value of the E.M.F. applied. The way in which the quickening is 
 effected is as follows : — The reduction in the resistance of the circuit 
 provides a higher E.M.F. with which to charge the cable. This higher 
 E.M.P. would charge the cable with a correspondingly greater quantity 
 of electricity if allowed to do so. The value of the capacity of the 
 signalling condenser is so adjusted that the heavy current lasts only 
 until the cable contains the same quantity of electricity which would 
 be stored by the normal current. When the rush of current is over 
 and normal current begins to flow, it at once produces an effect at the 
 distant end, as the cable is fully charged. The time taken by the 
 heavy current to charge the cable with an equal quantity of electricity 
 to that which would be produced by the lower E.M.F. fully charging 
 the cable is obviously less than the time which the normal current 
 would take to produce the same charge. The difference between these 
 times represents the saving in time affected by the arrangement. Any 
 saving in time means increased speed, hence the importance of the 
 device. The second signalling condenser must be exactly similar to 
 
Advantages and Disadvantages of System. 145 
 
 the first, in order that the balance may be preserved. They tend 
 to quicken the rise and fall of current strength. A shunted condenser 
 is frequently inserted in series with the relay or receiver as the case 
 may be, and its object is to neutralise the self-induction of the 
 receiving apparatus. 
 
 The advantage of the Bridge system over the differential is that 
 very little retardation is introduced by the sending apparatus. The 
 current is sent direct to line, and does not have to pass through the 
 coils of the relay, which always possess considerable self-induction. 
 Again, all the apparatus which is necessary for Bridge working, in 
 addition to the apparatus usually used on a circuit, is a supply of 
 rheostats. The apparatus need not be difEerentially wound. Frequently, 
 cables (with a repeater in circuit) are worked on the land side on the 
 differential system and on the cable side on the Bridge method. The 
 objection to the Bridge method for general use is that it requires 
 much more battery power than the differential system, and for 
 long serial lines possess no advantage over it. On the other hand, 
 where there is cable the Bridge system may mean a large increase in 
 the working speed of the cable. The calculation of the current through 
 the various branches of the circuit is somewhat complicated, requiring 
 thorough familiarity with Kirchoff's laws and somewhat advanced 
 algebra. 
 
 The current required is from 20 to 25 m.a„ and a suf&cient number of 
 Bichromate cells should be provided to supply this current. 
 
146 Electricity in its Application to Telegraphy. 
 
 CHAPTER XXIV. 
 
 Batteey Testing Instruments, 
 
 It is necessary that we should have a quick and accurate method of 
 testing batteries in order that the efficiency of the service may be 
 maintained. There are now two forms in use, both of which were 
 designed by Mr. Eden, the first for the larger offices and the second 
 for smaller offices. The latter is less expensive than the former, and 
 whilst not being quite so rapid is yet sufficient for all purposes where 
 a very large number of batteries have not to be tested. The tangent 
 galvanometer is used in conjunction with both instruments and must be 
 of the latest form with two coils. 
 
 The first mentioned instrument for the larger offices was designed 
 on the supposition that the E.M.F. of a Daniell cell was 1'073 BA. 
 volts. These units will be employed in this case and the alterations 
 which will be made when the instruments are returned to the factory 
 for repairs will be explained at the end of the chapter. The E.M J*, of 
 the Daniell cell is now found to be 1*08 standard volts, consequently 
 the values of the resistances employed will have to be raised by a 
 trifle over 2°/o. The tangent galvanometer will also be wound to 320 
 standard ohms. The alterations will thus merely entail an increase 
 of all the resistances employed. One Daniell cell through 1073 
 ohms gives a current of one milliampere, and this current produces a 
 deflection of 80 divisions on an accurately adjusted instrument. If 
 now we add another cell to our circuit we shall have doubled the 
 E,M.F. and we shall consequently get practically double the deflection, 
 but if when we insert this second cell we also add a second 1073 ohms 
 we shall have precisely the same deflection which we obtained with 
 one cell, viz., 80 divisions. That is to say if we have 10 cells we must 
 make the total resistance lOu x 1073U in order that we may obtain the 
 same deflection which one cell would produce through 1073u, Let us 
 now suppose that instead of the deflection being 80 divisions it was 
 only 75. That would indicate that the E.M.F, of the battery was less 
 than 1073 volts. The current due to a battery through a very high 
 resistance is practically directly proportional to the E.M.F. of the 
 battery. The effect of the high resistance is to make the internal 
 resistance of the cell a negligible quantity in comparison with it, as has 
 
E.M.F. Test. 147 
 
 been previously pointed out. The deflection in divisionson the galvau- 
 ometer is proportional to the current flowing through it, therefore, 
 
 80 : 75 : : 10-73 : the E.M.F. of the cells, i.e„ E.M.F. of cells = ^ 
 
 of 10' 73 = 10-059 volts. The fall of E.M.P. from the standard 1073 is 
 '671 volts. The Post Office system of stating this fact is to express 
 the fall as the percentage fall from the standard. The E3I.F. 10059 
 volts is 931 per cent, approximately of the standard, thus the fall is 
 6i°/o. If the fall on the ten cells is 6 J per cent., then clearly the 
 average fall on each cell must be 6J per cent. In short the method 
 amounts to inserting with each battery under test sufficient resistance 
 to obtain the standard deflection were the cells in perfect order. Any 
 fall from this condition is indicated by the reduced deflection, and is 
 read directly on the inner scale of the galvanometer. 
 
 This system, described above, necessitates the calculation of the per- 
 centage, and, moreover, would mean that when testing 40 Bichromate 
 cells 80 X 1073w would have to be inserted, owing to the fact that the 
 E.M.P. of a Bichromate cell is twice that of a Daniell. Now, 80 divi- 
 sions is 2-5ths of 200 divisions, so that if, instead of inserting 1073o> 
 per cell, we insert 2-5ths of 1073(o — i.e, 429'2m — ^we shall have a deflec- 
 tion of 200 divisions. As, however, only one coil of the galvanometer 
 is in circuit, 100 divisions only is obtained. 
 
 The internal resistance of any cell is averaged at Su, and this amount 
 is accordingly deducted from 429'2(o ; therefore, 426'2 is inserted with 
 each cell under test. The connections of the latest form of battery 
 testing instrument are indicated in Figure 40. The left-hand set of 
 coils B are arranged in multiples of the unit 426'2m. An arm is pivoted 
 at the centre of B, and is movable round the points indicated by small 
 circles. The first point is marked 0, but 373'5i<j is inserted with the 
 galvanometer for a purpose which will be appreciated later. The 
 second point is marked 5, and the resistance then added is Xiffl'bm, 
 We have thus 373-5w -f 1597-5W -1- 160w = 2131u-i.e., 5 x 426'2i.. in 
 circuit. This arm is placed at the number of cells under test, and 
 brings in circuit that many times 426'2(ii ; therefore the deflection is 
 brought down to the same deflection that one cell would produce 
 through 426-2<«. 
 
 If, now, we wish to test Bichromate cells whose E.M.F. is twice that 
 of a Daniell cell, we shall have double the deflection, and shall 
 not be able to read on the inner scale any longer. In order that 
 we may obtain the same deflection with a Bichromate cell that 
 we obtain with a Daniell we must shunt our galvanometer. Only 
 half the current is allowed to pass through the galvanometer, and for 
 this purpose IGOu is joined across the coil. In this case half the 
 current passes through the galvanometer and half through the IGOu 
 
148 
 
 Galvanometer Shunts. 
 
 shunt. Their joint resistance is 80u, and this is increased by 80u in 
 order that the resistance in circuit with the rheostat may not be 
 reduced by the application of the shunt. A skeleton diagram of the 
 connections is given in Figure 39. The key !Ki is at rest, and the path 
 of the current is from the positive pole of the battery to 2-circle and 
 1 of the galvanometer, through the 1-3 coil, through R back to the 
 negative pole. The 1-3 coil is shunted by &. This reduces the resist- 
 ance to 80iii, and B is also made SOu, so that the resistance of the 
 instrument remains unaltered. The connections of the shunt are 3 to 
 L to N through Ss to 2-circle and 1. The compensating resistance of 
 SOoi is inserted at B between K and L. In this case K and N are not 
 connected. 
 
 8 
 
 -AAAAAAAAAA/A- 
 R 
 
 A/ 
 
 
 Figure 39. 
 
 The Leclanche cell has an E.M.F. of 1*6 volts, i.e., one-and-a-half 
 times that of the Daniell cell. In this case the SOw is added to the 
 resistance of the galvanometer coil, and 480(u is joined across both ; 
 therefore, the current splits into three parts, two of which pass through 
 the galvanometer (160(o -I- SOu = 240(<i), and one of which passes 
 through Sz — i.e., through the 480(o coil. One-third of the current is 
 shunted away, and the same deflection is therefore obtained as with 
 a Daniell or Bichromate in equally good condition. The current 
 flows as before, but the position of B is different, K and Nare joined, 
 and N and L are unconnected ; S^ is now 480i<>. 
 
 It will thus be seen that Bichromate, Leclanche, and Daniell cells 
 can indiscriminately be tested by this instrument. The shunts are 
 brought into use by means of the switch at the left of the Figure 40- 
 When Daniell cells are in use the galvanometer is unshunted, but in 
 other instances the two sets of coils 80, 160, and 320 are applied to 
 either coU of the galvanometer. The arm of the rheostat R is placed 
 at the number of cells under test, and inserts that number of times 
 426*2w. The switch is placed at the form of battery under teat, and 
 thus the galvanometer assumes the required degree of sensitiveness. 
 
Internal Resistance Test. 
 
 149 
 
 The principle of the teat for internal resistance is equally simple, 
 and a simple method of demonstration will be adopted. When a 
 battery is connected to a circuit the difference of potential between the 
 «nds of that circuit is less than the E.M.F. of the battery. The battery 
 depicted in Figure 41 is connected between A and B. Let us suppose 
 that the battery has an E.M.F. of 1 volt, and an internal resistance of lOu, 
 If , now, we make B^ 10«i and insert the plug S, completing the circuit, the 
 difference of potential between A and B will only be half a volt, as 
 compared with 1 volt before S was inserted. The total resistance of the 
 
 FIG0BB 40. 
 
 circuit is 10 + 10 = 20<i>. The resistance of the cells is lOu, therefore, half 
 the E .M.F. of the cell will be lost in overcoming its own internal resist- 
 ance. Let us take another example : — E.M.F. of battery, 3 volts ; internal 
 resistance SOw ; Bs, lOu. In this case three parts of the E.M.F. will be 
 lost in overcoming its own internal resistance, and one part in over- 
 coming B2, i.e., the difference of potential between A and B, will be 
 
 5-^-j of 3 volts = I volt. Or to prove this in yet another way , — By 
 
 Ohm's law current in amperes X resistance in ohms = E.M.F. in volts. 
 
150 
 
 Shunted Battery. 
 
 Current through circuit = 
 
 = '075 amperes. 
 
 30 + 10 
 
 /. fall of potential in battery = 30 X -075 = 2-25 volts. 
 „ „ R2 = 10 X -075 = J75 „ 
 
 E.M.F. of battery = 3-00 „ 
 
 To take yet another example, suppose It: had been only one ohm. The 
 difference of potential between the ends of Bz (A and B) would be 
 
 30 I -^ of 3 volts. Similarly had Ba been 2bu, the difference of 
 potential would have been ■„ , „, of 3 volts. 
 
 30 -1-23 
 
 From the above example it will readily be seen that when a battery 
 is shunted the difference of potential between the terminals of that 
 battery falls to a lower value, which value is lower than the E,M.F. of 
 the cell in the same proportion that the battery's internal resistance 
 
 FiGUEE 41. 
 
 bears to the joint resistance of the internal resistance and shunt. Let 
 us take the above example. B.M.F. of battery 3 volts ; shunt or 
 resistance across the terminals of the battery 10<u j internal resistance 
 30u. Let El represent the potential difference between the ends of the 
 battery, 
 
 3 volts : B, volts : : 30w • ^^ ^° 
 3 volts : E, : : 30(.. : 7-5 ( 
 
 30 + 10 
 
 ohms 
 
 .-.E, =^^5-^ = -75 volts 
 
Internal Resistance Test. 151 
 
 This was the result which we obtained before. This relation is knowa 
 as Pollard's battery theorem.* 
 
 The principle of the internal resistance test consists in maintaining 
 the £.M.F. deflection through the rheostat It unaltered when B is re- 
 duced by the application of a shunt of fixed resistance. This reduces 
 the resistance of the circuit, so that in order that the deflection may 
 be unaltered the battery must be shunted. The shunt applied to B is- 
 made equal to 60 x 426*2w = 25,572w — the resistance of one coil of 
 the galvanometer ^ 25,412(o. Sixty cells is the highest number of cells- 
 which can be tested at one time by means of this instrument ; hence 
 the fixed shunt S is made equal to 60 units, in order that the arrange- 
 ment may be as sensitive as possible. When testing 60 cells B is- 
 25,412 ; hence the application of the shunt S reduces the resistance of 
 the circuit to one-half its former value, thus doubling the current 
 passing through it. In order to reproduce the original deflection the 
 E.M.F. of the battery must be reduced to one-half its former value. 
 In order to do this the battery must be shunted by a resistance equal 
 to the internal resistance of the battery, i.e., by a resistance equal to- 
 60 times the resistance of one cell. If the resistance per cell is lio- 
 then the shunt must be 60i» ; if 2«j, the shunt must be 120u. If, now, 
 we arranged a series of 60(o coils, and marked them from lu to lOu, they 
 would represent the resistance per cell. When 10 cells are being 
 tested B would be (as previously explained) 10 X 426'2w, i.e., S (which 
 is 60 X 426'2io) would be six times B. In this case the deflection would 
 be unaltered when the battery was shunted in the same proportion 
 i.e., by a shunt having a resistance equal to six times the resistance or 
 the battery. If the resistance of the battery is 2u per cell, i.e., Wo>, 
 total — the shunt applied would have to be six times 20w, i.e., i20<<). 
 Every 60(i) in the rheostat B is marked 1 ohm; therefore, 120a) is- 
 marked 2<<i, It will be seen that when we increase the number of cells 
 we increase B ; thus the fixed shunt S produces a greater relative reduc- 
 tion in the resistance of the circuit, but at the same time the resistance 
 of the cells is correspondingly increased, and a correspondingly greater 
 efEect is produced by the shunt B. 
 
 For instance, when 20 cells are joined up B is equal to three times- 
 B, If the internal resistance per cell is 3ai, a balance is obtained 
 when the shunt resistance B is three times the internal resistance of 
 the cells, viz., 180id, If, now, 60 cells are joined up S is equal to B, 
 and similarly B is equal to the internal resistance of the battery. 
 
 If the resistance by which the battery is shunted is not in the same 
 
 proportion to the internal resistance of the battery as is S to B, then 
 
 the current through the galvanometer when the battery is sending a 
 
 current through B only, and when unshunted, will not be the same ; 
 
 * For tuU proof see Eempe's Handbook of Electrical Testing, p. 50-^. 
 
152 
 
 Practical Arrangements. 
 
 conBequently the deflection will be altered in value when the key 
 joining up the shunts is depressed. First K is placed at the number 
 of cells under test. B is then altered until the current through the 
 galvanometer is the same whether the battery is shunted or not. If 
 the deflection on the galvanometer is decreased when the key is de- 
 pressed, the resistance shunting the battery is reducing its B.M.F. too 
 much — that is to say, it is too low- If the deflection is increased it is 
 too high. 
 
 Reference to Figure 40 shows that in addition to ten BOu coils there 
 are also ten 6ai coils. If, now, the insertion of COu is required for 
 shunting a battery, the cells composing which have Im internal resist- 
 
 FlGTTBE 42. 
 
 ance, 6ik will suffice if the resistance of the cell be l-lOthu. The 6u 
 coils are marked tenths of an ohm, and by their aid the resistance of 
 the cells may be measured to l-lOthw. 
 
 It has already been stated that the galvanometer must have two 
 coils. The objectjof this is that we may not have the resistance of the 
 galvanometer added to the joint resistance of B and S. For instance, 
 let us suppose 60 cells were under test, the joint resistance of B and 8 
 would be i B, but the resistance external to the battery would be i B 
 + the resistance of the galvanometer. A correction could be made for 
 this error by reducing the value of S, but this would not be made 
 absolutely accurate, as the value of S would be different for every 
 number of cells under test, and thus a mean value could only be 
 given to S. in the old form of this instrument this was done, as the 
 tangent galvanometer was not at that time double wound. The use 
 of double wound galvanometers has removed this cause of error, as 
 one coil is included with S and the other with B. The coils of the 
 galvanometer thus form simply parts of B and S. The coils are 
 joined up in parallel, so that the current in the 1 — 3 coil adds to the 
 
Connections of Instrument. 153 
 
 eSect produced in the 2-circle — 4-circle coD. If 60 cells are under test 
 the current from the battery arrives at 2-circle, where it splits, half 
 passing through each coil ; thus the effect is precisely the same as if 
 the whole of the current had passed through one coU only. 
 
 The shunts which are necessary when testing Bichromate and 
 Leclanche cells' do not in any way disturb matters, as the resistance 
 of the coils of the instrument is kept constant. It is obvious that if 
 for any reason it becomes necessary to shunt one coil of the galvan- 
 ometer, the other coil must be similarly shunted. If 60 Bichromate 
 cells were under test the current sent out by the battery would split 
 equally between the two coils. Half the current passing to each coil 
 would be shunted away by S, and Sa (Figure 39) ; consequently the 
 two coils acting together would register half the current sent out by 
 the battery. Had the 2-circle— 4:-circle coil been unshunted, that coil 
 would have produced twice the effect of the other coil ; thus a defleC' 
 tion would have been produced which was due to f of the current sent 
 out by the battery. The current sent out by Daniell cells in perfect 
 condition is 2i m,a., but that sent oat by Bichromate cells is double 
 this, owing to its higher E.M.F. 
 
 The connections of the instrument are given in skeleton in Figure 39. 
 It will be seen that on depression of K 1 the battery is shunted by B, 
 and at the same time E is shunted by 25,412cii. The key K 1 corre- 
 sponds to e and g of Figure 40. The right-hand set of coUs are joined 
 up by this key, and the resistance inserted is controlled by two arms, 
 as in the case of an ordinary rheostat. The arm of this rheostat should 
 never be placed at any value lower than the resistance of the cells 
 under test, as the coils are very likely to get fused by the heavy current 
 which would flow through them. The first two 60o» coils and the first 
 four 6<o coils are wound separately with a thicker wire than the re- 
 maining coils for this purpose. 
 
 This instrument is very rapid in operation. The placing of E at 
 the number of cells gives the E3I.F. deflection, which is read straight 
 off. The resistance is then obtained by the rheostat B. 
 
 It has now been found that the E.M.F.'s of the various cells are as 
 follows : — 
 
 Daniell 1'08 standard volts. 
 
 Leclanche 1'60 „ „ 
 
 Bichromate 2'16 „ „ 
 
 These values entail the use of a higher unit. A Daniell cell will 
 send a current of 1 m,a. through 1080 standard ohms ; therefore, in 
 order to obtain a deflection of 100 divisions with one coil of the gal- 
 vanometer, 2-5ths of 1080 standard ohms must be in circuit with every 
 cell. The internal resistance of cells is averaged at 3 standard ohms ; 
 
154 Internal Resistance of Standard Cell. 
 
 therefore, 432 — 3= 429 standard ohms must be inserted with each cell. 
 The galvanometer will also be brought up to 160 standard ohms per 
 -coil. This will mean that 429 standard ohms will be the unit in place 
 of 426-2 B.A. ohms and S will be (60 x 429)— 160 standard ohms. B and 
 -the shunts will also be wound to standard ohms. Thus the values of 
 all the resistances employed will be raised, but without in any way 
 Affecting the theory of the instrument. 
 
 The constant of the galvanometer must be accurately adjusted 
 l)ef ore the tests are taken. For this purpose it is necessary to know 
 -the internal resistance of the standard cell. The arm of the rheostat 
 'R is placed at zero. B is placed at 1-lOth. This joins the battery 
 up through one coil of the galvanometer. The key is then depressed, 
 and the battery is thus shunted by 6(«. If the deflection is thereby 
 halved, it is obvious that the resistance of the battery is equal to that 
 of its shunt ; if reduced to one-quarter, the shunt is equal to one- 
 third the resistance of the battery. This may be stated thus : — 
 
 FiGUEE 43. 
 
 Res. of battery = 6w X <"°°^'"'tg'^ deflection— shunted deflection) 
 ^ shunted deflection. 
 
 From what has been said in reference to the difference of potential 
 between the terminals of a shunted battery, the theory of this test 
 should be at once apprehended. 
 
 Having obtained the resistance of the standard cell, it is now only 
 necessary to consult the table given on Page 38. If the galvanometer 
 is marked " standard ohms " allowance must be made for this. The 
 battery testing instruments, unless marked "standard ohms," give 
 results slightly in the favour of the cells. 
 
 For the smaller offices an instrument has recently been designed by 
 Mr. Eden. It consists of two sets of resistance coils, B and R, Figure 42, 
 and a fixed resistance, S. The unit adopted is 429 standard ohms ; S 
 
Details of Winding. 155 
 
 is therefore 25,580u. B consists of six coils, marked 1, 2, 4, 8, 16, and 32, 
 The first coil is 369u, and cannot be plugged out of circuit ; the 
 remaining coils are 1, 2, 4, 8, 16, and 32 times 429(>>, and are so marked. 
 The number of cells to be tested minus 1 is unplugged in B. B consists 
 of six coils, marked '2(o, '4u, '8(o, l'6ii>, 3'2<<j, and 6'4(x, which are wound to 
 12(ii, 24U, 48ci>, 96u, 192(o, and 384w respectively. The principle of the 
 instrument is identical with that of the larger instrument. 
 
 Xo shunts are employed ; consequently the Bichromate deflection is 
 200 divisions, and that of the Leclanche 148 divisions- The percentage 
 falls ore read directly in the case of the Daniells, divided by two in 
 the case of the Bichromate, and by means of tables in the case of the 
 Leclanche cells. 
 
 The external appearance of the larger instrument as manufactured 
 by Messrs. Elliott Bros, is depicted in Figure 43, 
 
156 Electricity in its Application to Telegraphy. 
 
 CHAPTEB XXV. 
 
 The Concentrator. 
 
 In the larger offices there are frequently a number of circuits upon 
 which the traffic is under 100 messages per day. This amount of work 
 will not keep the instrument engaged continuously. If now we 
 have, say, twenty-five circuits of this character using the same class of 
 apparatus, we may use less than twenty-five instruments if provision 
 can be made so that an instrument is connected to a circuit only when 
 that circuit is in use for the transmission of messages. In brief, the 
 system is in reality an application of the Exchange system of working 
 telephones to telegraphy. It is interesting to note that the Telephone 
 Exchange was evolved from the Telegraph Exchange of Mr. 
 Heaviside, of Newcastle-on-Tyne, and that now we find a kind of 
 Telegraph Exchange built from the principles of the Telephone 
 Exchange. 
 
 Every circuit on the concentrator terminates in an indicator, i.e., an 
 indicator is connected between its line and earth. When the distant 
 station works his keys, the needle on this particular circuit at the 
 Head Office is actuated, thus calling the attention of the switch clerk 
 to the circuit. Each circuit is provided with special contacts, termed 
 switch springs, by means of which, on inserting a plug connected to an 
 instrument, the indicator is thrown out of circuit and the circuit is 
 joined to that instrument. 
 
 If we have 25 circuits each using needle or double plate sounder 
 apparatus, and upon each of which the daily traffic averages under 100 
 messages per day, we should terminate these circuits upon 25 indicators. 
 Twelve instruments would be found sufficient to take the traffic at 
 the busiest part of the day, viz., between ten and twelve in the morning. 
 These twelve instruments would terminate in plugs, and thus any 
 instrument could be connected to any line. The disadvantage of the 
 syttem is that if all -twelve instruments are in use and a thirteenth 
 circuit calls up, or traffic arrives for it, attention cannot be given until 
 one of the instruments is disengaged. Its advantage is that it saves 
 
Connections of Concentrator. 
 
 ^ — 0- 
 
 157 
 
 EAfrrif 
 
 PiGUHE 44, 
 
158 Principle of Concentrator. 
 
 space and staflE by concentrating the traffic and thus saves the time 
 occupied in moving from one instrument to another. 
 
 The instruments used as indicators are termed Non-polarised 
 Indicator Belays. The instrument consists of an electro-magnet with 
 a soft iron armature held away from the pole-pieces by means of a fine 
 spiral spring. The left-hand side of this armature makes contact with 
 an ivory tipped screw. The right-hand side is tipped with platinumi 
 and when currents, in either direction, are passed through the coils of 
 the instrument makes contact with a platinum tipped screw. This is 
 arranged to complete a local circuit and ring a trembler bell. A green 
 label covers up these contacts and outside this swings a soft iron needle 
 which is rendered magnetic by induction from a round bar magnet 
 upon which it is pivoted. It will thus be seen that the local contact 
 is closed whatever may be the direction of the current, but that the 
 direction of the deflection of the magnetic needle is changed when the 
 current is reversed. 
 
 The switch springs consist of two brass bars, with pins projecting at 
 right angles, so arranged that these pins rest upon a second two contact 
 pins fixed at right angles to the base board of the instrument. The 
 upper spring is slotted down the centre in order that it may not be 
 possible to insert a plug upside down. 
 
 The plugs consist of two strips of brass insulated from each other by 
 means of ebonite. One of these springs carries a brass projection 
 which fits the slot cut in the upper switch spring. The peg is 
 furnished with a black bone handle, and the connection is made by means 
 of two flexible conductors. The conductors are covered with cotton 
 and the whole is then wrapped round with steel wire, and this is again 
 covered with braided cotton. The connections are thus well protected 
 and will remain intact unless subjected to very rough usage, The other 
 extremities of the conductors are fixed to a cord connection strip. 
 The other terminals of these brass pieces are then connected to the 
 instruments. The pegs rest in holes made in a thick strip of leather. 
 The cords pass through these holes and are kept taut by means of weights 
 and pulleys. When the pegs are removed from the switch springs they 
 fall down into the holes provided for their reception, thus presenting a 
 neat and orderly appearance. 
 
 We are now in a position to consider the apparatus as a whole. Line 1 
 passes to the upper spring of A,, through the left-hand contact pin, 
 through the indicator A, then through top contact of the bottom right- 
 hand switch spring to earth. Line 2 passes from the switch-spring B,, 
 through the indicator B and through the switch spring marked " Time " 
 to earth. In fact the connections of the lines are all precisely similar, 
 viz., through the top contacts of the corresponding switch spring, through 
 the indicator to the common return wire through the upper contacts of 
 
Balancing Resistances. 159 
 
 the time springs to earth. The bottom contact of the switch springs 
 are joined to the left-hand contact pins for a purpose which will be 
 appreciated later. A peg is shown at the bottom of Figure 44. The 
 top side of the peg is the side which fits the top side of the switch 
 spring, and this is joined to the line wire from an instrument. It is 
 obvions that if the instruments are to be interchangeable, i.e., so that 
 any instmment can be used for any line, the instruments most all be 
 joined up alike, and that the distant stations must also be joined up 
 alike. The practice is to make all the instruments at the concentrator 
 end " up," and to make all the distant ends " down." The down Une 
 terminals are connected to the top sides of the pegs. The insertion 
 of a peg into a switch spring lifts the springs from the contact pins. 
 At the same time the top half of the peg is thus connected direct to 
 line. This cuts out the indicator. The bottom side of the peg is used 
 for "ringing off." When an office calls the plug corresponding to any 
 disengaged instrument is inserted in the corresponding switch spring. 
 When the telegraphist has taken the message it is necessary that the 
 switch derk should be advised in order that the -instrament may be 
 connected to another circoit. This may be done verbally by shouting, 
 but in order to avoid this a push is added to each instrament by means 
 of which a battery is connected to the underside of the peg. When a 
 peg is inserted into a switch spring the underside of the peg becomes 
 connected to the bottom switch spring which joins it to the left-hand 
 side to the respective indicator. The push has an earthed battery 
 consisting of four No. 1 Leclanche cells connected to the other contact, 
 thus its depression actuates the indicator corresponding with the 
 engaged switch spring. This serves as an instruction to the switch 
 clerk to disconnect the instrument. 
 
 The local contacts are joined up with a bell so that, at night, when 
 continuous attendance is not justified attention may be called to the 
 circuits. This bell may be cut out by means of a plug svritch when 
 oontinuons attention is given. , 
 
 Another difficulty which has to be dealt with was that when circuits 
 of unequal resistance are connected to a switch the cnrrent would be 
 very unequal if balancing resistances were not employed. Usually the 
 instruments are worked from two universal batteries, half the 
 instruments from each. These balancing resistances are inserted at the 
 terminal stations in the receiving circuit. In the case of a single needle 
 the resistance coils are inserted between the X and T terminals. The 
 resistance is thus not inserted in the sending circuit from the distant 
 station. This will not necessitate an increase in the battery power 
 employed at the terminal stations. In the case of intermediate stations 
 the ba4Hiery power must be increased as the current has to pass tiirongh 
 the added resistance when that station is working. 
 
 L 2 
 
160 Time Signal. 
 
 The only diflScnIty which has now to be dealt with is in the matter 
 of sending " time." It will have been noticed that the right-hand side 
 of all the indicators are joined together and pass to earth through the 
 switch spring labelled "time." By inserting a peg in this switch spring 
 we join an instmment to the whole of the lines bunched together, thus 
 the current splits between all the stations and each gets the signals sent 
 from this one instrument. These signals pass through the indicators 
 also. 
 
 The indicators are protected by a glass front when they are of the 
 kind described. These indicators are nsed for sounders, double plate 
 sounders, and needles. Where A.6.C .'s are used indicators of a different 
 form are employed. The currents sent out by an A.B.C. instrument 
 alternate too quickly to produce distinct movements of the needle of 
 an indicator relay, and an indicator without a needle is used. It 
 consists of an electro-magnet with an armature which holds up a 
 shutter. When a cnrrent passes through the indicator the armature is 
 pulled down, thus releasing the shutter, which falls down and calls the 
 attention of the switch clerk. 
 
 It is not usual to group upon the same switch circuits which are 
 operated by different forms of apparatus, but when it becomes 
 necessary for economical reasons to do so not more than two forms 
 should be joined to the same switch. 
 
 The system has been very widely adopted and has proved successful 
 wherever the traffic considerations have been carefully entered into. 
 
Electricity in its Application to Telegraphy. 161 
 
 CHAPTER XXVI. 
 
 CONSTEUCTION, 
 
 A line of telegraph may be carried on poles or may be placed under- 
 ground. Wherever possible, it is desirable to use open rather than 
 covered work owing to the amount of capacity which underground 
 work brings in. Open lines are placed either on the railways, on public 
 roads, or on private property. 
 
 A line of telegraphs upon the railway is usually more sheltered from 
 wind than other open wires. Its disadvantage is that the maintenance 
 of the wires upon the railway company's property is in the hands of 
 their own officials. It is not probable that wires upon the railway will 
 be disturbed. They suffer little from stone throwing, and cannot get 
 in the way of external building operations, 
 
 A line of telegraphs upon a public road is not so sheltered from the 
 elements as a railway line, but it presents the advantage that its main- 
 tenance is in the hands of the officials responsible for the satisfactory 
 working of the apparatus. The line is not likely to be disturbed by 
 building if the road is a long one. Even should the position of the 
 line have to be altered, such alteration will not involve the entire 
 removal of the line to another route. In the country near small town- 
 ships much damage is caused by stone-throwing, but this cannot be 
 entirely avoided. 
 
 Frequently it is quite impossible to obtain a route either on the 
 public highways or on the railways, and in this case it is obvious that 
 the permission of private individuals must be sought to allow the 
 Department to place poles upon their land. The great objection to 
 wires on private land is that in the event of building or other causes 
 they are in a position to demand the removal of the line from their 
 property. In this case a most serious difficulty maybe caused. Some- 
 times, where there are a number of landowners who are hostile towards 
 the Department near together the line may have to be diverted many 
 miles round. 
 
 In large towns it is not possible to take a liue of poles along the 
 main streets, and in this case the consent of the private individuals is 
 sought to place iron standards upon the roofs of their shops. Here 
 again we are confronted with difficulties in the matter of way leaves, 
 
162 Selection of Routes. 
 
 i.e., permission to place the fixtures. There is also a great liability that 
 the shops may wish to add another storey, or that from other causes 
 they will demand the removal of the line. A line of telegraphs erected 
 upon standards on house tops is termed an over-house route. 
 
 Underground routes are selected in the vicinity of large towns 
 where the number of wires is too large for over-house routes, and a 
 line of poles along the public road does not meet with the approval of 
 the Corporation. On account of the heavy capacity of underground 
 wires this class of route is avoided wherever possible. The great 
 advantage which it possesses over serial lines is its freedom from inter- 
 ruptions due to storms, etc. It is, however, liable to interruption 
 when the gas or water service requires attention. 
 
 The line wire is supported upon porcelain insulators in order that as 
 little current may escape to earth as is possible. The insulator which is 
 universally employed by the Post Office is that due to Mr. J. H. Cordeaux. 
 It is mounted upon a steel bolt, the upper extremity of which terminates 
 in a flange and a coarse screw (Figure 45). The lower extremity 
 carries a flange and a nut and a washer for the purpose of securing the 
 insulator to its support. The insulator itself consist? of a double 
 porcelain cup of glazed earthenware made in one piece. At the top 
 of the cup on the inside a coarse female screw, fitting the screw on the 
 bolt is provided. The insulator is screwed on to the bolt ; but between 
 it and the flange on the bolt an india-rubber ring is fixed, and this 
 allows the insulator to give slightly when a heavy strain is put upon it. 
 The object of the double shed is to obtain great insulation. A current 
 will escape over the surface of the shed to the bolt in wet weather 
 when the surface of the insulator is wet. The current has to pass 
 down the outer shed, up again inside, and down the outside of the 
 inner shed before it can make its way up the inside of the inner shed 
 to the bolt. The outer shed acts as a cover over the inner, thus keeping 
 it dry, and comparatively clean,, but itself becomes black by being 
 coated with the solid matter (chiefly carbon) which the atmosphere 
 contains. The corrosion of the wire also adds its quota to this 
 deposit ; the time which it takes to form depending entirely upon 
 the number of factories and their character in its vicinity. When 
 insulators get into this state they are removed, washed, and re-erected. 
 This coating of dirt forms a very good conducting surface in wet 
 weather, and hence on a wire where there are one or two thousand 
 insulators the decrement in the insulation of the line is most marked. 
 For telephone lines the outer shed is frequently corrugated with the 
 object of lengthening the outer surface of the shed. For points 
 where a very heavy strain is put upon bolt and insulator, a special 
 insulator is frequently used. These insulators are larger and heavier 
 than the ordinary form. The bolt is also heavier and penetrates much 
 
Insulators and Arms. 163 
 
 further into the solid portion of the porcelain. Shackle insulators 
 were formally used, but are now condemned by all Telegraph 
 Engineers on account of their low insulation. 
 
 The insulators are fixed to wooden arms by means of the nut and 
 washer at the base of the spindle. It is thus possible to remove a 
 broken cup without removing the bolt. This is a great advantage, as 
 the removal of a rusty bolt from an arm is well nigh impossible 
 without cracking the arm at the bolt hole. The arms are made 
 from well seasoned hard-grown English oak, felled at least twelve 
 months before being made into arms. The arms are usually from 
 2i to 3 inches square, and having a length to suit circumstances. 
 The arms are placed twelve inches apart, and all upon the same side of 
 the pole, viz., upon the London or up side. For telegraph work it was 
 usual to employ alternately arms 24 and 33 inches long, each of which 
 carries two wires, one on either side of the pole. The employment of 
 alternate short and long arm minimises the risk of contacts. As the 
 number of wires is becoming so large, it is now becoming impossible to 
 accommodate them by only placing two wires upon an arm. For tele- 
 phone trunk lines four wires are placed upon an arm. The insulators 
 are 12 inches apart in this case. 
 
 The arms are secured to the poles by means of arm bolts. The pole 
 is slotted to receive the arm, and when the bolt is fixed prevents the 
 arm from rocking, thus holding the arm at right angles to the pole. 
 Wooden poles are almost exclusively used on British road lines. The 
 round poles are usually cut from red fir trees which are obtained from 
 Norway and Sweden, care being taken that the selected timber is sound 
 and hard grown (that is with the annular rings closely pitched), straight 
 and free from large or dead knots. They are felled between November 
 Ist and February 28th following. 
 
 Rotting of the poles is caused by a sort of fermentation of the sap. 
 Seasoning slowly drys out the sap under conditions unfavourable to 
 fermentation. Anticeptics, i.t., substances which oppose putrefaction 
 are injected in order that the chemical actions set up result in the 
 formation of insoluble compounds not acted upon by air or water. 
 Sometimes, for very heavy lines, square pitch pine poles are employed, 
 and when the timber is fully dried it is usual to coat it with paint, and 
 not to use antiseptics. The antiseptic employed by the Post Office ia 
 creosote, as experience indicated that this was by far the most reliable 
 and satisfactory method of preservation. 
 
 Creosoting consists in injecting creosote, ten or twelve pounds being 
 allowed per cubic foot of timber. 
 
 Bumettising consists in injecting chloride of zinc. Poles can be 
 readily painted after this process, whereas it is not possible with the 
 creosoting process. 
 
164 
 
 Timber. 
 
 Boucherising consistB in injecting sulphate of copper. This process 
 is an exceedingly efficient one when the poles take the copper, but a 
 certain percentage do not do so, and conseqnently rot in some five or 
 «x years. 
 
 Three classes of poles are in use by the department, Tiz., light, medium 
 and stout. The light poles are used where the number of wires is fiye 
 
 FiGUBB 45. 
 
 or under. Medium poles carry from six to ten wires, and stout poles 
 over ten wires. The dimensions of a thirty foot pole of the three 
 classes are as follows : — 
 
 Light. 
 
 Medium. 
 
 Stout. 
 
 Diameter 
 at top. 
 
 Diameters ft 
 from bottom. 
 
 Diameter 
 at top. 
 
 Diameter 5 ft, 
 If rom bottom. 
 
 Diameter 
 at top. 
 
 Diameter 5 ft. 
 from bottom. 
 
 5i inches. 
 
 7i inches. 
 
 6i inches. 
 
 9 inches. 
 
 8i inches. 
 
 11 inches. 
 
 The poles are buried to a depth approximately equal to one-fifth of 
 their length, but are not often buried deeper than six feet. The holes 
 are dug in the direction of the wires and not across wires as the solid 
 earth is mnch firmer than the earth which has been removed, no matter 
 
Struts and Stays. 165 
 
 how well the punning may have been done. The pole is obviously 
 more likely to rock from side to side than in the direction of the wires, 
 
 Poles are usually erected by using ladders as levers, and thus by 
 raising the bottom end of the ladder forward. One arm is fixed upon 
 the pole and this arm prevents the ends of the ladder from slipping 
 oS the pole. Frequently where a large number of wires have to be 
 carried two poles are placed side by side, and are bolted together at the 
 top in order to afford additional strength. Again two poles are some- 
 times separately erected and connected by long arms, which arrangement 
 will provide for the heaviest road lines. The former arrangement is 
 known as an A pole and the latter as an H pole. 
 
 Where lines are light from twenty-two to twenty-four poles per 
 mile are fixed. Where, however, the line is heavy thirty or more poles 
 ore inserted per mile. The poles are placed sufficiently far from the line 
 of traffic and in such positions as to obviate danger due to vehicles and 
 pedestrians coming in contact with them. A strong wind will blow 
 the wires horizontally to a distance nearly equal to the sag, and suffi- 
 cient space must, therefore, be allowed to prevent the wires coming in 
 contact with buildings or trees. The wires should not cross the roads 
 unless absolutely necessary, and then the crossing should be at right 
 angles to the road, and should never make a long slanting cross. The 
 wire should be terminated at either side, so that in the event of the wire 
 breaking at another point it, the wire, may not run back and leave the 
 wire over the road slack. Guards, consisting of a curved rod of 
 galvanised iron, are fixed upon the insulator bolt in such a manner as 
 to prevent the wire falling ofE the arm in the event of the connection 
 between the wire and the insulator giving way. The guards are only 
 fixed ali very dangerous poiats as they tend to reduce the insulation of 
 the line. The wires should ordinarily never be less than twelve feet 
 from the ground, and when crossing roads at least twenty feet. 
 
 Where a line makes a bend considerable lateral strain is put upon 
 the poles in that bend, the tendency being to pull the poles over so as 
 to eliminate the curve. In order to prevent this, stays and struts are 
 used. The stay consists of a stranded wire, one end of which is fixed 
 to the upper part of the pole and the other end terminating in a rod 
 which passes through a heavy piece of wood buried some depth in the 
 ground. The stay wire and stay rod are connected by a swivel in order 
 that the stay may be tightened up after it has been fixed. The angle 
 between the stay and the pole should always be made as large as 
 possible, A strut is support in which the strain is compressive, instead 
 of tensile as in the case of the stay. 
 
 The best position for a support is clearly at the centre of the wires 
 as this is the centre of the strains. Stays are, if possible, fixed at this 
 point, known as the resultant point, but if not, on account of their 
 
166 
 
 Wires. 
 
 coming in contact with the wires, the stay is spliced and one wire is 
 taken to the top and the other to the bottom of the wires or other 
 convenient positions. 
 
 The strnt cannot always be fized in the centre of the wires, and in 
 that case it is fized below them. A stmt consists of a second pole fixed 
 to the first by means of bolts and standing at an angle to it. The 
 bottom of the stmt is fized to a piece of timber at right angles to it, 
 nsually a portion of an old pole. It will thns be seen that a stmt acts 
 also as a stay. Its unsightly appearance and lack of mechanical strength 
 prevents its adoption where a stay is possible. 
 
 On a heavy road line it is nsnal to stay the poles every half mile in 
 the direction of the wires, the object being to prevent the collapse of 
 one span from polling a mile or so of poles over on either side. Snch 
 
 SINGLE ST^tY 
 |=i@ STftUTTEO POi-E. 
 
 FlGUKE 46. 
 
 stays are indicated in Figure 46 at A by the short thick lines on either 
 side of the round circle indicating the pole. A, B, and C are in a 
 straight line, hence B does not ezperience any lateral pressure, unless 
 heavy winds across the wires prevail, in which case a stay or stmt must 
 be fixed. At C the line turns off to D and a stmt is placed at this 
 point. The pole at D is on the footpath, but the owner of the shaded 
 land wiU not permit the erection of a stmt upon his property. In this 
 case a pole is erected on the other side of the road and a stay wire is 
 connected from this pole to the pole requiring support. The stay pole 
 
Tests of Copper Wire. 167 
 
 is attached to a stay in the ordinary way. At E the line bends round 
 again and a stay is thus necessary here. 
 
 Where a telegraph line passes over hilly ground care is taken to keep 
 the height of the line as even as possible by the insertion of timber of 
 suitable length at each point. If this is not done we shall have a heavy 
 upward pull upon the top of any pole which is greatly below the poles, 
 on either side. 
 
 Iron wire was formerly used for all telegraph lines, but is now being 
 superseded by copper which is desirable on account of its superior 
 electrical properties. Iron, being magnetic, introduces a great deal of 
 retardation due to the increased self-induction*, and thus is not suitable 
 for very long Wheatstone wires, as retardation always means a decrease 
 in the working speed of the lines. Copper conducts electricity seven 
 times as well as iron, and hence an iron wire can be replaced by a copper 
 one of one-seventh the sectional area. 
 
 We have also to consider the tensile strength of the wires, as the 
 mechanical strength of the line is of as great importance as its electrical 
 condition. In regard to this quality there is but little to choose 
 between iron and copper. The thick 400 lb. iron wires can be replaced 
 by thin .100 lb, copper wires, which are much more sightly, without- 
 impairing the mechanical strength of the line. Further, copper has a 
 very much longer life than iron wire owing to its being less liable to- 
 oxidisation. A dark green skin forms on the surface of a copper wire 
 and prevents any further corrosion taking place. Iron wire when used 
 is galvanised in order to prevent oxidisation. The coating of zinc 
 afiords protection, inasmuch as the oxide of zinc formed is insoluble. 
 In localities where the air is charged with acid or other chemically 
 active fumes, the zinc covering speedily goes and the wire quickly 
 follows it. 
 
 Iron wire is made up in coils of about a hundred weight and the 
 quality of the whole quantity is judged from samples cut from any coils- 
 chosen at random. First the wire is tested for its mechanical condition.- 
 
 FoRM. — The wire must be cylindrical and have a diameter within 
 the limits prescribed in the table given on next page. It must be free 
 from scales, flaws, and other defects, and must be of uniform diameter. 
 
 Ductility. — The wire is gripped by two vices which are revolved,, 
 the wire thus being subjected to torsional strain. The required number 
 of twists which the wire must bear without breaking or splitting is given, 
 in the table on next page. This also tests the quality of the galvanising. 
 
 Tensile Strength. — The wire must lift a weight equal to nine- 
 tenths of the minimum tensile strength for the size of wire under test- 
 and the remaining tenth to be added gradually. 
 
 Electrical Properties, — The resistance per mile of the wire must 
 fall within the prescribed limits. 
 
168 
 
 Tests of Iron Wire. 
 
 Copper wire is tested in a similar manner. 
 
 FOBH. — Same as in the case of iron. 
 
 DtrcTiLiTT. — Same as in the case of iron with the additional test 
 'Of wrapping the wire in sis tarns ronnd wire of the same diameter 
 nnwiapping it, and again wrapping in sis tarns ronnd the wire in the 
 same Section. Shoold the wire break it is not sufficiently ductile. 
 
 Tensile Stebngth. — Same as in the case of iron. 
 Galvanised Ikon Wire. 
 
 Weight 
 per Mile 
 
 a 
 
 XD 
 
 Diameter 
 
 Strength and DnctiUt7 
 
 ■i 
 
 II 
 o 
 
 sa 
 
 
 Oa> 
 
 15 
 
 ^1 
 
 U O 
 
 pqo 
 
 h 
 
 =•9 
 
 as 
 
 go 
 
 "CO 
 
 ap 
 
 as 
 
 a:: 
 
 p. 
 
 lbs. 
 
 800 
 €00 
 450 
 400 
 
 200 
 
 lbs. 
 767 
 833 
 
 571 
 629 
 
 424 
 477 
 
 377 
 
 424 
 
 377 
 434 
 
 190 
 213 
 
 mils 
 
 242 
 209 
 181 
 171 
 171 
 121 
 
 mils 
 247 
 237 
 
 214 
 
 204 
 
 186 
 176 
 
 176 
 166 
 
 176 
 166 
 
 126 
 118 
 
 lbs. 
 
 2,480 
 1,860 
 1,390 
 1,240 
 
 1,026 
 620 
 
 lbs. 
 2,550 
 
 1,910 
 
 1,425 
 
 1,270 
 
 638 
 
 lbs. 
 
 
 obms 
 
 2,620 
 
 13 
 
 675 
 
 1,960 
 
 15 
 
 900 
 
 1,460 
 
 17 
 
 12-00 
 
 1,300 
 
 19 
 
 13-50 
 
 1,075 
 
 20 
 
 IS-OO 
 
 655 
 
 26 
 
 27-00 
 
 lbs. 
 
 90 
 
 120 
 
 90 
 120 
 
 90 
 120 
 
 90 
 120 
 
 90 
 ISO 
 
 40 
 65 
 
 NOTB. — ^The lower line of 400 lbs. gives the particulars lor charcoal wire. 
 
 Conductivity. — The resistance per mile of the wire most fall 
 within the prescribed limits. The particulars of the various classes of 
 wire used by the Post Office are given in the table above. 
 
 Galvanised Ibon Wire. — The wire having been approved is now 
 ready for use. The wiring is affected by passing the wires over the 
 ^arms of the poles upon which it is to be fixed. The wiring is begun 
 
Wiring. 
 
 169 
 
 from the first insulator, where it is bound right round the insulator, 
 i.e., terminated. This prevents the wire from slipping forward. The 
 wire is carried over the next four or five arms. At the last arm a 
 dynamometer is fixed to the arm by means of the hook attached to the 
 spring of the dynamometer or tension ratchet. The end of the dynamo- 
 meter terminates in a roller with a small hole cut through the centre of 
 it connected with a ratchet wheel and pawl, the other side of the wheel 
 being furnished with a crank. Through the hole in the tension ratchet 
 a long piece of wire is passed and is secured by a turn or two of the 
 crank. The other end of the wire is attached to a pair of draw-tongs 
 which are fixed upon the wire. By turning the crank the line wire is- 
 p ulled up, the tension being indicated by the scale affixed to the spring. 
 
 Copper Wire. 
 
 Weight per 
 Statute Mile. 
 
 Approximate equi- 
 valent Diameter. 
 
 
 
 Maximum 
 
 Besistance 
 
 per mile 
 
 ol wire 
 
 when hard 
 
 at60°F. 
 
 Minimum 
 Weight of 
 each piece 
 
 Standard 
 
 Bange 
 allowed. 
 
 Standard. 
 
 Range 
 allowed. 
 
 (or coil) 
 of Wire. 
 
 lbs. 
 100 
 
 150 
 200 
 400 
 
 lbs, 
 
 97J 
 102i 
 
 146i 
 153i 
 
 195 
 
 205 
 
 390 
 410 
 
 mils. 
 
 79 
 
 97 
 
 112 
 
 158 
 
 mils. 
 
 78 
 80 
 
 954 
 98 
 
 llOJ 
 113i 
 
 1554 
 160i 
 
 lbs. 
 330 
 
 490 
 
 650 
 
 1,300 
 
 30 
 25 
 20 
 10 
 
 ohms. 
 910 
 
 6-05 
 
 4'53 
 
 2-27 
 
 lbs. 
 50 
 
 50 
 50 
 50 
 
 The amount of tension to be put upon the wires now becomes a 
 matter for consideration. The tighter the wire is drawn the smaller 
 is the amount of sag or dip, but on the other hand the tighter the wire 
 is drawn the smaller is the margin of safety. In practice, wires are 
 pulled up to one fourth of their minimum breaking weight, thus 
 allowing a factor of safety of four. It will also be noticed that the 
 breaking weight is rather more than three times the weight per mile. 
 Tables are supplied showing the dip and strain to be put upon wires 
 at various temperatures, as it will be obvious that a wire, which in 
 summer has a certain pull upon it, will, in winter, when the wire has 
 contracted, experience a far greater pull. The relations between 
 
170 Binding -in. 
 
 len^ of the span, weight of wire, and stress, are indicated by the 
 following f ormnlae : — 
 
 d=g^ and » = 55 
 
 where I ^ length of span in feet ; 
 
 ui = weight of one foot of the wire ; 
 g ^ stress in lbs. ; 
 d ^ sag or dip in feet. 
 
 In other words the stress on any wire is eqoal to the square of the 
 leng^th of the span in feet multiplied by the weight per foot and 
 ■divided by eight times the dip in feet. Similarly the dip is equal to 
 the square of the length in feet multiplied by weight per foot and 
 divided by eight times the stress. These rules may be used to obtain 
 an approximate solution to any problem occurring in practice, bnt it 
 is not absolutely correct as the effects of elasticity, i.e., elongation of 
 the wire due to stress, are ignored. 
 
 Having put the requisite amount of tension npon the wire, the men 
 stationed on the top of the intermediate i>oles, at a preconcerted signal, 
 all take hold of the wire and bind it to the insulators. These men also 
 assist in passing the wire over the arms. Iron wire is bound to the 
 
 Figure 47. 
 
 insulator by means of binding wire which is mnch finer than the line 
 wires. Two lengths of wire are cut off and passed ronnd the insulator 
 and crossed over the line wire and are then twisted round the line wire 
 on either side. This binding wire is not soldered as the soldering 
 would destroy the mechanical strength of the binding. In the case of 
 copper wire a thin copper tape is rolled round the wire to protect it 
 and a binder consisting of a piece of round wire flattened out for some 
 distance is passed ronnd the insulator, and the flat portions are then 
 wound over the tape as shown in Figure 47. 
 
 When we come to the end of a coil a joint is made, this is also done 
 by means of fine binding wire. The two ends of the wire are laid side 
 by side, and are then bound round with thin wire, the binding terminating 
 ai>on the single wire at either side of the joint. The whole is soldered 
 in order to preserve the electrical continuity of the line. When solder 
 is not employed, or where the soldering is carelessly done, mnch trouble 
 
Weather Contact. 171 
 
 is caused by water entering the joint and forming a layer of oxide of 
 iron. In time this layer of oxide insulates the two wires causing a 
 disconnection. The disconnection may be in the middle of a span 
 and is, therefore, impossible to detect without testing. It will be 
 obvious that in wet weather when the oxide is wet the continuity 
 of the wire is preserved. The fault known as a dry joint will then 
 be in evidence in dry weather and not in wet. Thus the lineman has 
 io go up every pole to test. A single cell is put on at one end of 
 ihe wire and the fault is found with a detector. 
 
 In the past much trouble was experienced on account of the arms 
 carrying the insulators becoming damp in wet weather. The insulators 
 also become wet, and a small portion of the current leaks off on to 
 the insulator bolt. Before the days of earth-wiring this caused much 
 trouble as the current leaked on to the nearest wire via the arm bolt 
 and damp arm . If a number of wires travelled together for any con- 
 siderable distance they all showed slight contact in wet weather. This 
 difScnlty, known as weather contact, is now got rid of by connecting all 
 the arms together and running a wire to the bottom of the pole to 
 ■earth. Thus every arm is to earth in wet weather and any current 
 which escapes via the insulator gets to earth immediately, thus causing 
 no trouble on other circuits. It is true that by earthing the arms we 
 decrease the insulation of our line, but this is not a serious matter. 
 We can better afford to use a stronger battery and waste a larger pro- 
 portion of its current than we can afford to allow even a very small 
 proportion of thai of a smaller battery to get on to other wires and there 
 <!ause mischief. In practice the fact that the arms are earth-wired 
 makes no difference to the working value of the line even in damp 
 weather provided the insulators are all in good order. 
 
 Where the line goes over houses somewhat different construction is 
 adopted. Iron standards upon the roofs, stayed in all directions, are 
 •employed and the line wire is terminated at every insulator in order 
 to afford additional safety. The stays are so arranged that, in the 
 event of the whole of the wires on either side coming down, the stays 
 will be sufficient to hold the standard in position. It will be obvious 
 that these precautions are taken in order to minimise the risk of 
 interruption to the street traffic below. 
 
 Where it is necessary to place a line underground a conduit is formed 
 by laying cast iron pipes, having a diameter of three or four inches, 
 about two feet from the surface of the ground. The wires to be laid 
 in these pipes are insulated with gutta percha and made up into a cable 
 by binding them together every yard or so with tape. Cables containing 
 four gutta percha insulated wires are largely used in preference to the 
 ordinary single wires. The conductor is No. 18 gauge copper wire and 
 'Weighs 40 lbs. per mile, the covering bringing up the wire to No. 7i 
 
172 Underground Construction. 
 
 gange. Flnsh boxes through which the wires are made to pass are 
 placed every hnndred yards. These boxes come flash with the i)aTe- 
 ment, and the iron pipes terminate inside them, thus leaving aboat a 
 yard of slack wire in the box. The cables are made in lengths of about 
 f onr hmidred yards, and are pnlled into the pipes by means of an iron 
 wire which is threaded throngh the pipe as it is being laid. It wiU 
 thns be seen that we have a series of joints every four hnndred yards. 
 The wires are numbered by means of lead labels at every joint box. 
 The wires are identified and numbered by passing currents through the 
 wires, a man being stationed at either end, one with a battery and a 
 detector, and the other with a detector. The battery and detector is 
 connected, is joined to one of the wires, and the man with the detector 
 finds the wire by touching the wires in turn until he gets a deflection. 
 This is also shown on the distant galvanometer, and both label, that 
 wire No. 1, and the process is repeated untU all are numbered. 
 
 In order to renew wires the old wires are cut at a joint box and the 
 new wires joined in one by one so that the circnits are not interrupted. 
 When aU the new cable has been joined in, the circuits are working 
 through both the old cable and the new. The new cable is then drawn 
 in from the distant joint box by puUing the old one out. At this 
 joint box we have the old cable usually about four hundred yards long 
 lying on the ground, and it now becomes a problem as to how we are 
 to cut it out. This is done by baring every wire at the end of 
 the new cable and at the end of the old cable, and finfjing which 
 wires are connected by means of a battery, and the quantity coil of 
 the detector. To put in more wires a length of new wire having the 
 total number of wires is pnUed in, and the old wires thus pulled out 
 are augmented and pulled into the next section, and so on until the 
 route is complete. The great point to be observed in underground 
 work of this description is that on no account must wires be stopped. 
 
 Air space cable is now much used on accotmt of its low capacity and 
 high insulation combined with economy. It consists of pairs of wires 
 insulated with paper and protected by a lead covering. It is made in 
 various sizes from four to over four hundred wires. The joints are 
 made by passing a lead sleeve over one end of the cable and joining 
 and insulating the wires with paper tied with thin thread. When the 
 joints have aU. been made the lead sleeve is slipped over and secured 
 by wiped joints. The cables are terminated in boxes which are made 
 perfectly air tight. The wir^ are led out inside the box to terminals 
 which project through the iron of the box, and it is to these latter that 
 the leading wires are attached. Provided the cable is kept air tight 
 there is practically no limit to its life. In some cases arrangements 
 have been made so that dry air may be blown through the cable in order 
 to dry it, thus preventing the necessity for renewal. 
 
Electricity in its Application to Telegraphy. 173 
 
 CHAPTER XXVII. 
 
 The Making up ov Special Circuits. 
 
 In order to make up a special circuit an extensive knowledge of the 
 routes of the wires running through your test box is essential, and this 
 knowledge forms a part of the education of every Test Clerk. The 
 principal object to be gained is an increase in the number of circuits 
 to any particular office, with the minimum amount of disturbance to 
 other circuits. In order to make this clear, let us suppose that an 
 important football match is to take place at Blackburn, and that it 
 interests MR, LV, BM, TS, Hanley, Wolverhampton, Bolton, Bury, 
 and Preston. We shall require two circuits here, and it is necessary 
 that we should obtain them at the minimum cost in regard to other 
 circuits which will be stopped by portions being utilised to form these 
 special circuits. The only wire which passes through Bury is MR- 
 Blackpool one. This wire also passes through Blackburn and Preston. 
 We have a MR-CE spare wire which passes through Preston, but it 
 would not help us at all, as it does not pass through either Bury or 
 Blackburn. MR-Blackpool one cannot be spared, and it is the only 
 wire which passes through Bury and Blackburn. Clearly we must 
 have the wire, but we can only have it by replacing it by another wire. 
 We really want MR-Blackpool one from MR to PR. If we cross MR- 
 Blackpool one and MR-CE spare at MR and at PR we shall then have 
 a wire going through Bury, Blackburn, and Preston at our disposal, 
 and this without stopping a single working circuit. We have then the 
 MR-PR length of MR-Blackpool one, and can make it good with the 
 MR-PR length of the MR-Oarlisle spare wire. Now let BK cut this 
 wire in two and work each side as a separate circuit. This gives us 
 one wire with Bury and MR in circuit and one to PR. The MR wire 
 is now extended to LV by means of MR-LV one racing. The Preston 
 wire must be extended to BL, Hanley, WV, BM and TS. There 
 is a Manchester-Glasgow Stock Exchange wire which passes through 
 Bolton. The Stock Exchange closes at two o'clock, so that this wire 
 will be at our service. Preston then joins the MR side of the Stock 
 
174 Examples of making itp of Special Circuits. 
 
 sinaue rf£EDLE 
 
 A OQUBue ^l.JiT£. SOUNDBFi 
 
 -^ x=\ tvM£/rrs7-or/e aufz-EX 
 
 FlGTTEE 48. 
 
Replacement of Faulty Wires. 175 
 
 Exchange wire on to the other side of his instrument, and Bolton is 
 requested to come in circuit on MX-GHG-. Saw we have to extend 
 the wire to SSZ, WV and BM and T8. No, 24, an MR-TS wire, passes 
 through SSZ, WV, and BM. This and No. 82 are the only wires which 
 could be used, as no other wires pass through all these offices. BM 
 then comes in circuit on 24 with a repeater and a receiver. The relay 
 should always be placed as near the centre of the circuit as possible^ 
 Hanley and WV would also come into circuit on 24, thus it will be seen 
 that we have obtained two circuits from BK working to nine offices 
 without stopping a single working wire. 
 
 The instructions issued by MR would be as follows : — 
 
 Pkeston. — Cross MR-BIO,, and CE-spare and come in circuit on MX- 
 GrHG and MR-BIC, with receiver. 
 
 Bolton. — Come in circuit on MX-G-HG with receiver. 
 
 Blackburn. — Terminate both sides of MR-BIC, with transmitter. 
 
 Bury. — Come in circuit on MR-BIC, with receiver. 
 
 Birmingham. — Come in circuit on 24 with repeater and receiver. 
 
 Hanley. — Come in circuit on 24 with receiver. 
 
 Wolverhampton. — Come in circuit on 24 with receiver. 
 
 Blackpool.— Unaffected. 
 
 Liverpool. — Terminate No. 1 racing with receiver. 
 
 The wires taken are MX-GHG (between MR and PR), 24 (between 
 MR and TS), CE-spare (between MR and PR), and MR-LV, racing. 
 
 Figure 48 shows the circuits joined up, and by its aid the purpose of 
 the various crosses will be readily apprehended. The way in which 
 such arrangements would be carried out at the various test boxes was 
 explained in Chapter XVI. 
 
 To "make good" a wire means that a length of good line is substi- 
 tuted for the faulty portion. The principle upon which changes 
 are made is the sacrifice of less important wires to form important 
 connections. For instance, Manchester has three wires to Bradford, 
 the third of which can generally be spared. Again we should always look 
 at the abstract and see what wires are wrong, as we may then be able to 
 use a good length of a faulty wire. Sometimes it is impossible to find 
 a length, but the idea of making a wire good should not even then 
 be abandoned, as it may be possible to work round from the other 
 end. An example of this is shown in Figure 48 at B, where the 
 Manchester-Oldham-Halifax-Huddersfleld news wire is to full earth 
 between Oldham and Halifax. This is the only wire which passes 
 through Halifax and Oldham. It will then be seen that its replace, 
 ment requires a little consideration. There are two ways of doing 
 this, one by means of a fork at MR, and the other is by MR becoming 
 intermediate, i.e., working the upside of MR's instrument to HF by 
 means of another wire, thence to HX who terminates, and to OL from 
 
 u2 
 
176 Crosses. 
 
 the down-side MB is thus intermediate. MB may fork his np-side by 
 the second wire to HF and to OL with ME-HF,. Oldham would 
 then have to terminate and reverse his instrument. We should also 
 have to insert resistance against Oldham in order that he might not 
 take too much current. We could easily obtain an ME-HF length 
 by taking one of the four MB-LS wires, MB-HFj or by asking LV 
 to give MB either one of his LV-NT or L"V-LS wires, all of which 
 pass through HF. If we can spare MB-HF^ we should certainly 
 prefer to use this. The letters on either side of the instrument 
 indicate the up and down sides. If the wires are forked, MB would 
 have to reverse, earthing the down side, and forking the ap side along 
 MB-HF2. 
 
 As another example of forming circuit by going round, let us 
 suppose that the MB-Failsworth-Oldham-Lees circuit is to earth MB 
 and Failsworth. We cannot replace this length as we have no wires 
 passing through Failsworth. Oldham would be asked to terminate 
 and transmit for Lees, and put Failsworth through to MB by means 
 of MB-OL-spare. MB's Failsworth instrument would have to be 
 reversed and the Failsworth wire put to earth. Now the Failsworth 
 instrument is on a universal set, therefore, if we reverse it, we shall 
 short-circuit the battery every time either key is touched. Therefore, 
 we must employ a spare instrument with separate batteries. When 
 using instruments worked from a universal battery for spedal circuits 
 the line and earth terminals must never be touched. If it is desired 
 to reverse the instrument, the battery and relay must be reversed. 
 That is an important point. 
 
 In making up special wires we must always be very careful 
 not to take wires with long underground sections, otherwise the 
 speed attained will be low. Frequently a wire which is being 
 worked Wheatstone is too low as regards insulation for satisfactory 
 working. In this case we may be able to work simplex key 
 speed perfectly, and the wire should accordingly be crossed with 
 a perfect wire which is being worked simplex. Thus we improve the 
 Wheatstone circuit without effecting the simplex circuit. A rule for 
 finding the number of cells to give any required circuit through any 
 external resistance was given on Page 122. For Wheatstone circuits 
 of resistance 1,000«, 15 Bicros ; from 1,000 to 1,300, 20 Bicros ; 1,300 
 to 1,700, 25 Bicros. This provides for a current of 25 m,a. For double 
 current duplex sets large Daniells must be employed. To give a 
 current of 17 m.B. up to 500w, 10; 500-750, 15; 750-1,000,20. Large 
 Daniells are only made up in sets of 5 and 2, and the latter are never 
 used for line batteries. 
 
 The resistance, etc., of the various kinds of wire ordinarily employed 
 are given in Chapter XXYI, 
 
Constant for Iron and Copper Wires. 177 
 
 The weight per mile multiplied by the resistance per mile of any 
 kind of wire is a constant number. The constant for iron wire is 
 5,400. If we wish to know the resistance of 60 lb. iron, we have only 
 to divide 5,400 by 60, and we have our answer 90(d. The constant for 
 copper 890. By remembering these two numbers we can always 
 obtain the resistance per mile of any wire in use. 
 
178 Electricity in its Application to Telegraphy. 
 
 CHAPTEE XXVin. 
 
 Miscellaneous Notes. 
 
 Accumulators. — ^An ordinary or primary cell in return for chemical 
 energy yields electrical energy. A secondary cell or accomnlator is a 
 reversible primary cell in which the chemical energy may be restored 
 to the cell in return for. electrical energy. We have seen that a 
 series of chemical reactions taking place will furnish a current, there- 
 fore it is not surprising to find that, on passing that current back 
 through the cell in the opposite direction to that in which it came out, 
 the chemical componnds would again resume their original form. In a 
 Daniell cell we have zinc changed to zinc sulphate and copper deposited 
 upon the copper plate ; in return for this chemical energy, i.e., the 
 difference between the chemical energy possessed by the constituent 
 parts of the cell before and after degradation, we have electrical energy 
 in the shape of an electric current. A current passing through a suit- 
 able liquid results in chemical change, the liquid being split up into 
 its constituent parts. Such a substance is termed an electrolyte, and 
 its essential properties are that it must be a fluid, a compound body, 
 and a condaetor of electricity. By compound is meant a substance 
 formed by the chemical union of two or more elements, e.^ ., water which 
 is made up of oxygen and hydrogen. Mercury is a conductor of elec- 
 tricity and a fluid, but it is not an electrolyte as it is an element, i.e., 
 a simple substance which cannot be split up. If, now, on passing a 
 current through an exhausted Daniell cell in an opposite direction to 
 the E.M.F. of the cell, we can restore the chemical compounds to their 
 original places, i.e., split up the zinc sulphate into zinc and leave the 
 SO4 to attack the copx)er plate and remove the copper deposited upon 
 it, thus forming copper sulphate, we shall, provided the current be 
 continued long enough, have the cell restored to its former condition, 
 when we may again run it down by taking current from it, thus again 
 degrading the chemical compounds. We thus see that any and every 
 battery can be used as a secondary cell, and the only question which 
 we have to consider is, which is the cheaper, viz., to supply fresh zinc 
 and sulphate of copper or electrical energy obtained from the electric 
 lighting plant. In the case cited fresh materials, i.e., chemical energy 
 
Accumulaiors. 179 
 
 IS far cheaper. Hence we see that the term accumulator or secondary 
 cell is the name given to cells in which the chemical energy can be 
 ■eflBciently restored to the cell by the employment of electrical energy. 
 The cell used by the Department is manufactured by the E.P.S. 
 Company, and consists of seven lead grids, four of which are pasted 
 with litharge or lead oxide, and three with lead sulphate, which is 
 formed by adding sulphuric acid to red lead. The object of the lead 
 grid is to conduct the current to the active material. The grid is made 
 of an alloy of lead and antimony, as it is found that the electrolyte, 
 dilute sulphuric acid, has little or no effect upon it. 
 While charging — 
 
 Normal. 
 PbSOj + 2HjO + H2SO4. 
 
 One molecule of and two molecules of and one molecule of Sulphuric 
 Sulphate of Lead Water Acid 
 
 + PbSO, 
 
 and one molecule of Sulphate of Lead. 
 During charging — 
 
 = PbSO, + 02 + HjSO, + 2 m 
 
 Form one molecule and one molecule and one molecule and two molecules 
 of Sulphate of of Oxygen of Sulphuric of Hydrogen 
 
 Lead Acid 
 
 + PbSOi 
 
 and one molecule of Sulphate of Lead. 
 Fully charged — 
 
 = PbOa + SaSO, + Pb. 
 
 One molecule of and three moleculeB of Sulphuric and one molecule of Lead. 
 Lead Dioside Acid. 
 
 In the normal condition both plates are coated with lead sulphate and 
 immersed in sulphuric acid and water. Passing a current through the 
 ■cell splits up some of the water in the cell into oxygen and hydrogen. 
 The former replaces the SO, on one plate which joins with the hydro- 
 gen, thus forming sulphuric acid. The SO, from the other plate also 
 •combines with the remainder of the hydrogen. The oxygen combines 
 with the lead, thus forming lead oxide. When the cell is discharged 
 the normal conditions are resumed and the process may be again 
 repeated. 
 
 The plates measure 10 by 10 inches. The four negatives and the 
 three positives are connected together. The cells have an exceedingly 
 low internal resistance and on that account are particularly suitable 
 for universal battery working. The cells are charged with a current 
 of from 10 to 15 amperes. The various voltages are tapped ofE to meet 
 the average requirements, and as a general rule the nearest voltage in 
 excess should be used for any circuit which requires, or is fitted with, 
 a voltage other than that supplied on the universal system. In the 
 
180 Winding of Apparattis. 
 
 battery room there is a switch board with facilities for connectiiig 
 any of the batteries, either for charge or discharge. Thin lead or 
 platinoid wires are inserted in all the battery leads in order that, in the 
 event of a short drcait, the wire may become warm and break, thos 
 cntting the circnit. Accnmnlators are only possible where there is an 
 electric Ught plant from which the cells can be charged. 
 
 The condition of the cells in regard to quantity of energy they con- 
 tain can readily be ascertained by taking the specific gravity of the 
 solution with a hydrometer. When the cell is being charged the 
 specific gravity of the solution rises from 115 to 1*20 when folly 
 charged. 
 
 FiGTJBE OF Mebit. — By this term is meant the smallest current 
 which will produce a given effect upon a given form of apparatus. The 
 standard relay should work with half a milliampere, and half a miUi- 
 amx>ere is accordingly termed its " figure of merit." The single needle, 
 should deflect to either stop pin with a current of three milliamperes 
 and this current is thus its figure of merit. To test the figure of merit 
 of any particular instrument we have only to take a 5-cell large 
 DanieU and a couple of rheostats. Join all in series and increase the 
 resistance in the rheostats till the apparatus will just work, and then 
 calcnlate the current which the battery is sending through the 
 instrument. The result of your calculation is the figure of merit of 
 the instrument. 
 
 Winding of Appabatus. — Copper wire is used for winding all 
 electro-magnetic apparatus on account of its high conductivity. The 
 larger the resistance of the apparatus the more battery power shall we 
 have to provide to work it, but even supposing we found that we could 
 replace the thin copper wire by the thicker iron wire at a less cost it 
 would be most undesirable to do so. Firstly, it would not be so com- 
 pact and the instrument would be less el^ant. Secondly, and this is 
 the real objection, if we used the same number of turns of the thicker 
 wire the outer turns of wire would be further away from the electro- 
 magnet core and thus the lines of force due to these windings would 
 escape more into the surrounding air where they would produce no 
 effect upon our relay. Thus we see that there is a serious objection from 
 a theoretical point of view to this course. As a matter of fact the 
 thicker iron wire would not be a great deal, if any, cheaper than the 
 thin copper. To replace thin copper by thick iron we should have to 
 provide twice the length of iron wire and would not then get nearly 
 the same magnetic effect, provided the same current ciionlat^ through 
 both. In line wire it is quite another matter, but here again it is now 
 the practice to use 100 lb. copper in place of 400 lb. iron. The copper 
 costs little more than the iron, possesses less resistance, and is entirely 
 free from the retardation which occurs on iron wires, due to the fact 
 
Moment of a Magnet. 181 
 
 that the iron wire becomes magnetised thus giving it large self-induC' 
 tance. For wheatstone working a copper wire means an increment in 
 speed as against an iron wire. Eheostats and Wheatstone Bridges are 
 wound with wire possessing a high specific resistance in order that we 
 may, in the smallest compass, have the maximum possible resistance. 
 The smaller the amount of wire necessary to make a certain resistance 
 coil the less will be its cost. We must have also a wire which will not- 
 vary much in resistance when subjected to changes of temperature. 
 In rheostats this is unimportant as we merely wished to balance our 
 line and not to know its resistance. The coils in a rheostat are only 
 approximately correct. Now in a Wheatstone Bridge this is an exceed- 
 ingly important matter, as, if when the weather becomes warmer the 
 coils in our Bridge increased largely in resistance we should not then 
 be able to depend upon our results unless the temperature happened 
 to be that at which the Bridge was correct. German silver varies in 
 resistance to a very small extent with changes of temperature, so that 
 these slight variations are quite negligible. In more accurate bridges- 
 an alloy of platinum and silver is used. The coils are not wound in the 
 same way that an electro-magnet is wound as, were this done, we should 
 have currents due to self-induction upsetting our balance when the 
 current was started and stopped. Accordingly the wire is bent into a 
 loop, both wires are then wound on the bobbin side by side. A little 
 consideration will show us that the magnetic field set up by one wire is 
 exactly balanced and opposed by the other, and we hereby obtain a 
 resistance ooil possessing little or no self-induction, or as it is usually 
 called, a non-inductive resistance. 
 
 Moment of a Magnet. — The magnetic moment of a magnet is the 
 strength of either pole multiplied by the distance between them. This 
 corresponds to the strength of the magnet. A short magnet has the 
 North pole nearer to the South pole than a longer magnet. The action 
 of the magnet on a compass needle is proportional to the pole strength 
 of the nearer pole of the magnet, minus the action caused by the dis- 
 tant pole, which is obviously opposite in its action to the nearer pole. 
 If the magnet is infinitely short, then the North and South poles coincide 
 and no effect will then be produced on a distant compass needle, since 
 the action of each pole is equal and opposite. The action varies intersely 
 as the square of the distance, consequently the pole which is further 
 away from the compass is able to do little towards counteracting the 
 nearer pole if the compass is near and the maen^et is long. The respec- 
 tive moments of two magnets may be compared by suspending the 
 magnets one after the other in the earth's field and setting them to 
 vibrate between east and west. The respective moments will then be 
 proportional to the square of the number of oscillations executed in 
 any given time. For instance, suppose that one magnet makes 20, and 
 
182 Resistance of Wires. 
 
 -the other 40 oscillations in one minute. The respective moments are 
 ■then as W : 40«, i.e., as 1 : 4. 
 
 Heat Produced by Current. — The total amount of heat produced 
 in a wire is proportional to the total quantity of electricity which 
 passes through it multiplied by the E.M.F. employed, oris proportional 
 to CP X R X t where C is the current, R the resistance of the wire, and 
 t the time the current is ilowing. In working out any problem, re com- 
 paring the amount of heat produced in wires, it is best to insert the 
 quantities you know first, and then work out the question. Do not 
 attempt to solve suchquestions by remembering complicated formulae. 
 Compare the quantity of heat produced by a current of three amperes 
 ilowing through a wire whose resistance is lOu for five minutes with 
 that produced by a current of six amperes flowing through a wire of 
 5u) resistance for twenty minutes. 
 
 Heat is, in first case, proportional to 3^ X 10 X 5 ^ 450 
 „ „ second „ „ „ 6' x 5 x 20 = 3600 
 
 If the quantity of heat produced in the first instance is one unit the 
 heat produced in the second case is eight units. 
 
 Resistance Weight and Diameter of Wires. — The resistance of 
 any wire varies directly as its length and inversely as the square of its 
 diameter. For an example of a question on this law see Page 42. 
 
 The resistance of any uniform wire varies directly as its length and 
 inversely as its weight per unit length. 
 
 If two wires have the same total weight their resistances will vary 
 as the square of their respective lengths. 
 
 The above laws premise that the same kind of wire is used in each 
 case, i.e., the resistance of two yards of German silver wire is twice the 
 resistance of one yard of German silver wire of the same diameter. 
 
 Example : If a copper wire 80 yards long, weighing 20 lbs., has resist- 
 ance of l'2(<j, what is the resistance of a wire 405 yards long, weighing 
 50 lbs.? 
 
 80 yards of wire weighing 20 lbs. has a resistance of 1"2 ohms. 
 1 
 
 .-. 405 
 .-, 405 
 .-. 405 
 
 20 
 
 
 
 
 1-2 
 
 
 » 
 
 " 
 
 )) 
 
 
 SO 
 
 
 
 
 80 
 
 405x20 
 
 ^ 
 
 
 
 405X1-2 
 
 SO 
 
 
 
 
 80 
 
 1 
 
 t; 
 
 « 
 
 99 
 
 405x20x405x1-2 
 80x80 
 
 50 
 
 )» 
 
 ii 
 
 )) 
 
 4052x1-2x20 
 802X50 
 
 
 
 
 = 12-3a> 
 
 All problems of this character are far better worked out step by step 
 than by application of a formula, which is liable to be forgotten. The 
 only point which it is necessary to call attention to is the fact, that if 
 
The Polarised Sounder. 183 
 
 a wire is out into two equal parts the resistance weight and length of 
 such part is half that of the whole wire. The above problem is worked 
 out upon this principle. 
 
 The Polarised Sounder. — This is an instrument for use at local 
 offices in place of either single needle, double plate sounder, single 
 or double current sounder, simplex or duplex. It consists of a heavy 
 needle operated by the armature of an electro-magnet. The armature 
 is rendered magnetic by induction from a permanent magnet. Two 
 sounding pieces with projecting pins (as in Neale's sounder) are used. 
 The instrument is wound differentially. An arrangement is made by 
 means of which the needle, which is held vertical by means of a light 
 spring, may be moved from the centre against one of the tins. In this 
 position it is used for single current sounder circuits. The signals are 
 suflSoiently loud and clear for the busiest of local circuits. For double 
 current circuits the needle is kept midway between the tins, as the 
 spacing current will hold it over to one tin. As the instrument is 
 wound difEerentially, it may be used for double current duplex circuits, 
 and thus dispense with relay, sounder, and local battery. It is louder 
 than the acoustic needle, though not so loud as a double plate sounder 
 but the latter requires careful regulation. The instrument supplies a 
 long felt want, as the apparatus at local offices should always be of the 
 simplest possible description. 
 
184 Electricity in its Application to Telegraphy. 
 
 QUESTIONS. 
 
 1. What is meant by the molecular theory of magnetism, and in 
 what way do a piece of brass, a piece of iron, and a magnet difier 
 from one another t 
 
 2. Describe Ohm's law. Give practical units. The resistance of a 
 circuit is l,400(ii and the E.M.F. applied is 40 volts. What is the 
 strength of the current through it ? Answer: 28^ m.a. 
 
 3. If the current flowing through a circuit of l,500(i> resistance is 10 
 mUIiamperes what is theB.M.P. of the battery employed? Answer : 
 15 volts. 
 
 4. A battery of forty Daniell cells each having an E.M.F. of one 
 volt is worked through an entire circuit resistance of l,000(<i. What is the 
 current strength in milli amperes ? Answer: 40 m.a. 
 
 5. What is the resistance of a circuit through which a current of 
 20 m.a. flows when an E.M.F. of 10 volts is applied 1 Answer : 500w. 
 
 6. If the current through a circuit of 1,500(« resistance is 20 m.a,, 
 what is the E.M.F. employed ? Answer : 30 volts. 
 
 7. If the current flowing through a rheostat with l,000(<i inserted is 
 20 m.a., what is the difference of potential between the terminals J 
 Answer : 20 volts. 
 
 8. Sketch a Daniell cell and explain local action. Give chemical 
 symbols when the poles are connected, 
 
 9. Describe the Bichromate and Leclanche cells, 
 
 10. What is a Leclanche agglomerate cell ? Why are Leclanche cells 
 never used as local batteries 1 
 
 11. Describe the chemical actions which take place in a Leclanche 
 cell and say why it polarises more rapidly than the Daniell ? 
 
 12. What is the object of amalgamating the zincs in batteries and 
 how does it effect that object 1 
 
 13. Describe briefly any form of dry cell and state what advantages 
 you consider it possesses. 
 
 14. What is the object of the depolarising fluid in every form of 
 cell 1 What is the exciting fluid in the Leclanche cell 7 
 
Qvsstions. 185 
 
 15. What is Ohm's law? A battery whose E.M.F. is 95 volts is 
 employed to work a line ofEering 3,900aj resistance. The resistance of 
 the apparatus is 650w and that of the battery 320u. What is the cur- 
 rent sent out t Answer : 19^ m.a. 
 
 16. How would you measure the internal resistance of a cell ? 
 
 17. How is the constant of the tangent galvanometer taken! 
 
 18. A deflection of 90 divisions is produced on the tangent gal- 
 vanometer by a standard cell. How should you reduce this to 80 
 divisions ? 
 
 19. How would you test the E.M.P. of a battery if given a galvano- 
 meter, standard cell, and a rheostat 1 
 
 20. How would you compare the B.M.P.'s of two cells ? 
 
 21. It is observed that the pointer of the tangent galvanometer 
 fiwings about for a long time before coming to rest. What is the fault 1 
 
 22. How would you test the E.M.F . of a cell by means of the tangent 
 galvanometer 1 
 
 23. With a galvanometer of lOw resistance and 15«) in the rheostat a 
 deflection of 40 divisions is obtained from a battery. On raising the 
 resistance in the rheostat to 50u the deflection fell to 20 divisions- 
 What is the internal resistance of the battery ? Answer . lOw. 
 
 24. The standard cell with a 70u galvanometer and 330w in the 
 rheostat gives a deflection of 32°, a second battery with 7,930w in the 
 rheostat gives the same deflection, What is the E.M.F, of the second 
 battery 7 Answer : 21"6 volts. 
 
 25. A 10-cell Daniell (E.M.F., 1 volt ; internal resistance, 8u per 
 -cell) is used to work a line 8 miles long, whose resistance is Vim per mile, 
 J'ind the strength of the current if the resistance of the distant instru- 
 ment is 200w. Answer : 26"6 m.a. nearly. 
 
 26. What are the characteristic features of each of the three forms 
 of primary cell used by the P.O. 1 
 
 27. What would be the total E.M.F. and the total resistance of six 
 Daniell cells joined up, three in " series " and two in " quantity," the 
 Jj.M.F. per cell being taken as one volt, and the resistance per cell as 
 bm 1 Answer : E.M.F, 3 volts resistance 7'5i<j, 
 
 28. Ten cells are joined in parallel and send a current through an 
 external resistance of 15&i. If the E.M.F. is one volt and internal 
 resistance lOu per cell, what is the current produced t Answer : 
 «2-5 m.a. 
 
 29. Ten cells having an E.M.F. of 1 volt, and an internal resistance 
 •of 6(K per cell are joined in series with a resistance of 140u. What is 
 the current produced 1 Answer : 50 m. a. 
 
 30. Distinguish between grouping battery cells in " series " and in 
 ■" quantity," and state which arrangement would be best for working 
 ihrough a high resistance ? 
 
186 Qtiestions. 
 
 31. What is the current from 50 Daniell cells, each of 4m internal 
 resistance, and E.M.F. of 1 volt, through l,000w external resistance T 
 Answer : 42 m.a. approx. 
 
 32. By bow many Bichromates could the Darnells in the above ques- 
 tion be replaced in order to give the same current ? Answer : 26. 
 
 33. Electro-magnets have soft iron cores. Why is this the case, and 
 what would be the effect of replacing them by (a) hard steel, (6) by 
 brass cores ? 
 
 34. Describe the pony sounder and sketch the connections of a direct 
 sounder set. 
 
 35. A sounder is found to work badly from the armature sticking 
 after the current is taken off. What defect in the make of the instru' 
 ment would cause this, and how could it be overcome ? 
 
 36. Describe and explain the action of any form of lightning pro- 
 tector f 
 
 37. Draw a diagram showing three single needles joined up on one 
 circuit. 
 
 38. Describe the single needle apparatus in detail. 
 
 39. Describe the Seimen's or the P.O. standard relay, and explain 
 the difference between such relays and non-polarised ones. 
 
 40. Join up a relay (a) in series, (J) in parallel (in quantity), and say 
 what would be its resistance in each cas» ? 
 
 41. Why will not the tongue of a relay remain in a position midway 
 between the two contact points ? 
 
 42. What is meant by " giving the relay a spacing bias "? 
 
 43. A double plate sounder is found to be completely out of 
 adjustment, what steps and in what order would you take them in 
 order to put the instrument right? 
 
 44. Join up a single needle and a double plate sounder on one 
 circizit. 
 
 45. Sketch and describe the double plate sounder. 
 
 46. What is probably the fault if you are unable to stop the call 
 of the distant station on a double plate sounder circuit. 
 
 47. Sketch the connections of a single cxirrent sounder with relay 
 circuit. 
 
 48. In a set of Morse apparatus what parts belong (a) to the local 
 circuit (6) to the line circuit. 
 
 49. It is noticed that dots are missed and at the same time the dashes 
 stdck. What would you expect the fault to be ? 
 
 50. Describe the connections and action of a condenser. 
 
 51. What terminals do the up and down lines touch on a double 
 current sounder set ? 
 
 52. Sketch the connections of any form of telegraph apparatus in 
 use in the Post Office. 
 
Questions. 187 
 
 53. Why can a relay be worked with a smaller current on a doable 
 than on a single current circuit ? 
 
 54. A double current sounder set is completely out of adjustment. 
 How would you set about putting it into working order ? 
 
 55. If it was desired to short circuit a double current Morse circuit, 
 which two points should you connect 1 
 
 56. Why cannot a non-polarised relay be used for double current 
 working ? 
 
 57. What is the object of the double current system 1 
 
 58. How could the apparatus connections of a double current 
 sounder be verified ? 
 
 59. 8ketch the connections of a double current sounder set and say 
 how it could be made to work single current . 
 
 60. With the coils of the relay in parallel on a long double current 
 circuit the marks are good. If when placed in series none are received, 
 what and where is the fault ? 
 
 61. Describe the double current key and indicate the connections in 
 each position. 
 
 62. A double current key is only sending out one current. Where 
 is the fault 1 
 
 63. What fault would cause you to be unable to stop a call on a 
 double current sounder circuit 1 
 
 64. You are able to send to a distant station but when the switch 
 is turned nothing is received. In what part of the circuit is the fault ? 
 
 65. Explain the principle of differential duplex working as applied 
 to any form of apparatus. - 
 
 66. Sketch the connection of a single current sounder duplex, 
 
 67. Why is a condenser used on a duplex circuit 1 
 
 68. Show by means of a diagram how the current passes through la 
 double current duplex. 
 
 69. Sketch the connections of a double current sounder duplex. 
 
 70. Having balanced a duplex circuit it is found that the galvano- 
 meter needle is vertical. Why is this ? 
 
 (a) If the line is in normal condition as regards insulation. 
 (5) If the insulation is very low. 
 
 71. On a long wire it is found impossible to work duplex, and on 
 account of the current being too weak to work the relay, but that on 
 turning to simplex the signals and deflection on the galvanometer were 
 improved. Why is this the case ? 
 
 72. When both keys are at rest on a double current circuit the 
 batteries combine together. When one key is depressed the deflection 
 on the galvanometer is unaltered. Why is this, since the battery has 
 now only to send a circuit through one coil of the relay, whereas in 
 the former case it has to do this and also send a current to line I? 
 
188 Qtiestimis. 
 
 73. Explain the theory of balancing a duplex circuit. 
 74 What is the object of the condenser coils 1 
 
 75. A double current sounder duplex set will work duplex, but not 
 simplex. Where should you look for the fault ? 
 
 76. It is found impossible to balance a duplex circuit, whose resist- 
 ance is ordinarily about 800u, although from to 8,000w is tried. What 
 Saxjlt should you look for ? 
 
 77. Say what takes place on a double current duplex circuit (a) both 
 keys at rest, (6) both depressed, (c) when one key only depressed. 
 
 78. Trace tiie path of the current in the single needle, double 
 -current sounder simplex and duplex and qnadruplex CDiagram Book 
 supplied). 
 
 79. Explain the principles of the double current duplex and qnad- 
 ruplex. 
 
 80. What is the difference between a polarised relay and a non- 
 polarised one J 
 
 81. Explain in detail the principle and construction of a non- 
 polarised relay. 
 
 82. What is meant by decrement working ? 
 
 83. What is the object of the uprighting sounder in increment 
 working? 
 
 84. Sketch and describe the qnadruplex increment key and say why 
 an ordinary single current key could not be substituted, 
 
 85. What are the objects of the three resistance coils used in con- 
 nection with the qnadruplex instrument ? 
 
 86. A quadruples at rest sends a normal current to line. This cur- 
 rent is unaltered when the B key is worked. What is the fault ? 
 
 87. Describe the quadruplex A key, 
 
 88. How would you join up a Wheatstone receiver iu place of a 
 relay? 
 
 89. Shew by means of a diagram how you would fit a transmitter to 
 a key worked circuit, 
 
 90. A distant station's transmitter is sending a marking bias. What 
 does this mean and how could you prove it ? 
 
 91. What is the object of a shunted condenser on a Wheatstone 
 <!ircuit ? 
 
 92. What is the object of the condenser in duplex working, and in 
 what way does its application differ from its application to a Wheat- 
 stone circuit ? 
 
 93. A transmitter is found to send out a marking current only. 
 What is the fault and how may it be remedied 7 
 
 94. Describe a Wheatstone transmitter and receiver, 
 
 95. Give a dis^ram showing what terminals you would use in 
 crossing wires at the test box. 
 
Questions. 189 
 
 96. Two wires are in contact. How would you proceed to localise f 
 
 97. How could you come in circuit on a through wire ? 
 
 98. How would yon alter the test box connections in order to work 
 up to an office instead of down? 
 
 100. Show by a sketch how an instrument can be forked on to two 
 wires, one of which is an up station, and the other a down station. 
 What instructions, if any, should be issued to the terminal offices 1 
 
 101. How would you increase power on a circuit and change the 
 local? 
 
 102. Show by means of diagrams how to "earth," "dis,"' "put 
 straight," "loop," and "cross" two wires at the test box. 
 
 103. How would you cross two instruments, and what is usually the 
 object of such crosses? 
 
 104. It is found that the signals are reversed. What is the fault, 
 and how should it be remedied ? 
 
 105. How would you extend a wire (a) remaining in circuit (5) 
 going out of drcnit ? 
 
 106. What precaution should you adopt in order that two circuits 
 unequal in resistance may both receive good currents when forked 1 
 
 107. Sketch the connections and explain the use of the test box 
 galvanometer. 
 
 108. Explain the system of morning testing now in use. 
 
 109. What is meant by insulation resistance 1 The resistance of a 
 line 105 miles long is 4,050w when disconnected at the distant end. 
 What is its insulation resistance per mile ? Answer : 425,250w. 
 
 110. How would yon test the insulation of a line with a 10-cell 
 Daniell and tangent galvanometer ? 
 
 111. How would you determine the number of cells to be used for 
 the morning test ? 
 
 112. What is the object of the two 10,000(o blocks used for the 
 morning test ? 
 
 113. What is the joint resistance of two wires whose resistances are 
 20.U and 60w respectively ? Answer : 15w. 
 
 114. In what proportion would 12 m.a. split between the two wires 
 in the above question ? Answer : 3 m.a. through 60u, and 9 m.a. 
 through 20(0. 
 
 115. Sketch the skeleton connections of the morning test— (o) for 
 a loop, (fi) for a single wire. 
 
 116. What is meant by the perfect current, and how would you 
 calculate the perfect current for a line of given conductor resistance ? 
 
 117. If the resistance of a circuit be 1,000m, what is the resistance of 
 a fault at its centre if 100 m.a. is sent out and only 50 m.a. 
 Answer: SOOu. 
 
 118. Describe a galvanometer shunt and explain its object. 
 
 N 
 
190 Questions. 
 
 119. Thirty-two divisions are registered on the tangent galvano- 
 meter with l-40th shunt, what is the strength of the current 1 Answer : 
 16 m.a. 
 
 120. A current is received at the end of a telegraph line. Show how 
 its strength could be measured by means of the tangent galvanometer 
 if the current were about 20 m.a. What shunt should you employ 7 
 Answer : l-20th. 
 
 121. If the resistance of a galvanometer be nine times that of its 
 shunt, what will be the proportion of current flowing through each 1 
 Answer ; Shunt 9-lOth, galvanometer 1-lOth. 
 
 122. By what resistance must a galvanometer of lOOw be shunted in 
 order to reduce its deflection to one-tenth of its former value 1 
 Answer : 11 gw. 
 
 123. A wire AC passing through B is found to work badly. From 
 A to B, 20 nules, it is underground and shows an insulation re- 
 sistance of 3,000w. From B to C it is open 50 nules long, and 
 gives an insulation resistance of 5,000a), which, if either, is the faulty 
 section ? 
 
 124. What shunt should be employed on the tangent galvanometer 
 in order to measure 200 m.a. ? Answer : l-lBOth, 
 
 125. What would be the resistance of a shunt employed on a 90(o 
 galvanometer to reduce a deflection to 1-lOth of its former value? 
 Answer : lOw, 
 
 126. What would be the combined resistance of the galvanometer 
 and shunt in the above case ? Answer : 9ii<. 
 
 127. If the current through the galvanometer (in question 125) uu- 
 shunted were 20 m.a., how would this current divide between the gal- 
 vanometer and shunt 1 Answer : '2 m.a. and 18 m.a. 
 
 128. Why is the resistance of a tenth shunt one-ninth the resistance 
 of the galvanometer 1 Grive a nomerical example showing the current 
 originally flowing through the galvanometer and the currents flowing 
 through the galvanometer and shunt in the second case. 
 
 129. A tangent galvanometer has currents passed through it which 
 produce deflections of 30° and 60^ respectively. What are the relative 
 strengths of the two currents ? Answer : as 1 : 3. 
 
 130. Assuming the deflection of 60° to be obtained with the galvano- 
 meter unshunted, what should be the resistance of a shunt which would 
 reduce the current to 30°, the galvanometer having a resistance of 
 IjOOOui and the current flowing into the galvanometer being the same in 
 both cases? Answer: 500<o. 
 
 131. When a message is started the signals are good but gradually 
 fade away. What is the fault ? 
 
 132. Are wires ever looped for telegraph working ? Ifso, why, and 
 what is the disadvantage of the practice 1 
 
Questions. 191 
 
 133. Three stations, A, B, and 0, are worked on the double current 
 system. The intermediate station B closes at 8 p.m. and leaves his 
 switch to " send." How can A work to C 1 
 
 134. How would you localise a disconnection ? 
 
 135. How is a contact localised ? How should the battery and gal- 
 vanometer be joined up ? 
 
 136. How is an earth fault localised 1 G-ive an example. 
 
 137. Explain how an intermittent fault can be localised by crossing 
 successive lengths. 
 
 138. What are the symptoms of an underground fault and how does 
 it differ from a stay fault ? 
 
 139. How is an intermittent contact localised ? 
 
 140. How would you proceed in order to ascertain whether a wire 
 is in contact, and having ascertained this, how would you find the second 
 wire? 
 
 141. Explain the principle of the Wheatstone Bridge. 
 
 142. Grive a diagram of the Wheatstone Bridge arranged for taking, 
 loop tests. 
 
 143. Give a diagram of the Wheatstone Bridge and its connections. 
 
 144. Why is it necessary to depress the battery key before the gal- 
 vanometer key when balancing the Wheatstone Bridge 1 
 
 145 . What is the ob j ect of the peg switch on the Wheatstone Bridge 1 
 
 146. How would you test three lengths of line (joined together at 
 the distant end) for conductor resistance 1 
 
 147. If there are three wires, A, B, and C, and the resistance of A 
 and B is 2,180<o, B and C is 2,650u, A and C is 2,431w. What are the 
 separate resistances of A, B, and C ? Answer : A=980iw, B=l,199J<i), 
 C=l,450Jw. 
 
 148. State how a contact between two telegraph wires could be 
 localised when a third wire is available for the purpose. 
 
 149. How would you convert a double current sounder into a com- 
 plete wheatstone set ? 
 
 150. How would you measure a resistance of 100,000<o by means of 
 the wheatstone bridge ? 
 
 151. Three wires are looped two at a time. The resistance of A and 
 B is 2,100w, A and C is 2,145w, and of B and C 2,195w. What are the 
 individual resistances 1 Answer : A l,025<<i, B l,075w, C l,120io. 
 
 152. How would you localise an earth fault by loop test ? If the 
 conductor resistance of the two wires is 1,400(d, and a balance is 
 obtained with 500w unplugged, how many ohms is the fault from the 
 testing office ? Answer : 450w. 
 
 153. Having balanced a resistance by means of the wheatstone 
 bridge it is noted that, when the galvanometer key is depressed before 
 the battery key, a momentary deflection is obtained. How can it be 
 
 n2 
 
192 Questions. 
 
 ascertained whether the resistance possesses capacity or self-induction 
 from the direction of the deflection ? 
 
 154. If it is desired to work several double current sets from one 
 battery, how would yon join up the batteries, and why ? 
 
 155. Why is it necessary to provide a split battery for universal 
 working 1 
 
 156. Describe the principle of the universal battery system, and say 
 why it is not desirable to use Daniells or Leclanche cells for such 
 purpose. 
 
 157. Where should the resistance blocks be placed on a doable plate 
 sounder circuit for universal working, and why 1 
 
 158. Describe the standard DanieU cell, and say in what particulars 
 it differs from an ordinary Daniell. 
 
 159. Explain the general principles of the Delaney multiplex system, 
 and say why the number of instruments which can be worked from it 
 is limited to four between Manchester and London. 
 
 160. A wire which is worked duplex will not work hexode. Why is 
 this the case ? 
 
 161. What is the object of the reed in multiplex working ? 
 
 162. How is the speed of rotation of the trailer adjusted ? 
 
 163. By what means is synchronism obtained in the Delaney 
 multiplex ? 
 
 164. Why IS the relay used for multiplex working wound to so high 
 a resistance, and what is the object of the condenser which is joined 
 across its terminals ? 
 
 165. The signals are good on all arms of a hexode but one, where 
 none are received. What should yon expect the fault to be ? 
 
 166. If the marks sent on the distant first arm were received on the 
 second arm also, what fault would you look for ? 
 
 167. A multiplex sometimes works better in damp than in perfectly 
 fine weather. Why is this the case ? 
 
 ' 168. What is the object of the shunt placed across the terminals of 
 a sounder 1 
 
 169. In what position is it preferable to place the repeater on a long 
 telegraph line 1 
 
 170. What is the object of the automatic switch in a fast speed 
 repeater, and why are its coils shunted ? 
 
 171. How is the current obtained to work the local receiver on a 
 repeater i 
 
 172. What is the object of the leak resistance coils in a repeater f 
 
 173. Is it advantageous to the distant station's signals or not to 
 decrease the resistance in the leak circuit ? 
 
 174 How should the transmitting relays in a repeater be adjusted 
 to obtain the greatest speed t 
 
QussHohs. 193 
 
 175. If the contact points of the relays become oxidised, what will 
 be the result, and what steps would you take to prevent the points 
 becoming oxidised ? 
 
 176. Describe the action and object of the shunt which is placed 
 across the auto-switch magnet. 
 
 177. What is the purpose of the small condenser (spark condenser) 1 
 Describe its action in a repeater. 
 
 178. Describe how double currents are sent from the single current 
 keys afSxed to a repeater board. 
 
 179. Apparently signals are passing through all right to the distant 
 station, yet it is subsequently found that such is not the case. What 
 portion of the repeater apparatus is probably at fault and the 
 cause ? 
 
 180. Over a line of 400 miles long we get a speed of one hundred 
 words per minute, what would be the speed simplex if a repeater were 
 introduced at the centre of the line ? Answer : 400 per minute. 
 
 181. A wire will work duplex, but not quadruplex. A second wire 
 will work quadruplex but not hexode. A third wire will work simplex 
 but not duplex. What is the cause in each case ? 
 
 182. A current is received on a wire upon which there is a repeater. 
 What is the fault ? 
 
 183. Describe the general principle of the simplex and duplex fast 
 speed repeater. 
 
 184. Explain the principle of the bridge doplex and say what advan- 
 tages and disadvantages it possesses as compared with the differential 
 system. 
 
 185. Explain the object of the signalling condenser used on the 
 Bridge Duplex. 
 
 186. Describe the principle of the battery testing instrument. 
 
 187. Why is a difEereiitially wound tangent galvanometer required 
 for battery testing instruments 7 
 
 188. How should the constant of the galvanometer be taken ? 
 
 189. Explain the principle of the E.M.P. test employed in the battery 
 testing instrument. 
 
 190. How is the resistance of the galvanometer circuit kept at the 
 same value when the coils are shunted 1 
 
 191 Describe the general principle of the concentrator. 
 
 192. How is " time " sent on the concentrator J 
 
 193. What means are adopted in order that any instrument may be 
 used for any circuit no matter how short ? 
 
 194. Describe the arrangement required in order that the tele- 
 graphist may indicate to the switch clerk that he has finished. 
 
 195. Sketch a telegraph pole which is required to carry six wires, 
 the pole being in such a position that it requires staying. 
 
194 Qniestions. 
 
 196. Discnss the relative advantages of iron and copper for tele- 
 graph lines. 
 
 197. If yon know the dip and span of a telegraph wire, how coold 
 yon determine the dip required for another span so that the strain on 
 the wire may be the same in both cases 7 
 
 198. Discnss the relative advantages of srial (road, rail, and over- 
 hcnse) and underground wires. 
 
 199. How are wires erected and regulated ? 
 
 200. What is meant by earth wiring a telegraph pole and what is its 
 object ? 
 
 201. Describe the double shed insulator. 
 
 202. Elzplain how underground wires are renewed without interrupt- 
 ing the circuits. 
 
 203. How are copper wires secured to the insulators ? 
 
 204. What is meant by a strutt, an A pole and an H pole f 
 
 205. TS-BE 132 relayed at PB passes through GE is faulty between 
 Stafford and PB. TS-GW. 109 relay at LS passes through CE and is 
 faulty between CE and GW. How would you make one good wire 
 and show position of relay 2 
 
 206. How many cells and of what form would you employ to work 
 a double current sounder circuit fifty miles long 1 Answer : 22 small 
 Daniells (thirty would actually be used), 
 
 207. Why are wires crossed at the test box, 
 
 208. What point should be observed in selecting a wire for special 
 arrangements. 
 
 209. What is meant by "taking a wire," and " replacing it by another 
 wire," and what object is gaLaed 7 
 
 210. How many large Daniell cells (E.M.P, "96 volts internal resist- 
 ance 5iii) will be required to work a circuit requiring 16 m.a. 120 
 miles long. Besistance of wire 12tD per mile. Besistance of sending 
 galvanometer SOu. Besistance of distant apparatus 430(o ? Answer : 
 35 cells. 
 
 211. A circuit having a resistance of 680u requires a current of 
 16 m.a. How many large Daniells would be required and how many 
 would actually be used i Answer : 13 required, 15 used. 
 
 212. A wire passes through A, B, and, C, and is faulty between A 
 and B. How would you work a wire through A, B, and C, if a second 
 wire between A and (not passing through B) were available for the 
 purpose? 
 
 213. How could a contact be localised on a double current circuit 
 by requesting each office in turn to put his switch to " send " for a 
 short time 1 
 
 214. What is meant by dividing a through wire, and what is the 
 object gained ? 
 
Questions. 195 
 
 215. If you were able to chooae. between an open iron wire, an open 
 copper wire, and a covered copper wire for special work, which should 
 you prefer, and why ? 
 
 216. Give an example of a cross, and explain its object. 
 
 217. Why is iron used for telegraph lines in preference to copper ; 
 why is copper used for instrument coils ; and why are rheostats wound 
 with German silver wire 1 
 
 218. A wire fifteen miles long and 171 mils in diameter has a resist- 
 ance of 202u. What is the resistance of a wire six mils long whose 
 diameter is 121 mils ? Answer : 162w. 
 
 219. If a wire two miles long, weighing 6001bs. per mile has a resist- 
 ance of 18(n, what is the resistance of a wire ten miles long having a 
 total weight of one ton ? Answer : 241w. 
 
 220. Two wires each weighing 1501bs. are respectively half-a-mile 
 and one mUe in length. If the resistance of the finer wire be 6'o, what 
 is the resistance of the other 1 Answer . Vbia. 
 
 221. Two wires of equal length, weight respectively lOlbs. andl51bs., 
 what will be relationship between their resistance ? Answer as 15 : 10. 
 
 222. Describe an accumulator or secondary cell and state what 
 advantages they possess over primary batteries for universal working. 
 
 223. What is meant by the figure of merit of an instrument ? 
 
 224. What is the object of earth wiring ? 
 
 225 The signals sent on one instrument are received faintly on all 
 the other instruments at a small office. What is the fault ? 
 
 226. A current of 10 amperes flows through a resistance of 5 ohms 
 for 6 seconds, and another current of 6 amperes flows through a resistance 
 of 7 ohms ; during what time must the latter current flow in order that 
 the heat generated in the two cases may be the same? Answer : Hi 
 seconds nearly. 
 
 227. How would you compare the moments of two magnets ? 
 
EVERY TELEGRAPHIST, 
 
 And every person interested in Telegraphy and Telephony, 
 
 SHOULD BEAD 
 
 AND 
 
 Civil Service Recorder. 
 
 Organ of the Postal Telegraph Service, and the 
 ONLY Telegraph Paper Published in this Country. 
 
 THE BRIGHTEST AND BEST OF CIYIL SERVICE PAPERS. 
 
 Published every alternate Friday, Price 2d. 
 
 Special attention is paid to Technical Articles dealing in an 
 up-to-date fashion with Telegraphy, Telephony, and cognate subjects. 
 A series of Articles on the Baudot Telegraph System (French) is now 
 running. Although primary a Telegraph paper, the Chronicle devotes 
 a good deal of space to Postal matters, and it has readers and sub- 
 scribers in nearly every Post Office in the United Kingdom. 
 
 Post Free Subscription for Half Year (Tliirteen Issues), 2s. 6d. 
 
 Published by Page & Pratt, Ltd., 22, St. Andrew's Street, and 
 5, 6, AND 7, Ltjdgate CiRons Buildings, London, B.C.