bUSHCR & IMPORTER! i^Mt lrtuv nan* EM MEMORIAL George Davidson 1 Ql Professor of Geography University of California THE ELECTRIC TELEGRAPH. THE ELECTRIC TELEGRAPH. BY EGBERT SABINE, \\ FEL. SOC. ARTS, MEMB. BRIT. ASSOC., ETC. ETC. LONDON: VIRTUE BROTHERS & CO, 26, IYY LANE, PATERNOSTER ROW. 1867. LONDON : VIRTUE AND CO., PRINTERS, CITY ROAD. TK524.V TO 0. W. SIEMENS, ESQ., F.R.S., Memb. I.C.E., Memb. I.M.E., Etc. Etc. Etc. MY DEAR SIR, Allow me to dedicate this little Book to you. There are two reasons which urge me to do so : first, that there is no name more intimately associated with the progress of Telegraphy, both in England and on the Continent, than your own ; and secondly, that it gives me an opportunity of show- ing you how sensible I am of the many instances of kindness which I have experienced at your hands in affording me facilities for prosecuting my studies in this and other branches of applied science. Believe me, My dear Sir, Yery sincerely yours, EGBERT SABINE. London, January, 1867. JV1510967 PREFACE. THIS book was commenced with the intention of making it a purely Elementary Treatise. As he proceeded with his work, however, the Author became convinced that, in our language, the want is much more felt of such books as Schellen's, Dub's, Moigno's, Blavier's, and others, than of such as he had proposed to make his. He has therefore endea- voured to make it sufficiently elementary to come within the comprehension of every educated man, and, at the same time, sufficiently technical to be useful to electricians. The work is divided into Two Parts : the First being confined to a short history of the Electric Telegraph, and descriptions of many of the past and existing methods and apparatus ; the Second Part being confined exclusively to the more scientific matter relating especially to Cable "Work. JSTo new theories are started, nor has anything been introduced which expe- rience has not confirmed as having merits to recommend it. In conclusion, the Author begs to tender his most hearty thanks to Mr. Yarley, Professor Hughes, Mr. De Sauty, and others, for advice or information they have given him on PREFACE . points on which he was either doubtful or ignorant, and to Sir Charles Bright for the valuable curve of the resistance of gutta-percha at different temperatures. At the same time he has much pleasure in acknowledging the aid he has gleaned from the excellent works of Wiedemann, Dub, Schellen, Moigno, Blavier, De Castro, and others. CONTENTS. FIEST PART. SKETCH OF THE HISTORY AND PROGRESS OF THE ELECTRIC TELEGRAPH, WITH DESCRIPTIONS OF SOME OF THE APPARATUS. I. EARLY OBSERVATIONS OF ELECTRICAL PHENOMENA. Page 1. First Observations of Electricity by the Ancients ... 1 2. Early Systematic Inquiries 1 3. First Transmission of Electricity 2 4. Terms, " Electrics and Non-Electrics " 2 5. Terms, "Vitreous and Kesinous" 2 6. Franklin's Theory Terms, " Positive and Negative ". . 3 7. Winckler and Watson 3 8. Discovery of the Ley den-jar 3 9. Watson's Leyden-jar 4 10. Experiments in Transmission 5 II. TELEGRAPHS BY FRICTIONAL ELECTRICITY. 11. First Published Idea of an Electric Telegraph .... 6 12. Lesage's Telegraph 8 13. Pith-ball Electroscopes 8 14. Lomond's Telegraph 9 15. Reusser's Telegraph 9 16. Silva and Betancourt's Telegraph 10 17. Cavallo's Telegraph 10 18. Eonalds' Telegraph 10 CONTENTS. III. TELEGRAPHS BY GALVANIC ELECTRICITY. Page 19. Discovery of Galvanism 13 20. Yolta's Contact Theory Chemical Theory ...... 16 21. Volta's Pile 17 22. Properties of the Voltaic Pile , . . . 17 23. Decomposition of Water 18 21. Sommering's Telegraph . 18 25. Schweiger's Telegraph 20 26. Coxe's Telegraph .^ ...... 20 IV. TELEGRAPHS BY ELECTRO-MAGNETISM AND MAGNETO- ELECTRICITY. 27. Oersted's Discovery of Electro-Magnetism 21 28. Schweiger's Multiplier 22 29. Ampere's Telegraph . . . ..-'^-.i VI ..,.-.. 22 30. Eitchie's Telegraph . . . . . 23 31. Sturgeon's Discovery of the Electro-magnet ...... 24 32. Schilling's Telegraph * . . . 25 33. Volta-Electric Induction and Magneto-Electricity ... 25 34. Gauss and Weber's Telegraph . . . . .','..'.. 27 35. Steinheil's Telegraph .............. 32 36. Earth Circuit ...'".. . . 35 37. Wheatstone's Telegraph . . .:....,..,.,:...;. 36 38. ,, Alarm . . ... . ._,;...;. .... 38 39. Eelay ............... 40 40. Morse's Early Telegraph . . ... .,* ..,;. . . . 40 41. ,, Electro-Magnetic Telegraph . . ... . .., ;. 41 42. ,, Eelay . . ;..,..- ... ^ .: ; , .... . . . 42 43. Davy's Telegraph .. . . . 42 44. Wheatstone's Dial Telegraphs with Battery 43 45. ,, ,, Improved .... 45 46. ,, ,, Magneto-Electricity . 45 V. TELEGRAPHS NOW IN USE. 47. Cooke and Wheatstone's Single Needle Instrument . . 46 48. ,, ,, Double ,, , , . 50 49. Siemens' Self-acting Make-and- Break . . . '..'.... 51 50. House's Printing Telegraph . . . . . . ." . . . 53 61. Hughes' Eoman-type-printing Telegraph 55 52. Breguet's Electro-Magnetic Dial Instrument .... 65 53. Kramer's Pointer Telegraph 70 54. ,, Alarm Circuits 72 55. Siemens and Halske's Magneto -Electric Pointer Telegraph 75 56. ,, ,, Polarised Alarm . . 79 CONTENTS. XI Page 57. Wheatstone's Universal Telegraph 82 58. Simple Morse Circuit 84 69. Morse Eecorders 85 60. Siemens' (Morse) Embossing Instrument, with Movable Magnet 86 61. The Morse Alphabetical Code 87 62. ,, Transmitting Plate 90 63. ' ,, Apparatus with Eelay 92 64. Simple Morse Embosser for Two Stations with Eelay . . 93 65. Intermediate Station Commutators 95 66. Nottebohm's Commutator 96 67. Siemens and Halske's Commutator 98 68. Borggreve's 100 69. Commutator for Stations with Three or Pour Lines . . . 102 70. Battery Commutator 104 71. Translation . . . ., 104 72. Yarley's Apparatus for Translation 107 73. Static Induction in Submarine Lines Ill 74. Siemens' Submarine Key 113 75. De Sauty's 115 76. Varley's Switch for Submarine Work 117 77. John's Ink-printing (Morse) Apparatus 120 78. Digney's Ink-Eecorders 122 79. Direct Working Ink- Writers of Siemens and Halske . . 125 80. Arrangement of a Board with the Direct-working Ink- Writer 127 81. Morse Telegraph worked by Induction Currents . . . 131 82. Polarized Eelay 132 83. Plan of Connections 1 34 84. The Magneto -Induction Key 136 85. Siemens' Polarized Ink Eecorder 138 86. ,, ,, ,, used as a Submarine Koy 139 87. Complete Submarine Board 142 88. The Sounder 146 89. Burning Holes in Eecording-paper ........ 147 90. Morse Apparatus, worked by Closed Circuit . . . . . 148 91. Telegraphing in Opposite Directions in the same Line . . 154 92. Gintl's Method by Electro -Magnetism . . . V V . . i54 93. ,, ,, Chemical Telegraph . . .' ; . v '. . . 156 94. Frischen and Siemens' Method 158 95. Telegraphing in the Same Direction in the Same Line . . 160 96. Automatic Printing Telegraph 163 97. Siemens' Magneto -Electric Type Telegraph 166 93. Stoehrer's Double Style Telegraph 172 Xll CONTENTS. VI. ELECTRO-CHEMICAL TELEGRAPHS. Page 99. Bain's Chemical Telegraph 176 100. Pouget-Maisonneuve's Chemical Telegraph 179 101. Bonelli's Printing Telegraph 182 VII. OVERLAND LINES. 102. Posts . . . . . ...... 185 1 03. Line Insulators . . . .._... 189 104. Stretching Insulators . . . . . ^. ........ . 193 VIII. ATMOSPHERIC ELECTRICITY. 105. Lightning Guards Steinheil's 206 106. ,, Meissner's . 207 107. ,, Siemens' 208 108. ,, Breguet's. . , ;.-...;. . . 209 109. Fardley's . ........ . . 210 110. ,, Nottebohm's 210 111. Siemens and Halske's Point . . . 211 112. Breguet's Improved 211 113. ,, German . . . . . . .... 212 114. Kirchhoff's . . . ,. 213 115. Bianchi's Vacuum . 213 SECOND PART. ELEMENTS OF THE SCIENCE AND PRACTICE OF ELECTRIC TELEGRAPHY. I. ORIGIN OF THE GALVANIC CURRENT. 11. Polarisation of Metals in Liquids . . . . . .\. . . 215 2. Effects of the Galvanic Current , , '. . . . . . . 218 3. Batteries . . .' 220 4. Daniell's Constant Battery 221 5. Kramer's Modification of ditto 224 6. Meidinger's ,, 224 7. Siemens' ,, ,, .......... 226 8. Varley's ,, .'.-.-.-..- 227 CONTENTS. Xlll Page 9. Sand Batteries 228 10. Alum Batteries 228 11. Marie Davy's Proto- sulphate of Mercury Battery . . . 229 12. Terra- Voltaism - . ; 230 13. Amalgamated Zinc 231 II. MEASUREMENT OF THE GALVANIC CURRENT. 14. Voltameter 232 15. Galvanometers 235 16. Tangent Galvanometer . .'.'... '. . ... . . . 237 17. Sine ,, 241 18. Astatic Conditions of a Needle Pair 243 19. Magnetism of the Coils .' : .' .' . . ... . . . 245 20. Sine and Tangent Galvanometer 246 21. Weber's Reflecting Galvanometer 247 22. Thomson's ,, 249 23. Eheostats Wheatstone's . ;..'.. -._" . . . . . 251 24. ,, Jacobi's . .' : : ; . . . . . . . . 252 25. ,, Poggendorff's . . .'. . .' -;" . . . . 253 26. Siemens' Resistance Boxes : . . . . 253 27. Eisenlohr's Resistance Column 255 28. Ohm's Law , 256 III. CONDUCTING POWERS or MATERIALS. 29. Specific Conducting Powers . ..;.;...". . . . 264 30. Pure metals - *- 266 31. Influence of Temperature 266 32. Alloys ^ 268 33. .Effects of Annealing ' 269 34. Fused Metals . ^ . ..... . . . 270 35. Electric Permanency of Metals and Alloys . . . . . 273 36. Fluids . ... . . . . ... 276 37. Determination of Galvanic Polarisation 278 38. Insulating Substances 280 39. Influence of Temperature 282 IV. METHODS OF CABLE-MEASUREMENT. 40. Kirchhoff's Laws of Branch Circuits ...*'. ... 285 41. Wheatstone's Balance 290 42. Siemens' Apparatus for Testing Cables 294 XIV CONTENTS. Page 43. Copper Eesistance with both ends of Cable . . . ... 297 44. ,, one end only 298 45. Insulation by Bridge Method 299 46. ,, Deflection ,. . '., .. , . 299 47. Charge 301 48. Discharge 301 49. Constant of Sensibility of Galvanometer 303 50. Measurement of Electro -motive Force 305 ftl. British Association Balance 306 52. Balance by Bisected Wire 309 53. Determination of the Constants of Galvanic Elements . . 313 54. ,, of the Eesistances of ,, . . 313 55. ,, of the Electro- motive Forces of Elements . 315 56. Eegnault's Method 316 57. Fechner's ,, 317 58. Wiedemann's .... ., . ..... . . 317 59. Method with Galvanoscope 317 60. Wheatstone's Method . . . '. .....". .' . . . . 319 61. Ohm's ,, 320 62. Poggendorff's (Compensation) Method . "* '". . . . . 320 63. Comparative Electro -motive Forces of Metallic Pairs . . 324 Y. UNITS OF RESISTANCE. 64. Siemens' Mercury Unit 328 65. The Absolute Units 333 VI. SUBMARINE TELEGRAPH CABLES. 66. The First Cables ...... . . . . , V ; . . 341 67. The Old Atlantic Cable ...-.- 342 68. Eed Sea and Indian Cables . . . . 343 69. Malta and Alexandria Cable .344 70. Persian Gulf Cable . .- .- v .- -. .' ^V^i>'. 345 71. Copper- covered Cables .... .. * 346 72. New Atlantic Cable * * * . * - '&* - 348 73. Tests of Finished Core at the Gutta-percha Works . . . 350 74. Testing under Pressure . .-U '": ; '. 350 75. Measurements of Copper Conductor . . .- . . . . 353 76. ,, Insulation Eesistance 356 77. Insulation Test by Wheatstone's Bridge ... . . . 358 78. ,, Differential Method 358 79. ,, Deflection of Magnet 364 80. Insulation under Pressure 365 81. Electrification 367 CONTENTS. XV Page 82. Joints in the Core 368 83. Testing Joints 370 84. Self-heating of Cables 373 85. To Find the Locality of a Fault with both Ends .... 377 86. Eupture of Conductor . 379 87. De Sauty's Method of Finding Locality of Eupture . . 379 88. Varley's ,, ,, ,, ,, . . 381 89. Final Tests of Complete Cable 383 90. Insulating Materials 384 91. Absorption of Water by Insulating Materials 386 92. Siemens' Cables 387 93. The Cable in the Ship Paying-out Apparatus .... 388 94. Paying-out Taking Bearings 391 95. Soundings 391 96. Diagrams of the Bottom 392 97. Cable Descending to the Bottom 393 98. The Break 396 99. The Dynamometer 398 100. Electrical Operations during Paying -out 399 101. Determination of Place of Fault in Insulator .... 404 102. End Operations 408 103. Hipp's Method of Sealing Faults 408 104* Charge and Distribution along the Line 410 105. Inductive Capacity of Materials . . ., . . ... . 411 106. "Wippe," or Self-acting Make-and-Break 415 107. Eate of Working through Cables 419 THE ELECTRIC TELEGRAPH. PAET I. SKETCH OF THE HISTORY AND PROGRESS OF THE ELECTRIC TELEGRAPH, WITH DESCRIPTIONS OF SOME OF THE APPARATUS, I. EARLY OBSERVATIONS OF ELECTRICAL PHENOMENA. 1. THE phenomenon of electrical attraction produced by friction of bodies- was, in some instances, known to the ancients. It was first noticed about six hundred years before the Christian era, by Thales, the founder of Ionic philosophy. He observed that when amber was subjected to friction it acquired the power of attracting light substances, such as bits of feathers. On this account he was led to attribute to amber a species of vitality. The next mention we find is that of Theophrastus, who, three hundred years later, observed that a hard stone (supposed to be tourmaline), when rubbed, attracted straws and little pieces of sticks in its vicinity. Pliny, as well as other naturalists, both Greek and Roman, remarked, at different dates, the same phenomenon, which they regarded, in the spirit of the times, with superstitious reverence. 2. No systematic inquiry into the subject was undertaken until Dr. Gilbert, towards the close of the sixteenth century, 6 THE ELECTRIC TELEGRAPH. at the expense of much, pains, arranged and published a list of all those bodies in which he had observed the same property. Towards the middle of the seventeenth century Dr. Wall discovered the electric spark on rubbing a cylinder of amber with a piece of flannel. On approaching the cylinder with his linger, he obtained, for the first time, the spark, and noticed the noise which always accompanies it. Boyle and Otto Guericke added to the little stock of know- ledge then in hand, as well as Hawkesbee ; but their dis- coveries are a little out of the reach of Telegraphy. 3. The first discovery which we have on record of the power of transmitting the electric fluid to a distance through an insulated wire, is that of Stephen Grey, pensioner of the Charter House. Grey, having succeeded in electrifying a glass tube open at both ends, was desirous of finding out whether he could obtain the same result if he stopped up the ends with corks. This shows how at random the experi- ments were conducted at that date, and how little system had been introduced into these inquiries. But Grey's experiment succeeded, and he was surprised to find the corks also highly electrified. On presenting the corked ends of the tube to a feather, he found that the feather was first attracted and then repelled. This led him to infer that the electricity which the tube had acquired by friction passed spontaneously to the corks. From the communication of electricity from tubes to corks Grey was led to transmit it through strings and wires; and in 1727 we find him employing a wire 700 feet long, suspended in the air by silk threads, to one end of which he brought his excited glass tube, whilst another person at the other end observed the electrification. 4. After Grey, the subject was taken up by Desaguilliers, who instituted inquiries into the different conductibilities of bodies. The discoveries of Grey had caused the bodies operated on to be assorted into two classes, which Desa- guilliers proposed to distinguish by the names of "electrics" or non-conductors, and " non- electrics " or conductors. 5. On making experiments on the attraction of any light HISTORY AND PROGRESS. 6 substance by an electrified body, it had been observed by Grey that the former was repelled from the moment that it was itself electrified by contact. It was further remarked that when the electrified body was a rod of glass, the light body would be strongly attracted by a stick of resin also electrified by friction. It is not a settled question whether Symner it was, or Dufay, who in 1733 concluded, from the combination of these facts, the existence of two electricities. It was supposed that all bodies in their natural state con- tained an equal amount of each of these electricities in equilibrium ; but that from the moment this equilibrium was upset, and until it was re-established, the elements would divide themselves between the rubber and the rubbed body those identical with the electricity of a glass rod showing themselves in some bodies, and, in others, those of the same nature as the electricity of a piece of resin. This occasioned the former to be called vitreous electricity, and the latter resinous electricity. 6. Benjamin Franklin believed, however, in the existence of only a single fluid, and explained the phenomena by sup- posing that on exciting any substance till the equilibrium of the electricity was destroyed, an excess of it would be deposited on one side, and a deficiency, necessarily to the same amount, would occur on the other. Hence he gave the name of positive electricity to that which Dufay had called vitreous, and negative to that called resinous. Dufay, without the remotest idea of the transmission of signals for practical purposes, and with the pure curiosity of a physical experiment, made some capital attempts to ascer- tain the distance to which the electric attraction could be observed in an insulated wire. 7. Winckler, in Leipsic, and Lemonnier of Paris, in 1746, and Dr. Watson, Bishop of Landaff, in 1747, took up the same inquiry. 8. The discovery of the Leyden jar by Muschenbrceck, of Leyden, in 1746, came very opportunely for the experi- menters in the transmission of electric power. Muschenbrceck, struck by the escape of electricity into the B 2 4 THE ELECTRIC TELEGRAPH. air, which he attributed to the vapours and effluvia suspended in it, had determined on an experiment by which he sought to preserve some of the mysterious fluid, to keep it out of contact with the air. For this purpose he selected water as the recipient of the fluid, and a glass bottle as the best means of imprisoning it. On one occasion, happening to hold the bottle in his right hand, whilst he was charging the water contained in it by a wire leading to the prime conductor of a very powerful electrical machine, Muschen- broeck removed the wire with his left hand, and received a shock which his imagination probably led him to regard as much more terrible than it really was ; for, in a conversa- tion with Reaumur, he is reported to have said that he felt himself struck in his arms, shoulders, and breast, so that he lost his breath, and was two days before he recovered from the effects of the blow and the terror. " For the whole kingdom of France," added Muschenbroeck, "I would not take a second shock." 9. The Leyden jar such was the name given to it thence- forth was soon endowed with a more convenient form and Fig. i. became one of the chief instruments in the hands of the students of electricity. The form given to it by Watson resembled that shown in Fig. 1, in which a a is a coating of tinfoil upon the outer surface of a glass jar, b b an inner coating of the same material, and c a knob attached to a wire in connection with the inner coating. On charging the knob and inner coating HISTORY AND PROGRESS. 5 of the jar with, for example, positive electricity, the charge acts upon the natural electricity of the outer coating, which, during the operation, should be connected by a conductor with the earth, decomposes it, and repels the positive element, attracting and retaining the negative element on the outer coating. If the communication between the knob and the source of electricity be broken, the charge will remain accumu- lated on the inner coating of the jar ; and if a connection be then made between the knob and the outer coating by means of a wire or of a discharger, as shown in the figure, the opposite electricity accumulated on the coatings of the jar will rush towards each other through the conductor, producing, on its approach to complete communication, a spark of brilliant light. Suppose that, instead of the short discharger, a wire of several yards' length were employed, the effect would be the same. And it was virtually to ascertain the maximum length of this wire that formed the purpose of those re- searches of Grey, Desaguilliers, Watson, and others, to which we are indebted for the first suggestion of a telegraph. 10. "Watson, in 1747, stretched a wire across the Thames, over old Westminster Bridge. One end was fixed to the exterior coating of a Ley den jar, the interior coating being connected to earth through the body of the experimenter, and the other end held by a person who grasped an iron rod. The moment the latter dipped the rod into the river, both felt a shock. Subsequently, in the same year, Watson transmitted an electric discharge through 2,800 feet of wire and the same distance of earth at Stoke Newington ; and on the 14th of August, in the same year, repeated his experiments on a con- siderably larger scale, transmitting the electric impulse through 10,600 feet of wire suspended between wooden poles erected on Shooter's Hill. Franklin made similar experiments in 1748 across the Schuylkill, at Philadelphia, and Du Luc, about the same date, across the Lake of Geneva. But up to this time the experiments had been conducted 6 THE ELECTRIC TELEGRAPH. without a suspicion of the glorious results to which they were leading. And even in the hands of the ingenious and original Franklin, we do not find that the idea suggested itself to him to apply the power he found capable of being felt at the end of a wire of considerable length, to the com- munication of intelligence. II. TELEGRAPHS BY FRTCTIONAL ELECTRICITY. 11. In the Scot's Magazine for 1753* is a letter to the Editor, from a correspondent signing himself " C. M.," to whom we must give the credit of being the first who published the idea of applying electricity to the telegraph. This interesting communication is as follows : " To the Editor of the ' Scof* Magazine. 9 " Renfrew, Feb. 1st, 1753. " SIR, It is well known to all who are conversant in electrical experiments, that the electric power may be propagated along a small wire, from one place to another, without being sensibly abated by the length of its progress. Let, then, a set of wires, equal in number to the letters of the alphabet, be extended horizontally between two given places, parallel to one another, and each of them about an inch distant from that next to it. At every twenty yards' end, let them be fixed in glass, or jeweller's cement, to some firm body, both to prevent them from touching the earth or any other non-electric, and from breaking by their own gravity. Let the electric gun-barrel be placed at right angles with the extremities of the wires, and about one inch below them. Also, let the wires be fixed in a solid piece of glass, at six inches from the end ; and let that part of them which reaches from the glass to the machine have sufficient spring and stiffness to recover its situation after having been brought in contact with the barrel. Close by the supporting glass, let a ball be sus- pended from every wire ; and about a sixth or an eighth of an inch below the balls place the letters of the alphabet, marked on bits of paper, or any other substance that may be light enough to rise to the electrified ball ; and at the same time let it be so continued that each of them may reassume its proper place wiien dropped. All * Scot's Magazine, vol. xv. p. 73. The page is headed " An expeditious method of conveying intelligence." HISTORY AND PROGRESS. 7 things constructed as above, and the minute previously fixed, I begin the conversation with my distant friend in this manner: Having set the electrical machine a-going as in ordinary experiments, suppose I am to pronounce the word Sir : with a piece of glass, or any other electric per se, I strike the wire S, so as to bring it in contact with the ban-el, then , then r, all in the same way ; and my correspondent, almost in the same instant, observes these several characters rise in order to the electrified balls at his end of the wires. Thus I spell away as long as I think fit ; and my correspondent, for the sake of memory, writes the characters as they rise, and may join and read them afterwards as often as he inclines. Upon a signal given, or from choice, I stop the machine ; and taking up the pen in my turn, I write down whatever my friend at the other end strikes out. " If anybody should think this way tiresome, let him, instead of the balls, suspend a range of bells from the roof, equal in number to the letters of the alphabet, gradually decreasing in size from the bell A to Z ; and from the horizontal wires let there be another set reaching to the several bells ; one, viz. from the horizontal wire A to the bell A, another from the horizontal wire B to the bell J3, &c. Then let him who begins the discourse bring the wires in contact with the barrel, as before ; and the electric spark, breaking on bells of different size, will inform his correspondent by the sound what wires have been touched : and thus, by some practice, they may come to understand the language of the chimes in whole words, without being put to the trouble of noting down every letter. " The same thing may be otherwise effected. Let the balls be suspended over the characters as before, but instead of bringing the ends of the horizontal wires in contact with the barrel, let a second set reach from the electrified cable, so as to be in contact with the horizontal ones ; and let it be so contrived at the same time, that any of them may be removed from its corresponding horizontal by the slightest touch, and may bring itself again into contact when set at liberty. This may be done by the help of a small spring and slider, or twenty other methods, which the least ingenuity will discover. In this way, the characters will always adhere to the balls, excepting when any one of the secondaries is removed from contact with its horizontal ; and then the letter at the other end of the horizontal will immediately drop from its ball. But I mention this only by way ,of variety. " Some may, perhaps, think that, although the electric fire has not been observed to diminish sensibly in its progress through any 8 THE ELECTRIC TELEGRAPH. length of wire that has been tried hitherto, yet, as that has never exceeded some thirty or forty yards, it may be reasonably supposed that in a far greater length it would be remarkably diminished, and probably would be entirely drained off in a few miles by the sur- rounding air. To prevent the objection, and save longer argument, lay over the wires from one end to the other with a thin coat of jewellers' cement. This may be done for a trifle of additional expense ; and as it is an electric per se, will effectually secure any part of the fire from mixing with the atmosphere. I am, &c., " C. M." This is one of the most interesting documents to be found in the whole history of Telegraphy. The writer was, evidently, not acquainted with Watson's experiments, or he would not probably have suggested insulation by "jewellers' cement ; " but the suggestion was an ingenious one. The idea which we find of keeping his lines charged with electricity, and giving the signals by discharging them, as well as that of reading signals by sound of bells, both of which, long years afterwards, were brought, with certain modifications, into practice, deserve to be remembered to his credit. 12. To Lesage,* however, belongs the honour of having established, in practice, the first telegraph wire for the trans- mission of intelligible signals. His system was almost the realisation of the idea of the Scotchman, " C. M." He erected at Geneva, in 1774, a telegraph line of twenty- four metallic wires, insulated from each other. Each wire was connected at the further end to a separate pith-ball electroscope, and corresponded to one of the letters of the alphahet. In this way any letter could be indicated hy bringing to the end of the wire corresponding to the letter to be sent, a source of static electricity produced by friction, which would imme- diately cause the divergence of the pith balls of that par- ticular electroscope. 13. The electroscopes used in these experiments consisted of two small pith balls suspended from a common metallic support, by cotton threads or fine wires. It will, without * Moigno's " Telegraphic Electrique," p. 59. HISTORY AND PROGRESS. 9 further explanation, be evident from what has gone before, that in charging the system, shown in equilibrium at a, Fig. 2, the two pith balls, having the same kind of electricity, would repel each other and assume a position similar to that shown at b in the same figure. This would continue as long as the charge lasted. The balls would, in course of time, however, approach each other again by their own gravity, the escape of their electricity into the surrounding air diminishing the repelling force. This could, and perhaps was, effected by Lesage suddenly, by discharging his line as soon as he had given a signal ; in other words, by letting the line and pith balls reassume their state of elec- trical equilibrium. 14. Lomond, in 1787, by the em- ployment of a delicate electroscope, g> 2 ' and combinations of signals, given by the divergence of pith balls, succeeded in transmitting intelligence with the aid of a single line wire. A short account of this invention is given by Arthur Young,* in the following words : " M. Lomond has made a remarkable discovery in electricity. You write two or three words on a paper ; he takes it into a room, and turns a machine enclosed in a cylindrical case, at the top of which is an electrometer, a small fine pith ball ; a wire connects with a similar cylinder and electrometer in a distant apartment ; and his wife, by remarking the cor- responding motions of the ball, writes down the words they indicate, from which it appears that he has formed an alphabet of motions. As the length of the wire makes no difference in the effect, a correspondence might be carried on at any distance, within or without a besieged town, for instance, or for objects much more worthy of attention and a thousand times more harmless." 15. In 1794, Reusser proposed, in the Magazin de Voigtfi the construction of a telegraph by means of electrical dis- * "Travels in France," vol. i. p. 979, 4th edition. 1787. f Magazin de Voigt, vol. ix. p. 183. 10 THE ELECTRIC TELEGRAPH. charges passing over the parts of a broken conductor enclosed in a glass tube, or by letters formed by spaces cut out of parallel strips of tinfoil pasted on square plates of glass. Such letters are shown in Fig. 3. An electric dis- charge from the interior coating of a Ley den jar, being sent, for instance, through the double strips of tinfoil, from the end marked -f to the end marked connected with the outer coating of the jar, a spark would pass over each of the intervening spaces at the same time, and the letter would . appear beautifully illumin- , ^^^^^ | ated in the dark. This ex- i ^"^^ CL""" *""""" periment of Reusser forms a yery common p '-'-- ' illustration of tension elec- tricity in lecture rooms. Reusser further suggested to call the attention of the observer, at the distant station, by firing an electric pistol by means of the spark. 16. In Spain, about the same time, Don Silva read a paper before the Academy of Sciences of Madrid, on a system of telegraphing with a single wire, by means of continuations of sparks, said, by the Magazin de Voigt, to have been carried out, two years later> with no small success, by the Infanto Antonio ; and Betancourt stretched a single line in the air, over a space of twenty-seven miles, between Madrid and Aranjuez. He employed a battery of Leyden jars and received signals by observing the divergence of suspended pith balls. 17. Cavallo was the next who strove to attain the perfec- tion of a telegraph by means of frictional electricity. In 1795 he published his " Traite d-'filectricite," in which he gives descriptions of his systems of electric signalling and communication. He proposed to transmit letters and nu- merals by combinations of sparks and pauses. His electric alarm was based upon the explosion of a mixture of hydrogen and oxygen gases or of gunpowder by the electric discharge. 18. It is necessary here to depart a little from historical order, to mention the last and most ingenious invention of a telegraph worked by frictional electricity ; this was the HISTORY AND PROGRESS. 11 invention of Mr. Ronalds, of Hammersmith. For the pur- poses of experiment he erected a line, eight miles long, insulated by silk and dry wood, in his garden, and also buried a considerable length of wire, insulated in glass tubes, encased in pitch and wood, in the earth. This was in 1823. For the following description of the invention we are in- debted to Mr. E. Highton's* book : Ronalds employed an ordinary electric machine and the pith-ball electrometer in the following manner. He placed two clocks at two stations ; these two clocks had upon the second-hand arbour a dial with twenty letters on it ; a screen was placed in front of each of these dials, and an orifice was cut in each screen, so that one letter only at a time could be seen on the revolving dial. The clocks were made to go isochronously, and as the dials moved round, the same letter always appeared through the orifices of each of these screens. The pith-ball electrometers were hung in front of the dials. It is evident, therefore, that if these pith balls could be made to move at the same instant of time, a person at the transmitting station, by causing such motion in both those electrometers, would be able to inform the attendant at the distant or receiving station what letters to note down as they appeared before him in succession on the dial of the clock. This was accomplished in the following manner. The transmitter caused a current of electricity to be constantly operating upon the electrometers, except only when it was required to denote a letter, and then he discharged the electricity from the wire, and instantly both balls collapsed. The distant observer was thereby informed to note down the letter then visible. In this way letter after letter could be denoted, words spelt, and intelligence of any kind trans- mitted. All that was absolutely required for this form of telegraph was, that the clocks should go isochronously during the time that intelligence was being transmitted ; for it was easy enough by a preconcerted arrangement between the * " The Electric Telegraph," by E. Highton, p. 50. 1852. IxJ THE ELECTRIC TELEGRAPH. parties, and upon a given signal, for each party to start their clocks at the same letter, and thus, if the clocks went to- gether during the transmission of the intelligence, the proper letters would appear simultaneously, until the communica- tion was finished. The attention of the distant observer was called by the explosion of gas by means of electricity from a Ley den jar. Fig. 4 shows an elevation of the apparatus, in which D is an electrical machine, B a pith-ball electrometer, A the screen Kg. 4. hiding the letters on the dial-plate except the one seen through the orifice, F the gas alarm, and E the tube convey- ing the wires from the station. Fig. 5 shows the dial ; and Fig. 6 the same with the screen before it, and the pith-ball electroscope. HISTORY AND PROGRESS. 13 Too mucli credit cannot be given to all these men for their energy in struggling, with the imperfect means and small experience at their command, to realise an end which the nature of the electricity they employed rendered impossible. And if we can award to none of them the title of the in- ventor of a practicable telegraph, we must, at least, give them credit for having fully appreciated its importance, and Fig. 5. Fig. 6. for having dedicated their energies to the accomplishment of the task they set themselves, and persevered in the face of many sad difficulties and disappointments. III. TELEGRAPHS BY GALVANIC ELECTRICITY. 19. A new discovery that of galvanic electricity had for some years occupied the attention of the scientific world, and at the beginning of the present century the thoughts of students of electricity were directed to its application for the purposes of Telegraphy instead of the unmanageable frictional electricity with which they had hitherto had to content themselves. The first mention we find of this species of electricity is in the recital which Sulzer published, in 1767, of the follow- ing experiment. On taking two pieces of different metals, and placing one of them (zinc, for example) above, and the other (perhaps copper) underneath his tongue, he found 14 THE ELECTRIC TELEGRAPH. that, so long as the metals did not make contact with each other, he felt nothing ; but that when the edges were brought together over the tip of his tongue, the moment contact took place, and during the time it lasted, he experienced an itch- ing sensation, and a taste resembling that of sulphate of iron. If he changed the relative positions of the metals, he experi- enced a different sensation, which he found difficult to describe. Sulzer supposed that the contact of the two metals occa- sioned a vibration of their particles, which, acting on the nerves of the tongue, produced the taste in question. The next to whom chance afforded an opportunity of making the discovery of galvanism, but who let it pass with as little profit as Sulzer had done, was a student of medicine in Bologna. He was once occupied, we are told, in dissect- ing a mouse which he held in his hand, when, having touched one of the nerves with his scalpel, he felt a shock resembling that produced by electricity.* In 1790 Madame Galvani, wife of the professor of anatomy at Bologna, being attacked with a slight cold, her physician prescribed her//Y><7 broth. Frogs were provided for the purpose, skinned, washed, and laid upon a table in the laboratory of the professor to await the moment when they were to undergo the culinary operation. Whether this operation was to be performed in the laboratory, is not said ; it is certain, however, that Madame Galvani was there with one of the professor's assistants, who was at the moment engaged in some ex- periments with a large electrical machine which stood upon the same table. Whenever the assistant, in the course of his experiments, took sparks from the conductor of the machine, Madame Galvani was astonished to observe a twitch- ing resembling life in the limbs of the dead frogs. This circumstance excited the lady's curiosity in the highest degree, and she related her observation to her hus- band, who immediately repeated the experiment, and found * " Essai sur 1'Histoire G-enerale des Sciences, pendant la Revolution Francaise," par J. B. Biot, p. 9. HISTORY AND PROGRESS. , 15 the convulsions return whenever he took sparks from the machine.* The professor, who was luckily not more learned in elec- trical science than in some of the other branches of physics, was unable to give the explanation of the phenomenon, with which he would probably have contented himself, and have let the matter drop? had he been an experienced electrician. But he was struck by the novelty of the new fact, and he determined to follow it up. Galvani thenceforth prosecuted his studies and experi- ments on the electricity of animals, with perseverance, and chance rewarded him for his industry by again coming to his aid. In his experiments on the electricity of frogs, he had occasion to separate the lower parts of the bodies from the upper. Having prepared a frog in the ordinary manner, on one occasion he remarked that on hanging it up to an iron balustrade of his house, by a hook of copper wire passed through part of the dorsal column remaining above the junc- tion of the thighs, it all at once underwent a series of lively convulsions. The professor was more than ever astonished, for there was this time no electrical machine to account for the appearance, and he was compelled to take refuge in the hypothesis of what he called " animal electricity," supposing opposite kinds of electricity to exist in the muscles and nerves of the animal. In this hypothesis he regarded the muscles and nerves as the charged coatings of a Leyden jar. It is worthy of remark and regret that Galvani should have so thoroughly mistaken the important part played in the affair by the two metals, iron and copper. He regarded these, however, only in the light of a compound conductor, through which the opposite electricities, assumed to exist in the nerves and muscles, discharged themselves. It is not a rare thing in the annals of science that mere chance has suggested some great discovery. But we seldom * " Aloysii Galvani de Viribus Electricit. in motu musculari Com- mentar.'' p. 2. 16 THE ELECTRIC TELEGRAPH. meet with chances so favourable to advancement as those which directed the attention of Galvani to the study of animal electricity by a phenomenon of electroscopic sensi- bility in the nerves of a frog ; his fortunate ignorance, which, combined with his ardent imagination, caused him to form hypotheses only excellent because, in being stubbornly sup- ported by himself, he added to the stock of facts, through his many and varied experiments, and in occasioning a discus- sion in which his views were successfully refuted by the masterly intellect of Alexander Volta. "In der Beobachtung einer anfangs isolirt stehenden Erscheinung liegt oft der Keim einer grossen Entdeckung," says Humboldt ; * and Galvani found it so, and deservedly earned his title of a pioneer in science. 20. Immediately after the publication of Galvani's dis- covery and hypothesis, Alexander Yolta, professor of physics at Pavia, occupied himself with an inquiry into the causes of the frog phenomenon, and was not long in perceiving a want of basis in Galvani's theory. With much penetration Yolta recognised the intrinsic elements in the complicated appear- ance which Galvani had discovered, and sought, with success, to produce the same by substituting other materials for the frogs and other animal bodies. He contended that the two metals, copper and iron, in the experiment of Galvani, were the real electromotors, and that the muscles of the dead frogs only played the part of a moist conductor in completing the circuit. Yolta was of opinion that the simple contact of two dissimilar metals was sufficient to develop electricity, and that the strength of the electricity excited depended upon the nature of the metals. This was vigorously opposed by the partisans of Galvani, who held tenaciously to the doctrines of their master, and a scientific war of opinions ensued between the schools of Pavia and Bologna, out of which Yolta came victorious even before he had completely verified his sagacious conjectures by experimental proof. If the tongue be applied to the conductor of an electric machine which is being turned, an acid or alkaline taste * Cosmos. Einleitende Bemerkungen. HISTORY AND PROGRESS. 17 will be perceived according as the conductor is being charged with positive or negative electricity. The similarity of these results with those obtained by Sulzer was an analogy advanced by Yolta in support of his contact theory. Another theory, and that now very generally accepted, was first suggested by Fabroni, and is known in contradistinc- tion as the chemical theory. This theory regards chemical decomposition as necessary to the development of the voltaic current. A discussion of the arguments advanced in support of these two theories would be out of place here. M tiller has devoted an excellent chapter to the subject in one of his books.* The German physicists for the most part hold out for the contact theory, whilst the French and English generally accept the chemical as the most rational and as that affording the most satisfactory explanations of known phenomena. 21. Volta, in 1800, wrote a letter from Como to Sir Joseph Banks, President of the Royal Society, in London, stating that he had found a means of augmenting, at pleasure, the development of galvanic electricity. This he had accom- plished by placing upon a plate of glass first a disc of copper, then on this a disc of zinc, and over these a similar sized disc of damp cloth ; and by continuing to pile up discs of these materials, in the same order, copper, zinc, cloth, until he had a sufficient number. He then connected wires to the lower and upper plates. 22. This apparatus is known as the Yoltaic pile.f Its pro- perties are concisely stated as follows : 1st. It communicates a charge of positive electricity to a condenser in connection with the wire attached to the last zinc disc, when the last copper disc is connected with the earth. 2nd. It communicates a charge of negative electricity to the condenser when the poles are reversed, that is to say, when the zinc of the upper part is put into contact with the earth, and the condenser with the copper disc at the bottom. These * " Fortschritte der Physik," vol. i. p. 225. f The more recent forms given to the pile, as at present employed in tele- graphy, are explained in the second part. 18 THE ELECTRIC TELEGRAPH. experiments may be repeated ad infinitum even after the pile has been mounted some hours, provided the cloth retains some moisture. 3rd. It produces chemical effects with an energy propor- tional to the number of elements accumulated. The zinc disc which forms one of the extremities of the pile, being that which, in communication with the condenser, gives a positive charge, has been called the positive pole, and the copper plate at the bottom of the pile the negative pole. 23. Immediately after the receipt of Yolta's letter, by the President of the Royal Society, a pile was constructed on this principle by Mr. Nicholson (the conductor of Nicholson's Journal) and Sir Anthony Carlisle. A drop of water being, on one occasion, used by them to make a good contact between the conducting wire and a plate of metal with which they were experimenting, Carlisle observed a disengagement of gas from it. Further experiments discovered very shortly the decomposition of water by the electric current. Thus was chance once more on the stage in promoting electrical discovery, and this time the magnificent investigations of Humphrey Davy were the result. 24. In the year 1808, Herr S. T. Sommering communi- cated to the Academy of Sciences in Munich his invention of a system of telegraphing based upon the discovery of the British chemists, Nicholson and Carlisle,* that water is de- composed into its constituents of oxygen and hydrogen by the voltaic current. At the station which was to receive signals were arranged, in a narrow vessel of water, thirty-five glass tubes, each con- taining a gold point, twenty-five marked with letters of the alphabet, nine with numerals, and one with a zero. From each point an insulated wire was led to a metal terminal at the transmitting station. To send a signal it was only necessary to bring the two poles of a voltaic pile to two of the terminals in question. The current passing from one terminal traversed its line wire to the voltameter at the receiving station, where it * "Galvanism," by Sir W. S. Harris, p. 35. HISTORY AND PROGRESS. 19 passed between the gold points corresponding to the ter- minals touched by the poles, and returned through the other line wire to the terminal of the other pole of the pile. In doing this, bubbles of hydrogen appeared at the gold point in communication with the positive pole, and bubbles of oxygen at the other one. Thus two signals were given simul- taneously, to which that of the hydrogen took precedence. When it was desired to indicate only one letter, the positive pole of the battery was brought in connection with zero and the negative with the letter to be transmitted. The con- ABODE FGHIJKLMNOPQRS TUVVVYZ 1234567 89 OH ABCDEFGHIJKLMNOPQRSTUVW1Z12 3456 78 90 Fig. 7. nection wires were well insulated, and, at a little distance from the transmitting terminals at one end, and from the voltameters at the other, were bound up in the form of a cable. Sommering proposed to call the attention of the receiving station by liberating an alarm by means of accu- mulated gas. Fig. 7, a copy of that given by Sommering in his descrip- c2 20 THE ELECTRIC TELEGRAPH. tion of this telegraph, will illustrate the foregoing. A A is a sectional view of the glass reservoir. A B c .... 8 9 are the thirty-five gold points of the voltameter arrangement, passing through the bottom of the glass reservoir. The lower ends of the thirty-five gold points were soldered to insu- lated copper wires passing through the tube E to the trans- mitting station. Here the wires, marked with the respective letters and numerals, were insulated on a wooden support, K K. The ends of the terminals were furnished with holes, i, to receive the poles B and c of the voltaic pile p N. The construction of this telegraph the first in which voltaic electricity was employed involved an outlay which, in conjunction with the slowness of working, would have prevented its commercial utility even if the science had not advanced by any of those astonishing strides which marked its progress a few years afterwards. 25. An improvement on the system of Sommering was published by Schweiger, of Halle, in an appendix to his memoir of Sommering. He suggested that, for the alarm, it would be possible to employ a pistol by the connection of a battery to the pile. In addition to this, he proposed to diminish the number of wires used in Sommering' s telegraph to two, by using two galvanic piles of unequal power, so that the amount of gas given off in a certain time by the one battery would be much more than by the other, and by varying the time of development of the gas and of the inter- vals, he proposed to form a code of signals. 26. About the same time that Sommering invented his telegraph, the same system was suggested by Professor Coxe, of Pennsylvania, and described by him in a paper published in Thomson's "Annals of Electricity," 1810. Coxe had the idea also of telegraphing by means of the decomposition of metallic salts. His systems were, as he described them, considered, however, impracticable. f HISTORY AND PROGRESS. 21 IY. TELEGRAPHS BY ELECTRO-MAGNETISM AND MAGNETO- ELECTRICITY. 27. The power of lightning to weaken the magnetism of the compass needle, and even sometimes to reverse its polarity for a long time, suggested the suspicion of a near relation between electricity and magnetism. The definite discovery was, however, not made until, in 1820, Professor Oersted, of Copenhagen, found that a magnetic needle, sus- pended in the neighbourhood of a wire in which a current of electricity was passing, was deflected from its position of rest. Ampere made experiments, and found the law by which this influence was governed, and which he briefly expressed as follows : " Imagine a human figure in the direction of a conductor through which a positive current is flowing upwards, the figure will have the north pole on the left hand if its face be turned towards the needle." Thus, if a positive current pass along the upper wire in the annexed figure (Fig. 8) from a towards b, the magnetic needle N s, suspended in its neighbourhood, will be deflected, and take up a position indicated by the dotted line n s, at nearly right , angles to the wire. When the current is reversed the poles will be deflected a/ s ^^ 1' in the other direction, "&*& 8 - the pole n being where s is, in the figure. It will become evident on regarding the figure also that, if a similar current pass in the lower wire from b' to a', it must produce the same magnetic direction upon the needle as a current in the upper wire would from a to b. In the case of the upper wire the observer's head is supposed to be near b and his feet near a, the current passing upwards in the direction of the arrow, and the north pole is found on his left hand when he faces the needle. In the case of the lower wire, however, the direction of the current is reversed, and the 22 THE ELECTRIC TELEGRAPH. position of the observer must be supposed to be reversed also head at a, feet at b'. While . he faces the needle the north pole is still found on his left hand. When currents pass, therefore, in both wires at the same time in opposite directions, they act in the same sense on the magnetic needle N s, and, other things being equal, their combined force is double that of a single wire. The same would be reached by joining b and b' by a wire, and letting a current of equal strength pass from a to b, b' } and a. 28. Professor Schweiger, of Halle, the same who suggested improvements of Sommering's telegraph, soon after the discovery of electro-magnetism by Oersted, invented an apparatus based on this principle, which he made by coiling a wire several times round a magnetic needle, and found that the deflect- ing force increased with the number of turns. Such an apparatus, called an electro-mag- iietic multiplier, is shown in Fig. 9. It has since become one of the most essential instruments for the measurement and indication of galvanic electricity, 29. The brilliant discovery of electro-magnetism* was speedily followed by attempts to employ it for the telegraph, which made, from this time, gigantic progress towards its present state of perfection. The idea of substituting magnetic needles suspended in multipliers of wire in place of the voltameters of Sommering * It is said that the same discovery had already been made in 1802, by Grrandominico Romagnosi, of Trent, and made known in a book entitled "Manuel du Galvanisme," published in Paris in 1805. If it be true that Romagnosi discovered the deflection of a magnetic needle by the current, the discovery could have excited no interest whatever, and must have been known within a very limited circle, as the discovery of Oersted in 1820 was imme- diately hailed as a landmark in science. HISTORY AND PROGRESS. 23 first to have occurred to Ampere, who explains it in a paper read before the Academy of Sciences at Paris, in October, 1820.* He says that by means of the same number of magnetic needles and line wires as there are letters of the alphabet, and with the help of a voltaic battery whose poles could be brought in connection, one after the other, with the ends of the wires, a telegraph might be produced by which all possible communications might be made to a person at a distance off, who was charged to observe the needles. If a keyboard whose keys were each marked with a letter of the alphabet were adapted to the battery, so that on pressing down the key of any letter the circuit corresponding to that letter would be closed, correspondence could be carried on with ease, and would only require the time necessary to press down the keys at the one station, and to read off the letters from the deflected needles at the other. This telegraph, as imagined by Ampere, was, however, doomed to the same fate as that of Sommering, of never coming into practice, and for the same reasons, principally the number of line wires. Had Ampere combined his system with that which Schweiger proposed of reducing Sommering's telegraph to two wires, or with any other using a code of signals, the problem of the electric telegraph would have been solved from the year 1820. But two serious inconveniences the irregularity of the piles, and, above all, the rapid decrease of their intensity only permitted the application of this great idea on a small scale. 30. Ritchie, however, carried out a really excellent modi- fication of Ampere's invention by encircling thirty magnetic needles with coils of wire ; each needle was furnished with a small screen, so that when it was unaffected by a current, the screen covered over a letter of the alphabet, which was exposed as soon as the needle was deflected. This telegraph was first exhibited in public, some years later than the date of its invention, by Mr. Alexander, of * " Annales de Physique et de Chemie," vol. xv., p. 72. 24 THE ELECTRIC TELEGRAPH. Edinburgh. He divided his thirty wires into twenty-six letters of the alphabet, three signs of punctuation, and one asterisk for indicating the end of a word. The return circuit was formed by a single wire. 31. In 1825, Mr. Sturgeon, of London, discovered that when a soft iron bar is surrounded by a helix of wire, through which a galvanic current is passing, it acquires a considerable quantity of magnetism, which lasts as long as the current continues in the coil. In this way he constructed some powerful mag- nets. A form which he made, and which acquired an immense lifting power, is shown in Fig. 10. For this purpose, pieces of soft iron were bent in the form of a horse shoe, round the horns of which he wound spirally a length of well insu- 10 - lated copper wire. One end of the magnet so arranged became a north pole and the other a south pole if the spiral wire were wound in the same direction through- out, supposing the horseshoe to be bent straight. The positions of the poles depend, of course, upon the direction in which the spirals are wound, and upon the direction in which the current traverses them, according to the same law as that by which Ampere expressed the posi- tions of the poles of deflected magnets. In the coil, Fig. 11, for example, the positive current descending between the observer and the soft iron bar, the spiral being right-handed, the north pole would be on the right hand of the observer. 11 - In order to increase the power of electro-magnets the wire has to be wound several times round each of the horns. This is generally done by means of bobbins of wire which can be removed at pleasure. Such magnets have been made with iron cores 3 inches and more in diameter, and over a foot in length, which have carried nearly a ton weight. It is strange that this discovery, HISTORY AND PROGRESS. 25 which has since proved so important in practical telegraphy, was not made use of in any system until some years later. 32. The next important step towards the perfection of present telegraphy was that made in 1832 by Baron Schilling von Cronstadt. In some of the accounts Of his system we read that it consisted of a certain number of insulated platinum wires united by a silk cord, which set in motion, by means of a sort of piano, five magnetic needles placed in a vertical position within coils of wire. According to other accounts, he employed only one magnetic needle and multiplier, with two leading wires, and was enabled, by means of a combina- tion of the deflections of this needle to the right and left, by changing the poles of the battery at the ends of the two wires, to give all the signals necessary for a complete correspon- dence. His call signal was given by means of an alarm. Schilling executed models of his apparatus, which were ex- hibited before the Emperor Alexander, and still later before Nicholas. He, however, unhappily died before carrying out his invention in practice. The probability is that he sug- gested both the systems of five and single needle instruments ; the latter as an improvement on the former. 33. We owe to the enterprising genius of Michael Faraday the two discoveries not less important in physics than useful in relation to the telegraph Yolt a- electric induction, and magneto- electricity . These were no chance discoveries; and in this they differ from those made by Galvani, who stumbled over his pheno- mena ; they were results of profound consideration ; Faraday anticipated his discoveries before he made them. He says, in his " Experimental Besearches," * " Certain effects of the in- duction of electrical currents have already been recognised and described ; as those of magnetism ; Ampere's experiments of bringing a copper disc near to a flat spiral ; his repetition, with electro-magnets, of Arago's extraordinary experiments, and perhaps a few others. Still it appeared unlikely that these could be all the effects which induction by currents could produce. . . . These considerations, with their conse- * "Experimental Researches in Electricity," 1st series, vol. i. p. 1. xJt) THE ELECTRIC TELEGRAPH. quence, the hope of obtaining electricity from ordinary mag- netism, have stimulated me at various times to investigate experimentally the inductive effects of electric currents. I lately arrived at positive results, and not only had my hopes fulfilled, but obtained a theory which appeared to me to open out a full explanation of Arago's magnetic phenomena, and also to discover a new state which may probably have great influence in some of the most important effects of electric currents." The first successful experiment of Faraday was made with 203 feet of copper wire, coiled in one length on a wooden bobbin, and a similar length interposed on the same bobbin between the turns of the first coil. The wires were insulated from each other by twine. The ends of one of the coils were connected to the two ends of a multiplier with a finely sus- pended magnetic needle a galvanometer ; and the ends of the other coil with a battery of one hundred pairs of four- inch square plates with double coppers. When contact was made with the battery, there was, says Faraday, a sudden and very slight effect at the galvanometer, and there was also a similar slight effect when the contact with the battery was broken. But during the continuance of the voltaic current through the one helix, no galvanometer appear- ances, nor any effect like induction upon the other helix, could be perceived. Faraday began these experiments in 1831 and continued them during following years. He found that not only were currents induced in helices by induction of others in their neighbourhood in which currents were passing, but that on inserting the end of a permanent magnet into the middle of a helix of wire, a current of electricity was gene- rated whose direction depended upon the pole inserted and the end of the spiral with regard to the direction of its windings. He says, " If such a hollow helix as that described be laid east and west, or in any other constant position, and a magnet be retained east and west, its marked pole always being one way, then, whichever end of the helix the magnet goes in at, and, consequently, whichever pole of the magnet enters first, still the needle is deflected the same way : on HISTORY AND PROGRESS. 27 the other hand, whichever direction is followed in withdraw- ing the magnet, the deflection is constant, but contrary to that due to its entrance." 34. In 1833 Schilling's proposition of the manner of giving signals with a single needle was carried out in a more complete form by the Go'ttingen physicists, Gauss and Weber. Their telegraph consisted of a single magnetic needle surrounded by a multiplier of wire, the needle being moved, however, by magneto-electricity instead of galvanism. This was the first employment of Faraday's discovery in the service of Telegraphy. We read, in relation to this telegraph, in a report of the magnetic observations of these physicists,* the following : " There is, in connection with these arrangements, a great, and until now, in its way, novel project, for which we are indebted to Professor Weber. This gentleman erected, during the past year, a double- wire line over the houses of the town (Gottingen), from the Physical Cabinet to the Observatory, and lately a continuation from the latter build- ing to the Magnetic Observatory. Thus an immense galvanic chain (line) is formed, in which the galvanic current, the two multipliers at the ends being included, has to travel a distance of wire of nearly 9,000 (Prussian) feet. The line wire is mostly of copper, of that known in commerce as ' No. 3,' of which one metre weighs eight grammes. The wire of the multipliers in the magnetic observatory are of copper, 'No. 14,' silvered, and of which one gramme measures 2-6 metres. This arrangement promises to offer opportunities for a number of interesting experiments. We regard, not without admiration, how a single pair of plates, brought into contact at the further end, instantaneously communicates a movement to the magnet-bar, which is deflected, at once, for over a thousand divisions of the scale." And further on, in the same report : " The ease and certainty with which the manipulator has the direction of the current, and therefore the movement of the magnetic needle, in his command, by means of the communicator, had, a year ago, suggested experiments of * Pogg. Ann., 32, p. 568 ; and " Dingler's Journal," 55, p. 394. THE ELECTRIC TELEGRAPH. an application to telegraphic signalling, which, with whole words and even short sentences, completely succeeded. There is no doubt that it would be possible to arrange an uninter- rupted telegraph communication in the same way between two places at a considerable number of miles distance from each other." The purpose of setting up this aerial line was not for the study of telegraphy, nor for the perfection of telegraph apparatus ; but to enable these physicists to institute in- quiries into the laws of the intensity of galvanic currents, under different circumstances, on a large scale. At the same time the lines were used for regulating the clocks at the Cabinet de Physique and observatories. The telegraph apparatus consisted of three parts : 1. The apparatus for production of the currents ; 2. The receiving instrument ; and 3. The commutator, or instrument for reversing the currents. The arrangements used by Gauss and Weber for the production of magneto- electric currents at the transmitting station con- sisted of two or three large bar magnets, n s, Fig. 12, each weighing 25 Ibs., fixed together ver- tically, with their simi- lar poles in the same direc- tion, on a stool, PP. A wooden bobbin, a a, sup- plied with handles, b b, and wound with more than one thousand turns of in- sulated copper wire,* rested Fig. 12. on the stool and around the upper ends of the magnet, so that, on lifting up the bobbin by the handles, a current * In the apparatus constructed at a later date "by Gauss and Weber, they increased the number of turns to seven thousand. HISTORY AND PROGRESS. 29 would be induced in the coil in one direction; and on lowering it again, a current would traverse the coil in the opposite direction. The ends -(- and were connected to the commutator, and thence to the line wires. The coil a a was called the inductor. The receiving instrument placed at the distant station is represented in Fig. 13. It consists of a large coil or multi- N Fig. 13. plier, m m, of insulated copper wire, on a copper frame, the ends -\- and being in connection with the line wires. A permanent steel magnet, MM, 18 inches long and 3"' X 5'" transverse section, was suspended in the middle of the multiplier by a number of parallel silk fibres from the ceiling of the room. To enable the observer to read off with care the small deflections of the magnet, a mirror, N, was affixed to the shaft K, carrying the magnet, in which was seen, through a telescope, R, at a distance of 10 or 12 feet, the reflex of a horizontal scale, s s. The commutator, introduced for directing the currents in one direction or the other through the line, consisted of an 30 THE ELECTRIC TELEGRAPH. arrangement for bringing two points alternately in commu- nication with two others. Let a and c, Fig. 14, be two points in connection with the two poles of a battery, or other electromotive system, and b and d the ends of any other circuit ; if the metal bars e and/ be pressed upon the ends a b and c d respectively, the current will pass in the direc- Fig. 14. tion B -f- a e b R df c B. But if the bars e and f be removed from these positions and placed at right angles, that is to say, e between b and c, and /between a and d, as shown by the dotted lines, the current will go through B -f- d R (in the opposite direction) b c B. On lifting up the coil a a, Fig. 12, from the stool to the top of the vertical magnet-bars, a current was induced in the wire encircling them. This current passed by the com- mutator, placed as in Fig. 14, from a to b, through one of the line wires and the multiplier R of the receiving station, deflecting the magnet for an instant in one direction, and returned by the other wire over c and d of the commutator. When it was wished to deflect the needle of the receiving instrument in the opposite direction, this was attained by simply lowering the coil a a again to its original place, and the observer at the receiving station read off one deflection to the right, for instance, and one to the left. But, in con- structing a code of signals, it was necessary that two or more deflections to the right or left should frequently follow each other. This was done by means of the commutator. HISTORY AND PROGRESS. 31 Thus, on lifting the coil a a, if we suppose a deflection of the magnet was produced to the right, by reversing the commu- tator and then lowering the coil again, another deflection in the same direction would be observed. To produce a third deflection in the same direction it would be necessary, evi- dently, to reverse the commutator again before raising up the inductor. After this fashion Gauss and Weber were enabled, by an ingenious combination of deflections to the right and left, to form the following alphabet and numerals with a maximum of four elementary signals in a letter : r = a rrr = c,k Irl = m Irrr = w llrr = 4 I = e rrl ==. d rll =. n rrtt =. z lllr = 5 rr = i rlr = f,v rrrr =. p rlrl llrl 6 rl =. o Irr = g rrrl = r rllr = 1 Ml = 7 IT = u m = h rrlr = s Irrl = 2 rill = 8 11 I llr = I rlrr =. t Mr = 3 mi = 9 r represents the swing of the north pole of the magnet towards the right, and / the swing of the same pole towards the left of the magnetic meridian. Various lengths of the pauses between the signals indicated the conclusion of words and sentences. The copper frame around the needle was necessary in order to prevent the great number of oscillations which the magnet would have made across the meridian had no such check been introduced. The checking action of masses of metal in the vicinity of an oscillating magnet was discovered by Arago, and has been described by Sir William Snow Harris,* in whose ex- periments the oscillations of a freely suspended magnetic needle were reduced from 420 without a damper, to 14 with a damper of copper surrounding the needle. In the case of the magnet used by Gauss and Weber, its mass, and the minuteness of the angle which was necessary for the deflection to be read off with the aid of the telescope and mirror, must have assisted materially in bringing it back to the meridian line. * "Magnetism," p. 58 ; " Phil. Trans.," 1831, Part I. 32 * . THE ELECTRIC TELEGRAPH. 35. Ga.uss and Weber's line, as has been said, was erected between the Physical Cabinet and the observatories of Gottingen for other than telegraph purposes. It was for this reason that Gauss, unable to afford the time necessary to perfecting the system, which he believed capable, with modifications, of leading to brilliant results, requested Pro- fessor Steinheil, of Munich, to simplify the apparatus and endow it with a practical application. The perfection to which this ingenious inventor brought Gauss and Weber's telegraph has rendered it as much or more his than theirs. He studied thoroughly the subject of magneto- electricity, and made experiments, discoveries, and suggestions which earned for him the name of the founder of electro-magnetic telegraphy. A description of this telegraph by its inventor is to be found in Dingler's Journal, 70, p. 292. It consists principally of three parts the inductor, the receiving apparatus, and the line. The compound permanent magnet employed by Steinheil in his apparatus consisted of seventeen horseshoe magnets, weighing together 60 Ibs., capable of lifting about five times its own weight. Two induction coils, of together 15,000 con- volutions of insulated copper wire, turned on an arbor, and presented in rotation the axes of the coils to the poles of the magnet, so that when one coil was under the north end of the magnet, the other would be under the south end. The commutator in connection with these coils was so constructed that, in turning them from right to left, the alternate currents, or all those going in one direction only, passed through the line, and, on turning them back- wards or from left to right, only those currents in the opposite direction were let into the circuit, the others being cut off. Steinheil was careful to admit his currents for the shortest possible space of time, and for this purpose, allowed the contact springs of his commutator to make contact with the lines only at the moment when the induced current was at its maximum. Fig. 15 represents the contact plate to HISTORY AND PROGRESS. 33 which the leading wires from the receiving station were connected. The contact springs to which the ends of the wire coils were attached travelled in the white annular spaces, and made contact only at a and b, whilst in every other position the circuit was interrupted. The receiving apparatus consisted of an oblong coil of wire or multiplier of 600 turns, in the centre of which Fig. 15. were supported, on vertical axes, two magnetic needles, their neighbouring ends having opposite magnetic polarity. Fig. 16 gives a vertical, and Fig. 17 a horizontal section of this instrument, a b is the coil of wire, N s and N s, the magnet needles, turning on the axes m and m', and m Fig. 16. carrying on their neighbouring ends brass continuations with small ink reservoirs, c c'. These reservoirs were furnished with capillary tubes and filled with printer's ink, so that on coming in contact with the strip of paper travelling before them they each printed a dot. Two plates, h h', prevented the needles from being deflected in the direction opposite to that in which they were to print, as the deflection by the current would otherwise have caused them to swing, and perhaps mark the paper, not only when responding to signals, but also when oscillating. D 34 THE ELECTRIC TELEGRAPH. By means of the plates, therefore, a current sent through the coil deflected only one of the needles at a time, the other being held back ; and on the current being changed, the reverse took place the other needle only being deflected. Thus the signals on the paper were recorded in two lines those to the right marking the deflections in that direction, and those on the left indicating the left hand deflections. A paper strip was kept in uniform motion by means of clockwork, which, whenever a mark was made, moved the paper onwards, leaving a blank space before the next signal. These signals were necessarily only dots, because induction currents are only of momentary duration. Much nicety was required in obtaining magnets of exactly the right size. They could not, for example, be large, because their inertia would have been too great ; nor too small, because their mechanical force would not have been great enough to have effected the printing. Two small per- manent magnets in the rear of the printing needles retarded their inclination to be deflected when not under the influence of currents. The letters of the alphabet were constructed from combinations of, at most, four of the dots given in suc- cession by the pointers of the two printing needles. They were arranged by Steinheil as follows : A ..... .% L ---- .. ---- ... "R M 1 * . . . . j. . . , C,K . . .. 1ST .,.. 2 . . . . .* D . . . . O . . . . 3 .... t E . . . . P . . . . .. 4 .... . F ..... E . . . . .. 5 .... ... G ..... 8 .... .. 6 .... ... TT T 7 J. . . , . . ' Ch. ....... TJ.V..V 8........ ScL W Q kJL-Jl ' .. TT ** i/....' rw _ J . . . . Z . . . . .. HISTORY AND PROGRESS. 35 Messages were sent with this apparatus at the rate of ninety-two words in a quarter of an hour,* or over six words per minute. The line wires were in three parts ; the first included a length of 30,500 feet, erected in the air a few inches over the roofs of the houses between the Royal Academy of Munich and the Observatory at Bogenhausen. The weight of this section was about 200 Ibs. The greatest span between two poles was 400 yards. The second section of the line connected the residence of Professor Steinheil with the Observatory in the Lerchenstrasse, there and back a length of 2,000 yards ; and the third section, a length of about 400 yards, connected the Academy with the workshop of the physical cabinet. 36. When experimenting on the Nuernburg and Fuerther Railway, to ascertain if the rails could not be made use of as lines for the service of a telegraph, Steinheil made the important discovery that the earth might be used as part of the circuit of an electric current. This discovery, which ranks with those of Yolta and Oersted, was one of the greatest contributions ever made to the progress of the telegraph. Had the identity of the electricities been known earlier, return circuits other than the earth for voltaic currents would never have been used ; for in all the earlier experiments and attempts with frictional electricity, the earth was used as the return circuit. Steinheil took advantage of his discovery, and removed the halves of his lines, leading, in their stead, the corresponding connections of his apparatus to plates of metal buried in the earth. A communicator of peculiar construction enabled the operator to transmit to, and receive from, either Bogen- hausen or Lerchenstrasse, at pleasure. It was arranged that when the indicator was in circuit with one station, the wire of the other should be connected to the multiplier of the receiving instrument. The receiving instrument was not used exclusively to record the messages on the paper strips ; sometimes small hammers * Dingier' s Journal, 67, p. 370. D2 36 THE ELECTRIC TELEGRAPH. were substituted for the ink reservoirs, striking against bells of glass or metal of different notes. Thus Steinheil's appa- ratus formed also, upon occasion, an acoustic telegraph. The history of the subject so far shows us that no single indi- vidual can claim the distinction of having been the " inventor of the electric telegraph ; " but if there is one worker who deserves more credit than another for his energy, intelligence, and success in the service of his adopted science, that man is certainly Professor Steinheil. 37. The ingenious experiments with which Professor Wheat- stone occupied himself, in 1834, in his researches on the velocity of the electric wave in solid conductors, seem to have first directed his attention to the subject of telegraphy. Mr. Cooke had already employed himself with the construc- tion of telegraph lines for railway purposes before he joined Professor Wheatstone. Their first joint invention was a telegraph with five indicators and as many line wires. Its appointments were as follows: Five multipliers of fine insulated copper wire, with light magnet-needles, were arranged in a line across the back of a diamond- shaped dial-plate. The upper side of the plate, which served at the same time as a cover for the case contain- ing the multipliers, was marked with twenty letters of the alphabet, c, j, Q, u, and z being omitted, as capable of being replaced by others, at the expense, perhaps, of a little ortho- graphy, but at the saving of another line wire with its magnet-needle and multiplier. The margin contained the nine numerals and 0. Fig. 18 shows the upper side of the dial-plate, with pointers attached to the axes of the magnet- needles, broken away in the middle, to show the multi- pliers. Each pointer was deflected from its position of rest always under the same angle on each side, so that by observing the deflections of any two needles, and following with the eye the direction pointed out by their nearer ends, at the point of intersection of these imaginary lines would be found the letter intended to be transmitted. Thus, in telegraphing a letter of the alphabet, the deflections of two needles in con- trary directions were always necessary. In Fig. 18, for HISTORY AND PROGRESS. 37 example, the needles 1 and 4 are deflected, pointing to the letter Y. The numerals and were telegraphed by the deflec- tions, to the right or left, of single needles. One end of each of the multipliers was brought out on the lower half of the right-hand side, and continued to one Fig. 18. of the five-line wires ; the other ends were joined together and attached to the line used for a return circuit. The manipulator or key, Fig. 19, consisted of six metallic springs 6, 1, 2, 3, 4, 5 each of which was provided with two buttons with contacts working downwards upon two parallel metal strips, p and N, to which the poles of a voltaic battery were connected. With this arrangement, the operator, by pressing down, at the same time, the buttons of two springs, one over each of the strips p and N, could transmit a current through one line and back through another, deflect- 38 THE ELECTRIC TELEGRAPH. ing thereby the two needles at the receiving station for signalling a letter. As the numerals were indicated by the deflections of single needles, a sixth wire was provided for the return circuit. =: B Fig. 19. This telegraph, although extremely beautiful in detail, was inferior in a practical sense to that of Steinheil. It was put up on the London and Birmingham and Great Western railway lines, and tried fairly, but found, on account of the number of line wires, to be too expensive, and was accordingly given up. 38. The necessity of supplying the receiving station with some signalling apparatus for calling the attention of the observer to the commencement of a correspondence had been fully understood by every inventor since Reusser, who pro- posed to attain this by firing an electric pistol, and Sommer- ing, who proposed to do the same by the liberation of mechanism by accumulated gas. It was left, however, to the energy and persevering genius of Wheatstone to completely solve the problem. HISTORY AND PROGRESS. 39 The first alarm employed by him was an apparatus in which the attraction of a soft iron armature to the cores of an electro-magnet, whose coils were in connection with the line, released a wound-up mechanism. This alarm arrangement is given in Fig. 20. The tooth- wheel n was arrested by the end of the lever p. On the other end of p was a soft iron armature, ?, which, when a current passed through the coil u of the electro-magnet, was attracted, and released the wheel. The case contained a clockwork surmounted by a bell, which was struck by a hammer, moved backwards and forwards by means of an eccentric. At the sending station were a small battery, K, and a key, Fig. 20. s, consisting of two metallic springs, a b, of which the lower one, b, was fixed on a block of wood, and insulated from the other by a strip of ivory. The spring a could be pressed against b ; but when untouched, remained separated from it by its own elasticity. From the positive pole of the battery, a wire, 1, went to the electro-magnet, u, of the alarm, and was turned several times round the horns of soft iron ; a second wire, 2, returned to the transmitting station, where the line was connected 40 THE ELECTRIC TELEGRAPH. with b f whilst the spring a was connected by a wire with the negative pole of the element. 39. On long lines of telegraph, Wheatstone found that the current became much weakened by the resistance of the wire, the coils of the apparatus, and by indifferent insulation, so that he was obliged to employ a considerably augmented battery power in order to effect the attraction of an arma- ture, whereas, when he employed a magnetic needle, the latter was always easily deflected. This suggested to him a way to overcome the difficulty, which he succeeded in doing by closing the circuit of a local battery, by the deflection of a magnetic needle at the receiving station, by means of which the electro-magnet of the alarm described above, or an electro-magnet, whose armature formed a hammer and struck directly on a bell, was put in action. The battery at the receiving station was called the local battery, in con- tradistinction to the line battery, which was placed at the transmitting end. The local battery consisted of fewer elements, as its circuit was short, and the resistance of the coils of the electro-magnet not great. It was once a popular fallacy in England and elsewhere that Messrs. Cooke and Wheatstone were the original in- ventors of the electric telegraph. The electric telegraph had, properly speaking, no inventor ; it grew up little by little, each inventor adding his little to advance it towards perfection. Messrs. Cooke and Wheatstone were, however, the first who established a telegraph for practical purposes, comparatively on a large scale, and in which the public were more nearly concerned than in those experiments in which the ends of the wires were brought into laboratories and observatories. Therefore it was that the names of these enterprising and talented inventors came to the public ear, whilst those of Ampere and Steinheil remained comparatively unknown. 40. We read in Dr. TurnbulPs book, that in the latter part of the year 1832, Mr. Morse, an American artist of some notoriety, whilst on his homeward voyage from Europe, con- ceived the idea of an electro-chemical telegraph ; and that in HISTORY AND PROGRESS. 41 his constructions and subsequent experiments he was much indebted to the valuable aid of a fellow-passenger, Dr. Jack- son, of Boston, who had witnessed in Paris numerous experi- ments in telegraphy, and was besides versed in both electricity and chemistry. t The first invention made by Morse was a chemical tele- graph by the decomposition of acetate or carbonate of lead, or turmeric moistened in a solution of sulphate of soda on paper, by the galvanic current between platinum points. Morse proposed to work this telegraph by means of types forming a code of the letters of the alphabet, numerals, &c., to be set up in a sort of composing stick, and passed by mechanical means under a lever carrying a contact breaker, which would rise and fall correspondingly to the forms of the types. On arriving at New York, Morse took steps to get a set of types for his proposed telegraph, but was, it is said, prevented by press of business from doing much with it until 1836. In the mean time he gave up the idea of a chemical receiving apparatus, and determined on the adop- tion of electro-magnetism. 41. Morse's telegraph underwent, in the few following years, many important modifications. An idea may be formed of how different the later arrangements of Morse's apparatus were from the first, from the fact that the electro- magnets of the latter weighed 158 pounds, and that two men were necessary to lift it with its stand. The bobbins of the wire coils were 3 inches long and 18 inches diameter, and the iron core nearly 1 inch. Morse used No. 16 copper wire insulated with a coating of cotton threads, it being at that time his opinion that the coil should be made of the same sort of wire as that used for the line. The transmitting key used by Morse in his later apparatus is shown in Fig. 21. It consisted of a lever turning on the axis, supported by uprights which were screwed into a small block of dry wood. The screw on the longer arm of the lever was pressed upon the front contact, whilst the similar screw in the shorter arm was used only to regulate the play of the key. 42 THE ELECTRIC TELEGRAPH. One pole of the battery was connected to the front con- tact, called the anvil, and the other pole to earth. The line was connected through the supports with the lever, which Fig. 21. was kept on the back, or what has since been termed the reposing contact, when not in use, by means of a steel spring underneath the longer arm. The surfaces of the hammer and anvil of the contact were faced with platinum, in order to prevent oxidation by dampness from the atmosphere, or by being burnt by the action of the current. 42. Finding his instrument not sufficiently delicate for great distances, by reason of the line- resistance and loss of current by bad insulation, Morse had recourse to an expe- dient of relays or repeating circuits. The arrangements designed by him for this purpose were of a somewhat primi- tive form, but in principle they are the same as that known as translation, and used on all submarine lines at the present day. 43. In 1829, Edward Davy obtained a patent for an electro-magnetic chemical telegraph, in which he had inge- niously applied "Wheatstone's idea of combining an electro- magnet with a clockwork, in the construction of a receiving instrument. Davy's telegraph required at least four line wires, which, independently of its complication, would be reason enough to account for the fact that it never came into practice. It has, however, the merit of having been the first system in which the movements of a clockwork were governed by HISTORY AND PROGRESS. 43 an escapement worked by an electro-magnet, and probably, in its turn, suggested many of the subsequent inventions. With three batteries at the sending station, Davy was enabled, by reversing the currents, to deflect at pleasure the tongues or needles of six relays at the receiving station. This was done by putting in each line two relays, similar in construction to those used by Wheatstone in conjunction with his alarm, one being acted upon by positive currents and the other by negative only. Beyond the second relay coils, the three lines were connected together with the fourth line wire, which was used as a common return circuit. The receiving apparatus consisted of a sheet of cloth or other chemically-prepared material, drawn between a metallic cylinder and a series of six platinum rings, placed equi- distant on the outside of a wooden drum. Each of these rings was connected with the contact points of one of the relays, and a common local battery was inserted between all the tongues or needles of the relays and the metallic cylinder, so that when the needle of either of the relays was deflected, the current of the local battery passed through the chemi- cally-prepared cloth to the metallic cylinder, producing a dot. On the arrival of a current the metallic cylinder was moved forward a certain distance by means of the clockwork. The operation of successively opening and closing the circuit at the sending station, imparted to the cylinder at the receiving station a rotatory motion resembling that of the long hand of a clock governed by the pendulum and escapement. The cloth used for receiving the marks was impreg- nated with iodide of potassium and muriate of lime. Six longitudinal lines, intersected by transverse ones at similar distances, divided the whole surface of the cloth into regular squares, which facilitated reading off messages. 44. The system which we have next to notice is another invention of our ingenious countryman, Professor Wheat- stone ; this is his first dial instrument, patented in 1840. The apparatus in question seems to have undergone 44 THE ELECTRIC TELEGRAPH. several modifications in the course of a year or so. The principle remained, however, unaltered in all of them ; it was that of sending, from the transmitting station, a series of alternate currents through the line, which, passing round the soft iron of an electro-magnet, moved an armature, and regu- lated the motion of an escapement similar to that of a clock. It consisted of two parts : 1. The transmitter, and 2. The receiving instrument. The transmitting portion of the original apparatus con- sisted of a commutator, to direct the current of a battery alternately through two electro-magnets at the receiving station. The direction of the current was effected by means of a tooth- wheel, supported by a metal upright. The teeth of this wheel, to the number of fifteen, were so arranged that the teeth and the spaces in rotation represented thirty letters of the alphabet, numerals, &c. On each side of the wheel was a spring contact, only one of which made contact with the wheel at the same time; when the one pressed against a tooth the other was always opposite to a space. These springs were connected to two line wires, and a battery was inserted between the tooth- wheel and earth. From the circumference of the wheel protruded thirty spokes, and on the base of the upright was a bar, used as a stop for the hand of the operator when turning the spoke wheel, and it was wished to signal the letter opposite the spoke taken hold of. The receiving instrument or indicator was formed by a dial having 30 divisions corresponding to the letters, nume- rals, &c., of the transmitter. The index which moved over the dial was driven by a clockwork, the escapement of which was fixed in the axis of a beam supporting two armatures of soft iron, over the poles of two electro-magnets, in the circuits of the two line wires. As the tooth- wheel of the transmitter was turned round, currents were alternately sent through the side contacts, through the lines, and round the cores of the escapement magnet. Whenever, therefore, the tooth- wheel of the transmitter rested at any place, a current HISTORY AND PROGRESS. 45 circulated in one or other of the escapement magnets, the armature was held down on one side, and the index prevented from moving farther round the dial. 45. An improvement in the apparatus was made by dis- pensing with one of the line wires, as well as one of the contact springs, of the sending commutator, and one of the electro-magnets of the indicator. This was a material step in the right direction, and fulfilled the first condition of a successful telegraph that of requiring only a single line wire. In the improved indicator the duties of the one electro- magnet were fulfilled by a spiral spring with an adjusting screw for tightening or loosening it. This spring acted in the contrary direction to the single electro-magnet, but, of course, with inferior force. It had tension enough, however, to separate the armature from the poles of the electro- magnet, and to bring over the beam bearing the escapement, whenever the current in the electro-magnet was interrupted. 46. But the most important improvement introduced into the construction of this apparatus was in the substitution of magneto- electric currents for those of a voltaic battery. The sending apparatus, so modified, is shown in Fig. 22. It consisted of a permanent horse-shoe magnet, or combina- tion of magnets, fixed to the base board B, between the poles of which was placed a vertical shaft, supporting, on opposite sides, the coils c c of an electro-magnet. They were so arranged that, on turning the shaft, the cores of c c at the same moment approached, and left the poles of the perma- nent magnet. The ends of the coils of wire round the electro-magnet were connected,, by means of a sliding contact underneath, with the terminal screws e and /. On the top of the shaft was a pinion D, locking into the tooth- wheel w. The number of teeth of the wheel w were in relation to those of the pinion D, so that one complete revo- lution of w would cause D to revolve half as many times as there were letters, numerals, &c., engraved on the correspond- ing dials of the sending and receiving instruments. It will be evident, without further explanation, that a half-revolu- 46 THE ELECTRIC TELEGRAPH. tion of the pinion and the coils of the electro-magnet would produce a current in the line in one direction, and that the continued motion in the same direction another half-revolu- tion, would produce a current in the contrary direction. This arrangement required a slight modification also of the receiving apparatus ; but in principle it remained the same. Fig. 22. M. Froment, of Paris, has constructed some step-by-step instruments, on the system of Professor Wheatstone, in which he has succeeded in simplifying the mechanism. Y. TELEGRAPHS NOW IN USE. 47. Single- needle Telegraph of Wheatstone and Cooke. The single -needle telegraph of Messrs. Wheatstone and Cooke is a modification of the five-needle system by the same inventors, described above. The principle of the system depends upon the construction of an alphabetical code whose basis consists in two elementary signals the deflections of a vertical pointer to the right and to the left as in Gauss and Weber's telegraph. Fig. 23 represents the front view of a single-needle instru- ment a mahogany case with engraved metal face, in the HISTORY AND PROGRESS. 47 middle of which is a vertical pointer. In the lower part of the case is a handle, on the arbor of which, inside, a commu- tator is placed, so that when the electrical connections are made, the movements of the handle and pointer to right or left correspond. On the metallic face of the instrument are engraved I I 1 $ ' \ V| \ A S3 B 1181 S I? C 311 D 183 F SIS 1 110 13 N - 311 5 33 - A O 1183 H 113 I 31 J 3133 111 - S 113 - H 131 - U 183 - D K 1SS1 L 331 o= S !i 18 311 C -o 313 - F 331 - L 833 - R ? 51n 1313 333 1111 - P 1113- M 1131 - B 1133- O 8 111 T a V 131 V 1311 1311 - V 1313 - Q 1331 - ft 1333 - W W 1333 X 3113 Y 3111 Z 3531 8111 - Y 3113 - X 3131 - Z 3133 . J to Numbers 3311 toPriv.Sigs. 3313 Repeat ... 3331 Wait 3333 Code .... 1 lung Letter. ...3 3311 to Numbers 3313 to Priv.Sig.. 3331 . . . Repeat 3333 .... Wait 1 long . . . Code Fig. 23. the deflections corresponding to the various letters of the alphabet. Fig. 24 represents the interior view of the same. A A is a long vertical coil of fine silk-covered wire, in the middle of which a small magnetic needle plays, having at 48 THE ELECTRIC TELEGRAPH. the end of its axis the pointer seen on the outside face of the instrument. One end of the coil is connected permanently with the terminal screw L, to which the line wire is attached at the back of the instrument, and the other end with the commutator, Fig. 24. by which a battery is put between it and the earth, and its direction reversed according to the position of the handle. The commutator is of simple construction. The arbor of the handle is divided electrically into two halves, a and 6, insulated from each other by an intervening thickness of dry wood, p. The half a is connected per- manently, through a spring, s, with the copper plate of the HISTORY AND PROGRESS* 49 battery, and b with the zinc pole of the same, through the spring s'. On the lower side of a, and on the upper side of b, are attached projecting pieces of metal, i and *, which play between the springs t t' and T T', respectively. When the handle is vertical, the metal arms i and i' are also vertical, and the springs t and t' repose against the opposite sides of the rest k. On turning the handle to the right, however, the spring t' is lifted by the projection i from the rest k, and the projection i 1 makes contact with r, whilst t and r' remain unmoved. The spring r is not seen in the figure. The circuit is thereupon completed from the pole of the battery, through N, spring s, half a of the arbor, the arm *', spring r, terminal T, earth, and from the + pole of the battery, terminal p, spring s', part b of arbor, arm i, spring t', multiplier, terminal L, line, opposite station apparatus, and earth. The opposite springs are lifted and the current reversed on turning the handle the other way, The apparatus serves both as transmitter and receiver. When receiving signals, the handle remains vertical. The currents arriving by the line pass through L, coils of the indicator, spring i 9 metal stop k, spring t, terminal T, to earth, and the needle is deflected to the right or left, according as the arriving current is positive or negative. If signals are to be given by the apparatus, the manipu- lator has only to turn the handle to the right or left to effect corresponding deflections of the needle of his own instru- ment and of that at the station to which he is sending. The code of signals for the letters of the alphabet, &c., is engraved on the dial, either by means of arbitrary signals, as in Fig. 23, where the right-hand deflections are shown by the numeral 3 and left-hand by the numeral 1, or by means of strokes of different lengths. In the latter case the long and short strokes indicate the number and direction of the de- flections representing each letter ; and shorter strokes are to be executed before the longer ones to which they are attached. The letter A, for example, is indicated by two deflections to 50 THE ELECTRIC TELEGRAPH. the left ; N" by two deflections to the right ; I, by three deflec- tions consecutively to the right, and then one to the left ; Y, by a deflection to the left, then one to the right, then one to the left, and another to the right ; and so on. This instrument is used almost exclusively on some of the railway lines in the United Kingdom and on some of those in the East. Its great simplicity, inexpensiveness, and little liability to derangement, have obtained it already a long life. 48. Double-needle Telegraph of WJieatstone and Cooke. Another form of the needle telegraph, used also to some extent on some of the English lines, is the double-needle telegraph, consisting of two single-needle instruments com- bined in the same case. They are, however, totally indepen- dent of each other in so far as their electrical connections are concerned, each being worked with a separate line wire. The handles in front of the case are connected with two arbors inside, similar to the one shown in Fig. 24, each of which commutates the current of a battery through the galvanoscope coils surrounding the needle attached to one of the pointers. The discs over which the pointers move are provided with ivory pegs to limit the deflections on each side of the vertical line. The discs are sometimes made circular, and may then be turned round in the dial plate, enabling the operator to shift the pegs, in order to keep the pointers midway be- tween them when the magnet-needles are deflected by constant atmospheric currents. The alphabetic code adapted for this instrument is as follows : The left needle deflected alone, + Corl A B Eor3 Dor 2 F & \ V \^ \\ / V / The left needle deflected alone, Hor 4 L or 5 I K N or 7 M or 6 \ V HISTORY AND PROGRESS. 51 W Y or Both needles deflected at the same time, E or 8 U or 9 S T X Y Q Z /////#// /\ \/ The same rules are observed with regard to the short and long strokes, as with the single needle instrument. The employment of two needles in receiving the signs renders this telegraph very expeditious ; the rate at which it is worked being about double that of the common Morse telegraph. The necessity of two lines, however, prevents it taking any prominent place amongst the existing systems of useful telegraphs. 49. Principle of self-acting Make-and-break. A considerable step in advance of then existing systems was made by Dr. Werner Siemens, of Berlin, by the invention of his first beautiful step-by-step motion telegraph. The principle of this apparatus was that of the automatic Fig. 25. transmission of currents, or what has been called the self- acting make-and^break* Suppose the soft iron armature A A, Fig. 25, of the electro- magnet E, is supported on an axis, B, at one end, and by the spiral spring s in the middle, by which it is pressed E 2 52 THE ELECTRIC TELEGRAPH. gently upwards against the contact-screw c at the other end, and that an electric circuit is established, as is represented in the figure, in which the positive pole of the battery v is connected by a wire, w, with one end of the wire-coil of the electro-magnet E, the other end of the wire-coil being connected with the contact-screw c, by a wire, w', whilst a third connection- wire, uf', joins the axis of the armature A, with the negative pole of the battery. It is obvious that the current circulates in the coils of the electro-magnet, magnetises them, and causes the armature to be attracted to the poles, leaving the contact- screw c, and thereby interrupting the battery circuit. The instant this occurs, and the battery cur- rent ceases to circulate in the coils, the soft iron cores of the electro-magnet lose their magnetism, and have no longer the power to retain the armature which is consequently lifted up again by the spring s. On reaching its position of rest, it makes contact again with c, and re-establishes the battery cir- cuit. This is followed by an immediate interruption ; and the same play must necessarily be repeated, the armature being Fig. 26, always re-attracted to the poles of the electro-magnet, break- ing thereby the circuit, and being again let go, and making the circuit again, and so on ad infinitum. The manner in which Dr. Siemens applied this method of interruption to the service of his telegraph system will be easily seen from the accompanying plan, Fig. 26. If we were to take the apparatus just described and attach to the end of the armature, in any way, a continuation with HISTORY AND PROGRESS. 63 an escapement, so that when the armature is moved up and down, the escapement, moved by it, works round a little scape- wheel, e, carrying a light pointer, p, around the circum- ference of a dial, D ; it is evident that when the apparatus is connected by the wires as shown in the figure, the armature will always be moving up and down, and the pointer, there- fore, always running round the dial. If we now insert into the same circuit another electro- magnet, E, with a similar armature, escapement, tooth- wheel, and dial, the electro-magnets being both magnetised by the same currents and demagnetised at the same time, the move- ments of the armatures will be synchronous ; and if the pointers have been started from the same places on each dial, they will stop at the same points on the dials whenever the armature E is arrested in its upward or downward motion. This apparatus is employed to a considerable extent on the Prussian railway lines. It has, however, not found any ex- tensive employment in England. Apart from its somewhat complicated mechanism and costliness, it is undoubtedly the nearest approach to perfection in a telegraph apparatus of anything we have yet seen. 50. House's Printing Telegraph. This telegraph, the work of Mr. House, of New York, was the subject of an applica- tion for patent in 1845. It belongs to the class of step-by- step motion telegraphs, and consists of two separate parts : the transmitter or commutator, and the receiver or printing instrument. The transmitter is composed of a contact-wheel in every way resembling that used by Wheatstone in one of the modi- fications of his dial instrument, which, in turning, sends a series of currents from a battery at the transmitting station. As each make-and-break of the circuit indicates a letter, whenever a letter is to be transmitted, the contact- wheel is arrested at a certain point by which either the current is allowed to flow, or is interrupted during the continuance of the indication according as a contact-spring happens to rest upon a tooth of the wheel or opposite a space, at the moment of stopping. 54 THE ELECTRIC TELEGRAPH. For the purpose of stopping the contact-wheel at its proper place for each of the letters, House employs a key-board like a piano, with twenty-eight keys, representing twenty-six letters of the alphabet, a dash, and a dot. The contact- wheel is of brass, four or five inches in dia- meter ; its circumference is divided into twenty- eight equal spaces, alternately indented to the depth of a quarter of an inch, so as to expose on the surface fourteen shallow teeth. A spring of metal, insulated from the contact-wheel and shaft, is placed before the former of these, so that when it revolves, the top of the spring comes in contact with each of the teeth in succession, but has not the power to penetrate into the spaces and make contact there. The plan adopted by Wheatstone to hold his current on or to interrupt it, during the reading of a signal, was to stop his commutator opposite an index or pointer at the letter to be indicated. House does this otherwise, with the aid of his piano keys : he puts two rows, of each fourteen pegs on the outside of a cylinder, each peg turning with the cylinder underneath one of the keys of the piano. The latter are held up by springs, and furnished with hooks or cams which, when depressed, catch hold of the pegs of the revolving cylinder and arrest its motion. The pegs of successive letters follow each other round the circle in a spiral at distances of one twenty-eighth of the circumference, and therefore, when the cylinder is turned from one letter to another, just so many contacts and inter- ruptions are given as will bring the pointer or wheel of the receiving apparatus round the same distance. The receiving apparatus is rather complicated ; it is started by an electro-magnet of very novel construction. Above the movable armature, on a common shaft, is a hollow cylindrical slide-valve, in connection with a chamber of compressed air, filled by means of a pump, and supplied with a safety-valve to permit the escape of superfluous air when the pressure becomes greater than is required for work- ing the apparatus. The piston moved by the compressed air let into the HISTORY AND PROGRESS. 55 cylinder by the slide-valve, moves horizontally, and is in connection with the lever of an anchor- escapement engaging with the teeth of the scape- wheel of the printing machine. The scape- wheel has fourteen teeth, and requires, there- fore, twenty-eight movements of the lever to complete a revolution. A steel type- wheel revolves with the same shaft as the scape- wheel, its circumference being furnished with twenty- eight equidistant projections, on which are engraved the letters of the alphabet, a dot, and a dash. The shaft also carries a little drum with letters painted on it in the same order as those on the type- wheel, by which the operator may read off the message when the type- wheel is not printing. On the upper surface of the type- wheel, at the extreme edge, are twenty- eight teeth, against which plays a small steel arm, attached to a metal cap, turned by friction on a shaft revolving in the reverse direction. When the type- wheel is in motion, this arm plays over the teeth, but as soon as the wheel is stopped, falls in between them, which it has not time to do during the revolution. By falling in the teeth of the type-wheel, the arm allows the cap to revolve with its shaft, and by means of two pins, to release a detent, which, in its turn, permits an eccentric to revolve. A con- necting-rod from the eccentric, pulls the paper strip to the type- wheel and prints the letter. An ingenious arrangement is also made for the progres- sion of the paper, by means of a ratchet-wheel and clicks attached to a notched drum over which the paper passes. The line from Philadelphia to New York was the first on which this instrument was used. It found very general adoption on the American lines, after the year, 1849, and is still to be found at work. It is said to be much less liable to get out of order than would be judged, at first sight, from the complication of the receiving apparatus. 51. Hughes' Roman-type printing Telegraph. The essen- tial principle of this highly ingenious system is the syn- chronous movements of type- wheels at two or more stations, and of the power to press a strip of paper at each of the stations simultaneously against the types on the correspond- 56 THE ELECTRIC TELEGRAPH. ing parts of the wheels, by the action of a single electric wave or impulse. A clockwork at each station turns, with a continuous and uniform motion, an axle, at the extremity of which the type- wheel is supported. The synchronism is attained by the aid of a vibrating spring and anchor escapement. The rotation of the type-wheel is transmitted to a vertical arbor, fur- nished at its lower extremity with a horizontal arm tra- velling over a circular disc, in which is arranged a series of contact pins, in number corresponding to the types. Each : ; 4.. ft- 1 9 Fig. 27. pin therefore represents a letter, and is raised when it is wished to telegraph this letter along the line. The hori- zontal arm, which travels round the disc with a motion uniform with that of the type- wheel, comes in contact with the pin just at the moment when the corresponding type is HISTORY AND PROGRESS. 57 at the lowest point and closes an electric circuit, by which the paper is lifted up against the type- wheel, and the letter printed. The key-board used to elevate the contact pins is shown in Fig. 27. It consists of twenty-eight keys, alternately white and black, marked with the twenty^six letters of the alphabet, a full stop, and a blank, corresponding to an empty space in the type- wheel. Below each of the keys is a mov- able lever, whose fulcrum is at K", and which terminates at the bottom of one of the contact pins K K, arranged in a circle in the metal box A, in the top and bottom of which are holes for the ends to protrude the upper holes being long, to allow of a radial motion. Each pin is held down by the pressure of a small spring, but may be elevated by pressing down the corresponding key of the piano-board. Fig. 28 gives a vertical section of the printing instrument and key-board. The section shows a white key, hinged at K", connected to its lever K', a contact pin, k, on the right, and also to a black key, whose lever reaches to a contact pin on the left of the box A. The contact pins are provided with shoulders to limit their movements in each direction. The horizontal arm, which travels over the circle of con- tact points, is attached to the bottom of the vertical arbor Q, to which motion is imparted by the bevelled wheel G 2 , on the shaft G. It is made up of three principal parts the arm r, jointed at a ; the resting piece, or earth-contact, r ; and the shovel r. The vertical shaft Q is of brass, and is divided electrically into two parts by an insulating ring of ivory, q. The lower part is supported by the central pedestal, which is insulated from the box A by a non-conducting ring. The continuation of the jointed arm r, which is held by the portion of the shaft above the insulating ring q, is pressed down by a spring, which keeps a small screw in the middle of the continuation in metallic contact with the second piece r , supported by the portion of the shaft below the ring. The shovel r is of steel. "When a key is depressed, the corresponding contact pin is elevated, and if the arbor Q is in motion, the extremity 58 THE ELECTRIC TELEGRAPH. of the arm r mounts upon the elevated pin, by which con- tact between r and ^ is interrupted, and that of r with Jc established. The arm r having made contact, the shovel r , which immediately follows it, pushes the pin k in its slot _ HISTORY AND PROGRESS. 59 outside the circumference swept over by r ; so that if the latter make another revolution whilst the finger is kept down upon the key, no second contact is made, and the same letter is not repeated. The operator feels a vibration of the key as the shovel passes by the pin, and is thus made aware that the letter has been printed. The type-wheel H contains on its circumference, in twenty- eight equal spaces, twenty-six letters of the alphabet, a dot, and a blank space ; it is fixed to the extremity of the axis cc, which is put in motion by means of the hollow axis G, enveloping it in the greater part of its length. The connection between c c' and G is made by the mediation of a fine rachet- wheel, G 5 , attached to the axis G, the click m l being on the axis c c'. On the latter are supported, besides the type-wheel and click, a corrector, H', or wheel with long narrow teeth, equal in number to the types, serving to establish precision between the movements of the horizontal arm r and the type- wheel. On the same axis is a wheel, HJ, having a notch at one part of its circumference for stopping the type-wheel when the blank space is opposite the printing press, in case it should spring forward. The hollow axis G is turned by a clockwork moved by a weight, a wheel of which engages with the pinion G X , and supports, besides the racket G 5 and bevilled- wheel G 2 already mentioned, the escape wheel G 4 and a tooth wheel G 3 , which locks into the pinion ij (Fig. 30) of the printing shaft I. The printing shaft turns seven times as fast as the type- wheel, and carries a fly-wheel, i", at one extremity, in order to overcome the inertia of a small shaft, whose duty is to lift the paper up to the type-wheel at the other extremity. This is shown partly in section in Fig. 30. The printing shaft i and its continuation i are locked together by means of a ratchet-wheel, i lf and click, i'. At the end of the continuation shaft i is a cam, h lt for lifting the press and the paper against the type- wheel. The printing press is shown in Fig. 29. Underneath the type-wheel is a small cylinder a, over which the paper is led, its axis being in the middle of a bent lever, b, turning at a ; 60 THE ELECTRIC TELEGRAPH. attached to it is a ratchet-wheel, in the teeth of which catches a click affixed to a movable piece, b 1} terminating in the rectangular arm & 2 , which is forced upwards by a spring attached to the frame of the apparatus, but is stopped against the axis i. When i makes one revolution, the cam lifts the arm b of the lever, together with the cylinder a and paper strip up to the lowest tooth of the type- wheel by which the paper strip is impressed with the print of the type, kept inked by an inking roller, M. The cam being very sharp, the movements of ascent and descent are pro- portionally rapid, and the paper touches the type during only an infinitely short space of time. The axis continuing to turn, the cam meets the arm b and depresses it, causing the click to draw round the cylinder and advance the paper a certain distance. By the side of the ratchet-wheel i the printing shaft t MM afa lll'limyi - APPAREIL. ._; i ) b Fig. 31. carries an escapement h ti, arrested by a continuation of the lever L L', moving with the armature of the electro-magnet. The armature is of soft iron, supported at the extremity of a lever D over the poles of the electro-magnet Fig. 31. The lever turns between supports on the axis, and tends to rise by the force of a spring regulated by the adjusting screw D'. The screw d' (Fig. 32) on the end of the lever L L', turning on the axis /, sits over the armature ; the other end of the HISTORY AND PROGRESS. 61 lever engages with one of the pallets of the escapement h ti, and governs the motion of the axis i. When a current traverses the coils of the electro-magnet the armature and lever are depressed, the click is put in gear, and the pallet h of the escapement, released, turns with the axis i. At the moment when the pallet h' passes under the lever, it relifts Fig. 32. it, and depresses the screw d', returning thereby the armature to the poles of the electro-magnet, and, at the same time, throwing the click out of gear. The magnet B is of novel construction. It consists of a permanent horse-shoe magnet, with soft iron cylindrical con- tinuations on the poles. These continuations are each en- circled by a coil of wire. When no current passes through the coils, the armature is attracted to the poles by the magnetism distributed in the iron. This force is opposed by the adjusting spring, which is so regulated that, the armature being in contact, a very weak current is able to neutralise the attraction. The printing shaft has also the duty of correcting the move- ments of the type- wheel, and of insuring always that, at the moment of printing a letter, the type is in its proper position. This is effected by means of a curved cam, h 2) on the axis i. The instant the cam 1i lifts the arm b of the frame carrying the printing roller, the projection h 2 locks into the teeth of the wheel H', and adjusts, if it be necessary, its position. If, on entering the teeth of H', the cam has to push the wheel forwards, or, to accelerate the motion of the axis c tf, the click m is pushed onwards, passing over one or more of the 62 THE ELECTRIC TELEGRAPH. teeth of the ratchet-wheel G 5 * If, on the contrary, the cam has to retard the motion, the click pulls the ratchet-wheel backwards, for which purpose the latter is not made rigid on the axis, but is formed of a disc held between leather washers supported by two plates of metal, fixed on the hollow shaft G-. The electric circuits of the apparatus are very simple. The bottom of the vertical shaft Q is connected to earth, and the upper part to one end of the coils of the electro-magnet, the other end being to line. One pole of a battery is connected to the levers k of the contact pins, the other pole to earth. At two corresponding stations the plates of the batteries must always be looking the same way, because the home apparatus is intended always to work as well as that of the distant stations, and the armature of its magnet is only liberated by currents in one direction. When a current arrives, therefore, from the line, it passes first through the coils B of the magnet, then through the vertical shaft Q, which it descends, and goes over from the screw in the jointed arm v to the resting piece r y and from this to earth. When a current is to be transmitted, the operation consists principally in interrupting the earth circuit, and in inserting the battery into the break. This is done by the contact pins and jointed arm of r. A key being depressed, the arm r in its journey rides over the pin, and its screw is lifted up from contact with r, which breaks the direct earth circuit. At the same time the contact of r with the pin k, which is in communication with a pole of the battery through the lever K, sends a current from the battery (K k, r Q), through the coils of the magnet into the line, &c. Suppose two such apparatus, properly adjusted, at the extremities of a line of telegraph, the clockwork wound up, the electrical connections properly established, and the type- wheels locked. The employe who desires to transmit presses down the blank key of his instrument ; this pushes up the corresponding contact-peg in the circle K, and when the chariot arrives over the pin, the extremity of the piece r rides over it, separating the earth contact and introducing the battery into the line circuit. The current passes through HISTORY AND PROGRESS, 63 the vertical shaft, the coils of the magnet, and line wire to the other station, where it circulates in the coils of the magnet, the vertical shaft, &c., and goes to earth. In traversing the coils of the magnets of both instruments, the current weakens the attractions of the armatures to the poles of the electro-magnets ; the former are forced off by the spring, the screws d' are raised, and the levers L at the same time depressed. The pallets h of the escapements h ti, are thereupon released, the axes i put into gear with I, and the type- wheels released. During the revolution made by the axes *, the cylinders a are raised by the cams, and lift the paper up to the printing-wheels at the moment when the latter are unlocked. No letter is printed, because the blank space in the type- wheel occurs just there. The paper strips and cylinders descend again ; the former advancing a step. The clicks are then disengaged from the ratchets, and the pallets h recaught by the levers L', which were lifted up, causing the armatures to be pushed down again to the poles of the magnets. If a key answering to any letter be now pressed down, the current is repeated the moment the chariot passes over the raised contact pin ; the printing axis is put in motion, the letter printed, and the paper pushed on as before, and so on, until the message is completed. It sometimes happens that the apparatus do not agree when one of the stations sends its message. In this case, the employe at the receiving station advises his corre- spondent of it by giving him a signal ; both then arrest their type- wheels, and the transmission is recommenced, begin- ning always with the blank. To avoid the inconvenience of irregular working, which might arise from changes in the battery power, Professor Hughes has adopted a method of short circuiting the coils of the electro-magnet the instant after the armature is released, that the current, whatever may be its intensity, comes into play only long enough to effect the required weakening of the magnetic attraction. This is done by con- necting one end of the electro-magnet coils with D, and the 64 THE ELECTRIC TELEGRAPH. other end with L, in addition to the other connections, and by adjusting the screw d', so that when at rest the armature, reposing on the poles, does not touch it ; but as soon as the neutralisation occurs, it is lifted up by the force of the spring, and the coils short circuited by contact of D with d'. The speed of transmission attained with this apparatus is very great. The chariot and type-wheel revolve about 120 times in a minute, and an expert manipulator can trans- mit on the average two letters during a single revolution of the shaft. The word " telegraph," for example, is completed in six turns, as follows : 1st turn blank and t. 2nd . . * , . e and I. 3rd ' . , 7 1 . e. 4th ....'. g and r. 5th a and p. 6th . . "V . . h. The French word " bonte " is done in four turns : 1st turn blank. 2nd .... , b and 0. 3rd n and t. 4th . .... . e'. Another example is the word " dintz," more fortunate than either, being transmitted during a single revolution. This invention was brought by Professor Hughes from America, before the submersion of the old Atlantic cable, on which he made his first important experiments on the speed attainable in working his apparatus on submarine lines. Since then important improvements have been made in the construction and mechanical execution of the apparatus, in the atelier of M. Fromont of Paris. The principles have, however, undergone no change. The system has been adopted by the United Kingdom Telegraph Company, under an arrangement giving the Company the exclusive right to use the apparatus them- f s HISTORY AND PROGRESS. 65 selves, and grant its use to others in this country. On the company's line between Birmingham and London, on which messages and press matter are constantly passing to and fro with the aid of this apparatus, the average speed, as stated by the company, is forty messages per hour, and is believed to be capable of considerable augmentation when the employes have had more practice. The company intend introducing this system on all their lines, and have reason- able hopes of its ultimate success. In France the system is gaining daily a wider employment ; in Russia and Germany the administrations of telegraph are likewise disposed to adopt it. The following is a fac-simile of the printing : BY HUCHES'S TELEGRAPH INSTRUMENT, 52. Brfyuefs Electro- Magnetic Dial Instrument. The prin- ciple of Breguet's apparatus is that of alternately making and breaking at the transmitting station, the circuit of a voltaic battery. At the receiving station is an electro- magnet, whose armature is correspondingly attracted and let go. The armature acts on a pallet, which interposes itself between the teeth of two scape-wheels turned by clockwork. The apparatus consists of three parts : The transmitter, The receiving instrument, and The alarum. The transmitter is shown in Fig. 33. It consists of a metal dial, supported by three pillars on a wooden base. The whole dial is divided into twenty-six equal sections, separated by two circles. In the inner circle are engraved twenty-five letters of the alphabet and a +, and in the outer circle, the numbers from 1 to 25, and a +. Opposite each letter, on the periphery of the dial-plate, is an indentation for drop- ping the handle into. Underneath the dial-plate is a disc, G, with a serpentine groove on its underside, turning on the 66 THE ELECTRIC TELEGRAPH. axis of the handle H, which moves above the dial. A small peg with a friction wheel runs in the groove, and imparts a vibrating motion to the contact lever. The further end of this lever is faced with platinum on each side, and makes Fig. 33. contact alternately with the screws p p, with p, when the handle is over the even numbers, and with p' when it covers the uneven. The terminal c is connected with the positive pole of the battery. It is also connected by a wire, /, to the contact screw p '. The terminal R, on the left, is in permanent con- nection by a wire, v, with the contact p, and is intended for the wire leading to the receiving instrument. The terminal L of the left line- wire communicates with the contact lever, N. The point of N touches, at pleasure, either the contact s, to which the alarm is connected, or r, which is in communica- tion with the revolving disc, and through this and the friction wheel and 1 1' with battery and earth, or, lastly, it may touch the end of a metallic strip marked " communication directed On the other side of the dial is a similar lever, N, connected with the terminal L of the line on the right. N may be placed on s, the alarm, or on r, which, like r, is connected HISTORY AND PROGRESS. 67 with 1 1, or lastly, it may be placed on the other end of the metal strip " communication directs" Fig. 34 represents the interior of the receiving instrument seen from the back ; M M is a horizontal electro-magnet, whose armature, suspended between screw points, carries on its upper side a metallic rod q, which is limited in its play by adjusting screws in the frame/. At right angles to q, near the top, is a peg, q, working in a fork F, fixed to Fig. 34. one end of the horizontal shaft a. At the other end a pallet, g, engages alternately with two parallel scape-wheels, im- pelled by a clockwork in the case above, and placed so that the thirteen teeth of the front and back wheels alternate when looked at from the front. When the apparatus is at rest and the armature held back by the spring, the pallet locks into the teeth of the back wheel ; but on the attraction of the armature to the poles of the magnet, the pallet springs into the teeth of the front wheel, which, being half a tooth in arrear, allows the wheels and pointer to turn one twenty- sixth of the whole circle. As soon as the armature is released the pallet leaves the teeth of the front wheel, and re- enters between those of the other. The latter being half F2 68 THE ELECTRIC TELEGRAPH. a tooth behind, the clockwork turns the wheels and pointer another one twenty-sixth. Thus every time the circuit is made or broken, the pointer advances one of the twenty-six divisions of the dial. The spring for drawing the armature from the poles is adjusted by means of a bent lever, t, on the end of which it is hooked. The lever itself is fixed to the under side of a disc, s y turning on the vertical axis h. On the opposite side of the disc s is a long vertical pin, i y which is gradually turned with the disc in a small angle by an inclined plane on the rim of the drum u. The shaft a, with its pallet and fork, is supported by a frame or lever, c D, turning on the centre c on the left, and on the right held up by a spiral spring which forces it against a pin, passing through the guides M and N. On pressing upon the button E of the pin, the frame is moved down- wards and releases the escapement-wheels from the control of the pallet, but carries down with it a check which prevents their unlimited run. As soon as they are free to rotate, the wheels turn round with the pointer until their further pro- gress is arrested by the check which catches hold of a pin at the back of one of the scape- wheels, corresponding in posi- tion with the zero or + of the dial. The alarm generally used with this instrument contains no new principle whatever. The attraction of an armature liberates a clockwork, which turns a disc with an eccentric crank. The latter in revolving moves the hammer of a bell to and fro. It is similar in construction to Wheatstone's first alarm. Another alarm which Breguet has supplied with some of his telegraphs is constructed on the principle of the self-acting make-and-break employed in Dr. "Werner Siemens's first dial-telegraph. The connections of the apparatus for a station will be seen by reference to Fig. 35, which shows the various pieces of the apparatus. When neither of the stations is using the line the switches N and N' of both the apparatus are placed on s and s', so that HISTORY AND PROGRESS. 69 a signal arriving from either side, L or L', will be given notice of by the alarms A and A'. In this way the current from the left goes from L, past a lightning guard, e, through a galvanoscope, G, L, N, s, alarm A, earth, and back LJ Fig. 35. again. Arriving by L', a current passes the lightning guard e'j G', L', N', s', alarm A', earth, and so on. When a signal is given from L, the employe turns his switch N on r. The current passes then from L through the galvanoscope G, the needle of which it deflects, L, N, r, disc, contact-lever, p, R, receiving instrument, earth, &c. Breguet has also made use of an idea used previously in the construction of telegraphs in Germany. Instead of the movable armature and stationary electro-magnets, he sometimes employs a cylindrical electro-magnet made to turn on its longer axis on screw points. The poles of the soft iron core are furnished with soft iron continuations, which hang down between the opposite poles of two permanent horseshoe magnets. "When the current magnetises the soft iron core arid its continuations, the coil with core and con- tinuations are deflected to one side or to the other, attracted by one of the permanent magnets, and repelled by the other. 70 THE ELECTRIC TELEGRAPH. When the current is reversed the polarity, and therefore the deflection, is also changed. 53. Kramer's Pointer Telegraph* The different telegraphs which have hitherto been mentioned are worked by sending currents of electricity from the transmitting station, either from a galvanic battery or from an induction apparatus. As soon as the signal is given, or the work done, the current is cut off, and the line becomes inactive. The reverse of this mode of operation was introduced by Kramer, in his dial- telegraph, and subsequently by Frischen, for working the Morse instruments on the lines under his charge. In both these systems the current of a galvanic battery circulates continually in the line, and attracts, when at rest, the arma- ture of an electro-magnet at the receiving station. On breaking the line at any point the armature falls off, and remains off until the battery circuit is closed again. With the system of currents transmitted for each signal, it is obvious that a separate battery is required for each station. This is not the case when the system of closed circuit is used ; because an interruption in any point must be followed by the same effect on the armatures of all the electro-magnets in the circuit. The exterior of Kramer's apparatus differs in appearance very little from that of Siemens and Halske's. It consists of a round dial, with thirty keys on the circumference, numbered from to 29, inclusive. An inner circle is marked, in cor- responding sections, with the letters of the alphabet irregu- larly placed, and a third circle, concentric with the others, contains a double row of numerals from 1 to 9, and some other figures and blanks. The interior of the apparatus is shown in Fig. 36. Two circular plates of metal are connected together by means of three pillars near their periphery; through their centres passes the axis c of the pointer z z seen on the dial. This axis carries a scape- wheel, r, and a tooth- wheel, R; it is turned by the tooth-wheel H engaging with its pinion 3. On the axis of the wheel H is a pinion locking into a tooth- * Der Elektro-magnetische Telegraph. ScheUen, p. 195. HISTORY AND PROGRESS. 71 wheel attached to the barrel, on which is wound a cord passing over a pulley and carrying the weight G. The escapement r is formed by a horizontal wheel, on the rim of which are sixty vertical steel pins thirty projecting downward, and the same number upwards arranged alter- nately at equal distances from each other. The prongs of a Fig. 36. steel fork lock into the pins, and are just so far apart, that when one prong touches the rim on one side, the next pin can pass under the other prong. The fork is supported by a bell- crank lever, h, turning on the axis w, between upright bear- ings. An armature, A, of soft iron is supported by the lower arm of the lever, opposite the poles of an electro-magnet, M, which regulates the movements of the fork and scape-wheel. The poles of the magnet are covered with thin pieces of German silver soldered to them, to prevent the armature making close contact with them. The figure shows the arma- ture attracted to the poles, which is the position of rest of the apparatus. 72 THE ELECTRIC TELEGRAPH. The wheel commutator R, the tongue d, and contact anvil x, are provided to regulate the interruptions of the currents. The wheel has on its circumference thirty teeth, which, in the course of one revolution, lift the hammer, therefore, thirty times from its contact with the anvil. When the armature of the electro-magnet is attracted to the poles, a tooth of the wheel R lifts up the hammer from the anvil, and interrupts the circuit, the electro-magnet immediately becomes demagnetised, and the armature falling off again, depresses the fork of the escapement, and allows the scape-wheel to advance six degrees together with the wheel commutator, R. By this means, contact is re-established between d and x y the tooth which separated them passing by, and the end of the hammer d falling into a space. The electrical circuit is shown in the figure by wires, and in following the motions of the scape- wheel, contact- wheel, R, &c., it is necessary to imagine a battery inserted between T' and m. The current, then, goes from the + pole of the battery, following the arrows along the circuit T', n t d, x y 5, coils of M, to the > pole of the battery. In this way the pointer keeps on running round the dial as long as the maintaining power of the clockwork lasts. The motion may be arrested either by breaking the battery circuit, in which case the armature falls off, and the fork- escapement rests on the upper surface of the scape- wheel, or by arresting the pointer itself. The latter is the method employed in telegraphing ; the key over the dial, corres- ponding to any letter, on being pressed down, interposes a peg which bars the further progress of the pointer, and, in this respect, resembles the arrangement of Siemens and Halske's pointer telegraph. When it is wished to advance or to put back the pointer on the dial, at any station, without the assistance of the current, this is done by pressing on the button of the lever T > P> which presses the armature against the electro- magnet, and lifts the fork. 54. An alarm used with this telegraph is rung by the release of a soft iron armature from the poles of an electro-magnet. HISTORY AND PROGRESS. 73 At an intermediate station, during the correspondence between two end or distant stations, when the continued interruptions of the circuit would cause the alarm to sound whilst the correspondence lasts, which would become annoying to the employes, an arrangement is made for so weakening the currents in the coils of the electro-magnets by means of a shunt, that the alarm does not sound. This Fig. 37. shunt is shown in Fig. 37, by a wire, w, w ', and resistance coil, R, which has about five times the resistance of the coil of the electro-magnet, allowing therefore only about five- sixths of the current to pass through the legitimate route. This shunt circuit has another object. At the moment a current is sent through the coils of the electro-magnet m m, the induced current tends to weaken the effect of the battery current. This would not be perceptible if the induced current were obliged to traverse the whole line. But in going round the shunt L, R, L', with little resistance, it exercises its full opposing force on the magnet, and prevents the armature being attracted. The plan, Fig. 38, shows an ingenious arrangement for providing against the errors frequently arising in systems based on the principle of closed circuits, from bad insu- lation of the line. Should there be, as is sometimes the case, a battery at each of the end stations, and the line, in some points intermediate, in imperfect contact 74 THE ELECTRIC TELEGRAPH. with the earth, it is evident that when either of the inter- mediate stations interrupts the circuit, the current at the stations near the ends will not be interrupted, but only weakened. Kramer, unable to realise a perfectly insulated line, and having to provide against emergencies, makes his Fig. 38. alarm work as well by a weakening as by a complete inter- ruption of the current. For this purpose he short-circuits one of the coils of his electro-magnets in the following manner. In front of the armature a contact screw, c, is pressed upon by a metal spring, b, so that when the armature is attracted towards the poles of the electro-magnet, at half-way, it comes in contact with b, and separates it from c. The axis on which the armature turns is connected by a wire with the middle e of the electro-magnet coils ; the contact screw c is in connec- tion directly, by a wire, with the line L' ; and between the back of the spring b, and the same line- wire L', a resistance, io t equal to that of the half, m, of the electro-magnet coils, is inserted. A current, arriving by the line L (disregarding the shunt R), passes from L by m, e, m', b, c, to L'. Very little goes through w, because its resistance is very great in proportion to the resistance of c. The armature is thereupon attracted, and as it descends, carries down the spring b with it, and HISTORY AND PROGRESS. 75 interrupts the short-circuit c L', but makes, at the same time, contact between itself and the spring 6, and in so doing, shunts the current from the half m of the electro- magnet coils, by the circuit e, d, b, m. The short-circuit to line c L' being broken, the current must pass from b through IK, whose resistance is equal to the resistance m' ; thus the total resistance of the circuit remains unaltered. The arma- ture is now held by only the one pole m, and the spring /is so adjusted that on the slightest decrease in the magnetism of m, it exerts force enough to pull back the armature. The idea of attracting the armature a certain distance by two poles, and this done, of holding it there by one, is as novel as it is ingenious, and answers its purpose very well in this instance. The plan is, however, complicated, and the apparatus, of course, requires very nice adjust- ment. 55. Magneto-electric Pointer Telegraph of Siemens and Hakke. More generally employed than their pointer telegraph with voltaic currents, and than Kramer's, is the convenient and trustworthy instrument which Siemens and Halske some years since constructed for the Bavarian tele- graph lines. The system is almost exclusively employed on the lines of the Grande Socie'te des Chemins de Fer Busses, by the London, the Danzig, and the Koenigsberg fire brigades, and on various other lines in England and elsewhere. The apparatus consists of a battery of permanent magnets, between the poles of which a coil of insulated wire on a revolving armature of soft iron develops, on being turned on its axis, alternately positive and negative currents. These currents traverse the line one after the other, and passing through the coils of an electro-magnet at the receiving station, cause its armature to vibrate and turn an escapement- wheel and pointer. Fig. 39 represents an external view of the complete apparatus. A A is a cylindrical case, containing the trans- mitter ; c, a handle which, in operating with the apparatus, is turned round from letter to letter, marked on the hori- zontal dial-plate, stopping always against the tooth opposite 76 THE ELECTRIC TELEGRAPH. the letter to be indicated ; and B, the receiving instrument, supported by a bracket at the back. The latter has in front a small dial, corresponding in its arrangement with that of the transmitter, and a pointer whose motions follow Fig. 39. faithfully those of the handle of the instrument which is working. Fig. 40 shows the internal mechanism of the transmitter. The metal disc J, with inclined teeth on its rim, is supported by the back #, and by two square pillars y. On the back, which consists of a stout plate of soft iron, are screwed a series of several pairs of permanent magnets, G G ; those on one side with their north, and those on the other with their south poles projecting. Between the poles of this system is a cylinder of soft iron, E, which serves as keeper of all the magnets. It is cut out longitudinally in two deep, broad grooves, on opposite sides, in which a spiral of fine well- insulated copper wire is coiled. The whole armature is supported by the brass caps F F', in pivots above and below. Above the spiral, the pinion T locks into the tooth- wheel L, turning on an arbor A A, by means of the handle H above the disc J. The pro- portion between the teeth of the wheel and those of the pinion is such that one revolution of the wheel causes the pinion to revolve thirteen times, changing the magnetism along the whole length on each side of the armature, and HISTORY AND PROGRESS. 77 thereby inducing in the coil, if the circuit is closed, thirteen positive and thirteen negative currents of equal magnetic effect. When the coil is turned half round on its vertical axis, the polarity of the armature is also changed, and a magneto- electric current induced in the coil w in one direction, and Fig. 40. on turning it half a revolution farther round, so that it takes up its former position, a current in the opposite direction is induced. The ends of the coil of wire, which are attached to the brass caps, are connected, one to an earth plate, the other to the coils of the electro-magnet of the indicator, and thence to the line. In employing a series of separate magnets, arranged at a distance from each other, a greater inductive effect is produced than if only a single permanent magnet were used, or if the 78 THE ELECTRIC TELEGRAPH. separate magnets were combined in the form of a battery ; because the poles, acting at the same time upon the armature, produce along its whole length on each side an uniform magnetism, so that the wire coil receives, in every point, the same magneto-electric impulse. The interior of the indicator is shown in Fig. 41. The top s s of a permanent magnet of hard steel, bent in a rectangular form, and having a space cut out of its upper end, protrudes through the side of a circular brass plate, A A, one- eighth of an inch thick. In the slit in the upper end s s of the magnet is the axis of a movable tongue of soft iron, having at its extremity the same polarity as the end s, and vibrating between the poles n and s of a polarised electro-magnet, m m. By polarised is meant that the soft iron cores around which the wire is wound receive polarity from a permanent magnet. In this case the cores of the electro-magnet are attached to a stout piece of soft iron, resting on the north pole of the angular magnet which distributes to the whole system Fig. 41. above the point of contact north polarity. The tongue or armature of soft iron is, therefore, attracted by both the poles of the electro-magnet with equal force, and if not exactly balanced in the middle between them, will rest on one side or on the other by the superior attraction of the nearer pole. HISTORY AND PROGRESS. 79 When a current is sent through the coils of the electro- magnet in the direction which increases the magnetism of the more distant pole, whilst it reverses the magnetism of the nearer, the latter forthwith repels the armature with nearly the same force with which the former pole attracts it, and the armature in consequence goes over to the former pole, and rests there, not only as long as the current lasts, but after- wards, when no current is circulating in the line. It is this which renders this indicator so sensitive for induction currents which are only of momentary duration. At the end of the armature is a German silver fork, carrying two horizontal arms thin steel springs with hooks which catch into the teeth of a ratchet-wheel. The ratchet- wheel has thirteen teeth, so that when the armature oscillates from right to left, or from left to right, the wheel advances half a tooth ; and in order that the pointer may march over the whole dial, the armature must oscillate thirteen times in each direction. This is provided for by exactly that number of reverse currents being transmitted from the sending station, when the handle of the transmitter is turned once round the dial. Behind the hooks are two screw stops which limit their motion and prevent the skipping of the wheel. 56. The alarm is generally a separate piece of mechanism ; it was, however, sometimes in the earlier forms of this apparatus combined with the indicator by attaching a hammer to a continuation of the armature of the electro-magnet, and letting it strike on two bells placed just within its reach ; but this plan has been almost entirely abandoned in favour of the alarm shown in Fig. 42. The principle is the same as that on which the indicator is constructed : a polarised electro-magnet, m m, is supported by the upper pole of the angular bent permanent magnet M on a wooden base, b. From the lower pole of M springs a movable armature or tongue of soft iron which takes polarity from it, and plays between the projecting cores of the electro-magnet, which are also polarised. A continuation of the armature forms a hammer, h t and strikes upon the bells b b'. 80 THE ELECTRIC TELEGRAPH. A commutator is sometimes employed for directing the arriving currents, at pleasure, through the indicator or through the alarm ; but, more usually, the alarm is inserted in the same circuit, and when it is not required to give notice of the arrival of a current, its coils are short-circuited by a contact peg inserted in a hole between the terminals s s' of the electro-magnet coils. The plan according to which two such instruments are connected up for two stations is shown in Fig. 43, where the line L is connected at each station to an alarm, A, which can be short-circuited at s, if required ; the other side of the alarm is connected with one side of the coils of the indicator i by a wire, w ; thence a connection, ?*/, leads to the coil of the transmitter or inductor T, and to earth. On turning the handle of the transmitter, the currents pass, first of all, through the indicator of the home apparatus, the pointer of which turns correspondingly with the handle ; HISTORY AND PROGRESS. 81 thence over the terminals of the alarm, through the line, to the alarm of the distant station, which it rings or not, according as the contact peg is out or in, through the indi- cator coils, moving its pointer simultaneously and corre- Fig. 43. spondingly to the movements of the first indicator, and, finally, to earth a short circuit over the coils of T being established when the handle stands at zero of the dial. The older forms of the apparatus were a little clumsy and noisy in manipulating; but the vast improvements made lately in its construction in the London establishment of Messrs. Siemens Brothers have reduced these inconveniences to a minimum, without in the least lessening the power or trustworthiness of its indications. It is spoken very highly of by the employes, and may be looked upon as one of the best existing pointer telegraphs. 82 THE ELECTRIC TELEGRAPH. 57. Wlieatstone* s Universal Telegraph. This is another form of step-by-step telegraph, the invention of Professor Wheatstone. During the last few years it has obtained con- siderable employment on private lines, being found at nearly all the ends of the network which Professor "Wheatstone has helped to spin over the metropolis. It consists of two parts the " communicator " and the " indicator." The communicator is contained in a small square box, on the upper surface of which is a raised dial-plate, surrounded by thirty equidistant keys radiating from the same centre. Upon the dial-plate are marked the twenty-six letters of the alphabet, three points of punctuation, and an asterisk ; in an inner circle are the nine numerals and a cross on each side. A hand or pointer, turning on an axis in the centre of the dial, rotates in connection with the handle in the front, and may be arrested at any letter while the handle is being turned, by depressing one of the keys or buttons. Inside the box is a fixed permanent horse-shoe magnet, placed horizontally, carrying, on its poles, four soft iron cylindrical cores with their coils of wire, arranged at equal distances from each other in the circumference of a circle. On an axis passing through the centre of this circle, in connection with the handle, revolves a soft iron armature whose breadth is a little greater than the distance between two adjacent cores. When the armature revolves, therefore, it approaches one pole as it recedes from the one diagonally opposite, and thus induces simultaneously in the two coils currents in the same direction. A small circular chain is placed horizontally underneath the keys, and becomes slightly bulged to the extent of its slack when a key is pressed down. As soon as another key, however, is pressed down, the chain straightens itself underneath the first key and lifts it up. The purpose of the keys is to arrest at pleasure the march of the pointer round the dial, and to short-circuit the currents at any letter. This is done by an arm, attached to the axis which carries the pointer on the dial, coming in contact with the HISTORY AND PROGRESS. 83 lower part of the depressed key. Motion is imparted to the axis of the pointer and carrier arm from the handle by means of a bevilled wheel, which engages with a pinion fixed to the axis carrying the armature of the electro- magnet. The proportion is so adjusted that, for every current induced in the coils, the pointer shall advance the distance of one letter on the dial. If, therefore, the pointer and arm start freely from zero when the handle is turned and any key, as D for example, depressed, the armature is rotated, producing alternate electric waves whilst the pointer will pass over A B and c respectively. Arriving at D, the carrier-arm comes into contact with the depressed key and cuts off the passage of the subsequent currents until another key is depressed, by which the carrier-arm is released and travels on with the pointer. The face of the indicator is divided into thirty equal spaces exactly similar to the dial of the communicator, with a double circle of letters and numerals. On an axis in its centre is a pointer like the minute hand of a watch, to which motion is given by a small escapement-wheel with fifteen teeth. Two magnetic needles, or bars, fixed to an axis, lie parallel between two small electro-magnetic coils with soft iron cores, and are so arranged that, when currents pass through the coils and magnetise the cores, the latter exercise mutual attractions and repulsions on the poles or extremities of the magnet-needles, imparting a backward and forward motion to their axis. The scape- wheel is carried by a short vertical arm fixed to the end of this axis, and is rotated by working to and fro against stops or pins. The electrical connections of the apparatus are simple. The coils of the communicators and indicators of all the apparatus are connected up in a common circuit. When the coils of one of the communicators are turned round by means of the handle, if the pointer is free to move round the dial, a current traverses the line at every letter which the pointer passes over, moving the hands of the indicators correspondingly; but as soon as the carrier-arm 84 THE ELECTRIC TELEGRAPH. attached to the axis of the pointer is arrested by coming in contact with a depressed key, the currents which follow are short-circuited. The hands of the indicators therefore stand still upon the same place on the dials until the key is raised and the short circuit removed. In the front of the indicator a contact lever is moved between stops marked A and T. When the lever is placed on A the arriving currents ring the alarm, the telegraph being thrown out of circuit ; when it is placed on T, the alarm is out of circuit and the indicator works. The working of this instrument, as well as the neatness with which the whole is constructed, cannot be too highly spoken of. The great advantage which it possesses over the other step-by-step telegraphs, in which the coils or arma- tures are stopped and started at every letter, is that its currents are uniform, whereas in the other systems, at starting and stopping, the operator cannot avoid moving his handle slower than when driving it midway between two letters, which very frequently gives rise to " skipping " of the pointer. 58. Simple Morse Circuit. In its simplest form the Morse telegraph consists of a transmitting key and a recording instrument, with intervening line wire, battery, and earth connection. The purpose of the key is to close the circuit of the battery conveniently for the formation of arbitrary signals. The signals representing the letters, &c., consist of combinations of two elementary marks, a dot and a dash. The former is given by the momentary closing of the circuit, and the latter by closing it for a longer time, by means of the key. The signs are received at the distant station by the corresponding attractions of the armature of an electro- magnet, which marks them on a strip of paper in its vicinity. The plan by which these arrangements are made at two stations is represented in Fig. 44. At each of the stations, B is a battery of voltaic pairs, connected between the point 1, underneath the metallic lever K, and the earth. When the lever K is not being manipulated, it is held by a spring upon the metal point 2, between which and the earth are inserted HISTORY AND PROGRESS. 85 the coils of the electro-magnet M, whose armature is employed to mark the paper and record the signals given from the distant station. On pressing down the key K, for example, the contact between the lever and the point 2 is interrupted and that at 1 established, the current of the battery B goes from c through the contact point 1 and front tine, Fig. 44. part of the lever to 3, where it enters the line L. Arriving by L' the current passes over K', from the middle 3, to the back contact point 2, and from this, the key being at rest, it traverses the coils of the electro-magnet M', and then goes through the earth back again to the battery B. 59. In construction, the Morse instruments are very various, nearly every maker having peculiar arrangements of his own. The earlier apparatus in America and England were homespun, and of little mechanical merit as specimens of art, nor did they advance very considerably beyond this until within the last few years. In the hands of the French and Germans mechanicians, however, the instrument reached a high degree of completeness, and has secured for 86 THE ELECTRIC TELEGRAPH. itself in every country where the telegraph is to be found an employment exceeding that of any other system. The adop- tion of the Morse instrument by the French Administration of Telegraphs in 1857, gave an impetus to inventors, and many real improvements were the result. 60. Embossing Instrument with movable Magnet. This is a construction of the Morse by Messrs. Siemens and Halske, of Berlin, once extensively used on the Russian, Danish, and some of the German lines, but at present replaced to a great extent by newer constructions. The movement of the writing-lever is effected by the attraction of the opposite poles of two electro-magnets rendered active by the same currents. Fig. 45 gives a perspective view of the instrument, m m' are two straight electro-magnets. The core of m is fur- nished with a facing of soft iron, r, on each of its poles ; the core of m is supported between two screw-points, and is furnished at each end with a continuation, p, ending in a facing opposite to and of the same size as r. Between the two continuations p p, is a frame carrying the printing lever. a b and a b' are the ends of the coils of the electro-magnets connected electrically with the terminals A and B. When a current traverses the coils it polarises their cores in reverse directions, that is to say, when the end and continuation p of the core in front are south-polar, the corresponding end of the other core, with its continuation r, will be rendered north-polar the reverse polarities being, of course, at the further end and the faces of p and r on each side, having opposite magnetism, will attract each other. The attraction of these four poles in the same sense renders the instrument extremely delicate, and the force with which the poles tend to approach each other being very great, the instrument is well adapted for recording signals by scoring the style into the paper strip. When the current ceases the printing-lever is brought back by means of the spring /. w w' are the rollers between which a paper strip is drawn. The style, carried at one end of the beam, enters a groove in the middle of the roller w, when HISTORY AND PROGRESS. 87 the other end of the beam is depressed. The play 01 the beam, which turns on the axis c, is limited by the adjusting screws u and z. The style is carried on the end of a screw which enables the operator to regulate its position in the groove so as to indent the paper more or less legibly. Fig. 45. When the current passes through the coils of the electro- magnets, the armatures are attracted to each other and the style forced into the paper strip underneath the groove, where it is held while the current lasts, and as the paper during this time continues to be drawn through, a long or short score is produced, according to the time which the transmitting key is held down. 61. The Morse Code. The elementary signs of the Morse telegraph are two, a dot and a dash, produced by the record- 88 THE ELECTRIC TELEGRAPH. ing instrument according to the time which the key at the transmitting station is held down. The general adoption of this system on the Continent some years back occasioned the establishment of certain rules for the settlement of the letters, numerals, &c., which have been subsequently almost universally adopted. The formation of so many letters, &c., out of two elementary signs required the greatest number of varia- tions with the given number of elements. The number of variations, with repetitions of two elements, are: With single signs 2 = 2 With two signs 2 2 = 4 With three signs 2 3 = 8 With four signs 2 4 = 16 Therefore, by using variations of from one to four of the two elementary signs, we have at our command 2 + 4 + 8+ 16 = 30 variations for the formation of an alphabet. These- variations, with some others, have been disposed as follows : Letter. A A B C D E $ F G H I Sign. I. ALPHABET. Letter. J K L M 9 o 6 p Q B, S Sign. HISTORY AND PROGRESS. 89 Letter. T U tJ V W Sign. Letter. X Y Z Ch Sign. II. NUMERALS. Numeral. Sign. Numeral. Sign. 1 . . . __ 6 . . .... 2 . . ..___ 7 . . ... 3 . . -.- 8 . . . 4 . . .... 9 . . - 5 . . ..... . . __- III. PUNCTUATION, &c. Sign. FuU stop Colon ....... . Semicolon . Comma . . . ... .. Interrogation . . . . ... . . Exclamation ..;.*. . . i ^ Hyphen . . . ..... ...... Apostrophe Fraction line* fin verted commas ..... _ f Parenthesis . _ t Italics or underlined . . ... . ^ IY. OFFICIAL SIGNALS. Sign. Public message . . . .... Official (Telegraph) message . * To be placed between the numerator and denominator of a vulgar fraction. t To be placed before and after the words to which they refer. 90 THE ELECTRIC TELEGRAPH. Private message . . . . . CaU . . . .'...; ..._. Understood . . . . .' ^ Interruption Conclusion Wait .... . . . _... Receipt B The length of a dot being taken as a unit, the length of a dash = 3 dots. The space between the signs composing a letter = 1 dot. letters = 3 dots. words = 6 dots. The formation of the numerals is ingenious ; they are each represented by five of the two elements, and so that, disregarding the dashes which stand on the right hand, and giving a value of unit to a dot and two to the dashes on the left, the value of the numeral represented is expressed. The signs of punctuation, official signals, &c., are either higher variations or arbitrarily chosen letters of the alphabet whose single appearance is a sufficient indication that they are not to be construed as forming parts of words. 62. Morse's Transmitting Plate. Soon after Morse's inven- tion of the transmitting key his attention was directed to the fact that some people find great difficulty in manipulating the arbitrary combinations with uniformity in the length of the marks and spaces. He, therefore, constructed an arrangement for facilitating the transmission. On a metal plate, B B, Fig. 46, are soldered series of raised rectangular pieces of metal, whose lengths and distances apart correspond with the arrangement of the Morse alphabet. Between these pieces, strips of ivory of equal thickness are inlaid, making the whole surface, A A, level. These metal pieces are shown black in the figure and the ivory white. From a binding screw, c, attached to the plate B B, and therefore in electrical connection with each of the metallic rectangles, a wire, m, is led to the receiving instrument and a battery, the further HISTORY AND PROGRESS. 91 pole of the latter being to earth.. The line wire w ends in a spiral of insulated wire fastened to a style, G, with blunt platinum point and insulated handle. Line I I I I I I I I I Fig. 46. This apparatus is intended to replace the key. In order to transmit a message with it the operator takes the insulated handle in his hand and scores the point with an uniform speed over the signs of the letters, one after the other, which 92 THE ELECTRIC TELEGRAPH. he wishes to telegraph. In doing so the circuit is closed as soon as the point of the style touches any of the metal pieces, and is broken again when it moves over the ivory. The arrangement has never enjoyed an extensive employ- ment, and is now, perhaps, entirely out of use. The reason of this is probably to be found in the fact that the imperfect appreciation of time which prevents some acquiring uni- formity in manipulating the key renders them as unable to move the style with an equal velocity over the plate, time being a factor of velocity. 63. Morse Apparatus with Relay. When the line con- necting two stations is long, it is impossible sometimes, even with very great battery power, to move the armature of the electro-magnet with force enough to impress the paper legibly. It was on this account that Morse employed a relay in working his recording apparatus. The principle of the relay has already been explained in conjunction with Wheatstone's alarm. The form of relay used with the Morse instrument differs, however, from that invented by Wheatstone. Instead of the magnetic needle and mercury cups, the local circuit is closed by the contact of the armature of an electro- magnet, with a metal anvil, both being inserted in the local circuit. A common form, known as the American relay, from its general employment on the American lines, is shown in Fig. 47. The electro-magnet M M is fixed horizontally on a board, having before its poles the soft iron armature a, supported by a tongue turning on the axis b. The armature is held back by a spiral-spring f, stretched between the tongue and an adjusting screw, g. The coils M M of the electro-magnet terminate in the binding-screws L' L", to which are brought respectively the line and earth wires. The local battery and Morse apparatus are inserted between the terminals L' L". The former of these is in per- manent connection with the axis b by a wire, x, and the latter with the body of the bracket k i, which carries two screws, h d, with a platinum point, and c, whose point is insulated with agate. HISTORY AND PROGRESS. When at rest the tongue leans against the agate point, and the local circuit is open ; but when a current circulates in the line and coils of the electro -magnet the armature is attracted towards the poles, and the tongue strikes against Fig. 47. the point of the contact-screw h d, and closes the local circuit, the current of which passes from /' (#, b, d, i, k, g) to I". 64. Simple Morse Embosser for two Stations with Relay. Fig 48. represents a plan of connection of a Morse embosser with relays and local batteries for two stations. G is the line galvanoscope connected, on the one side, with the line, on the other, with the lever of the key K. Its purpose is to show the presence of current in the line, and to give a rough idea of its strength. The front or working contact 1 of the key is connected with the pole c of the line-battery B, and the other pole z with the earth-plate. The back, or reposing contact, of the key is connected with one end of the electro- magnet coils of the relay R, the other end being in commu- nication with the earth-plate. Lastly, between the contact- point 2 of the relay and its tongue or armature are inserted the coils of the Morse M and the local battery L B. When in repose the levers of both keys are on the contacts 2, and the line, therefore, at both ends to earth through the coils of the relays. On pressing down either of the keys 94 THE ELECTRIC TELEGRAPH. the current passes direct from the z-pole of the battery to the earth-plate and earth, and from the c-pole through the line galvanoscope, line, key of opposite station, and relay, to earth. The deflection of the relay-tongue, from contact 1 Line Fig. 48. to contact 2, closes the local circuit, and the armature of the receiving instrument works in conformity with the motions of the. key at the sending station. Another method of connecting up the same instrument for two stations is shown in Fig. 49. In this method the lever of the key is in permanent contact with earth. The c-pole of the battery is connected with the front contact of the key, and the z-pole with the point of junction between the galvanoscope and relay, the latter being inserted between the galvanoscope and back contact of K. The local circuit is arranged as before. A current arriving by the line while the key is at rest passes through the galvanoscope, coils of relay, back contact and lever of key to earth. When the key is pressed down on the contact 1, the c-pole of the battery is put to earth HISTORY AND PROGRESS. 95 through the lever of the key, and the circuit being thus completed, the current from the z-pole passes through the galvanoscope into the line. Xuie Fig. 49. In the former method (Fig. 48) the operation of the key consists in shifting the line from relay to battery. In the other method, the battery and relay have the same fixed con- tacts 1 and 2 of the key ; but the earth and line change places, the line taking the place of the earth in the former, and the earth being shifted by the key from relay to battery. This is by no means so good as the former method, because it necessitates a good insulation of the battery, without which a current, depending on the magnitude of the fault, will not only pass always through the line, but also through the coils of the home-relay; and any accidental contact of the battery with earth will give a signal at the relays of both stations, whilst, with the former method, a similar accident would be entirely without effect further than weakening the currents sent on to the line, notice of which is amply given by the galvanoscope. 65. Intermediate-station Commutators. Where intermediate stations occur, which are supplied each with only one Morse 96 THE ELECTRIC TELEGRAPH. instrument, it becomes necessary to employ a commutator or current director to put the apparatus at pleasure in the cir- cuit of the up or down line in order the meet the require- ments of the service. At such a station the apparatus must be so arranged as to be able to assume either of these three positions : (1.) When the intermediate station is entirely cut out of circuit, and the end or distant stations on opposite sides correspond directly through the line. (2.) When two end or distant stations on opposite sides correspond with each other, and the intermediate station receives the despatch, at the same time. (3.) When the intermediate station wishes to communicate with a station up or down the line whilst it has notice of currents arriving from the other side. To avoid the inconvenience of altering continually the con- nections to suit these various positions of the apparatus, com- mutators are employed. Various forms of these instruments are given by Siemens, Nottebohm, Borggreve, and others. 66. One of the completest is that of Nottebohm. It consists of six bars of metal screwed on to a wooden base, cut out in seven holes to receive contact pegs between them, so as to bring them in metallic contact with each other. Fig. 50 gives the commutator in half-size, and Fig. 51 the contact peg in full-size. L and L are the terminal screws to receive the line wires coming from the galvanoscopes. R and RU the ends of the electro -magnet coils of the relay. T is connected with the back contact, 2, of the key, and E with earth. Between the front contact of the key and the bar R O the line-battery is inserted. The beam of the key is also to earth, according to the second plan mentioned above (Fig. 49) for arranging the Morse system. In the first position, when the intermediate station is to be cut out of circuit to let two other stations correspond direct, the contact peg is put into the hole 3. The current coming from the left-hand side passes over G', L, 3, L, G 2 , and so on, to the other line. The employe can see, by the deflections of the needles of his two galvanoscopes, GJ and G 2 , HISTORY AND PROGRESS. 97 when the stations are corresponding, and when both have done, at a given signal, he re-arranges his commutator for regular work. This signal is given by both stations simul- G. taneously, and consists in pressing down their keys for the space of one minute, by which the needles of the galvano- scopes are deflected steadily in one direction for that length of time. In the second position, by which the intermediate station participates in the messages transmitted by the distant stations, and which are, for the most part, official instructions, time, or information for the employes, contact pegs are put into holes 1 and 6. A current then arriving from the same side will traverse G L , L, 1, R , 2 (coils of relay), 1, R,,, 6, L, G 2 , to the other line. The relay then performs its functions of closing the local circuit and setting the recording instrument in motion. Position No. 3 is attained by two different arrangements of the contact pegs, according as the one or the other of the H 61t 98 THE ELECTRIC TELEGRAPH. lines is to be used. When the apparatus ' at the inter- mediate station is to correspond with a station on the right, whilst the line on the left remains in circuit with the board, holes 2, 4, and 7 are provided with pegs. If the intermediate station now works the key, currents circulate from the c-pole of the battery through 1, lever of key, 3, earth ; and from the z-pole to R O , through the peg in hole 4, L, G 2 , to the line on the right. In receiving signals from the same side the key remains at rest, the currents arriving pass over G 2 , L, peg 4, R O , 2, coils of relay, ,1, R U , peg 7, T, back contact of key, 2, 3, earth, &c. During both transmission to and reception from the station on the right-hand side, if a current arrive from the opposite direction, it must pass through the coils of the galvanoscope G L (deflecting the pointer), over peg 2, to earth. The deflection of the galvanoscope pointer is observed by the employe, who takes his measures accordingly. When the reverse is to take place, that is to say, the intermediate station is to correspond with a station lying to the left, whilst that on the right remains in circuit with the galvanoscope, the pegs are inserted in holes, 1, 7, and 5. The signals given by the key take the following road : z, of battery, R O , peg 1, L, G L , to line on the left, apparatus at the opposite station, and earth, and from the c-pole of the battery to 1, lever of key, 3, earth. Arriving currents from the same direction come over G lt L, peg 1, R , 2, coils of relay, 1, RU, peg 7, T, key, 2, 3, earth. Those arriving from the other side deflect the pointer of G 2 , and to pass to earth by G 2 , .L, peg 5, earth. In both cases the employes have to pay attention to the galvanoscope, as in cases of emergency it is sometimes necessary to postpone the transmission or reception of a message on one line until the more pressing one from the other side has been disposed of. 67. Siemens and Halske's Intermediate Station Commuta- tor. The commutator represented in perspective in Fig. 52 is much simpler in construction than that of Nottebohm, and answers all the requirements of an intermediate station HISTORY AND PROGRESS. 99 where a single apparatus only is used. The apparatus is put in circuit between the screws 1 and 2, to which the lines L L L-2 Fig. 52. and L 2 are respectively connected, while the earth-plate is brought to one of the screws, 3 or 4. When the contact-cone is out, the current passes through the apparatus in the circuit L 2> 2, A, 1, L!, that is to say, the intermediate station receives the signals in common with a distant station. When the contact-cone is in hole in the apparatus is short-circuited, and the current passes through L 2 , 2, contact-cone in in, 1, LI. When the contact-cone is in hole I the apparatus is used as terminal of the line L 1? and the currents from LJ take their way through L I? 1, A, 2, i, earth, &c. Those from L 2 through i to earth. If the contact-cone is put in n the instrument receives from and transmits to the line L 2 on the other side, in the same way. In both positions the currents arriving by the line which is not being corresponded with pass H 2 100 THE ELECTRIC TELEGRAPH. through the galvanoscope inserted in the line circuit and give notice to the operator. 68. Borggreve's Commutator for Intermediate Stations. The many inconveniences arising from delay in the reception and transmission of messages, combined with the possibility of mistakes in stoppering the holes of Nottebohm's commu- tator for the different positions of the apparatus, has shown the necessity of furnishing intermediate stations with two apparatus. This enables them to correspond with the stations on both sides at the same time. M. Borggreve, Inspector of Telegraphs in the Prussian service, has arranged a commutator for the use of inter- mediate stations with two apparatus, in which, as in that of Siemens and Halske, only one stopper is required. It is composed of five brass slabs screwed on an insulating base of vulcanite. The way in which the commutator is connected with the two Morse-apparatus, as well as its form and appointments, is shown in the plan, Fig. 53. The line wires L 1 and L 2 are connected to the upper screws, while the lower screws on the same bars are connected with the levers of the transmitting keys K 1 , K 2 . To the binding- screw of the middle bar is attached the earth-wire, and to the screws of two intermediate bars, the cross-commutators c 1 and c 2 , whose opposite points go to the back contacts of the keys. R I and R 2 are the relays of the two Morse-appa- ratus M 1 and M 2 . The two local circuits are supplied with a common local battery, B, and the two line circuits with a common line battery, L B. When the contact plug is inserted in hole 1 of the commutator u the apparatus are both short- circuited, and currents pass through L 1 , G 1 , stopper 1 of com- mutator, G 2 , L 2 , &c. The employe sees by the deflections or otherwise of his galvanoscope-needle when the direct corre- spondence of the end or distant stations is concluded, usually by an agreed signal. If the stopper is in hole 2, apparatus 1 can correspond with the line L 1 , and apparatus 2 with the line L 2 , inde- pendently of each other. The currents arriving by L 1 go HISTORY AND PROGRESS. 101 through o 1 , u, K 1 , c 1 , and coils of R 1 , c 1 , u, 2, earth, &c. The cross-commutators are put between the back contacts of the keys and bars of the commutator u, in order to enable the operator to invert the coils of the relays, in case the residuary irth\ Fig. 53. magnetism in the soft iron cores, from continued currents in one direction, should interfere with their delicacy. The currents transmitted from the station towards L 1 , by pressing down the key K 1 , take the direction from L B, copper pole c, front contact and lever of key, first bar of commutator u, G 1 , and L 1 . The operations with apparatus 2 and L 2 are similar. When the contact peg is put in one of the holes, 3 or 4, the apparatus on the opposite side to that in which the peg put in, is ready for participating in any messages that 102 THE ELECTRIC TELEGRAPH. may be passing through the line in either direction. Suppose the peg to be inserted in the hole 3. The currents coming from the direction L I? go then from L X through' G 1 , peg 3 of the commutator c 2 , coils of relay on the right, back contact of K 2 , lever K 2 , commutator u, G 2 , L 2 , &c. The relay R 2 will be set in action and work the Morse M 2 , which will print all the signals passing from the line L 1 into the line L 2 . 69. Commutator for Stations with three or four lines from different directions. Where more than two lines from differ- ent directions meet at a station, it is required to employ a commutator by which, whilst one or two lines are being corresponded with, the remainder can be connected up, two and two, for circular correspondence. When three lines meet at a station a commutator, arranged by Borggreve, is usually employed, by which any two of the lines may be connected together in circuit with a galvano- scope, or complete recording apparatus, while the third line is open for telegraphic communication from the same station as terminal. When four lines meet, it is necessary to arrange the com- mutator so that, upon occasion, they may be connected up two and two with intervening receiving apparatus, by which the corresponding stations are in direct communication, and the intermediate stations able, at the same time, to participate in the information transmitted. Fig 54, a, gives a plan of connections of Borggreve's com- mutator for three lines, by which the following combinations are possible : 1. L 1 circular with L 11 , L m with apparatus. 2. L 1 L m , L n 3T n T in T i -L* j> -L* > -L* )> )> In using this commutator three of the holes are always stoppered at the same time. Fig. 54, 5, gives the position of the stoppers in the holes answering to the positions 1, 2, and 3, above. In the first position the circular current traverses L 1 , 1, HISTORY AND PROGRESS. 103 circular apparatus, 2, commutator, 3, L n , &c., and the currents received or transmitted by the intermediate station, L m , commutator, 4, back-contact of key, relay, 5, earth, &c. Fig. 54. In the second position, the circular- cur rent goes through L 1 , 1, circular apparatus, 2, commutator, L m , &c., and the station currents through L n , 3, commutator, 4, back-contact of key, relay, 5, earth, &c. In the third position, circular currents go over L n , 3, commutator,- 1, circular apparatus, 2, commutator, L 111 , &c., and station- currents over L 1 , commu- tator, 4, back-contact of key, relay, 5, earth, &c. The plan of connections for a station with four lines com- bines both Borggreve's commutators ; and the combinations in which the operator can arrange the lines are as follows : 1. Li with Ln, and Lm with Lrv 2. Li Lin Ln Lrv 3. LII Lin Li Liv In each combination the positions of the lines with regard to the apparatus may be : (1.) Direct, the station not receiving the message. 104 THE ELECTRIC TELEGRAPH. (2.) Circular, the station participating in the despatches. (3.) Corresponding, the station transmitting and receiving by the lines. 70. Battery commutator. When the insulation of the line varies, or by any other reason as, for instance, when an end station has to transmit to a near station with little line resist- ance in the circuit it becomes necessary to alter the strength of the current, a battery com- mutator is inserted, by which one-third, two-thirds, or the whole of the elements, may be brought into service by simply changing the place of a con- tact peg. Such an apparatus consists of four slabs, as in Fig. 55. The copper pole of the battery is connected to the screw of the bar, B, and connections from elements one-third and two-thirds of their number from the copper pole, are brought to the screws of the bars c and D respectively ; the zinc pole being con- nected in the usual way with earth. The place of the stopper in J,-fife <$->..- Fig. 65. back the tongue of the relay against its insulated contact, thereby breaking the local circuit. 77. Thomas John's Telegraph* The great force necessary to press the style against the paper strip in order to leave visible marks, and the consequent employment of a relay with local battery, were obviated by the invention of an Austrian telegraph engineer, Mr. John, as early as 1854. * Brix. Journal,'vi. p. 5. HISTORY AND PROGRESS. 121 In his apparatus the marks were made upon the paper strip by means of a small circular disc of metal kept revolv- ing in a dish of coloured fluid, and pressed gently against the paper, when the armature of the electro -magnet was attracted. Beyond this, all the rest of the Morse arrange- ments of clockwork, &c., remained unaltered. The object of the invention was solely to diminish the force necessary for marking the paper, so that the electro-magnets might be able to work the beam when inserted directly in the line. And in this the method has signally succeeded, as the almost universal adoption of modifications of it has proved. The apparatus, as constructed by the inventor, was by no means a piece of elaborate workmanship, nor were his Tig. 66. arrangements of levers and paper- guides quite so commodious as were desirable, but the happy fate of the Morse system is, probably, in no slight degree indebted to this idea. In Fig. 66 the clockwork is left out. E is the electro- 122 THE ELECTRIC TELEGRAPH. magnet, whose armature, D, is affixed to one end of the beam A c, and plays between adjusting screws. The beam is supported at z, and has on its shorter arm, A, a connecting rod, B, hinging at h on the horizontal arm of the bent lever h I M, which turns on an axis at x. At M the lever carries the printing-disc, the lower segment of which is immersed in a dish of indian-ink, R. On the axis of this printing disc is a small pulley, p, with a cord passing over, and receiving motion from the pulley p' below. P' is attached to a drum, Q, which revolves with it. The paper strip, shown in the figure by dotted lines, is led to the apparatus from v under- neath a guide-drum, H, at the back of the drum Q, which it rubs against and turns round (thereby imparting a rotary motion to P, P*, and to the printing disc), behind the drum T, over a metal edge, s, and across the stage L o, where the message is read off. The purpose of the metal edge s is t to present a sharp corner of the paper to the printing disc in order that it may receive well-defined marks. "When the electro-magnet is in action the armature D is attracted, the connecting-rod lifted up, and the inking disc, which needs not be at a greater distance than half a milli- meter, pressed gently against the paper on the part which is travelling at the moment over the edge s. During this time the motion of the paper keeps the printing disc revolving in the ink, which causes a freshly inked, surface to be always presented. 78. The Direct Working Ink- Recorder of Beaiidoin and Digney* The difficulties in the way of the arrangement proposed by M. John were well considered by Digney, who modified it accordingly, with the view of rendering it simpler in its construction and surer in its effects. Instead of making the printing disc approach the paper strip, Digney lifts the paper up to the disc, which he keeps rotating on a fixed axis, moved by the same mechanism which draws the paper through. Over the top of the disc he places a roller of felt or cloth, moistened with oil colour, * " Revue des Applications de 1' lectricit en 18578," par Du Moncel, p. 169. HISTORY AND PROGRESS. 123 which turns by the friction of the printing disc, and keeps the periphery of the latter always freshly inked. The paper strip passes underneath the disc, over a knife-edge, forming the continuation of a beam, carrying, at its other end, the armature of the electro-magnet. When the latter is attracted, therefore, the knife-edge presses the paper against the revolv- ing disc. John's idea is thus reversed, but the principle remains the same. Fig. 67 represents an elevation of the apparatus, the clock- Fig. 67. work being in the interior. The parts directly turned by the clockwork are the roller R, and the printing disc D. The paper strip is drawn from the drum p through the slit G, under the guide-pulley u, between the printing disc D, and knife-edge K, between the rollers R and B I? and across the 124 THE ELECTRIC TELEGRAPH. horizontal stage s. The jockey- roller R X turning freely within the frame F, and pressed down by the spring Y, holds the paper strip tight upon the roller R, so that, as the latter turns round, a progressive motion is imparted to it. The jockey R x can be lifted up from the paper by turning the lever H to the left. The force with which it presses upon the paper on the roller R is regulated by means of the adjusting screw s, against which the end of the spring Y abuts. T is the feeding roller of felt, kept moist with fresh oil-colour, and turning freely on its axis in a frame sup- ported on a regulating-screw on the axis B. The purpose of the screw is to move the roller in or out a little to pre- vent its surface always riding over the disc in the same line. The axis B consists of a pin fixed in the side of the appa- ratus, from which the frame containing T can be easily removed. T rests, when at work, a little obliquely, by its own weight only, on the top of the printing disc, with which it revolves. M is the electro-magnet, A its armature, supported by the beam L L, turning on its axis I, and is held in its position of rest by the spring /, stretched between a right-angled arm, /, of the beam and the adjusting screw, S L . When the armature is attracted to the poles of the electro- magnet, the knife-edge, forming the end of the continuation K K of the beam, is raised, and lifts the paper against the disc D, which revolves in a reverse direction to that of R and to the passage of the paper, against which it rubs so long as the armature is kept down. On the cessation of the current the spring pulls back the beam, and the paper-strip falls off the disc. This instrument is worked without relay and local battery, and has become a great favourite with the employes; the deciphering of inked letters being infinitely less fatiguing than that of embossed. The renewal of ink, when the ap- paratus is in full work, is not required more than once a day Thus the principal difficulties in the way of the Morse apparatus are removed. The Digney instrument has been HISTORY AND PROGRESS. 125 found by the French Administration of Telegraphs to work so well, that they have adopted it for use on the Government lines in France and the colonies. This success is in a great measure due to the nice discrimination between the sizes of the movable and fixed portions of the apparatus having reduced the vis inertia of levers, armatures, &c., to a mini- mum, whilst amply sufficient strength is insured to effect the complete marking of the paper. M. Guillemin, in 1862, made a series of interesting expe- riments with this apparatus to determine the maximum number of elementary signals, and, consequently, how many words it was capable of recording in a given time. The transmitting apparatus he employed consisted of four wheels of twenty-five centimetres diameter on a common axis : one of them made dots, the second dashes, whilst the two others served to discharge the line after every elementary signal. The words France and Paris, which in the Morse alphabet represent a mean of the French words, were repeated on a line of 750 kilometres, in fine weather, thirty times per minute, and in wet weather he easily attained the rate of forty words. On a line of 450 kilometres, passing by Le Havre, the reception was augmented to seventy-five words per minute six times that which the employes are able to attain with the hand. 79. Direct Working Ink- Writers of Siemens and Halske. An important modification of Digney's instrument is made by Siemens and Halske, who substitute a small inverted bottle containing ink, and secured by a felt stopper, for the inking roller of felt, described in the preceding paragraph. This arrangement is represented in Fig. 68. B B is a small inverted glass bottle containing the printing fluid ; its neck is cemented into a brass ring, c, fitting into a collar, D. At the back of the collar is a horizontal hollow axis, E, supported by a pin fixed in the side of the apparatus, on which the whole thing turns, and from which it may be removed in the same way as Digney's felt roller. A stopper of thick felt, //, is put into the mouth of the bottle to allow the colouring fluid to come through very gradually. The bottle presses, 126 THE ELECTRIC TELEGRAPH. by its own weight, upon the printing disc, which is in con- nection with the clockwork, and performs the same functions as the corresponding member of Digney's instrument print- ing on the upper side of the paper strip p p, which is lifted by the knife-edge, K. Both this method and that of Digney are not entirely Fig. 68. without objection, however, on account of the printing disc and paper being underneath the reservoir of ink, from which, when the apparatus stands inactive some time, the colouring fluid frequently runs down and makes a blot on the paper ; besides this, they are both liable to the objection that the surface of the felt quickly dries up in warm weather. To remedy these defects, Siemens and Halske have made a second and still more valuable improvement in the inking process. It consists in again reversing the order of things, in making the printing disc revolve with its lower half immersed in a dish of colouring fluid, and in lifting the disc up against the paper, which runs above it, instead of pressing the paper against the disc. This is the perfection of the mechanical arrangements which M. John was able, only in an incomplete way, to carry out. This modification is shown in Fig. .69. A is a glass phial HISTORY AND PROGRESS. 127 cemented into a brass neck, a, supported between screw- points at n, and by the point of the levelling screw, b. The forepart of the brass neck a is cut out in a curve, and forms a gutter, in which the lower part of the printing disc c dips. At d, a round hole in the upper part of the neck, usually covered up by a metal cap, facilitates the filling of the phial. The index e is intended to guide the operator Fig. 69. in adjusting the niveau of the printing fluid in the phial, so at to cover the requisite segment of the disc c. 80. Arrangement of a Board with the Direct-working Ink- writer. The direct- working ink-writers of Messrs. Siemens and Halske are mounted on mahogany boards with the trans- mitting keys, galvanoscopes, terminals, and connections. The accompanying plan, Fig. 70, gives the arrangement of two boards fitted up with all the apparatus, and properly connected together, as they ,are used at intermediate stations for translation, terminal work, &c. On each of the boards are mounted an ink -writer, H, a simple transmitting key, D, a line galvanoscope, B, and twelve terminals for the internal and external connections. The five terminals on the right-hand side are filed out, so as to admit a contact peg between their ends and a common HISTORY AND PROGRESS. 129 terminal bar, F, connected to one side of the galvanoscope. By supplying one or more of these holes with contact pegs, the apparatus may be arranged in different positions. When the two apparatus, I. and II., are to work as ter- minals to the two lines L X and L 2 respectively, contact pegs are inserted in the holes s, and between and E, of each of the boards. The circuit is then complete for receiving from L!, through L, B X , F, s, 1 and 2 of D X , coils of H X , > E, earth ; and for transmitting, when the key is pressed down, through Earth, E, ^, z, Battery, c, ^, and 3 and 1 of T> L , s, F, B I? L, LJ, &c. For translation between L X and L 2 , contact pegs are put in the holes between T and F, and between ^ and E, of each of the apparatus. The currents from L L then pass through L, B I? F, T, i, to App. II., i , 1 and 2 of H 2 , n, back to App. L, ii, Coils of H L , g, E, Earth, &c. The beam of H L is deflected, and closes the circuit, including Earth, E, , z, Battery, c, to App. II., n r , back to App. L, in, 3 and 1 of H 1? i, to App. II., T, F, B 2 , L, L 2 , &0. Both apparatus are intermediate when a contact peg is inserted in the hole between s and F of each of the boards. In this case both apparatus are moved by the same currents, and record simultaneously. Either of the apparatus may be used alone as intermediate or circular apparatus, when its contact peg is between s and F, while the other apparatus has a peg in the hole between D and F. Both apparatus are out of circuit when the pegs of both are put in the holes , F. The through-circuit, L X , L, B X , F, D, z, to App. II., z, p, F, B 2 , L, L 2 , is then established. The insertion of pegs in the holes between L and F of the boards short-circuits the galvanoscopes, and between F and L and F and E, at the same time, puts the lines directly to earth a position always advisable during a thunderstorm. When the ink- writer is used with a relay, the arrange- ments of the board undergo some modification. This consists in the addition to each board of a relay, G (Fig. 71), and of two terminals, c z, for the poles of a common local battery. K 130 THE ELECTRIC TELEGRAPH. The internal connections of the board with regard to the ter- minals are the same, with the exception that the coils of the relays are substituted for those of the ink- writer between n. and z, and that the local circuit, shown in dotted lines, is added. HISTORY AND PROGRESS. 131 81. Morse Telegraph worked by Induction Currents. In numerous instances has magneto- electricity been pressed into the service of the telegraph, but always in conjunction with the step-by-step or needle systems. For the Morse it was considered useless, as the currents developed, being only of momentary duration, are only capable of themselves of giving successions of dots, whereas the Morse alphabet requires also an elementary signal of longer duration. This difficulty was removed by the ingenious invention of Siemens and Halske, in the construction of a relay, the tongue or armature of which would remain of itself on either contact, when once deflected, until a current different from the one last sent through removed it to the other side. If, therefore, when the tongue was in a state of rest on the insulated contact, a momentary current of magneto -electricity were sent in the right direction through the coils of the relay, the armature would move to the local contact, and would remain there, closing the local circuit, notwithstanding the current which deflected it had long since vanished, until a current in the opposite direction brought it back to the reposing contact. In this way either lines or dots could be produced at pleasure by regulating the interval between the succeeding currents. This principle is the same as that used at a later date in the indicator of the magneto-electric telegraph of the same inventors, which has already been described. The solution of this problem has placed at the command of the telegraphist a source of electricity of much greater intensity for working the Morse instruments through great distances, than the voltaic current, and which he is able to produce at a considerably less expense. The complete apparatus consists of : - A transmitting key, An induction apparatus, A polarised relay, and A Morse recording instrument worked by a local battery. The induction apparatus sometimes used consists of an iron core a bundle of soft iron wires surrounded by convolu- K2 132 THE ELECTRIC TELEGRAPH. tions of thick copper wire, forming the primary, and by a long fine wire outside this, forming the secondary coil. The primary coil is put in circuit with the key and with a battery of large surface and little internal resistance. The secondary coil is connected at one end with the earth, at the other with the line. It is sometimes divided into two parts, which may be connected parallel or in series, according to the resistance of the line. 82. The polarised relay differs in its construction from all the others. There is no spring employed to pull back the armature after it is let go by the poles of the electro- magnet. Fig. 72 is a sectional view in the direction of the armature, and Fig. 73 a top view of the relay. The perpendicular electro-magnet E is composed of two cores of soft iron united below, in the ordinary manner, by a cross-bar, A, also of soft-iron. The coils of wire terminate at the screws 1 and 2. The north end N of an angular bent permanent magnet, N s, is screwed on to the cross-bar A, to which it communicates north polarity beyond the point of contact, and also to both the cores and poles of the electro -magnet E. The soft iron tongue c is supported on an axis in a slit in the south end s of the permanent magnet, and thus receives south polarity. This tongue is so placed that it may oscillate between the north poles N and N' of the electro-magnet. Its play is limited by the contacts D and D f . D is used as a contact for closing the local circuit, in which are included the printing instrument and the local battery, when the tongue c strikes against it. D' is furnished with an agate point, and while the tongue rests against it, the local circuit is open. A and B are the terminal screws of this circuit. Whilst it is situated equidistant from both the north polarised ends, N and N 7 , of the electro-magnet, the south polarised tongue c is attracted towards each of them with equal force. When, at the sending station, the key is pressed down, the current of the local battery circulates in the primary wire of the induction coil. A momentary induced positive current HISTORY AND PROGRESS. 133 passes through the line and relay, which has the effect of Fig. 72. Fig. 73. magnetising the pole N of the electro-magnet north, and the 134 THE ELECTRIC TELEGRAPH. pole N' south ; but as both poles were previously north by the influence of the permanent magnet N s, the effect of the current is to strengthen the north magnetism of N, and at the same time to weaken only that of N'. The tongue c is, there- fore, attracted to the pole N with double force, and remains on that side after the cessation of the current, attracted by the pole N, whose distance from c is then less than that of N'. The platinum contact of c remains against D, and closes the local circuit until the key at the transmitting station is let go, and the cessation of the current in the primary wire of the induction apparatus induces a negative current in the secondary coil, line, and relay, which has the reverse eifect of the last current, strengthening the north magnetism of N', and correspondingly weaking that of N. The pole N' there- upon attracts c against the insulated point D', where it rests until another positive current passes and throws it off again. The Morse recording instrument is of the usual construc- tion of Digney, or Siemens and Halske, described above. 83. A plan of this admirable system is shown in Fig. 74, arranged for two stations, c and c are the induction coils, of which e e are the soft iron cores ; the limits of the primary and secondary coils are shown by concentric rings. K is a transmitting key, which closes two working contacts in front ; R the polarised relay ; B the local battery ; and i the receiving instrument. At each of the stations the middle contact of the key is connected to line, and also to one end of the primary wire of the induction apparatus. The battery is included in two circuits : first, between the remaining end of the primary coil and the second contact b of the key ; and, secondly, in the ordinary local circuit of the relay and recording instrument. One end of the secondary wire of the induction-coil is to earth, the other connected with the first contact a of the key, the back or reposing contact c leading through relay to earth. The key differs slightly from that used in the ordinary Morse circuits, having, as we have seen, two working con- HISTORY AND PROGRESS. 135 tacts. The lever is furnished with a spring, which presses upon the contact a, by which, when the key is lifted up, the contact with a is interrupted an appreciable time after that with b. This is necessary, because if they were both interrupted at the same instant it is evident that no induction current could arise in the secondary coil, its circuit being broken. J \Eartl\ Fig. 74, By the contact a continuing an instant longer than b, how- ever, the induction current which follows the interruption at b has time to pass over a and through the line. The Morse telegraph has been worked by this system of induction currents to- a considerable extent on the lines in Eussia, Bavaria, and Hanover. Sibeller says that messages have been sent direct, without translation, by this method, on a line of 200 German miles, equal to nearly one thousand English. Compared with the methods of working the Morse tele- graph by voltaic electricity, that of induction currents offers many advantages ; the line batteries are opened, and spaces between the signals are given by reversed circuits, which 136 THE ELECTRIC TELEGRAPH. work always cleaner than those given by making and breaking the same current. The polarised relay above described is also profitably employed on lines worked only with galvanic currents, with which it is found to be far more delicate than the relays with springs. It is, however, necessary to give the armature c (Fig. 73) a bias on the side D', which is done by advancing the soft iron continuation of the pole N' of the electro-magnet a little nearer to the armature than N, by which, when no current passes, the tongue is held against the insulated contact, and the distances may be so finely adjusted that a very weak current suffices to move it. 84. The Magneto- Induction Key. Instead of the Morse key, induction coil, and local battery, Siemens and Halske use also an instrument arranged in the form of a key, by Fig. 75. which a coil of wire, wound on a soft iron armature, is oscillated between the poles of a permanent magnet, and develops alternate currents for working the polarised relay. The magneto-induction key is shown in Fig. 75, in per- spective, s and N are two rows of permanent bar-magnets ; the upper ones with their north ends, and the lower ones HISTORY AND PROGRESS. 137 with their south ends in contact with a stout plate, p, of soft iron in the same way as in the transmitter of the magneto-electric pointer telegraph of the same inventors. Between the poles of this system, and oscillated in an angle of a few degrees by means of a handle, H, in the frame be- tween two screw points, is the soft iron armature, as long as the magnet system is wide, cut in deep longitudinal grooves, on opposite sides, as is shown by the sec- tional sketch Fig. 67. In these grooves the coil c of fine insulated wire is wound. The play of the handle is limited by two adjusting screws in the frame A. When at rest, the handle is held against the upper screw by a spiral spring, s, stretched between the handle and front of the triangular piece D on the top. One end of the coil of wire on the armature is attached to the screw k, on the terminal K, from which one connec- tion goes to line and another to the screw w, at the foot of the frame A. The other end of the coil is connected with the metal frame supporting the armature, and through the axis/, to the upright support Q, from which a leading wire goes to terminal t and earth. "When a current arrives while the instrument is in circuit with the line, it goes from L over R, w, upper adjusting screw in A, through handle H, axis /, Q, t, earth, without traversing the coil. This is the purpose of the connection between R and w. When the handle is pressed down, the polarity of the armature is reversed, and a positive magneto-electric current induced in the coil, which circulates also in the line wire, and deflects the tongue of the polarised relay at the receiv- Fig. 76. 138 THE ELECTRIC TELEGRAPH. ing station, from the insulated point, and closes the local circuit so long as the key is held down, and no negative current induced by letting the key go back to its position of rest. 85. Siemens and Halske's Polarised Ink Recorder. A novel and very useful invention, where the attention of the employe is not invariably to be relied on, or he is occupied with other duties besides his instrument, was introduced by Messrs. Siemens and Halske, in the construction of their polarised ink recorder, in making the clockwork which is used to unwind the paper strip, self- starting. That is to say, as soon as a current arrives, and so long as it lasts, the clock- work is allowed to run and to draw the paper strip over the knife edge of the printing lever underneath the printing disc ; but when the signals stop, the clockwork is arrested also. In principle the electro-magnet is the same as that of the polarised relay described above. A strong angular permanent steel magnet polarises the two cores of an electro-magnet, which partake both of north polarity ; while between their ends, the printing beam of soft iron, moving on an axis in the other end of the permanent magnet, has the opposite polarity. The clockwork does not in any material point differ from that of the ordinary instruments. A hollow drum is turned by means of a mainspring in its interior, which also puts in motion the entire train of wheels, as well as the printing and driving rollers. A fly regulates the motion of the whole. The self-starting apparatus is arranged as follows : Close to the electro-magnet of the printing lever is. a smaller electro-magnet, A B, Fig. 77, called the releasing magnet, the coils of which are in the same circuit as those of the larger one. When a current passes, therefore, through, both their armatures are attracted at the same instant. The armature c of the releasing magnet is carried by the releasing beam, turning on the axis H. At the other end of the releasing beam is a friction spring, o E, which, when the armature is in its position of rest, presses upon the ivory break- wheel F by means of a weight. The last wheel of the HISTORY AND PROGRESS. 139 system is carried upon the axis on which the drum F and a fly are fixed. The clockwork is therefore stopped when the armature is at rest, or when no current passes. When, however, the arma- ture is attracted, the friction-spring E is raised from F, and the clock- work starts ; a boot, T, hanging from the beam and resting on the rim of the revolving drum M, is lifted up, and continues to j^ 7^ dance upon the rim by the friction of the drum in revolving. After the current ceases the armature is released, and the boot, descending on the drum, is carried off by the rotation, not allowing the spring E to stop the clockwork until after the last current has ceased for some seconds. The starting of the clockwork may be effected at pleasure by the operator, by raising the friction- spring E, and it may be stopped by pressing it against the ivory break- wheel. 86. The Polarised Ink Recorder used as a Submarine Key. Translation. The manner in which the ordinary Morse apparatus is connected up for translation has already been explained. It is performed with the polarised apparatus as follows : When the tongue of the relay G 15 Fig. 78, of the receiving apparatus is deflected against the local contact, the local battery is put into circuit, and the printing instrument draws the beam H X from the screw 2 to the screw 3, the former being connected with the counteracting battery K, and the latter with the line battery + K. When the printing beam is connected by s 2 and I with the line L 2 , supposing the printing lever H L to be resting against the contact 2, a negative current enters the line L 2 in the following way : K, earth, opposite station apparatus, L 2 , /, S 2 , H X , 2, K. But when the printing lever H t is attached to the contact 3, 140 THE ELECTRIC TELEGRAPH. a positive current enters the line L 2 as follows : 4- K, 3, H t , S 2 , /, L 2 , opposite station apparatus, earth + K. It is therefore evident that the printing instrument will translate any signals it may receive to the next station. As there is no communication between the line and the Fig. 78. point s l9 the discharge current does not pass through the relay. Suppose now the relay G 2 placed in circuit for recep- tion of signals, by transferring the switch from s 2 to s lf the arm I, in going over, rubs against the earth contact s 3 , and consequently the line will be discharged before being con- nected with the relay. At the back of the apparatus, the axis supporting the releasing lever carries also a commutating beam, which, when at rest, makes contact with s l9 in connection with the relay ; but when the armature of the small electro-magnet is attracted, the commutating lever makes contact at the point s 2 , in communication with the printing lever. The discharge of the line is effected by means of the boot T, Fig. 77. This boot is insulated at the toe and heel, but not in the middle of the sole, so that either in a state of repose or when dancing, no electrical connection exits between the boot, that is to say, HISTORY AND PROGRESS. 141 the lever c H o, and the drum M. As soon, however, as the current ceases, and the boot is pushed sideways, the conductor, let into the middle of the sole, comes into contact with an insulated platinum ring on -the edge of the drum, and com- municates through the spring S 3 with earth. The course of the current is shown in Fig. 79. In repose it would be as follows : L 2 , /, s 1 , G 2 , earth, opposite station instrument, L 2 . Fig. 79. In this position the relay is in circuit. If, however, the printing lever H^ be attracted (through the agency of an inward current through L X ) the releasing magnet m will at the same instant attract the lever I to the contact 82, and thus break the relay circuit at S L . The course of the current will then be as follows : + K, 3, printing lever, S 2 , /, L 2 , opposite station instrument, earth, + K ; and as soon as the printing lever makes contact at 2, as follows : K, earth, oppo- site station instrument, L 2 , I, s 2 , printing beam, H L , 2, K. Lastly, if the conducting sole of the boot be in contact with the platinum ring o^ of the drum, the following will be the manner of discharging the line : L 2 , 1, n, o lt S 3 , earth. The Translating Spring. In translating, the introduction of the main and counteracting batteries into the line is effected by means of the printing beam, in order that when in- 142 THE ELECTRIC TELEGRAPH. termediate stations connect up for translation, the signals may be passed through the whole with no interference on the part of the employes. When the printing beam is drawn from the upper contact 2, to the lower one B, by the galvanic current, a certain interval is necessary, and this interval could be subtracted from the time the armature is actually held down, and consequently from the lengths of signals on the paper. The printing beam would therefore actually be attracted for a shorter space of time than the key is held down at the sending station. The diminution would be repeated at each following station, so that, in fine, if there were several stations translating, the primary signals would have to be transmitted very slowly, in order that they might be legibly received at the terminal instrument. This is remedied, in an ingenious way, by Siemens and Halske's translating spring, which is situate under the printing beam, and immediately above the contact point 3. As soon as the beam commences its downward motion, following the attraction of the armature, the spring touches the contact point 3 ; and when the beam leaves the point, the spring still presses upon it for a time, and only separates from it at the last moment, so that the manipulator need not give any special attention to the length of his signals in working the key, as they will be transmitted exactly as he sends them, provided the line be properly discharged. 87. Complete Submarine Board. For use on submarine lines, Siemens and Halske have had the polarised ink recorder, polarised relay, submarine key, and other apparatus, set up on slate slabs, the connections of a permanent character between the various parts being made underneath the board. The binding screws of the local circuit are marked with italic, and those of the line circuit in Arabic numerals and letters. Fig. 80 gives a theoretical plan of the connections of two slabs at an intermediate or translating station, and Fig. 81 the arrangement of the various parts of the apparatus in one of the slabs. B B (Fig. 81) are the galvanoscopes, c the translation commutator, D the submarine key, F the current commutator, G the relay, H the printing instrument, and M HISTORY AND PROGRESS. 143 the circuit breaker. Behind the galvanoscopes B B are fourteen terminal screws, to which the wires leading from the apparatus Fig. 80. are brought, their arrangement being as follows : L is the line wire, E the earth wire, c copper pole, and z zinc-pole of battery. The local circuit, indicated by dotted lines, shows the terminals between which the local battery is connected. HISTORY AND PROGRESS. 145 The remaining terminals, T, i, n, n, &c., are for the connec- tions between the two slabs. The following are the different positions of the commu- tators, &c., for different uses of the apparatus : 1. THE APPARATUS AS TERMINAL STATION. POSITION I. Both translation commutators are stoppered at S. Both current commutators are at 1. Contact cones in both the circuit-breakers. A) Apparatus I. (left] receives signals. The current coming from Z t passes through the screw, Z,, the galvanoscope (1, B P 2), the translating commutator, c,, contact cone in S, key (1, D P 4), galvanoscope BJ current commutator (1, p,, 3), relay (1, G P 2), current commutator (2, F P 4), screw Z, circuit-breaker, M n screwZ', and earth, and through the earth back to the battery of the sending station. The tongue of the relay, 'G P is attracted against the metal contact point, and the local circuit closed as follows : Local Battery (Z t i, C7), the screw, A, of the printing instru- ment, a, through the releasing magnet, b, at the same time through both coils of printing magnet, , to the relay, A, tongue, metal contact, , and back to the Z of the local battery. Thereupon both the magnets of the instrument are made to attract ; the releasing magnet sets the clockwork in motion, and the printing lever of the other magnet is held down until an opposite current coming from L l repels the tongue of the relay from the metal contact. .5) The Apparatus I. is made to transmit signals. The key D l is drawn sideways, so that the spring * a is pressed against the contact, 2. The counteracting battery, K, is then in circuit as follows : (Z, K, C) Z, MJ, E, earth, opposite station apparatus, Z p lightning guard A, Apparatus I. (1, B,, 2), (C,, ), (1, D,, *g, 2), back to the battery, K. A negative current, therefore, passes through the line and the L 146 THE ELECTRIC TELEGRAPH. relay of the opposite station, the tongue of which is consequently pressed firmly against the stone. 2. APPARATUS I. AND IT. TRANSLATE. POSITION II. Both translation commutators in T. Both current commutators at 1, and both circuit breakers stoppered. The self-releasing clockwork is in action in both in- struments. In this case a positive current from the opposite station would take the following direction : Z, (1, Bp 2) (C p T] i, i, (1, 2 , *p 4) n, n, 4, BI (1, P,, 3) (1, Gp 2) (2, F p 4) Z, Mp JE, PL, and through the earth to the opposite station battery. The relay G, completes the local circuit, and therefore both magnets of the printing instrument, Hp become active ; the releasing magnet allowing the clockwork to run, and the printing magnet working the beam. When, in so doing, the printing lever touches the contact, 3, the battery -f-K is put in circuit with the line, Z 2 , and when it touches the contact, 2, the counteracting battery K will be similarly put in circuit with the same line. 3. APPARATUS I. INTERMEDIATE BETWEEN BOTH LINES. POSITION III. a) For receiving legible signals from L l the commutators of Appa- ratus Z must be respectively in S and 1. In Apparatus II. the translating commutator is in D, and com- munication with the earth cut off at M l and M 2 . b) For receiving signals from Z 2 , on Apparatus Z, the com- mutators are in /Sand 2 ; in Apparatus ZZ, translating commutator in D, communication with the earth being cut off at M, and M 2 . When the commutator of the Apparatus ZZ is stoppered in Z>, the latter instrument is entirely out of circuit, by the connection (c 2 , D) Z ; while a current coming from Z, will pass through Z p l (c,),!, 1, f y 4 of D, 1 and 3 of F lt Z, Z y D of <7 2 , and so on toZ e . 88. The Sounder. In America the method of reading by sound has almost entirely superseded that of recording. The apparatus used is called a sounder. In consists simply of an electro-magnet erected on a wooden base board, with an HISTORY AND PROGRESS. 147 armature attached to one end of a lever, at the other end of which is a spiral spring for drawing back the armature when the current ceases, the oscillation of the lever being limited by anvils. When a current circulates round the cores the magnetism induced attracts the armature, by which the end of the lever strikes on the top of the lower anvil, and produces a sharp noise ; on the cessation of the current, the armature is let go, and the lever drawn back by the tension of the spring strikes with a less intense noise on the upper anvil. Adjusting screws attached to the spring enable the operator to regulate the sounds of the beats on the two anvils. With this excep- tion the whole arrangements of the Morse system, of relays, keys, &c., remain the same. Prescott says : " It was soon discovered after the introduc- tion of the Morse system of telegraphs that words could be read by the click of the magnet ; but paper was used upon which the arbitrary alphabet of dots and lines was indented by the instrument for all matters of business up to 1852, and by many lines even later ; but at the present time there is scarcely an office of any importance in the United States where the paper is used to receive the record." The same author says that since the abolition of the paper upon the Morse lines errors rarely occur ; that the ear of the employe is found to be a much more reliable organ than the eye ; not one error being made in reading by sound, while at least ten were made formerly in reading from the paper. The system has, nevertheless, the disadvantage that it leaves no record for the justification of the operator. In France the messages are invariably recorded by the Digney instrument, but it is not unfrequent that the employes read the message by ear before looking at the paper. In addition to the methods already mentioned of recording messages by the Morse on paper strips by the decomposition of salts, by scoring or embossing, and by inking, it has likewise been attempted to attain the same object by burning holes in, and by scorching the paper. Home, for example, suggested the employment of a bent L 2 148 THE ELECTRIC TELEGRAPH. piece of platinum wire kept at a white heat by the passage of a voltaic current, in place of the inking apparatus or style. Messrs. Farmer and Batchelder of Boston constructed a recording telegraph in which they only scorched the paper. A platinum point was connected by a lever with the arma- ture of an electro-magnet, and brought into contact with tissue paper by opening and closing the circuit. The platinum point was kept red hot by a spirit lamp under- neath. 90. Morse Apparatus worked by Closed Circuit. The method adopted by Kramer, and also by Morse in an early telegraph of his, of working by interruptions of a current instead of by occasional currents, has been taken up by Frischen, and used by him on the Hanoverian railway lines for working the Morse instruments. " A great advantage of this arrangement/' says Frischen,* " is that, on lines with several intermediate stations, only the terminal station requires to be provided with a line battery, whilst a local battery is necessary at each intermediate station. By this the cost of batteries is considerably reduced ; besides which, the relays, by reason of the uniform current, do not require often to be adjusted; and the employe is enabled to place confidence in the call signal without con- tinually having the apparatus under his eye. The last point is of particular importance when the employe entrusted with the care of the apparatus has other business to attend to, which is often the case on railway lines. In arranging a Morse line for closed circuit between two stations the line current must traverse the galvanometer, relays, and keys in such a way as to hold the tongues of the relays on their reposing or insulated contacts, and the galva- nometer needles permanently deflected. When a signal is given by interrupting the circuit, the force of the adjusting spring of the ordinary relay, or the superior attraction of the nearer pole of the polarised relay, must be sufficient to overcome any residuary magnetism which may be in the cores of the electro-magnet, and by pulling it against * Brix. Journal, v. p. 214. HISTORY AND PROGRESS. 149 the working contact, close the local battery and work the Morse. Fig. 82 shows the connections of the apparatus for two stations with the Morse recorder. R and R' are the polarised relays, the coils of which are connected with line and with the levers of the keys T and T' respectively. To the back Station,! Stu.tion.JI. *-' Line Fig. 82. contacts of the keys are brought the opposite poles of the two line batteries ; L B having the zinc and i/ B' the copper pole to earth. The front contacts of the keys are used only as stops or anvils without electrical connections. Between the working contacts of the relays and their armatures, the local batteries B B' and the Morse apparatus M M' are inserted. Whilst the keys repose on the back contacts, as shown in the figure, the currents of the two line batteries circulate, in the same direction, through the line and coils of the relays : that 150 THE ELECTRIC TELEGRAPH. of L B goes from z to E (earth), E' of Station II., c, battery L B' (whose current adds itself to that of L B), c, back contact of T', lever, coils of R', line, coils of R, over the lever T, and back to c of L B. The armatures of R and R' are therefore continually attracted to the insulated contacts, and the local circuits are open. When one of the keys is pressed down upon its front contact and the circuit interrupted, the armatures of both the relays are simultaneously released, falling upon their working contacts, closing the local circuits, and putting the Morse machines, M and M', in motion. Frischen, who has more than any one else given his attention to the application of closed circuit methods for railway and other lines, had constructed plans of connections for station apparatus, for translation between two lines worked by closed circuit, and also for translation between a line with closed circuit and another worked with intermittent currents. Fig. 83 represents a plan of connections for translation with closed circuits. RI and R 2 are the relays of the two apparatus at the intermediate station ; K s T two switches, which, when the arms are in the middle between T and s, establish contact between 1 and K ; when the arm of a switch rests on T, contact is made between it and T, and that between K and 1 interrupted by means of a cam of ivory on the arm, which lifts it up ; and, lastly, when the arm rests on s, contact is established between them, whilst that between K and 1 is also made. M and M' are the two Morses of the usual construction, but with an additional contact at the end of the lever which makes and breaks contact between T of the switch and the line battery. K and K' are the manipulating keys, B the local battery common to both local circuits, and L B the line battery of the station. For station work, the handles of the switches on both sides are placed on s ; and the circuits of these and the corre- sponding terminal stations are established through the line, coils of the relay, arms and contact s (of switches), 1 and 2 of keys, L B, earth, &c. The tongues of the relays are there- HISTORY AND PROGRESS. 151 fore held on the insulated contacts. On breaking the line circuit at any point, the local circuits are closed, and the current of B goes through the coils of the printing instru- ments, which are set in action. For translation, the arms of the switches are shifted to Fig. 83. the spring terminals T ; the circulating current then takes its way through L', R', K and T (of switch), over to the other side, beam of Morse I/, D, line battery L B, earth, &c. The current of L B goes also to the Morse on the other side, switch T, K', relay R 2 , and line L 2 . 152 THE ELECTRIC TELEGRAPH. When the line L' is interrupted anywhere, the temporary magnetism disappears from the cores of the relay bobbins the armature falls off against its reposing contact, and the left side of the local circuit with Morse M is closed. The depression of the printing beam separates the contact spring from the screw point, by which the line L" is also interrupted ; and armature of the relay R 2 falls against its reposing current. On the deflection of the beam G of Morse M, the contact with H having been broken, the local circuit is interrupted ; and, notwithstanding the action of the relay R', the Morse M' does not move. Thus, in translating, both relays are in motion, but only one Morse apparatus that on the side from which the message comes. Sometimes another method of translation is adopted, by which the messages arriving at an intermediate station by a line with continuous current, are translated into aline worked by intermittent currents, and vice versd. This is often found useful in shunting despatches between lines already arranged with different systems. Fig. 84 gives a plan of connections for this operation, in which the single parts of the apparatus on the right-hand side are the same as in Fig. 83. Those on the -left-hand side are supplied with the additional contacts of the switch and Morse. The line battery is divided in halves. It is supposed that Line I. on the left, L, is worked by con- tinuous, and Line II. on the right, by intermittent currents. For station work, the continuous current circulating in Line I. (R, switch, 2, s, K 1, 2, L B, &c.) and Earth is interrupted. The armature of R then goes from the insulated to the working contact, and closes the local circuit by which the Morse M is moved. On the other side, the currents arriving go from Line II. (switch 2, s, K', 1, 2, R' E) to earth, and close the other local circuit setting M' in motion. To transmit a message from either side, the arm 2 of each of the switches rests on s, and the keys K and K are simply manipulated. In translation, the arms of the switches rest on the ter- HISTORY AND PROGRESS. 153 minals T. The continuous current circulates in Line L, R, 2, T of switch, over to the right hand, F, D, L B, Earth, &c. An interruption in this circuit caused the relay to work, and therefore the Morse M, which directs a current, corre- sponding to each interruption of Line L, in the circuit E, L B, M, 2, G, to T of switch, 2, to Line II. A current from Line II. passes through 2 and T of switch to Morse M, G, 1, coils of relay, R', to earth. The Morse M', 154 THE ELECTRIC TELEGRAPH. thereby put in motion, interrupts the current of Line I. between the screw F, and spring D ; and, by the separation of the beam of M' from the upper contact H, divides the local circuit of M, which, therefore ; in spite of the movement of its relay, remains passive. 91. Methods of Telegraphing in Opposite Directions at the same time in a ftingle Wire. This feat was for a long time considered to be an impossible one. Judging from the plans employed for ordinary circuits, it was urged that on sending currents of equal intensity in opposite directions from the ends of a single wire, they would eliminate each other, and no indications, could be observed at the relay or other receiving apparatus. 92. The 'problem was first solved in the year 1853 by Dr. Gintl,* an Austrian telegraph director, a plan of whose Fig. 85. arrangement is shown in Fig. 85. The conditions which it was necessary to observe were, that the relay or other receiving instrument at each of the stations should remain always in circuit with the line, and that the currents trans- mitted from either station should nevertheless not affect the relay of that station. These two conditions are fulfilled by Gintl's plan by the * Schellen, p. 310. HISTORY AND PROGRESS. 155 employment of a relay with coils wound with two separate wires, in one of which the current of his line battery cir- culate, and in the other that of an equating battery. These coils, wound in opposite directions on the cores, have equal and opposite magnetic effects on the relay when connected up in their proper circuits ; so that, on pressing down the key, although the whole of the current of the line battery passes through the relay, the latter remains perfectly unaffected. For convenience of closing the circuit of these two batteries at the same moment, Gintl employs a double key, a b c, and d b' (?, consisting of two separate levers insulated from each other, being connected together by an insulating cross- piece, and having in front a common knob. In the circuit of one series of the coils of the relay (usually the outer and thicker) are inserted, by means of the leading wires I and n, the equating battery, and the front and middle contacts, a and b, of the right side of the key. The front contact of the other side of the key is connected to the positive pole by the line battery, the negative pole being to earth; the middle contact, or lever, is connected with re- maining coils of the relay, and thence, on the other side, to the line wire ; and the back contact of the key to earth. On pressing down the knob of the key, the current of the line battery, L B, goes from the + pole over the lever a b', leading wire, terminal 1 of relay, the interior coils, terminal 2, through the line to B, where it passes from 3 through the interior coils of the relay, 4, key b' d, 5, E, earth, and, at station A, E to the pole of the battery. The current of the equating battery, at station A, goes, at the same time, through its circuit : +, key, a, b, I, the outer coils of the relay, 11, &c., neutralising the effect of the line battery upon the relay. Suppose now that, while the key of station A is depressed, that of station B is also pressed down, the line current from station B will pass from + of the battery through ', V, of the key, 4, coils of relay, 3, B to station A, where it will enter the coils of the relay at 2, and go from 1 over the key, b', a, through L B to earth, &c. Thus the equili- 156 THE ELECTRIC TELEGRAPH. brium, previously established by the equating battery, is destroyed ; and the relay of A will give a signal corresponding to the length of time which B keeps down his key. During also the whole time that A keeps down his key, the relay of B will be affected, whether the key at station B be pressed down or not, because, as we have seen, the effect of his own current on his relay is neutralised by his equating battery. If, therefore, both stations work their apparatus at the same instant, signals will be given properly by the respective relays. There is only one position in which a perfect reception of the signals transmitted from one station is not attained by the other. It is when, during the manipulation at either of the stations, the lever of the key is removed from the back contacts, c c', until it touches the front contacts, a a, or vice versa. In these cases the line circuit is interrupted for an instant at b, and the signal which should be given by the relay of the same station is disturbed. This is, however, a small evil compared to the great diffi- culty in retaining the compensation of the line and equating batteries for any length of time. The plan adopted by Gintl, of using a thicker and shorter coil on his relay for the equating circuit, occasioned the equating battery to expend itself quicker than the line battery, which encounters con- siderably more resistance ; and this continued diminution of the intensity of the compensating current, whilst the line battery kept nearly constant, caused a corresponding effect on the home relay, which gave the operator often some of his own signals back again, if he does not continually see to the strength of the currents. 93. This system was first used on the line from Prague to Vienna, but difficulties soon induced Gintl to forego the attempt to work Morse instruments by this method, and to adopt instead a chemical telegraph, by which he obtained much better results.* The plan of this modification is shown in Fig. 86, in which' a b is the line wire from one station to the other, con- * " Dub's Auwendung, &c.," p. 461. HISTORY AND PROGRESS. 157 nected at each end with a metal style, which rests on a strip of chemically prepared paper, supported on the under side by a metal contact. The latter are connected to the -f. poles of batteries E, the poles being to earth. Between the metal styles and contacts are inserted resistances, and secondary or compensating batteries, whose currents traverse the paper strips in the reverse direction to those of the line batteries, and prevent the decomposition of the salts con- tained in the paper. Let the negative current of the line battery at station A go from E to earth, and the positive current from E through the metal supporting the paper strip, through the paper and the style to the line a b, in the direction of the arrows ; at station b it will go through the style, paper, contact rest, Fig. 86, and line battery, to earth; In traversing the paper at station A, a decomposition of the salts would take place were it not for the counteracting battery E', whose current, of equal strength, passes from E' through w' t style, paper, &c., preventing the chemical action, until a current, arriving from b, or some such disturbance of the balance, causes an appreciable difference of the currents enough to affect the paper. The value of the resistance w, which is inserted in the circuit of the counteracting battery to balance the currents, may be calculated by the aid of Ohm's law, which will be explained in the second part. Gintl subsequently employed a single key with five con- tacts, instead of the double key just described. 158 THE ELECTRIC TELEGRAPH. 94. Plan of Frischen and Siemens- Halske. About the same date (1854) these celebrated engineers invented, independently of each other, an improved system of telegraphing in opposite directions in a single wire at the same time ; their plan, by which the counteracting batteries and double keys both sources of difficulty are entirely dispensed with, possesses important advantages over the methods of Gintl, and brings the problem of telegraphing in opposite directions as near to perfection as is possible with the conditions of so delicate an arrangement. Fig. 87 represents the plan of connections of two stations, A and B. The negative pole of the battery E is connected to earth, and the positive pole to the working contact of the Station A. Station S Jfcg. 87. ordinary transmitting key K; the back contact being, as usual in the Morse plan, connected to earth. Instead of the common arrangement of putting the relay in the earth circuit from the back of the key, it is inserted above the lever of the key. The relay consists of two coils, r and p t of equal and opposite magnetic effects. The coil r is con- nected between the lever of the key and line ; and the other coil, p, between the lever and a resistance, R, to earth. When the resistance of r and p are equal to each other, and R equal to the sum of the resistance in the circuit of the line L L! and of one side, r lt of the relay at station B, &c., to earth, then, on pressing down the key, the current of E will be equally divided between the coils r and p, which HISTORY AND PROGRESS. 159 having equal and opposite magnetic effects on the needle or tongue of the relay, will produce no effect at A. ; but it will deflect the tongue of B'a relay by passing through the coil r t . The same arrangements being made at station B, when the key K L is pressed down also, it is evident that the deflection of the armature of relay B will not be dis- turbed, because the magnetic effect of the home circuit is neutralised, as in the case of A. But the current from A can now no longer pass so directly to earth, in consequence of the interruption at the back contact of the key iq. It has, however, two paths open to it : the one through E D and other through p l and R X to earth. During the manipulation of the key in Gintl's apparatus, the circuit is interrupted during the instant which elapses between the breaking of one contact and the making of the other by the key. With the method before us this cannot be the case ; the current passes from the line L L , through both the coils r L and j^ of the relay, and n lt to earth. The current encounters, therefore, twice as much resistance that is to say, that of the line, &c., and that of RJ also, which are equal, and has, in consequence, only half the intensity it formerly had. The effect on the relay remains, however, the same, because the current has to pass through both the coils r x and p l9 which being wound in opposite spirals, work now in the same sense and with double force upon the armature. At A the relay is also deflected, since the balance between the currents in r and p has been destroyed by the opposing current from B, which passes, as in the case of the current arriving at station JB, through both the coils r and j? of the relay and the resistance R. When at this moment the key K is let go back on to its reposing contact, the arriving current is shunted from p and R to the back contact of the key and short-circuit w. Only half the resistance now opposes the current, whose intensity is, therefore, doubled, but to balance this, as before, only half the relay is traversed by the current. One of the greatest benefits to be derived from this method of telegraphing in opposite directions is a system of repetition 160 THE ELECTRIC TELEGRAPH. and control very necessary on some lines by an arrange- ment of translation, by which, a message transmitted by the employe from station A, for example, is not only received on the relay and Morse at station B, but also retransmitted by the Morse apparatus at B to station A$ where it can be examined at once to be sure of its correctness. 95. Methods of Transmitting Two Messages along a Single Line in the Same Direction at the Same Time. The first success attained in this direction was by Stark of Vienna in 1855. His method consists of sending from the transmitting station, by two keys, two currents of different intensities, which, on arriving at the receiving station, each set a relay in motion. The relays are arranged in such a way that when the weaker currents traverse the line, only one of the relays is put in motion ; when the stronger current traverses the line the other relay is affected ; and lastly, when both currents go together, both the relays respond to them. At the sending station Stark arranged two keys, as in the plan Fig. 88 ; K being a simple Morse key, and K' a similar lever, supplied at the back with an insulated earth contact, which it moves against the two anvils 5 and 6. The usual front and back contacts of the keys are marked in the figure 1 and 3 respectively, and the levers 2. The battery, which is connected up in series, or one element after the other, is used in two unequal parts, a and I, the number of elements represented by b being double that of a. The battery a is put into circuit with the line by pressing down the key K ; b y by the key K' ; and both together by depressing both the keys at the same time. The copper-pole of a is, therefore, connected to the contact 1 of K, the zinc-pole of same to 5 of K'. Copper-pole of b is connected with 1, and zinc-pole with 6 of K. Lever of K' is in connection with the back contact 3 of K ; line is brought to the lever 2 of K, and the back contact of K' goes to relay, &c. When K alone is depressed, the currents of a pass from z (5 and 4 of K') to earth, and from c (1 and 2 of K) to line. HISTORY AND PROGRESS. 161 When K* is depressed alone, the currents of b pass from z (6 and 4 of K' ) to earth, and from c (1 and 2 of K') to line. When both K and K' are depressed, the united currents of b and c pass from zinc of b (6 and 4 of K') to earth, and from copper of c (1 and 2 of K) to line. Relay Fig. 88. By the depression of one or other or both the keys at the sending station, three currents are therefore produced, whose intensities are in the relation of 1, 2, 3. These currents we will call s, s 19 and S 2 . At the receiving station all currents pass through two relays, i and IT, Fig. 89. A common local battery E 1 serves both these instruments ; its zinc-pole being connected with the tongue of each of them, and its copper-pole with their metal contacts. The relay u is furnished with outer coils, which are put into circuit with another local battery E and a resistance R, by means of the tongue of relay I. The tongue of relay i is held on its insulated contact by a spiral spring, whose force is adjusted that the currents s, or those of the portion c of the battery, are unable to move it ; but that it is easily moved by s x and s 2 the currents of section b and the whole. Relay u, on the contrary, was adjusted delicately, so as to be deflected by the weaker currents. When, therefore, the key K at the sending station is M 162 THE ELECTRIC TELEGRAPH. pressed down, the current of c is sent through the line, and passes through the coils of relays I and n to earth. Helay i is unaffected, but relay 11 is put in action, and the Morse M I in the local circuit (E t z, relay n, 3, 2, M 1? c, &c.) prints whatever signals are given by K. When K' is depressed at the sending station, current Si is Fig. 89. transmitted, and the tongue of relay i deflected against the local contact. Thus two local circuits are closed ; the first is that including the battery E 2 , R, and the extra coils of relay n, by which the action of the line current in this relay is counteracted, and the tongue held still against the insulated contact ; therefore M! does not respond to these stronger cur- rents. The second local circuit is that of the Morse M and battery E^ The intensity of the counteracting battery E 2 , whose mag- netic effect upon the armature of relay n we will call S 3 , is regulated by the interposed resistance R until it balances the magnetising power of the line current sent by K'. The third case is that in which, during the manipulation of the two keys, both happen to be pressed down together. When this occurs the current S 2 of the whole battery goes through both the relays I and n. Relay i is put in action as before, and closes its printing circuit, and that of the HISTORY AND PROGRESS. 163 counteracting battery E 2 . But as the opposite magnetic effect s 3 of the extra coils of relay n is only equalto that of s t , and since s a is equal to the sum of s and S D it is evident that the relay n will be acted upon by the difference of the magnetic effects due to the line and the counteracting currents, or by s, which is precisely the same as that produced when K alone is depressed. The Morse M I will therefore also be set in motion. Combinations have been made, also, by which in a single line, at the same moment, two messages could be sent in one direc- tion, whilst two were being received from the opposite direc- tion ; that thus four independent communications could be kept up. Other arrangements have also been made for telegraphing in the same direction at the same time to different stations along the line, both directly and by translation. Kramer, Bosscha, Maron, Edlund, and others have invented also many similar and equally beautiful methods, all of which have been tried, but none of which have found their way to any extent to practical application ; and the reason is very simply to be found in the varying resistance of telegraph lines, and in the varying electro-motive forces of the batteries, which occasion the inconvenience of having to adjust the systems by means of resistances to compensate these disturbances. Both these systems of telegraphing in opposite directions, and of telegraphing in the same direction more than one message at a time, must be looked upon as little more than "feats of intellectual gymnastics" very beautiful in their way, but quite useless in a practical point of view. 96. Automatic Printing Telegraph. Professor Wheat- stone, to whose inexhaustible fund of invention this modifi- cation owes its being, described it in a paper read before the Paris Academy in January, 1859. It consists principally in the mechanical transmission of signals by means of contacts given by series of perforations in bands of paper previously prepared and drawn through the manipulator ; the signals being printed by the recording instrument. M2 164 THE ELECTRIC TELEGRAPH. Three separate apparatus are required : the perforator, by which groups of holes are printed in the paper ; the trans- mitting or contact key, which is worked by the perforated paper strip ; and the recording instrument of peculiar construction. The perforator has a guiding groove, through which a paper strip passes. At the bottom of the groove is an opening to admit of the to-and-fro motions of the upper end of a frame containing three punches, the extremities of which are in a line transverse to the direction of the paper. A separate lever or key is connected with each of the punches for the purpose of elevating them ; the two external ones forming the groups of perforations which make up the message, and the middle one, which is smaller, marking the spaces between the letters and words. On pressing down one of these keys its punch is raised in order to perforate the paper ; at the same time, a clip which holds the paper firmly in its position is lifted up, and the frame containing the punches advances in the groove, the paper being carried forward by the punch which has perforated it. On letting go the key, the paper is first secured by the clip, and then the frame falls back into its normal position. The inventor has adopted the Morse alphabet by making the upper line of perforations represent the dots, and the lower the dashes. The transmitter receives the slips of paper prepared by the perforator, and transmits voltaic currents corresponding to the holes : positive by the holes on one side, and negative by those on the other. An eccentric in the interior produces and regulates the occurrence of three distinct motions : 1st (says the in- ventor) the to-and-fro motion of a small frame which con- tains a groove fitted to receive the slip of paper, and to carry it forward by its advancing motion ; 2nd, the elevation and depression of a spring-clip, which holds the slip of paper firmly during the receding motion, but allows it to move freely during the advancing motion ; 3rd, the simultaneous elevation of three wires placed parallel to each other, resting HISTORY AND PROGRESS. 165 at one of their ends over the axis of the eccentric, and their free ends entering corresponding holes in the grooved frame. These three wires are not fixed to the axis of the eccentric, but each of them rests against it by the upward pressure of a spring, so that when a light pressure is exerted on the free ends of either of them, it is capable of being separately depressed. When the slip of paper is not inserted, and the eccentric is in action, a pin attached to each of the external wires touches, during the advancing and receding motion of the frame, a different spring, and an arrangement is adopted by means of insulation and contacts properly ap- plied, by which, while one of the wires is elevated and the other remains depressed, the current passes from the voltaic battery to the telegraphic circuit in one direction, and passes in the other direction when the wire before elevated is de- pressed, and vice versa; but while both wires are simul- taneously elevated or depressed, the passing of the current is interrupted. When the prepared slip of paper is inserted in the groove and moved forward, whenever the end of one of the wires enters an aperture in its corresponding row, the current passes in one direction, and when the end of the other wire enters an aperture of the other row, it passes in the other direction. By this means the currents are made to succeed each other automatically, in their proper order and direction, to give the requisite variety of signals. The middle wire acts only as a guide during the operation of the current. In the recording instrument a paper strip is drawn off a paper drum by revolving rollers, which are turned by means of a clockwork inside the case. The paper strip passes underneath a shallow reservoir about an eighth of an inch deep, containing ink. At the bottom of the reservoir are two holes, so small as to prevent, by capillary attraction, the escape of the ink through them. The ends of the print- ing styles are placed immediately above these holes, and, when deflected by means of the electro-magnet, they are pushed down through the holes, and carry with them suffi- cient ink to produce legible marks upon the paper. 166 THE ELECTRIC TELEGRAPH. Professor Wheatstone, in his description, shows also how the apparatus may be arranged for translation, and also how it may be used with a magneto- electric machine instead of a voltaic battery, as the source of electric power. This excellent method is said to combine the advantages of a five-fold speed in transmission, with a considerably greater security for correctness and legibility. The great difficulty which many people find in acquiring dexterity in manipulating by the present system of Morse, would vanish were such a system of automatic transmission in general use, in which the demand for skill from the employes is reduced to a minimum. 97. Siemens and Halske's Magneto- Electric Type Telegraph. Towards the end of the year 1861 Messrs. Siemens and Halske succeeded in the construction of a transmitting apparatus which is, at the present time, one of those instru- ments by which an immense amount of work may be got through, with little trouble, in a very short space of time. The transmitter, which is automatic, like that of Morse's first electro-magnetic telegraph, is used in conjunction with the polarised ink-recorder already described. The transmitter consists of a long insulated wire wound upon a soft iron armature, revolving between the poles of a number of permanent magnets like that of the magneto- electric dial instrument of the same inventors. Of the alter- nate currents thus generated, those which are required to form the signals go through the line to the receiving station ; the others are cut off by an interruption. This is effected by the motions of a contact-lever, raised by the teeth of a series of metal types drawn under the lever by the same mechanism which is used to turn the coil. Fig. 90 represents a plan of the complete apparatus. L, L 1 is the angular contact lever turning on the axis c. When its point, L, is lifted up, the platinum face on the right side of the upper end, L 1 , comes in contact with the metal screw 2 ; and when it falls, the back of L 1 rests against the agate point 1 and breaks the circuit. I, I 1 is the inductor or revolv- ing coil of wire in close proximity to the poles of the perma- HISTORY AND PROGRESS. 167 nent magnets M, M 1 . c is a switch, the terminals a, b, and c of which are respectively connected with the transmitter, receiving instrument, and line. One end of the coil i, i 1 is in permanent connection with the centre c of the contact- lever L, and the other end with the terminal a of the switch. The contact-point 2, against which the lever L, L 1 plays, is to earth ; and, lastly, the side b of the switch is connected by a wire to the coils of the polarised Morse, and thence to earth. If the inductor i be turned round between the poles of the Fig. 90. permanent magnets M, M 1 , whilst L 1 rests against the insulated point 1, and the arm of 'the switch c is on a, the currents in- duced in the wire will meet with an infinite resistance between L 1 and 1, and no impulse will be transmitted through the line ; but if L be lifted up, and the contact between L 1 and 2 established, the currents will pass from the coils i, I 1 (c, L 1 , 2) to earth, and on the other side from i, i 1 (a, c, I) to line, and at the distant station, through the Morse instru- ment, to earth. When the arm c is put on b, the transmitter is cut out of circuit, and the currents arriving pass from line (7, c, 5, A) to Morse and earth. The types, which are set up in a composing-rule in order, like printing-types, are made of- thin pieces of metal cut in teeth, in forms resembling those shown in Fig. 92, represent- 168 THE ELECTRIC TELEGRAPH. ing the letters a and b. The bottom of the composing-stick is provided with a row of teeth, which lock into the worm of an endless screw on the shaft which turns the inductor- coil. Thus, while the inductor-coil is turned, and the alternate currents generated, the types are moved forward with a corresponding velocity, and make and break contact by lifting and letting fall respectively the point of the contact lever. Fig. 91 gives an elevation of the contacts, with the lever and part of the composing-stick containing the three letters Fig. 91. a, b, and c. c', c'' is the composing- rule, moved along in the direction indicated by the arrow, by the rotation of the screw T, and is held in its place by the roller r. The lever L, L 1 is held back against the contact 1, when not raised by the types, by means of the spring s. The figure, however, shows the point L lifted up by the broad tooth of the letter A, and the arm L 1 , therefore, pressed against the contact screw 2. The types are straightened, in the event of getting shifted upwards, by passing under the spring s 1 . The transmitter is fixed upon a table, underneath which is a fly-wheel and pulley, whose strap turns the shaft of the inductor and screw T. A common crank and treadle, like that of a lathe, imparts motion to the system. The composing-rules are about two feet long, and consist each of a straight, rigid piece of metal one- sixteenth inch thick and one-half inch broad, with a thin elastic piece, not quite so broad, screwed on lengthways. The types are put HISTORY AND PROGRESS. 169 in between the two, and are held, in some measure, tight by the elasticity of the first piece. The rigid bar is cut at regular intervals with vertical grooves or ribs, in which a corresponding elevation at the back of each of the types fits, to avoid shifting along the stick. When the arm L 1 is against the contact 2, or the hook L lifted up, as in the figure, whilst T is revolving with the coil, currents, alternately positive and negative, traverse the line, and the polarised Morse at the receiving station gives a series of dots. But when the contact L I with 2 is inter- rupted, the armature of the printing magnet remains on one of the contacts, upper or lower, until a reverse current removes it. The last current before the interruption, positive or negative, determines on which contact the printing beam shall repose. On the upper contact a blank space, and on the lower a line, is the result. It is the duty of the types, acting on the point of the lever L 1 , L, to provide these interruptions after the different currents, so as to produce the required letters of the Morse alphabet at the receiving station. Suppose a, b, Fig. 92, to re- present a series of alternate posi- tive and negative waves, produced by the revolutions of the coil be- tween the poles of the permanent magnets, whilst L 1 , connected with earth, completes the circuit. During each complete revolution a positive and a negative current within the space c are developed, each succeeding half-revolution sending a different current into the line. At the receiving station, whilst this continues, the armature of the polarised Morse will vibrate up and down, and print a series of short lines corresponding to the intervals between the transmission of the positive currents and the negative which follow them. If the composing-stick be not so far advanced along the Fig. 92. 170 THE ELECTRIC TELEGRAPH. stage as in Fig. 91, so that the types A, B, c are all on the right of the lever L, the arm L 1 will rest on the insulating contact 1, and the beam of the polarised Morse, at the receiv- ing station, on the upper screw, by which no mark is made upon the paper. As soon as the hook L, however, touches the first tooth of the type A, it will be pushed up, and L 1 thrown on 2. A positive current will instantly afterwards traverse the line, and make a dot at the receiving instrument. A negative current succeeds, which makes a space ; and then another positive current, which re-attracts the armature. Before the next following negative current is produced, the tooth has passed by, and L dropped into the space which corresponds to a whole revolution. The Morse-beam therefore prints a dash which continues till the next tooth pushes L up in time to complete the circuit for the negative current, which draws up the armature again, and produces a space which lasts until the first tooth of the letter B pushes up L, and allows a positive current to go to the Morse. This tooth is narrow, to break the circuit before the succeeding negative current is developed, so that the Morse prints a dash until the broad tooth of B pushes up L in time for the circulation of a negative current, which begins a space. The tooth in question is just so broad as to include three positive currents, and ends at a negative current. Each letter of the alphabet requires for its transmission an even number of waves. The shortest letter is the single dot representing the letter e t requiring two complete revolutions of the inductor-coil, equivalent to two positive and two negative currents, of which only one positive is used for marking. J, o, and Y require each eight complete revolutions, or the time and space of sixteen currents each, some of which are interrupted at the proper places. Each of the types begins with a depression which cuts off the first positive and negative currents, so that the first tooth of a letter invariably sends a positive current into the line. The types each end with a tooth after a negative current, so that the printing-beam is always drawn off by the types themselves before another type comes into play. HISTORY AND PROGRESS. 171 In this manner the indentation on 'the left of each type will always form a continuation of the space due to the last nega- tive current. The advantages, as stated by the inventors, of this beauti- ful system over the methods of automatic transmission by perforated paper bands and the ordinary Morse, consists, principally, in the greater speed with which a practised type- setter can set up a message, than an expert clerk could either manipulate his transmitting key or punch out the holes in the paper band ; whilst the opportunities which it offers of controlling the correctness of each message when set up, by simply reading off the plain Roman letters which are engraved in the fronts of the types, gives it an important advantage over other systems. The mechanical part of the transmission of a message con- sists in nothing else than in laying the composing-sticks, set up with the consecutive parts of the message, one after the other, on the stage appointed for their reception, removing those which have gone under the lever L, and in treading the lathe during the time. This is accomplished so fast that the machine can transmit the work of six or seven type- setters ; and, as the work of setting the types is much easier and requires less time than in manipulating the Morse letters with the ordinary key, such a machine will send comfortably eight times as many despatches as the ordinary key. The work of type-setting demands also little practice or intelligence. The types are marked with letters of the alphabet, and are put into their places in the composing-sticks from the boxes in which they are kept. It is true that, to keep the apparatus constantly at work, more employes would be necessary than are required for an ordinary Morse ; but this is profitable on lines doing much business. The almost mathematical precision of the signals facilitates immensely the work of reception by preventing the con- fusion and mistakes which sometimes arise in using the Morse, from irregular transmission. The apparatus cannot be arranged for translation as it 172 THE ELECTRIC TELEGRAPH. has been described. But this is scarcely necessary, as it can be worked easily through, an overland line of from 500 to 750 miles. It is found to give marvellously good results on the line between Hamburg and Berlin 370 miles even when an artificial resistance, equivalent to 1,000 miles, is added to the circuit. By the employment, however, of two batteries and a corn- mutating arrangement, translation would be easy. The direct working polarised ink-recorder, constructed ex- pressly for this system by Messrs. Siemens and Halske, is furnished with electro-magnets of somewhat larger dimen- sions, and contains a greater length of insulated wire than is necessary in ordinary circuits. 98. Stoehrer's Double-style Apparatus. Stoehrer sought to remedy the inconveniences arising from the multiplicity of signals required in forming letters when only two elementary signals the dot and dash, as in Morse's system are em- ployed, by the employment of two electro-magnets, with separate printing-beams acting upon the same strip of paper. This method puts four elementary signals, instead of two, at his disposal for the construction of an alphabet ; and thus places his method, in point of speed of working, on a level with the needle-apparatus of Wheatstone. The two beams of the Morse are not moved by the currents of two batteries, but by that of a single local battery directed to the one or other electro-magnet by a delicate relay which differs in its construction from those generally used ; the armatures being formed by two light permanent magnets whose opposite poles are alternately attracted or repelled according to the direction of the current in the coils. It is of necessity very delicate in its action in changing the local current from one to the other electro-magnet of the recording instrument. The recording apparatus consists of a Morse with two electro-magnets and printing-beams. The styles at the ends of the beams press upon the paper strip in the same trans- verse line, about a quarter of an inch apart, underneath a HISTORY AND PROGRESS. 173 common roller, in which two grooves are cut to receive the points. A double transmitting key is used, each lever being of the ordinary construction. A plan of the arrangement is shown in Fig. 93. K, K 1 are the two levers of the key. The lever K is to earth, and K 1 connected with a circuit breaker, u, thence to end 1 of the relay coils ; the other end, 2, of the coils is connected to the opposite side of u, and thence to line. The front and back contacts of K and K 1 are made with metal bars, 1 and 2, com- mon to both levers, and between 1 and 2 is inserted the line battery L B. The pole c of the local battery B goes to D, and Zine Fig. 93. is therefore in permanent connection with the poles of the electro-magnet ; the other pole, z, goes to both the printing magnets M and M 1 , and from the further ends of their coils to the armatures E and F of the relay magnets. When the apparatus is to be used as transmitter and correspond with another station on the line, the contact peg is put into the hole of the circuit breaker u. The levers K and K 1 of the key are then pressed down, one at a time, according to arrangement of alphabet and words to be trans- mitted. When the knob of the lever K is pressed down, the current of L B goes from c (copper over 2 and lever of K) to earth, and from z (zinc over bar 1, K 1 , u) to zinc, and through the apparatus at the other station to earth. K being let go, and K 1 pressed down, the current of L B goes from c (bar 2, 174 THE ELECTRIC TELEGRAPH. K l , 1, u, contact peg, 2) to line, &c. ; and from z (through 1, K) to earth. In case it is wished that the Morse apparatus print the message, for the sake of control the contact peg of u is not put into the hole. On pressing down the lever K, the current of L B passes from c (2, K) to earth, and from z (1, K 1 , 1 of u, coils of D, 1, 2, u 2) to line, &c. The other lever K 1 being pressed down, whilst K is at rest on the back contact, the positive current goes from c (2, K 1 , 1 of u, coils of D, 1 and 2, u 2) to line ; and the negative current from z (1, K) to earth. In this way the station apparatus M M l prints the message as well as that of the receiving station. The apparatus is ready for the reception of messages when the contact stopper is out of u. The currents arising from the line have to pass u 2, coils D, 2, 1, u 1, K J, bar 1, K, to earth, and back to the sending station. The poles of the electro-magnet D become magnetic, and according to the direction of the current in the line attract one or other of the keepers, E, F. When a positive current arrives, F is attracted and E repelled. The result is that the local circuit of M 1 is closed ; the current of the local battery B moves in the circuit c, cores of p, F, coils of M 1 z. The beam of M 1 is attracted, and the style impresses the paper with marks on the lower side. When the sending station reverses the direction of the current by pressing down the other lever of his key, the keeper E of the relay is attracted, and F repelled. By this the current of M 1 is interrupted, and that of M established ; the local current of B, now circulates in c, soft iron cores of D, E, coils of M, z, by which the beam of M is acted upon, and the style marks the paper on the upper side. The elementary signs, dot and dash, in each of the rows marked by the styles, give four elements for the composition of an alphabetical code. The consequence is that fewer signals are required for the formation of letters, &c., than in Morse's code. t* HISTORY AND PROGRESS. 175 The alphabet, numerals, signs of punctuation, &c.; arranged by Stoehrer are as follows : abed e f g h i k I m n o p q r a t u v w x y z ch v - -- .. 012 3 5 89 176 THE ELECTRIC TELEGRAPH. An idea of the time saved by this system may be gleaned from the signals representing the word " telegraph " by the Stoehrer and Morse codes respectively : Stoehrer:'. ...._. .. t e I e g rap h Morse: o t e leg Taking the dot as unit of time and the dash as equal to two of them, whilst the spaces between the dots and dashes of" a letter are equal to one dot, and those between the dif- ferent letters equal to the length of a dash, the time required for transmitting the word " telegraph " by Stoehrer's code will be 44 units, whereas by Morse's it will occupy 61. VI. ELECTRO- CHEMICAL TELEGRAPHS. 99. Bain's Chemical Telegraph. Mr. Alexander Bain has been, since the year 1845, the author of several improve- ments and inventions in telegraphy. The most important of these is his electro-chemical telegraph, patented in England in 1846. The sending apparatus is a simple Morse contact key. The receiving apparatus consists of a circular disc of chemi- cally prepared paper stretched upon a similar disc of metal kept in rotation by clockwork. On the upper surface of the paper rests the point of a style, which, while the disc revolves, is drawn towards the periphery so as to travel over the paper in a spiral curve. When the circuit is closed, the salt with which the paper is prepared becomes decomposed under the style. On a base board are erected the clockwork or driving portion of the instrument, and the recording disc which it turns round. The clockwork is regulated by the fly, and is connected with the recording disc by the shaft, to which it I''- HISTORY AND PROGRESS. 177 imparts motion by the friction of a small roller pressing upon its under surface. When no current is moving in the line, the style rubs on the surface of the paper without producing any mark ; but as soon as a current in the right direction is established, the salt solution with which the paper is saturated becomes decomposed, and leaves a blue mark upon the surface. When the circuit of the current is made and broken repeatedly, a series of dots or dashes is imprinted upon the receiving disc, the lengths and succession of which depend upon the manipulation of the key or contact maker. To render these dots and dashes intelligible, Bain has adopted an alphabetic code like that of Morse. The paper is rendered sensitive by being saturated in a mixture of prussiate of potash dissolved in water, to which are added two parts of nitric-acid and two of ammonia. An interesting rather than practical modification of this apparatus consists in substituting a revolving disc with style travelling in a spiral curve, exactly similar in form to that just described, at the sending station, for the key; only, instead of being covered with prepared paper, it is left naked, and letters or words written upon its surface, within the limits of the journey of the style, in some insulating material ; so that, when the style passes over the insulating writing, the current is interrupted. When both the discs are made to revolve synchronously, which is very difficult, it is evident that the paper at the receiving station will be marked with a dark spiral curve broken in just those places where the letter, written down on the transmitting dial, interrupts the passage of the current ; and a facsimile of the writing will be obtained. The same idea has been carried out, in a more convenient form, by the employment of short cylinders instead of the discs, on the sides of each of which are printers or styles running to and fro upon long screws as the cylinders are turned round. But the apparatus by which Bain has earned most credit, 178 THE ELECTRIC TELEGRAPH. is that with which Leverrier and Lardner experimented before the Committees of the Institute and Legislative Assembly at Paris. A band of paper, punched with groups of holes forming letters, conventional like those of the Morse alphabet, is passed between a metal roller and contact point in such a way that the point falls through the holes and comes in contact with the top of the cylinder, thereby closing the line. The messages are received upon a strip of chemically- prepared paper passed between a style and metal cylinder. Lardner thus describes his results* : " Two wires, extending from the room in which we ope- rated to Lille, were united at the latter place so as to form one continuous wire, extending to Lille and back, making a total distance of 336 miles. This, however, not being deemed sufficient for the purpose, several coils of wire wrapped with silk were obtained, measuring in their total length 746 miles, and were joined to the extremity of the wire returning from Lille ; thus making one continuous wire measuring 1,082 miles. A message, consisting of 282 words, was then trans- mitted from one end of the wire. A pen attached to the other end immediately began to write the message on a sheet of paper moved under it by a simple mechanism, and the entire message was written in full in the presence of the committee, each word being spelled completely and without abridgment, m fifty-two seconds, being at the average rate of fire words and four- tenths per second. By this instrument, therefore, it is practicable to transmit intelligence to a dis- tance of upwards of 1,000 miles at the rate of 19,500 words per hour." The alarm used with Bain's telegraph consists of two round plates of glass of different sizes, struck by a hammer which vibrates between them. The glass discs are supported from their centres by two horizontal arms of an upright. The hammer consists of a vertical tongue of brass turning on a horizontal axis, and carrying, half-way up, a cross-bar of soft * " Museum of Science and Art," vol. iii. p. 117. HISTORY AND PROGRESS.. 179 iron, which serves as armature for the poles of an electro- magnet. Bakeivell's Copying Telegraph. This is, properly speaking, only a better mechanical construction of Bain's electro- chemical telegraph. A clockwork is supported between the plates M, M' (Fig. 94), which puts the cylinder c in motion simultaneously with Fig. 94. the screw s, of the same length as the cylinder, and carrying a nut, q, with style and connection, r. When the cylinder is turned by the clockwork, therefore, the style travels up or down the cylinder according to the direction of rotation of the latter, thereby marking a spiral line whose convolutions are close together. Bake well used an electro-magnetic governor to attain synchronism in the movements of the two apparatus, with- out which it would certainly have been impossible to have obtained any dependable results. 100. Pouget-Maisonneuves, a native of France, writing in the Comptes Rendues in 1856, describes a method of electro- N 2 180 THE ELECTRIC TELEGRAPH. chemical telegraph with which he proposed to supersede the Morse recorder. Instead of the style at the end of the printing beam moved by the electro-magnet, Pouget takes a simple fixed style, and records the messages by chemical decomposition instead of by embossing, thus dispensing with the printing magnet and beam : he, however, retains the relay and other arrangements of the Morse system. His paper is prepared by being soaked in a mixture of 150 parts crystallised nitrate ammonia ; 5 ferro-cyan. potassium ; and 10 water. Before being used, the paper is moistened with dilute sulphuric acid sufficiently strong to make it conduct, but not to attack, the metal of the style. This paper is said to be cheap and easily prepared ; the salts are easily decomposed and the traces permanent. Gintl,* a German, in his method, dispenses with the relay, and records the messages on the prepared paper by the line current direct, in the same way as Bain. His paper is prepared with a solution of 1 part iodide of potassium and 20 starch paste, in 40 water. The results of this process are very satisfactory, and recom- mend the apparatus as more convenient, in some respects, for long lines than Morse's. An experiment, on a line between Amsterdam and Berlin, made in 1853, with six Daniell's elements, gave very legible signals ; and even with four ele- ments the marks, although weak, were readable when no reliable signals could be read with a Morse. The noiseless operation of the electro-chemical telegraphs may have assisted in keeping this method of recording out of more general use. It is always indispensably necessary to combine an alarm with the system, to call the attention of the manipulator ; not so necessary with the Morse, which is, * Brix's Journal, vol. i. p. 4. HISTORY AND PROGRESS. 181 in working, always accompanied by the rattle of the beam and armature. Another drawback from which the chemical telegraphs suffer is in the want of an arrangement of trans- lation which shall not, at the same time, weaken the current. 182 THE ELECTRIC TELEGRAPH. Otherwise the electro- chemical telegraphs are more conve- nient in manipulating, and much more simple and inexpen- sive as far as the apparatus goes, than the Morse. 101. Bonelli 9 s Chemical Telegraph. The Chevalier Bonelli has succeeded in the construction of a chemical telegraph by which messages are transmitted automatically, and facsimiles received at the station corresponding. Bakewell seems to have considered the employment of one style as preferable to that of many ; but he, nevertheless, mentions in his patent the possibility of using several. In the notice of his patent, published in the Mechanics 1 Magazine* he says : " Instead of one style for each cylinder, any con- venient number may be employed, each isolated from the others, and fitted with separate wires having their ends inlaid in an ivory disc, so as to be isolated from each other." The apparatus of Bonelli consists of a long stage or rail- road on a table, as shown in Fig. 95, on which travels a waggon containing, on the left side of the lower half a box of raised metal types, and on the right side of the upper half Fig. 96. a strip of chemically-prepared paper. Over the middle of the railway is a bridge (shown on an enlarged scale in Fig. 96), under which the waggon has to pass when transmitting or receiving a message. A A' are two buffers for receiving the waggon at the ends of its journeys. Just in front of the * " Mechanics' Magazine," vol. 1. p 544. HISTORY AND PROGRESS. 183 buffer A', and level with the rail, is a hook, D, which engages with an eye at the upper end of the waggon, and holds it until a current traverses the line, and releases it by means of an electro-magnet. On the left-hand side of the bridge, over the raised types, is a type- comb consisting of five movable teeth, insulated from each other, which are connected to the ends of five wires going to a similar number of styles at the receiving apparatus. As the waggon, with the types looking upwards, passes underneath the type- comb, the teeth come lightly into contact with the raised portions of the types, and close the circuits whilst the contact lasts. Thus letter after letter is transmitted. The right side of the bridge, which spans the middle of the rails, is appropriated to the reception of messages. It consists of a writing-comb composed of five teeth, made of platinum-iridium alloy, which is not liable to corrosion, insulated from each other, and pressing lightly upon the paper-strip. This comb would produce, if each tooth were traversed at the same time by an electric current, five lines something like the lines on music-paper. As, however, they are each only traversed by a current during the time some portion of a type is underneath the corre- sponding tooth of the type-comb at the sending station, they can only give lines at such intervals and of such length as is determined by the form of the type. If the teeth of the type and writing-combs be equally far apart at each of the stations, and the waggons travel over the rails at the same speed, it is evident that a dotted, or rather lined, facsimile of the types on the transmitting waggon will be received on the paper carried by the waggon at the receiving station. Any deviation from synchronism is, however, of very small importance, as the difference in one way or the other will only make the letters printed either a little narrower or broader than those of the fount from which the types have been taken, and in this consists the great advantage of the Bonelli arrangement over those of Bain and Bakewell. c is an ordinary galvanoscope, and m a mercury trough, 184 THE ELECTRIC TELEGRAPH. in which plunge five amalgamated contacts for short-circuiting the batteries when the type-box has passed through, and the prepared paper is going under the bridge. This is done by a catch on the waggon itself, which, passing by, turns the shaft carrying the five contacts. The waggon when at rest is held at the top of the rail- way by the catch, which, being in communication with the electro-magnet, is released by the first current which passes through the line. This circuit is closed by means of the key, K. Thus the waggons of both stations are made to start together. They are impelled by similar weights, and their speed regulated by means of fans which enable the operators to adjust the two instruments to practical synchronism. The waggons occupy from ten to twelve seconds in passing under the bridge. In this time, therefore, the message, set up in the type-box, is transmitted, and another received on the upper half on the prepared paper. In the instrument shown in Fig. 95, the type-box passes first under the bridge ; when the waggon has got half way, it short-circuits the batteries and leaves the line clear for the reception of a message on the paper on the farther half. About twenty words are on the average set up in each type-box. Thus a speed of transmission and reception of twenty words in six seconds, or of two hundred words per minute, is easily attained, not counting the time lost in changing the type-boxes and removing the paper. The paper intended for receiving permanent printing by the Bonelli instrument for distribution to the public, is pre- pared by being saturated in a solution of nitrate of manga- nese, which yields, under the action of the current, a light brown-coloured precipitate. That which is termed " fugitive printing," as for press work, by which the impressions are not necessarily of a permanent character, is done with paper prepared with a solution of iodide of potassium, which gives letters at first an iodine colour, but which in course of time lose their intensity. The speed said by the operators to be attainable in perma- nent work is 300 words per minute, and the fugitive printing HISTORY AND PROGRESS. 185 is stated to be got over at the almost incredible rate of 1,200 words in the same time. The following facsimile of Bonelli printing is cut from a strip printed at the 1864 Conversazione at the Institution of Civil Engineers : BY THE BQNELL! INSTRUMENT: s '.'. VII. OVERLAND LINKS. 102. The overland line wire, stretched between two stations, is suspended by insulated hooks from posts in the ground. The first line of this nature, which was put to any useful purpose, was the double line- wire of Gauss and Weber, erected principally for researches into the laws of the galvanic cur- rent, between the physical cabinet and the Observatory at Grottingen, a length of 3,000 yards, suspended between the towers of the city and on cross-pieces on poles sunk in the ground. The insulation of the wire from the poles was effected by means of felt wrapped round the cross-pieces on which the wires were twisted. The insulation of this line was of course very imperfect. The posts generally pressed into the service of the tele- graph abroad are young firs (pwws sy/^m). They are selected from 25 to 30 feet long, and at ike top seldom less than 5 inches diameter. The bark is stripped off and the posts smoothed, champhered off, and either impregnated or the lower ends charred up to about 8 feet from the bottom. Every tenth post is, or should be, a stretching post, stronger than the others. The wooden posts mostly in use here are of English larch ; but foreign timber, although dearer, is preferable on account of its greater durability. Impregnation with a solution of sulphate of copper is the invention of a Frenchman, Dr. Boucherie. His process seems to possess important advantages over others, accom- plishing as it does, at the same time, two essential objects 186 THE ELECTRIC TELEGRAPH. that of expelling the sap, and that of filling the pores of the wood with the preservative solution. In his experiments on the impregnation of timber, Dr. Boucherie has made the important discovery that no con- nection exists laterally between the tubes of a tree, and that by applying, under moderate pressure, a coloured solution to certain tubes at one end of a tree, the same tubes at the other end, and only these, are coloured. In this way, at one end of a felled tree, he applied a coloured solution to certain tubes forming the name " Faraday." The name was transmitted to the other end, and was perfect at every intermediate section. When the tree is cut down and trimmed, a solution of sulphate of copper is forced into it from one end to the other by a moderate pressure. The sap and fermenting matter are thus expelled, and their place taken up by the solution. The small cost of the apparatus, ease of manipulating it, and the increased durability which it imparts to the wood treated by it, highly recommend the process. It is necessary, how- ever, to take care that no ungalvanised iron comes in contact with wood so impregnated, otherwise the copper of the preservative solution will be reduced. Chloride of zinc is also used in Germany with some suc- cess. The posts are put into wrought-iron cylinders of 4J to 6 feet diameter, and 34 to 60 feet long, closed at one end, and covered at the other with tightly-fitting tops. The cylinders are provided with manometers, safety-valves, &c. ; and connected with air and pressure pumps, and a reservoir of zinc- solution. The wood is prepared by being subjected to a great pressure of steam, which, penetrating into the interior, not only tends to displace the sap from the pores and prepare them for the preservative solution, but also to coagulate the albumen in the sap, and in this way to retard the subsequent rotting. After this the cylinders are exhausted, and immediately filled with a solution of one part of chloride of zinc and thirty parts of water, which is kept on, under a pressure of eight to ten atmospheres, for about three hours. HISTORY AXD PROGRESS. 187 But it is questionable if this method is so good as that of Boucherie, as it is necessary to force the solution into the wood at right angles to its tubes, thereby injuring its strength and letting the sap, which is the immediate cause of decay, remain ; the coagulation of the albumen in the sap, to any material depth below the surface, being a matter of doubt. The method adopted by Sir Charles Bright is to have the poles well charred from the lower ends to a foot above the depth to which they are destined to be fixed into the ground, and the charred parts soaked in gas-tar for about twelve hours, the poles standing in tanks of tar, in a timber framing. The sap ingredients being the prime movers in the rotting of dead wood, the idea has occurred to put up insulators on the stems of living trees a method which has been found to answer well in Switzerland, America, and in some parts of Germany, where trees are to be found at convenient distances. The only drawback to this system is, the violence with which trees are sometimes moved in heavy storms. To obviate this difficulty, Lieut.-Col. Chauvin has constructed a swing- ing insulator, which will be described afterwards. Wooden posts invariably decay first at the ground level "the wind and water line" where the surface is moist and in contact with the air. A method of retarding the decay by sheathing the post at this part has been tried in India with comparative success, the lower end, to a certain height above the ground, being covered by an iron casing. In Bengal such a line was erected, the posts being of large bamboo canes and the protection of the lower parts cast-iron sockets. This brings us very near to a suggestion which has been much advocated that of dispensing with wood, and con- constructing the posts entirely of iron, whose durability is so superior. The greater cost of such posts is the only objection to them. Pillars of stone, or mason's work, would undoubtedly not only last longer, but would be less liable to accidents by 188 THE ELECTRIC TELEGRAPH. violence of the weather. Such supports have repeatedly been constructed, hut their cost has always been a bar to their further employment. In India, in the early days of telegraphy, many such pillars were erected, and in 1852 a line from Treviso to Tagliamento was entirely supported by obelisks 4J metres high, as shown in Fig. 97. In Switzerland they have begun in good earnest the use of iron posts, the line from Olten to Sissach, lately erected, being supported entirely by iron posts. In Prussia also the necessity has lgf been fully comprehended of discarding wood and taking to some more durable material. Mr. Borggreve has employed, on the line between Grera and Weissenfels, a pillar constructed of a wrought-iron tube, 1| inch diameter, fixed with lead into a socket on ike top of a freestone pillar, 6 feet high and 8 inches square. The iron post of Mr. "W. Siemens is coming Very generally into use abroad, and will no doubt find employment in England also, when the necessity of a durable post becomes thoroughly appreciated. This post is formed of two tubes, one set upon the other, and tlie bottom of the lower one made fast to a bent plate of iron buried in the ground. One of them is shown in Fig. 98. The base consists of one of Mr. Robert Mallet's patent buckled wrought-iron plates, 1 foot 9 inches square, 98> , HISTORY AND PROGRESS. 189 bent in a dish form. The buckled plate a a is secured by four bolts to the socket b b a cast-iron cylindrical tube 7 feet long and 4 inches outside diameter. Near the top, inside, the socket is furnished with a flange, upon which the bottom of the upper or main-post, as it is called, rests. This upper post, c c, is of wrought iron with welded joints ; it stands 12 feet high out of the socket, and is somewhat conical. At its upper end an iron ring is welded in to carry an iron rod, d, 20 inches long, forming the lightning guard. The stretching posts are of the same height, but of larger dia- meter and stronger than the ordinary ones. Mr. Siemens' post derives much of its merit from the role played by the buckled plate at the bottom. These buckled plates are things of great engineering utility. They are squares of sheet-iron, which Mr. Mallet by a simple process presses into a form very slightly different, but endows them with a strength immensely superior \ so strong indeed are they that if one of these posts were pulled up bodily out of the ground, it would bring up the superincumbent ground with it, and the buckled plate would not be deranged unless the bolts gave way ; while the same piece of iron, as a simple sheet, would bend' under a much less weight, offering no resistance worth speaking of against the strain. There can be little doubt that, in course of time, only metal posts will be employed, on account of their superior durability, solidity, and freedom from damage by accidents. In some climates wooden posts require to be renewed every two or three years, and, in the most favourable, rarely last over six years ; while an iron telegraph post is as durable as a lamp- post, and would certainly last ten times as long, and not cost five times as much as a wooden one ; so that in the end an immense saving would be effected by their employment, although the first cost is so much greater. 103. Line Insulators. Cook's insulator was the first used in England. It consisted of a body of earthenware the size and form of an egg, slightly flattened at the ends : the wire was passed through a hole in its longer axis. Bright's insulator, used by the Magnetic Company, consists 190 THE ELECTRIC TELEGRAPH. of a porcelain bell, Fig. 99, provided at the top with a notch for the reception of the line-wire, which is secured by a pin in a hole at right angles to it. One end of a bolt is cemented into a hole in the under-side of the insulator-cap, and by this it is secured to a cross-bar of wood screwed on to the post. Fig. 99. When a post has to carry a number of wires, the cross-bars are of different lengths, the longest being at the top, and each succeeding one shorter than the one above it, that, if a wire should break, it would fall clear of the insulator- caps beneath it. In 1852, Siemens and Halske invented their bell-insulator, which is the strongest and one of the best- construe ted supports. It consists of a cast-iron bell, a a, Fig. 100, with a flange, b b, by which it is screwed against the post. Inside the bell is cemented a porcelain cup, c c, ribbed inside and out to give a good hold to the cement. The cup, c c, in turn, carries the stalk or hook, d d, which supports the line- wire. The parts are put together, while hot, with a cement composed of sulphur and oxide of iron. As a further mode of insulation, the iron stalks or hooks are covered with vulcanite before HISTORY AND PROGRESS. 191 being cemented in ; sometimes the porcelain cup is replaced by a cup of vulcanite. These insulators are a little heavy, but their superior solidity and insulation is ample compen- sation, the iron cap forming at once a perfect protection against injury and a screen against the deposit of dew on the porcelain. In 1856 Mr. Clark patented an insulator in which he increased the length of surface of the porcelain over which the current escapes, without increasing its section. He attained this by a double bell insulator is supported by a stalk, D, Fig. 101, cemented into the interior of the inner cavity ; the line- wire is car- ried through a deep groove on the top, and is tied to the bell by a binding wire. Lieut.-Col. Chauvin, Direc- tor of Prussian Telegraphs, has adopted this style of insu- lation for many of the lines under his charge. He has also made numerous experi- ments on the most favourable proportions between the length formed in one piece. The and section of the cups, and has given them a form differ- ing slightly from Mr. Clark's only in an increased depth and narrowness of the inner Flg> lolt cavity, by which the deposit of dampness from the atmo- sphere is still further guarded against, as well as the sudden cooling of the insulator bell. Lieut.-Col. Chauvin has also constructed an insulator for attaching to the stems of living trees when these are used in- stead of posts for the support of telegraph lines. The insulator is hung upon a hook, free to swing about ; and the stalk, or wire-carrier, bent in a curve away from the stem of the tree, 192 THE ELECTRIC TELEGRAPH. that, when the latter is deflected by the wind, the line-wire, in swinging, may not come into contact with it. The hook Q, Fig. 102, held in the loop p of the bracket M M', is Fig, 102. twisted so that, in case of a sudden jerk, the line cannot be thrown upwards and the insulator disengaged from the bracket. The carrier o is also bent over the wire, to prevent the line jumping out. The bracket is formed so that the insulator hangs quite free of the stem. The Spanish insulator consists of a porcelain bell, b, Fig. 103, supported by a strap of hoop-iron fitting into the groove g, and screwed to the post. The line-wire is carried by a stalk cemented into the inner recess of the bell. The chief merit of this insulator is its cheapness. In climates like that of Spain it answers well enough, but would be utterly useless in England, where the atmosphere is always charged heavily with water vapour, which, condensing on the surface, would soon occasion a material loss of current. Varley's insulator is that most commonly employed in England. It consists of two separate red -earthen ware cups, a and , Fig. 104, cemented together with sulphur. The HISTORY AND PROGRESS. 193 outer cup a is provided with a groove to which the line wire is bound ; in the recess of the inner cup b, a wrought iron bolt, c, is cemented, by which the insulator is attached to the bracket d, on the post. A further insulation is obtained by Fig. 103. T Fig. 104. coating the stalk with vulcanite. The rim of the outer cup a is rounded off inside. The purpose of this is to avoid the sprinkling of the interior with rain-water, when a drop, hanging upon the bottom rim, is blown off by the wind. When a strong current of air separates a drop of water from a sharp corner, the drop is never carried bodily off, but bursts in the direction of the current. With the form given to the rim by Mr. Yarley, however, when a drop happens to hang on that side from which the wind comes, it is driven a little way up between the two cups, and does not burst. 104. Stretching Insulators. The weight of the wire in the space which it makes between two posts, assisted by the occasional pressure of the wind against it, causes it, after a time, to stretch and curve lower towards the earth. When a single wire is suspended, this is of no importance; but when several wires are supported between the same posts, by stretching, they are in danger of touching each other and causing interruptions of the service. To avoid this, the line is provided at intervals with insulators of larger and stronger make than the ordinary ones, to which the wire is made fast, thus giving, at intervals of half a mile or so, fixed points to the suspended Ijne, while the intermediate insulator hooks 194 THE ELECTRIC TELEGRAPH. serve only to support it, without resisting any horizontal strain. Siemens and Halske's stretching insulator is made with a stronger and larger cast-iron bell than the ordinary one. The porcelain boss or cup carries a stalk with two notches (Fig. 105), through which the wire is drawn and wedged on each side, leaving a loop between them. In cold weather, when the line con- tracts, this loop allows the wire between the posts to be slack- ened, and also, in case of a rupture, gives sufficient spare for making a joint. Kohl's stretching insulator is shown in Fig. 106. It is cemented upon the vertical bar or stalk a in its centre, and is turnable in the supporting bracket b, b', with the aid of a ig. 106. Fig. 107. lever, /. The top of the porcelain bell is cut out in a deep groove, into which the line wire is placed and then bound up tightly by turning the lever. This insulator finds employ- ment principally in Germany. HISTORY AND PROGRESS. 195 Another method of stretching with the insulator itself is by means of a wrought-iron winch, w, attached to the stalk, as in Fig. 107. The ends of the line wire on each side are passed through holes in the two drums d d, and wound up tightly, the drums being prevented from running back by clicks and ratchet-wheels on their axles. The Wire. The line wires are of iron ; very rarely of copper. The superior conducting power and durability of copper recommends it for employment, but the danger to which such a line is constantly exposed of being cut and the wire stolen is an argument against it. Were it not for this, a copper line would be much cheaper in the end than an iron one ; iron having a comparatively small conducting-power, and, to give the same resistance for the same length, must have a section at least seven times as great as would be required for copper. This increased section increases, of course, the weight of the line, which, as a consequence, neces- sitates stronger posts and stronger insulators, in order to allow each wire to be strained tightly between the supports to keep it from touching the others. Various plans have been proposed for coating iron wires in order to protect them from oxidation. The plan which has met with most favour in reward of its merits, is that of painting the wires with tar, or, as proposed by Romershausen, of varnishing them from time to time with a good coat of boiled linseed oil. By this method a thick crust is gradually formed, which protects the wire completely. The only objection to the process is its repetition, which renders it difficult in climates like that of England, for the painting can only be done successfully in fine weather. The other methods consist of covering the wire with a coat of some metal which oxidises slowly. Zinc is frequently used for the purpose ; it is applied by a process called " galvan- ising," by which the surface of the iron is covered with a thin film while the zinc is in a state of fusion. Dr. Siemens says that the " galvanising " of iron wire is only useful when the zinc is really melted together with the iron while the two metals are in contact under heat, which is the onlv 196 THE ELECTRIC TELEGRAPH. security that the coating will not spring off or crack in bending the wire. When the wire is badly covered it rusts at those points most exposed to wear and tear, as at the points of suspension, just as soon as a naked wire. In the neighbourhood of manufacturing towns also, the best " gal- vanising " is of no use, as the sulphurous acid gas in the air quickly attacks the zinc, with which it combines, and the salt, washed off by the rain, leaves the iron exposed to the weather and the further action of the acid. In addition to this, the process of galvanising is said to alter the molecular structure of the iron and to render it brittle. Instead of zinc, an alloy may be used, as proposed by Callan, which not only protects the wire against the attacks of acids and weather, but the coating is ductile and bends with the wire, a condition essential to its success. The alloys which Callan tried and recommends are composed of one part of tin with from one to eight parts of lead. A proposal has also been made by Mr. Bucklin, of New York, to dip the galvanised wires into molten copper or brass, by which means a protection, that may be increased to any required thickness by repeating the operation, can be obtained. This would be a cheaper way than that suggested by Professor Brix, the editor of the German Telegraph Journal to cover a bar of iron in this way with brass or copper, and then to draw it down to the required guage. Joints in land line wire are made by bringing the ends together and wrapping them with a binding wire, or by twisting them round each other. Fig. 108. Figure 108 represents a joint made by the former method, called the Britannia joint. The wires to be joined are bent at right angles, about half an inch from the ends, as at a a. HISTORY A1VD PROGRESS. 197 They are then laid together and wrapped or bound with gal- vanised iron binding wire and soldered. By the other method the two ends are laid side by side for about 5 inches, and each turned four or five times round the other, with a space between the two helices of about three-quarters of an inch. To make this joint, how- Fig. 109, ever, it is necessary that the wire should be quite soft at the ends, a condition which must be seen to beforehand. The ends are cleaned with emery paper or with a file, and twisted together by means of a lever arrangement made for the purpose. This consists of two bars of steel, a a Fig. 110. and//, Fig. 110. In the middle of a a is a clip, b, with a vice- screw, c, for holding it down upon the wire. In the middle of // is a block, perforated with a hole in the direc- tion of the lever. The ends of wire to be joined together are laid into the half-circular cavity in the bed of the clip from opposite sides, each end projecting about 8 inches beyond the lever. They are secured in this position by screwing the clip down upon them by the screw c. The projecting end of one of the wires is then bent upwards at a right angle and put through the long hole in the block of the hand-lever. While one man holds the handles of the clip a a, another turns the lever // and with it the end of the bent wire, round the straight one as many times as its 198 THE ELECTRIC TELEGRAPH. length will permit, keeping the hand-lever as close upon the joint as possible. When one end is completely twisted, the wires are taken out of the clip, the twisted part placed in the larger hole d, and screwed tight as before, The remaining end is then bent up at a right angle and twisted round in the same way. The complete joint, after being moistened with a solution of chloride of zinc, is dipped into tin- solder, care being taken that the solder adheres firmly to the wire and fills up all the spaces between the twists of the joints. Erection of Overland Lines, -The erection of land lines embraces very little which affords scope for the display of any- thing beyond mere manual labour. The only work for the engineer is to point out where the line is to cross roadways and rivers, and when it is to make long spans and sharp angles. "When the line has been measured off and the materials distributed to different points, the posts are carried to their places for erection in the ground. It is sometimes preferable to affix the insulators to the posts before the latter are put up ; but this depends upon the kind of insulator used. The posts are planted to a sufficient depth to give a firm hold five or six feet for an ordinary wooden-^post -and the earth well rammed down round it, with stones if they are to be had. The posts which occur at angles, where a greater strain is exerted on them, are strengthened by stays or by struts. The stays are of iron or of steel wire fastened by a ring or bolt to the top, or near to the top, of the posts, and to wooden pegs in the ground, fifteen or twenty feet from the post in the direction opposite to the strain which is to be counter- acted. The method of forming a strut by coupling two similar posts together is that preferred in France. The ordinary post has a notch about a foot below the top, on the side on which the strain is directed ; into this notch is put the top of another similar post, planted about a foot and a half or two feet from the first one. The two are fastened together by an iron collar or by a bolt. When Siemens' iron posts are used, they do not require to HISTORY AND PROGRESS. 199 be planted so deep. The holes are dug about two feet and a half deep and two feet square, the bottom being levelled and rammed to make a firm foundation for supporting the buckled plate ; and when the post is put up, the earth is rammed in, if possible with stones, above the level of the surrounding ground. In putting up these posts the lower tube, or socket, is first fixed, and afterwards the conical main, or upper tube, and the lightning guard. This post is strengthened in points of unequal strain by means of stays. The stay consists of a length of steel wire held to the upper part of the post by a collar of wrought iron and looped at the lower end to a hook at the end of a stay- rod, attached to a plate of iron, buried in the earth at a distance of twelve feet or so from the post. The stay is tightened by pushing the iron collar up the post. When the posts are erected and the insulators fixed, the wire is hung up and stretched. In open country the wire is coiled upon a drum mounted upon a carriage, and is paid out from post to post ; but when the line is much obstructed by trees, etc., the coils of wire are set upon drums on three- legged stools. Another arrangement for this purpose consists of a skeleton iron drum, one side of which is removable to admit of the coils being slipped on. The axis of the drum is hollow, for a pole to be put through it, that it may be carried by two or four men one or two at each end of the pole. The end of the wire being made fast at starting, it is allowed to unwind as they walk along. The wire is laid down along the line at the bottom of the posts, in which position it is examined and suspicious places repaired. The wire is made fast at one end, then lifted into the hooks of the insulators, or tied to the bells, according to their form, and stretched by means of the winch, shown in Fig. Ill, made fast to the next ordinary post beyond the stretching post to which the wire is to be fastened. The end links of a chain are hooked on to the two vertical pins a a, at the sides of the winch, the curved frame between them and the foot a' resting upon the post. The wire is grasped by 200 THE ELECTRIC TELEGRAPH. the " dutch tongs/' or " devil's claw," b, attached to the leather strap c, which is wound upon the drum d of the winch by turning the handle h. A pair of pulley- blocks and line may be used for stretching, but the winch is more convenient. Fig. 111. When the wire is sufficiently stretched, it is made fast to the insulator on the stretching-post, and the following length to the next stretching- post served in the same way. Phenomenon of Charge in Overland Wires. When a gal- vanic battery is connected to one end of an overland line, the other end being insulated, the wire becomes charged with electricity, whose tension depends upon the strength of the battery, just in the same way as a Leyden jar, or condenser, or submarine cable would under the same circumstances. The wire, stretched from post to post, forms the inner coating of a jar, the air acts as dielectric, and the earth, the neigh- bouring houses, trees, &c., form the outer coating. Dr. Siemens has determined the distances of faults in overland lines by measuring the discharge currents. To determine the capacity of a jar formed by a line of telegraph, he erected, in the yard of his factory at Berlin, an iron wire, 121 metres long and 2 lines diameter, at an average height of 8 metres above the ground. The points of suspension were carefully insulated ; one end of the wire was carried directly to the instrument with which the measurements were made, and the other insulated. In comparing the charge of this wire with that of a condenser formed by a glass plate, HISTORY AND PROGRESS. 201 1 millimetre thick and 2 -25 square decimeters of coated surface, it was found that a length of 1 metre of the sus- pended wire had the same capacity as a plate of glass 1 millimetre thick, with 100 square millimetres or O'OOOl square metre coated surface; or that an English mile of wire would have a jar-capacity equal to that of a glass plate condenser of 1 millimetre thickness, with a coated surface of the 0*16 part of a square metre ; or of such a plate, 1 metre long and 0'16 metre broad. Although the wire in this experiment was suspended much higher than is usual with line wires, the result cannot be far short of the truth, as the place where the experiment was made is in the immediate neighbourhood of tall buildings and trees, which also played their role in the phenomenon. The Earth- plate. A proper earth connection is as essential to the working of a telegraph line as the line itself. The earth connection is customarily obtained by a plate some six or eight square feet of sheet copper, buried in the earth at a depth that will insure it being always damp ; it is con- nected with the apparatus by a stout insulated copper wire. In a large station it is well to employ several different earths parallel to each other. For instance, a wire soldered to the gas-pipe gives an excellent earth; a wire soldered to the water-pipe is still better. Both these give earth-plates of large surface, being connected with the gas and water mains of the town. The want of a good earth connection can cause serious interruptions in the service. In temporary stations, as for instance those in military service, difficulty has sometimes been found in burying the plates, and in getting other means of earth. Some of the Prussian and other military stations have therefore been supplied with earth posts which are more portable, and are said in the end to be cheaper than buried plates of metal. The earth-post consists of a wr ought-iron tube, 12 feet long and 1 inch outside diameter. The lower portion of the iron is covered with a copper tube, soldered to it to pre- vent an insulating coat of oxide being formed. To the 202 THE ELECTRIC TELEGRAPH. extreme end is fastened a cast-iron screw, or bore, which is screwed into the earth ; the upper end is surmounted by an ornament and lightning discharger, and is furnished with binding screws for receiving the wires leading to the apparatus. VIII. ATMOSPHERIC ELECTRICITY. The experiments of Franklin, and various physicists since his time, have proved that the atmosphere is always more or less charged with electricity ; that in some parts the charge is positive, in others negative. Accumulations of atmospheric electricity occur particularly in the clouds, which become charged in their formation, their passage, or otherwise, with high tension. It is this charge of electricity which probably tends in a great measure to prevent the clouds falling readily in the form of rain. It is well-known that bodies charged with the same kind of electricity repel each other ; and it must be the same with the water-particles of which a cloud is composed ; when they are charged with electricity they repel each other, and this repulsion prevents them combining and falling to the earth. But when the electricity is discharged and the repulsion over, the water*particles are free to unite, and to descend to the earth in drops. This phenomenon we call a thunderstorm. It is the discharge and passage of such clouds which often prove destructive to telegraph lines and stations, and still more often disturb the regular service. The most terrible effects are to be attributed to the electrical discharge into the line, or direct stroke of the lightning. This occurs when a charged cloud passes over and attracts, in the earth's surface immediately underneath it, the opposite electricity ; and where points occur over the surface, or any object stands up high, these points and objects are charged with greater tension during the passage of the cloud. The induction between the cloud and the earth then resembles that of a Ley den jar, the dielectric being represented by the HISTORY AND PROGRESS. 203 thickness of air intervening between the cloud and the nearest points above the surface of the earth. As soon as this thick- ness is so diminished that the tension of the electricity is able to overcome it, a neutralisation ensues in the form of lightning. The telegraph poles and line wires are especially favourable for the discharge of atmospheric electricity, being extended over an immense surface and offering numerous points. As a cloud approaches, it induces an opposite charge in the earth's surface and in the line wire ; this charge becoming greater and greater as the cloud nears the line. While this goes on, electricity of the same kind as the charge of the clouds is driven along the line in the form of a current to the earth-plates of the stations at each end. When the cloud is near enough, it discharges itself into the line, and, more than neutralising the electricity of the latter, leaves the line some- times heavily charged with the fluid, which seeks the nearest road to earth. When the charge is very intense the lightning not unfrequently melts the line wire in its passage along it, and splinters several of the posts. Should it succeed in entering a station and reaching the apparatus, the usual con- sequence is that it melts the coils of the receiving apparatus, and perhaps shatters the station. Schellen, who has written a beautiful chapter on this sub- ject in his book,* says that the destruction of the posts is probably caused by the water, which during rain creeps into the numerous pores and cracks of the wood, being decomposed by the lightning, the sudden expansion following its conversion into gas bursting them asunder. At times the earth plays no part in the discharge of the clouds. This is the case when two clouds charged with opposite electricities meet in the air and neutralise each other. A line of telegraph being underneath one of them is charged by induction, as we have already seen, in com- mon with the surrounding earth's surface, although to a greater extent, with accumulated electricity. This charge does not change so long as the clouds remain tranquil ; but * Der Elektromagnetische Telegraph, p. 341. 204 THE ELECTRIC TELEGRAPH. the moment the discharge takes place the electricity with which the line is charged is suddenly set free and seeks the earth. It is not seldom that it takes its way over the posts and through the station, where, unless sufficiently protected, it does serious damage. There are, of course, many ways in which the positions of the clouds and their motions, with respect to the direction of the line, modify the conditions of the current. The passage of the electric discharge between two clouds over the line is alone able -to induce a powerful current, under certain circumstances, independent of the liberation of the static charge. The remaining phenomena of atmospheric electricity con- fine themselves to the production of currents of more or less intensity in the line. These are also dangerous to the apparatus, but not to the same extent as the stroke of lightning. These currents are produced in different ways. The atmosphere is everywhere electrical, either positive or nega- tive ; its electrical state depending partly upon the height above the surface of the earth, partly upon the hour of the day, and upon other causes of which our knowledge is limited. When a telegraph line runs through a region which, by reason of great difference of level or any other cause, at one end is positively electrical and at the other end negatively, the line taking the electrical tension of the atmosphere at all points, a constant current must pass through it so long as there are opposite electricities to combine in the circuit. Baumgartner, Henry, and others have made an especial study of this subject, and observed the phenomena under different conditions of height, weather, hour of the day, &c. The passage of a single charged cloud over the line occa- sions sometimes also a considerable and long- continued cur- rent through the apparatus. As the cloud approaches the line it induces in it an opposite charge. To do this the natural electricity of the line must be decomposed ; the electricity of the same kind as the cloud is repelled to earth through the HISTORY AND PROGRESS. 205 apparatus, and a supply of the opposite kind fetched by the same road out of the earth, and held by induction. The one going to and the other coming from the earth form a single current. This continues as long as the cloud nears the line. If it stood still for any length of time the current would cease, but the line would retain its charge. As soon as it moves off, on the other side, the static charge of the line becomes gradually liberated, and endeavours to establish equilibrium with the earth through the coils of the appa- ratus and earth-plates. Then arises a current in the opposite direction to the previous one, which lasts until the electricity of the line is neutralised. The last-mentioned phenomena are causes of annoyance rather than danger, but have been known frequently to interrupt the operations of an overland line for many hours in succession. Atmospheric electricity is the great enemy of overland lines ; and were it not for the protection which the present system of lightning-'dischargers in some measure affords, it is probable that repeated sacrifices of apparatus, stations, and even the lives of the employes, would long since have com- pelled the rejection of the overland system and the adoption of the subterranean and submarine only. The latter are always free from danger, and can receive no injury from atmospheric electricity, so long as they are not in electrical connection with overland lines. When this is the case, and the latter are not supplied with lightning-dischargers, or they fail to do their duty, the lightning enters the insulated wire, and, bursting through the dielectric in its struggle to reach the earth, ruins the insulation in one or more places. This is always to be guarded against, particularly in dealing with long submarine wires, which may be irreparably injured by want of sufficient foresight in enabling the high-tension electricity to go to earth soon enough. The way this protection is provided is by opening a way for the atmospheric electricity from the line to earth, which offers it less resistance, and which it therefore sooner strikes into than the legitimate circuit of the galvanic current. 206 THE ELECTRIC TELEGRAPH. 105. Steinheil 1 s Lightning Discharger. Steinheil* seems to have been the first who supplied the receiving apparatus with an arrangement for protecting it from the effects of atmos- pheric electricity. The method of doing this probably was suggested to him from observing that sparks sprang over from convolution to convolution of the multiplier coils of the apparatus employed on the line between Munich and JSTan- hofen in preference to going through the whole lengths of vhe coils to earth. He concluded justly from this that atmospheric electricity which charged his line resembled, in its disposition to spring over short distances, the better known frictional electricity, and differed in this respect from galvanic electricity. The behaviour of the two electricities is in no way more contrasted than in their choice of circuits. If, for example, a galvanic battery be inserted between the points a and I, Fig. 112, the same being already joined by the long spiral wire to, the whole current will pass through the latter and none will go over the space between a and b, how- ever near they may be, if they do not make absolutely metallic contact with each other. But if, instead of the galvanic-battery, a charged electric-battery or Leyden jar be substituted, the inner coating, for instance, being connected to a and the outer to 5, it would discharge itself immediately over the small space between a and b, and very little, if any, would pass through the coil w. Thus the way which for galvanic electricity offers an infinite resist- ance, is for static electricity a short circuit. Steinheil based the construction of his lightning- guard on this physical law. Instead of bringing the two ends of the line wire into the station, he fixed each of them to a plate of metal 6 inches square, erected over the bureau in which the apparatus was contained. These two plates were insulated from each other by an intervening layer of silk-stuff, which offered an almost infinite resistance to the passage of a voltaic * In 1846. Dingler's Journal, 109, p. 302. HISTORY AND PROGRESS. 207 current, but was at the same time so thin that a spark of static electricity of moderate tension could easily spring from one plate to the other. From the corners of each of the plates a conducting wire went to the apparatus below. The whole was fixed on the roof by an insulated support, and, of course, protected by screens from rain and wind. In the event of the passage of atmospheric electricity of high tension along the line on one side, for instance, it would spring over from plate to plate, rather than traverse the fine wire forming the coils of the apparatus, and would go on by the other line to the next station ; Steinheil's intention being to conduct the high tension electricity along the line, and to allow the apparatus to escape. 106. Meissner's Lightning Discharger. SteinheiPs dis- charger had been constructed with the sole view of affording the static electricity a short circuit across the apparatus whilst the fluid passed in the same line circuit as the galvanic current, from end to end. Meissner introduced the method of conducting the electricity directly from the discharger to earth a method much more in accordance with the nature of the electricity itself. Fig. 113 gives a perspective view of Meissner's lightning discharger. The upper plate, A, is of copper, 8 inches long, 4 inches broad, and three^eighths of an inch thick ; Fig. 113. it is fastened to a second plate, B, of somewhat larger dimen- sions, by means of screws, n n n n. The two plates are, 208 HISTORY AND PROGRESS. however, insulated from each other and from the coupling screws by the latter being contained in cylinders of ivory, and by the insertion of insulating rings of gutta-percha one- eighth of a line thick between the plates, outside each of the ivory cylinders. The coupling screws serve also to fasten the two plates to a base board, which is nailed or screwed against the wall. The end of the line wire L is attached to the corner of the upper plate A, by means of a binding screw. From the opposite corner a finer wire, I, goes to one end of the wire forming the coils of the receiving apparatus ; the other end of the coils is connected by the wire E with the terminal b on the plate B, which is put in communication with earth by means of a thick wire, e. The two thin wires / and E are covered with silk and are carried from the lightning discharger to the board twisted round each other. Should the tension electricity in the line, therefore, escape by any chance a passage across the plates A B, it will certainly pass from the wire I through the silk covering to the wire L, before it reaches the apparatus. The galvanic currents from the sending station arriving by L, cross over the plate A to the wire /, by which they reach the apparatus ; from the apparatus they come to the plate B, through the wire E, and, after crossing over B, go to earth by e. The intervening stratum of air between the plates offers an infinite resistance to the galvanic currents, which are there- fore not weakened by the lightning guards in the circuit ; but electricity of greater tension finds this air resistance infinitely small in comparison with that of the wire of the coil, and, therefore, on its arrival at A by the line L, it immediately springs over to B, and goes to earth through the thick wire e. 107. Siemens and Halske's Plate Lightning Discharger.* Fig. 114 gives a perspective view of a lightning- guard com- monly supplied by Messrs. Siemens and Halske. Over a cast-iron plate, a, called an earth-plate, are placed as near as possible, but without making metallic contact, two smaller plates, b , called conductors. Each plate has two screws for HISTORY AND PROGRESS. 209 attaching wires. To the screws/! and/ 2 are attached the up and down line wires respectively ; to the sere ws/j and/ n the wires leading to the apparatus i and n ; and, lastly, to the screw g, the earth wire. The upper plates, b b, are kept in their places by small knobs, c c, fixed to them, and by the buttons d d on the Fig. 114. earth-plate. In order to prevent metallic contact, these buttons are covered with vulcanite, as are also the sides of the earth-plate. The atmospheric electricity, on reaching the conductors springs over to the earth-plate, and thus escapes the apparatus. Another form of lightning discharger used on the Con- tinent, especially on railway lines, is a modification of the plate dischargers. It consists of three brass cylinders with platinum faces, supported in an insulating frame. The line wires are attached to the two end cylinders, between which the receiving apparatus is inserted. The middle cylinder is connected with earth. 108. Breguefs Wire Lightning-guard. An excellent idea was carried out by Breguet in the construction of his first lightning discharger, in the employment of the power of the spark to melt a fine wire, and of making the lightning itself cut 210 THE ELECTRIC TELEGRAPH. the apparatus out of circuit and save it from injury in its own endeavour to force a passage through it. He was first re- minded of the necessity of protecting the instruments from the effects of atmospheric electricity, by an accident which happened to the apparatus at the Yesinet station, on the 5th May, 1846, when all the wires were fused and the apparatus rendered useless. His paratonnerre consisted of a piece of fine wire, fifteen to twenty feet long, which he carried from the apparatus to the termination of the line outside the telegraph bureau. The insertion of this fine wire in the circuit did not appreciably weaken the current by increasing the resistance of the line, and when a stroke occurred the electricity melted the wire on leaving the line before it could reach the apparatus. 109. Fardley's Lightning Discharger. Fardley combined both the systems of discharging between plates of metal and of melting a fine wire, for better security for protecting a line of 65 miles, in 1847. He divided the line wire at the station and brought the ends within a distance of half a millimetre of each other at the side of a stout wooden post, which supported also a wooden roof to shield the ends from wet and dirt, instead of inserting plates of metal in the line, as Steinheil and others had done. To each of the parts of the divided line he joined some twenty feet of fine metal wire, which formed the leading wires to the apparatus. The line on either side, if struck by atmospheric electricity, would then, in all probability, discharge itself across the small space between the divided line wire to the line on the other side ; but, should this not be the case, and the fluid find its way along either of the leading wires, the wire would be fused before the electricity could reach the apparatus. The leading wires also terminated in a commu- tator, by which the operator, during a thunderstorm, could cut the apparatus entirely out of circuit, establishing at the same time another connection between line and earth whereby the line circuit remained entire. 110. Nottebohm's Lightning Discharger. Nottebohm has constructed and introduced a lightning discharger for use on HISTORY AND PROGRESS. 211 the Prussian state telegraph. It consists of a double cone, supported by a stout metal bar, in connection with the earth. The points are in close proximity with the points of two metal cones, which are supported on a common base, and severally connected with the lines, and also with the two ends of the wire coils of the receiving apparatus. When the line con- tains free static electricity, the latter springs from the cone on the side on which the line is struck to the little double cone in the middle, and thus avoids the apparatus. 111. Siemens and Halske's Point Lightning Discharger. A lightning- guard of beautiful construction, for protecting submarine wires which are connected with overland lines, is used by Messrs. Siemens and Halske, which, while sufficiently serving the same purpose as those already described, has the advantage of being fixed upon the instrument board, where no difficulty can arise in mounting it, and any disorder in which it may get will be more readily discovered. It is formed of a cube of brass connected with the earth, to three sides of which are presented sharp points of metal protruding from an arch in the circuit of the line wire. The brass prism is faced with three metal plates, carrying agate cones on the top and on each side under the arch. The purpose of the agate cones is to prevent the points being adjusted too close to the earth-block. The points are formed by three screws, the axes of which lie in the same plane, one running vertically through the top of the arch and the other two horizontally on opposite sides. Within each of the screws, along its axis, is a second screw of small diameter, terminating in a conical point of platinum. 112. Breguet's Lightning Discharger. Breguet has con- structed another lightning discharger, now used on all the French lines, in which he increases the means of discharging the static electricity from the line by increasing the number of points over which it can pass, and diminishing the resist- ance to its passage. For this purpose he arranges two plates of copper insulated from each other upon a common board. The opposite edges are cut out in the form of sharp teeth, so that the point of each tooth on one plate is opposite the p2 212 THE ELECTRIC TELEGRAPH. point of a tooth of the other. The plates are put as near to each other as possible without any of the teeth making contact with their neighbours opposite. The two line wires are attached to the plates by means of binding screws. A i fcC I modification of this lightning discharger includes both the arrangements of Breguet the saw teeth and fine wire. 113. Lightning Discharger of the Prussian and Austrian Telegraph Lines. This is described in detail in Brix's Journal. It is on the points principle, and is arranged for being placed on the instrument board. From a bar of metal, c, Fig. 115, on the left-hand side, project two points, HISTORY AND PROGRESS. 213 c and c', with adjusting screws. Opposite these points are two similar ones, projecting from the terminals A and A', which are connected by the spirals w and w with B and B', respectively. On the margin of the board are five terminals for receiving the wires of lines, apparatus, and earth ; the bar c is connected by a wire underneath the board, shown by dotted lines in the figure, with the back terminal E, destined for the earth connection ; the terminal A is similarly connected with a on the one side, and A' with ' on the other ; B and B are also in permanent connection with b and b on opposite sides. The two line wires are brought to a and a, and the apparatus is inserted between B and B'. 114. Kerekkoj/T* Lightning Discharger includes arrange- ments for discharging the line both between surfaces and points. A hollow brass cylinder, supported by a bracket, is in permanent connection with earth. Inside the cylinder is a second metal cylinder, insulated from it by short ivory tubes at the ends, and held in its place by the screw-points. The annular space between the cylinders does not exceed one- fourth of a line, so that electricity of moderately low tension can easily spring over. The line wire is connected with one of the two terminal screws of the inner cylinder, and the wire leading to the apparatus with the other. The outer extremities of the screws are furnished with points by which, should the electricity fail to leap over the annular space between the cylinders, it will in all probability dis- charge itself to the opposite points on two screws connected to earth through their uprights and the common base-plate. 115. Bianchi's Vacuum Lightning Discharger. Du Moncel describes this apparatus, which is of a novel but somewhat inconvenient construction. It consists of a brass globe inserted in the line circuit, covered and protected by two hemispheres of glass cemented into a broad metal ring, which is provided with radial spikes pointing inwards to within a very short distance of the surface of the globe. The latter is held in its place in the centre by an axis passing air-tight through the poles of the glass hemispheres 214 THE ELECTRIC TELEGRAPH. and ending in terminal screws. A tap is also inserted in the copper ring, through which the air is pumped out. The ring is supported by a metal plate at the back, connected to earth, and by which the lightning discharger is fixed against the outside of the building. The foregoing are some of the best arrangements invented as yet for protecting the apparatus from the effects of atmospheric electricity. They all, in a more or less com- plete degree, fulfil their purpose, but none with entire certainty, as we from time to time observe in the damaging effects of a thunder-storm to the apparatus at stations which are supplied with dischargers of the best constructions. On other occasions, however, the dischargers do their duty in saving the stations and apparatus from serious damage. Such a case was mentioned in the Electrician in a letter from one of the officials of the Levant submarine tele- graph line. The writer says that a heavy thunderstorm, passing over the Island of Metelin, completely destroyed eight poles, and was only prevented from going into the cable by a plate lightning discharger, the iron plates of which were fused together. PART II. ELEMENTS OF THE SCIENCE AND PRACTICE OF ELECTRIC TELEGRAPHY. I. ORIGIN OF THE GALVANIC CURRENT. 1. Any piece of metal, when partly immersed in a liquid, according to Pfaff's experiments, becomes polarised the part above the liquid being negative, and that in the liquid positive. The strongest polarisations are those set up by zinc and tin in solutions containing free nitric or sul- phuric acids ; therefore zinc and tin are termed the most powerful electro-motors. The polarisation of the metal, a piece of zinc for instance, is communicated to the particles of the liquid which tend to arrange themselves in order according to their component atoms : thus, if the liquid be pure water, each of whose atoms is composed of one atom of hydrogen and one atom of oxygen, these atoms of water will probably so arrange themselves that the oxygen and hydrogen sides will stand alternately with regard to the zinc ; for, according to Grotthuss, the atoms of oxygen and hydrogen composing the water are held together by electri- city, the oxygen being negative and the hydrogen positive, to such a degree that they exactly balance each other, and produce no free electricity. This idea is shown in Fig. 117 with five neighbouring atoms of water, of very exaggerated dimensions of course. The electro-motor is partly plunged into the water. It becomes polarised ; the part immersed is positive, and 216 THE ELECTRIC TELEGRAPH. attracts the negative or white components (the oxygen) of the atoms of water immediately in contact with it, repelling the black component (the hydrogen) which is positively electrified. The black, or positive component of atom No. 1, in its turn, attracts the white side and repels the black side of atom No. 2, and so on through the whole mass of the liquid. Let us now suppose another plate of a different metal, a good conductor, but a less powerful electro-motor than zinc, say copper, similarly immersed in the liquid at a little distance from the first plate. On entering the liquid its natural tendency is to polarise the atoms of water between itself and the zinc in the reverse direction to that done by the latter ; and it would do so if it were as strong an electro-motor as zinc, and would depolarise all the atoms already polarised by the Fig. 116. Fig. 117. zinc. But not being so powerful an electro-motor, the feebler polarisation which it takes itself and tends to communicate to the atoms of water is overpowered by the stronger polarisation of the zinc, and they retain nearly the position given them by the latter. The black or negative sides of the atoms all face the copper plate, and induce a negative state in the part below the surface, while the natural electricity of the copper, being decom- posed, its positive component takes refuge in that part which is above the surface of the liquid. An electrometer of sufficient delicacy, brought to the upper end of a, would indicate a negative tension, and in contact with b would show a state of positive tension. This is what is called the open SCIENCE AND PRACTICE. 217 circuit. If we close the circuit by connecting the upper end of a by a wire, c, with the upper end of b, a combina- tion of their opposite electricities takes place in consequence. The positive electricity goes from the copper pole to the zinc, and the negative from the zinc pole to the copper, outside tl\e clement. At the same time the component atom of oxygen is separated from the first atom of water and combines with an atom of zinc ; the atom of hydrogen of No. 1 combines with its neighbouring atom of oxygen of No. 2 ; the hydrogen of 2 with the oxygen of 3 ; and so on, until the copper plate is reached, where an atom of hydrogen is liberated and, being positive, is attracted by the copper. The oxygen liberated from atom No. 1, combining with the more powerful electro^motor, zinc, forms oxide of zinc ; and the hydrogen liberated at the copper collects on the plate in the form of gas until the bubbles are large enough to rise to the surface of the liquid. An essential condition to the formation of such an element is, therefore, mobility of the medium in which the plates are plunged. Were they, for example, contained in dry ice, or water in a solid state, no electro-motion could occur. The polarisation of the atoms of water depends upon the difference in the degree of polarisation of the zinc and copper plates ; and the rate of transfer of the atoms, or the galvanic current, depends upon the affinity of the more powerful electro-motor for the liberated oxygen. The oxidation of the zinc plate is therefore a measure of the current as is also the volume of hydrogen liberated at the copper plate. The hydrogen liberated at the copper plate collects here by degrees until it totally covers the immersed surface, attracted to it by its opposite polarity. Now Buff has proved that hydrogen is more positive than zinc ; the con- sequence is that, after a time, the positive polarity of the zinc plate is opposed on the other side of the element by the greater positive polarisation of a large surface of hydrogen gas. The atoms, 1, 2, 3, &c., cease, therefore, to 218 THE ELECTRIC TELEGRAPH. arrange themselves with the same tension in the order shown in the sketch, and the combination of the oxygen atoms with the zinc is retarded. The current of such an element, therefore, becomes gradually weaker from the moment of closing the circuit until the maximum collection of hydrogen gas on the copper plate is attained. 2. The current of a galvanic battery manifests itself in various ways, by some of which we are enabled to measure its intensity and to study its relations under different con- ditions. The most important of these manifestations are : its produc- tion of light, its property of heating bodies, its physiological, its* chemical, and its magnetical effects. At the moment when the poles of a single pair of plates of large surface are separated, the experimenter observes a beautiful bright spark pass between them. If the poles have been amalgamated beforehand, the spark is very intense. This is, on a small scale, the celebrated electric light. By taking forty or fifty elements of large surface and high electro-motive force, such as the platinum- zinc battery of Mr. Grove, or the carbon-zinc battery of Robert Bunsen, having first furnished the poles of the system with carbon points, if we touch these points together for an instant and then separate them, we have an electric light too brilliant to be regarded, without danger, by the naked eye. By separating the points steadily for a short distance, the particles of carbon are built into a bridge by the current through which it circulates, and keeps it incandescent with a peculiar brilliancy until it is broken. The heating power of the current may be observed with an element of large surface by inserting between its poles a short piece of platinum wire, which becomes, in a few moments, red hot. A battery of thirty Bunsen's elements easily melts pieces of platinum and of every other metal which are brought between its poles. So great indeed is the heat known to be generated by the galvanic current, that it has even been attempted to gasify carbon by it. The physiological effects of the galvanic current were first SCIENCE AXD PRACTICE. 219 observed by Sulzer and Gralvani, and led to the discovery of galvanism. The experiment published by Sulzer of putting a piece of zinc under and a piece of copper upon the tongue, and letting them come into contact with each other, is the simplest arrangement by which we become acquainted with this effect. When the poles of a battery of a number of elements are taken hold of in the hands, the latter being moistened, a very unpleasant sensation in the arms and chest is felt whilst the current continues, and a still more un- pleasant one when the circuit is interrupted and re-made. This effect is said to have been often used with advantage in medical cases, and has been employed in some highly interesting experiments with dead bodies, members of both human and brute creation. A use to which the practical electricians of the gutta- percha factories put this property of the galvanic current is in the detection of faults in insulated wires. For this purpose the wire is connected at one end with one pole of a battery, the other pole of which is to earth. The further end of the wire is insulated. The wire is then drawn through a wet sponge or cloth held in the hand of a workman. The moment a bad place passes through his hand the water enters the fault, closes the circuit of the battery through the man to earth. Notice of the fact is given unfailingly when the battery is strong enough. The epidermis is a little insensible to the action of the current, and becomes quite so after repeated shocks ; but the tongue and wounds in any part of the body are highly electroscopic. The physiological effect of the current has, we believe, only once been enlisted into the ranks of the telegraph. In 1839, Yorzelmann de Heer carried out a telegraph in which the operator received the signals in his fingers. Ten leading wires connected the corresponding stations. At the receiving end the operator placed his fingers and thumbs upon the ten metallic terminals of the lines. The signals were given by sending currents at the same time through two of the wires, and were observed 220 THE ELECTRIC TELEGRAPH. (a) in one finger of the right hand and one finger of the left; or (>) in two fingers of the right ; or, lastly, (c) in two fingers of the left hand. To the remaining two effects of the galvanic current the chemical and the magnetical we are indebted for the oppor- tunities which we have of measuring and studying more intimately its laws. To the chemical effects, besides the galvano-plastic art and all the ramifications of industry which it has been the means of introducing, we owe the decomposition of water and the discovery of many of the metals. The decomposition of water, made by M. Sommering the basis of his telegraph, has been found to be a just measure of the current producing it, and has therefore been used as such. The decomposition of salts is employed in the chemical telegraphs of Gfintl, Bonelli, and others, and promises to come ultimately into more extended use. The magnetic effects of the galvanic current are manifested in the deflection of a magnetic needle suspended near a wire in which a current is moving; in the magnetisation of a soft iron, when a current circulates round it ; and in the induction of currents in close circuits in the neighbourhood of a current, at the moment of its appearance and disappearance. 3. Galvanic Batteries. Cruikshank remodelled the pile of Yolta, and gave it the form of a trough divided into several compartments, each of which contained a pair of zinc-copper plates immersed in dilute sulphuric acid, instead of the cloth discs moistened with acidulated water, as in Yolta' s pile. Wollaston improved this form of battery, first, by entirely surrounding each zinc plate by the copper plate, and secondly, by making it a plunge battery, the electro-motors being arranged at distances along a non-conducting bar, which, when lowered in its frame, plunged each of the plates into its proper vessel of acidulated water. These elements had a con- siderable electro-motive force when first set up, but went down very soon afterwards, through polarization of the copper plates, by the hydrogen gas collecting on their surfaces. (I . SCIENCE AND PRACTICE. 221 Many other improvements have been made in the forms of zinc- copper elements, the chief of which are those of Sturgeon and Daniell. Sturgeon recommended the use of amalgamated zinc as being more electro-positive than common zinc, and only dissolved when current passes ; and Daniell succeeded in constructing a battery of more constancy than those hitherto employed, 4. Daniell' s Constant Battery. The principle of the electro- motive combination arrived at by Professor Daniell, as giving the best results, consists in immersing each of the electrodes or metallic plates in a different solution, by which the polari- sation, rendering the working of the previous piles of so short duration, was in a great measure prevented ; the element retained its electro-motive force for a longer period, and thence obtained the name, not absolutely correct, of a constant battery. To prevent their mechanical mixture, the solutions in which the plates are placed are divided from each other by a porous diaphragm, whose pores are not so close, however, as to prevent the necessary transfer between the atoms. In the earlier forms of his battery, Daniell employed a diaphragm of ox-bladder ; but this has been long since replaced by a cylinder, or vessel of unglazed porcelain. A very ordinary form of Daniell' s element at present in use is the cylindrical. The copper-plate is bent round to fit inside a cylindrical porous pot, filled with a saturated solution of sulphate of copper. A cylinder of amalgamated zinc is placed between the porous pot and outer glass vessel in a space filled with dilute sulphuric acid. Crystals of sulphate of copper are kept in the inner chamber, to insure the necessary degree of concentration, and for this purpose the copper cylinder is sometimes partly closed at the bottom to form a cup to contain the crystals. The dilute sulphuric acid in which the zinc is immersed contains usually about five per cent, of commercial sulphuric acid. When these elements are used in numbers, in the form of a battery, the copper cylinder of one element is generally connected permanently with the zinc cylinder of its neighbour, by 22 THE ELECTRIC TELEGRAPH. casting tie zinc on to a continuation of the copper plate. This saves trouble, and avoids the danger attending the coupling of the elements by metallic binding screws, from oxidation of the ends, or loosely made connections. When the current passes the zinc is dissolved, and the copper receives an equivalent increase in weight. In the chamber containing the zinc and acidulated water, the oxygen of each atom of water decomposed unites with an atom of zinc, forming an atom of oxide of zinc, which in its turn combines with an atom of sulphuric acid, forming sulphate of zinc, which is dissolved in the water. The atom of hydrogen released is transferred, by means of decompositions and re- compositions, towards the copper cylinder. In the interior of the porous pot an equivalent atom of sulphate of copper is decomposed into one atom of copper, one of oxygen, and one of sulphuric acid. The atom of copper is deposited upon the plate by the current ; the atom of oxygen, moving to- wards the zinc plate, meets the atom of hydrogen travelling from the other compartment of the element, and combines with it, forming together an atom of water ; while the atom of sulphuric acid goes to the zinc compartment to renew the supply there for the formation of sulphate of zinc, as that metal is dissolved. When the circuit of such a battery is kept closed for a length of time without addition to its constituents, the sulphate of copper in the porous pot becomes all decom- posed, and the water in the concentric chamber saturated with sulphate of zinc. In such a state it is evident that the advantages of the system are lost ; the hydrogen liberated from the water, not meeting with the oxygen liberated from the copper salt, polarises the copper plate, and lessens the electro-motive force of the element. The periodical addition of crystals of sulphate of copper, however, keeps up the saturation of the water in the inner vessel, and, consequently, the constancy of the element. But the water in the outer vessel should never be allowed to get too saturated. A great inconvenience is always found in these elements arising from the deposit of metallic copper at the bottoms SCIENCE AND PRACTICE. 223 and on the sides of the porous pots. Sometimes, after a battery has stood unused for a while, the porous pots are found to be completely impregnated with metallic copper, filling up their pores, and forming short circuits between the solutions, reducing thus the action of the elements to almost nothing, while the consumption of zinc is, at the same time, increased. This deposit of metallic copper is not the result of galvanic action, but of cementation. It would not occur if the plates were made of pure metal. As the common zinc is dissolved in forming the currents, the particles of iron and other metals mixed with it, fall to the bottom, and separate the copper from the solution of its salt as the latter comes through the pores of the diaphragm. A small local element is thus formed which goes on reducing the metallic copper and adding to the bulk of the deposit. The method suggested by M. Place which is now employed to lessen this damaging effect, is to saturate the bottoms of the porous pots to the height of a quarter of an inch with hot wax or parafine, and, in setting up the battery, to fill the dilute sulphuric acid into the elements four or five hours before putting in the sulphate of .copper solution. The pores of the pots become well filled with acidulated water before the sulphate comes into contact with them. No precaution can, however, entirely prevent this detrimental property. Another cause of inconstancy in the action of Daniell's element arises from the solution of copper entering the chamber appointed for the solution of zinc. When this occurs the trespassing copper is precipitated out of the sulphate, and adheres to the zinc cylinder, the colour of which changes to red and black, and its electro- positive condition becomes weakened. The electro-motive force of the element is lessened in proportion as the copper covers the surface of the zinc ; and the quantity of sulphate of copper and metallic zinc consumed represent a much greater strength of current or length of time that the circuit has been closed, than the operator has really had the benefit of. Both these processes of destruction of the element take 224 THE ELECTRIC TELEGRAPH. place faster during the time the circuit is open than when it is closed. 5. Kramer's Modification of Darnell's Battery. Kramer has had these difficulties in view, and has succeeded partly in overcoming them by interposing, between the zinc and copper plates, two porous diaphragms, the space between them being filled with diluted sulphuric acid, and containing a copper plate of large surface in connection with the ordinary copper plate of the element. The duties which the copper plate has to fulfil in the ordinary form of Daniell's element are in this way divided, the interposed copper plate acting, principally by reason of its large surface and proximity to the zinc plate, in con- ducting the current from the other elements when set up with others in a battery j while the other copper plate, immersed in the salt solution, fulfils mainly the functions of a copper electro-motor for the element itself. The interior copper cup is first filled with crystals of sulphate, and then all three compartments are filled up to within an inch of the top with sulphuric acid diluted with 100 times its weight of water. The distance of the sulphate of copper from the zinc plate being considerable, little or none of the salt can reach it, and hence little or no loss of materials can take place by diffusion. 6. Meidinger's Modification of Daniell's Battery. Professor Meidinger has introduced a still bolder modification of Daniell's battery, in which the members are so differently arranged as almost to claim the title of an original invention. Instead of separating the solutions of the sulphates by a porous diaphragm to prevent their mechanical mixture, Meidinger depends wholly for their separation upon the difference of their specific gravities, by which the detrimental precipitation of metallic copper upon the zinc pole is pre- vented. A condition is, however, necessitated in the employ- ment of Meidinger's element, which is not in that of any of the others ; it is that, when once set up, the battery must remain entirely undisturbed, otherwise the evil which it is proposed to obviate is only increased. SCIENCE AND PRACTICE. 225 Fig. 118. Meidinger's element is set up in a cylindrical glass jar, A A, Fig. 118, on the bottom of which is cemented a glass cup, d d. The diameter of the outer vessel is larger above than below, being provided, at about a third of its height from the bottom, with a shoulder, b b. A cylinder, z z, of amalgamated zinc, sits upon this shoulder, and is of such _ dimensions as to fit comfortably into x! the upper part of the vessel. The interior of the cup d d is covered similarly with a cylinder, e, of copper, to which a copper wire, g, insulated with gutta-percha, is attached, and leads out of the element. The glass vessel A A is covered by a wooden disc, provided with a hole for the wire g to pass through, and one also for a glass tube, h, formed like a test- tube, pierced at the bottom with a few small holes. The glass .tube h is fixed in the centre of the wooden cover, and reaches to about half-way down inside the cup e. The element is charged by filling it to the top of the zinc cylinder with Epsom salts solution, and the tube h with crystals of sulphate of copper. The sulphate of copper dissolving, saturates the water containing Epsom salts, and the solution being specifically heavier, descends through the hole at the bottom of the tube into the cup d d, which it about half fills. So long as the element remains unshaken, the fluids retain their respective places, and the diffusion of the sulphate of copper into the solution of sulphates of zinc and magnesia above takes place so slowly that, after a battery has stood a month or two, scarcely a trace of copper is to be observed on the zinc ring. Further, a great advantage is offered by the circular space between the cup d d and the sides of the vessel A A, into which the particles of iron and other foreign metals, divided from zinc, fall without coming into contact with the sulphate of copper. 226 THE ELECTRIC TELEGRAPH. The chemical action being identically the same as that of a common Daniell's element, the electro-motive force is necessarily the same. The resistance of the element is, how- ever, a little greater, resulting from the inferior conducting power of the solution of Epsom salts, and the distance which the plates are apart ; but even for local batteries this resist- ance does not interfere materially with its applicability to telegraphic purposes. 7. Siemens and Halske's modification of Darnell's Battery. MM. Siemens and Halske of Berlin have set themselves the same task .-as Meidinger and others, and have produced the battery which bears their names, the constancy of which has been tried for some years, and found to answer their expectations. These inventors directed their attention particularly to the improvement of the diaphragm, and sought for some substance which would prevent the mixing of the solutions and the unnecessary con- sumption of zinc and sulphate of copper. The substance recommended by them is paper-pulp. An element constructed with this material for a diaphragm is shown in section in Fig. 119. A A is a glass jar, at the bottom Fig. 119. p , . , . & , , J , 01 which is placed a cross of sheet copper, K, attached to a copper wire,/+, and over this a tube, c c y of unglazed porcelain, the lower part of which is widened out bell-fashion at c c. Between the porous bell and the glass vessel, the paper-pulp is put to the height e, well stamped down. The papier-mach6 as obtained from the paper-mills is prepared first by being well pressed, then treated with a quarter of its weight of English sulphuric acid, which is worked together with it until the whole acquires a homo- geneous glutinous structure ; after which four times as much water is worked up with it for a few minutes, and, lastly, the superfluous water is squeezed out by a press. SCIENCE AND PRACTICE. 227 The pulp is put loosely into the space between the porous pot and glass jar, and then hammered down with wooden tool and mallet, until it forms a compact mass. When the element is filled to a sufficient height with pulp, an annular disc of brown paper or cotton cloth is put over it. The zirc ring z surrounds the tube c c, and rests upon the paper disc e e, over the pulp, reaching to within an inch or so of the top of the vessel. It is cast with a neck, g, which prevents the local action that would set in between the copper wire h and zinc, were the acidulated water to reach the former. As it is, the whole ring is immersed in the liquid ; the connection with it is made by means of a copper terminal screw, ti, on the top of the wire h. To charge the element, small crystals of sulphate of copper are dropped down the chimney c c, into the copper compartment c' c ; both compartments are then filled to the same height with water, which may be fresh, acidulated, or contain table salt. When once set up, these elements require only to be supplied with sulphate of copper from time to time, and the water, which evaporates from the zinc compartment, re- plenished. When new, these batteries have a greater resistance than the common form of Daniell's, but not so much as to incapacitate them from employment in local circuits. Messrs. Siemens have during some years em- ployed them to a great extent in measuring the insulation resistances of cables, and have found them, when regularly supplied with sulphate of copper and water, and cleaned occasionally, to possess ninety per cent, of their original electro -motive force at the end of six months. 8. If, instead of paper-pulp, oxide of zinc be used for forming the diaphragm, the zinc ring will remain for months free from traces of copper, whether the element be in closed or open circuit. This plan of employing oxide of zinc was communicated to the author by Mr. Yarley, its inventor. An element properly constructed upon this principle affords, in our opinion, the best form of Daniell's element which exists. The particles of foreign metals, separated from the Q2 228 THE ELECTRIC TELEGRAPH. zinc, fall upon the oxide of zinc and remain there unaltered ; and the sulphate of copper which, by diffusion, makes its way through the porous pot, can only reach the zinc ring after traversing the whole body of the oxide of zinc ; but the acid of the sulphate of copper, on entering the upper compartment, combines with the oxide, forming sulphate of zinc, which is dissolved in the liquid, and releasing the oxide of copper, which remains in the form of a black powder. So long, therefore, as there is oxide of zinc in the diaphragm, through which the sulphate of copper must pass, the zinc will retain its original lustre and its electro-posi- tiveness unimpaired. Of course such a diaphragm gives rise to a great resistance perhaps twenty times as much as an ordinary porous pot but this is of no consequence when the resistance of the circuit outside is great. 9. Sand Batteries. For the use of the Needle telegraphs, the element mostly in use consists of plates of amalgamated zinc and copper, the spaces between them being filled up with sand moistened with sulphuric acid and water. Such batteries are mostly made up in boxes containing each ten or a dozen elements; they have a considerable resistance, and after being set up a short time, a very in- considerable electromotive force. Such as it is, however, it lasts pretty constant for many weeks without requiring attention, and is by no means a " dirty " battery. 10. Alum Battery. Stohrer says that the most constant pile he has met with is that composed of carbon and un- amalgamated zinc, when both plates are immersed in sand moistened with a saturated solution of alum in water, and separated by a porous diaphragm. Nine such elements which he used as a battery for telegraphic purposes, retained for two years very nearly their original strength. A disadvantage is found with these batteries which is common^ to all those which liberate hydrogen gas, that after the circuit has been closed for a time without much resist- ance, the battery exhausts itself. On breaking the circuit again, however, it recovers very speedily ; and as the cir- cuit in most telegraphic systems is open the greater part SCIENCE AND PRACTICE. 229 of the time, and the periods of closed circuit are not of long duration, this inconstancy is of little importance, for which reason Stohrer considers it well adapted to the general work of telegraph lines, particularly as it continues in good order for a very long time if treated with a little fresh alum water about once a month. 11. Marie Davy's Proto-sulphate of Mercury Battery. Davy discovered the greater electro- motive force resulting from the decomposition of a mercury salt at a carbon plate than that of a copper salt at a copper plate, as in DanielPs arrangement. His battery is extensively used now in the French and in some of the English telegraph bureaux, where it is preferred to Daniel!' s on account of its greater electro- motive force, and the comparative ease with which it is kept in order. The element is composed of a zinc-plate, immersed in pure water, separated by a porous diaphragm from a carbon plate immersed in a paste of proto-sulphate of mercury and water. The internal action is similar to that of Daniell's ; the mercury salt is decomposed by the current, metallic mer- cury being precipitated upon the carbon. These elements are commonly arranged in France in boxes of ten ; the zinc cylinders packed in sponge, which holds the water and retards its evaporation ; and the porous pots covered with cork, by which the carbon plates are suspended in the salt. The battery requires very little attention, and according to the inventor's statement, will retain its original force for six months. The intensity which a battery of these elements may have is limited in practice, because if the strength of the current exceeds a certain amount, the dissolution of the zinc goes on faster than the decomposition of the salt, and therefore there is an insuf- ficiency of oxygen liberated to combine with the hydrogen, which, collecting upon the carbon, weakens the proper polarisation of the system ; or, the mobility of the sulphate of mercury paste being small, the salt becomes quickly exhausted in the immediate vicinity of the carbon plates, and is replaced by a stratum of water. This is a phenomenon 230 THE ELECTRIC TELEGRAPH. observable more or less with every kind of element when a great number are connected up in series. 12. The Earth-element, or Terra- Voltaism. Steinheil says, speaking of this method of telegraphing, " On the repetition of the experiment of using the earth as conductor, M. Gauss provided the ends of the line wire at one station with a copper, at the other with a zinc plate. When these came into contact with the earth a powerful galvanic current traversed the circuit. The place of the acidulated cloth in the common voltaic pile was in this case taken by a thickness of 3,000 feet of earth." Bain, in 1844, arranged a similar system in the hope of being able to work his telegraph with it ; but the success attained was only moderate. He, as well as Robert Weare, however, succeeded better with the earth battery for working their electric clocks. Two years later Steinheil employed an earth-pair for working his instruments on the serial line between Munich and Nanhofen a distance of more than twenty English miles for both railway and general service. A copper plate of 120 square feet surface was buried in the earth at Munich, and in Nanhofen a plate of sheet zinc of the same dimensions. The current of this earth-pair was found to be amply sufficient to give the necessary signals by deflecting the magnet needles. In later years, Mr. Septimus Beardmore, of London, has taken up the advocacy of this system, which he has christened " terra- voltaism," and although he has not pushed it much farther than it was when it left the hands of Steinheil, he has expended much time in its study and made some interesting experiments. A very excellent suggestion of this gentleman to increase the electro-motive force of the pair, which he has found insufficient to overcome the resistance of a long line, is by the employment of an amalgam of potassium, moistened with diluted sulphuric acid, for the positive metal, and platinised graphite for the negative element, immersed in a solution of bi-chromate of potassium and sulphuric acid. The great SCIENCE AND PRACTICE. 231 cost of the potassium was found, however, to be an objection to its employment in practical telegraphy, and he conse- quently used, in some experiments made at Greenwich in 1860, a highly electro-positive alloy of sodium and zinc in porous pots in the earth. The system never had, and probably never will have, a chance of being employed on a line of any length. It costs more than a single ordinary battery, is not so easy to control, and its electro-motive force can never be increased beyond that of a single element ; whereas both the Morse and the pointer telegraphs require a current of considerable intensity to set them in motion. The sole advantage to be gained by its employment would be its constancy, if very large plates were buried deep enough in the earth to ensure them being uniformly moist ; the current would last until the whole of the electro- positive metal became converted into oxide a process which would take a very long time. An economical element is arranged by burying a copper plate in the damp earth, connected to an insulated wire, for the positive pole, and a wire connected to gas or water pipes for a negative. An iron-copper pair is thus obtained, which will continue active as long as the pipes last. 13. Amalgamated Zinc. Sturgeon* it was who suggested the amalgamating of the zinc plates of galvanic elements. Two important advantages were proposed and obtained by it : first, amalgamated zinc is not soluble in dilute sulphuric acid when no galvanic current passes from the metal to the liquid, and then only to an amount which is exactly equivalent to the strength of the passing current : there- fore, when the circuit is open the zinc is not wasted, as is the case when unamalgamated zinc is placed in acidulated water ; and, secondly, amalgamated zinc is considerably more electro-positive than unamalgamated. To these advantages may be added that the zinc of commerce contains always metallic impurities, amongst which iron, lead, cadmium, and manganese are the foremost. When unamalgamated, these * Researches, 1830. 232 THE ELECTRIC TELEGRAPH. metallic impurities, after a time, collect on its surface, and form an insoluble crust, which lessens the electro-motive force of the pair by preventing the dissolution of the zinc and by its different electric condition ; while amalgamated zinc, on the contrary, separates the particles from its surface, and allows them to fall off to the bottom. The way in which zinc plates or cylinders are amal- gamated is by dipping them first, for a minute, into a vessel containing dilute muriatic acid, and then by plunging them into a bath of metallic mercury. After remaining here for a minute, they are taken out and thrown into a tub of clean water, where the superfluous mercury is allowed to drain off. Berjot, in order to save the quantity of mercury which this method entails, has suggested a process of amalga- mating without metallic mercury. He dips the zincs into a solution of mercury in nitro-muriatic acid for a few seconds only, to completely amalgamate them. He recom- mends the process as easy, cheap, and certain.* II. MEASUREMENT OF THE GALVANIC CURRENT. 14. The Voltameter. When two platinum wires con- tinuations of the poles of a galvanic battery are plunged into water, bubbles of gas are observed to form on each of them, rising to the surface when their bulk becomes sufficiently great to overcome their adhesion to the metal. These gases are hydrogen and oxygen the former is developed at the zinc electrode, and the latter at the copper. The various forms of apparatus constructed for collecting and measuring the volumes of these gases are called volta- meters. One of the handiest for measuring large quan- tities of gas developed in stated intervals is the following : * The solution is made by dissolving 200 grammes of mercury, at a mode- rate heat, in 1 kilogramme of nitro-muriatic acid (1 part NHO d to 3 parts H 01.), and after complete solution, by the addition of another kilogramme of nitro-muriatic acid. This quantity of solution is said to be sufficient for amalgamating from 150 to 200 zincs. SCIENCE AND PRACTICE. 233 A wide-mouthed bottle (Fig. 120) is three-quarters filled with sulphuric acid (sp. gr. 1'3). The mouth is filled up by a leaden stopper, A A, through which, in small glass tubes, two well-insulated copper wires, c and d, are led, their ends being soldered below to two plates of platinum foil, and protected by a coating of varnish or resin against Fig. 120. Ffc. 121. the corrosive action of the acid. The upper ends of the wires c and d are furnished with binding screws, by which they may be brought into contact with the poles of the galvanic battery. On a glass tube, b, also cemented into the cover A A, an S-shaped glass tube of the same diameter is attached by means of a short piece of india-rubber pipe. The lower curve of this S-shaped tube is placed in a pneumatic trough underneath the opening of a glass mea- sure, graduated in cubic centimeters (Fig. 121). The poles of a battery being connected to c and d respectively, the current passes from one of the platinum plates to the acidulated water, and from the latter to the other platinum plate. In 234 THE ELECTRIC TELEGRAPH. its passage the particles of water are decomposed and their constituents given off. Of these the bulk of hydrogen gas given off at the negative electrode, in a certain time, is double the volume of the oxygen given off by the positive. The bubbles rise to the surface of the acidulated water, mix, and form an explosive gas, and at the same time an equal volume is forced out of the bottom of the S-shaped tube, and rises up into the graduated measure (Fig. 121). In calculating the intensity of the current from the volumes of water decomposed in given times, it is necessary to make allowance for temperature and barometric pressure. The common air is first expelled from the apparatus by letting the gas bubble out a few minutes before putting the graduated measure over the tube. The latter should have as small a bore as possible, without having too much resistance. According to Marriotte's law, the volume of a confined gas is reciprocally proportional to the pressure upon it, and therefore to its tension. Thus, if under the mean pressure of the atmosphere 760 millimetres of mercury the volume of the gas is v, under a pressure of two such atmospheres it would be only half as much, or . More generally, if h express the height of the barometer column in millimetres, when the measurement of the volume v h is made, and v t the volume of the same gas correspond- ing to the mean pressure, 760 mm , the temperature being constant, or, Physicists are generally agreed that the expansions of all the gases are very nearly, if not absolutely, equal, with equal increments of temperature ; in other words, that one co-efficient of expansion is common to all, and that this co-efficient for every degree of the centigrade scale, in the SCIENCE AND PRACTICE. 235 neighbourhood of the zero point, is equivalent to the 0,003665 part of the volume. Therefore, v t being the volume under mean atmospheric pressure, at the temperature t, its volume v at the tem- perature will be 1 -|- 0,003665 t As a numerical example, let the quantity of explosive gas developed in a minute be 25*64 cubic centimeters, measured in the graduated tube, the temperature of the gas 21 C., and the height of the barometer, at the moment, 775 -5 mm , we have then 775-5 x 25-64 centimetres 760(14-0,003665 x 21) for the volume of the explosive gas which would have been observed had the temperature been 0and the barometer 760 mm> Faraday has proved that the chemical and magnetical effects of galvanic currents are proportional to their strengths and to the volumes of water decomposed by them in stated times. It is therefore only necessary, in order to compare the intensities of two currents, to compare the respective volumes of gas developed in the same time and under the same conditions. Unfortunately, however, we can only measure in this way the currents in circuits of which the measuring apparatus itself forms part, and as the con- ducting power of water, even when highly acidulated, is very small, the resistance of the voltameter quite overpowers it, when the electro-motive force is feeble, so that no decom- position, or very little, occurs, and no satisfactory result can be obtained. 15. For this reason we are obliged to have recourse to the deflection of a magnetic needle suspended within the coils of a multiplier ; and when we know what function of its deflections is proportional to the currents producing them, the method affords us the most delicate measure we can desire. The object of all measurement is the comparison of some unknown with some known magnitude ; and this known 236 THE ELECTRIC TELEGRAPH. magnitude is termed the unit of comparison. Jacobi, of St. Petersburg, compared the volumes of gas developed in his voltameter with an unit of volume developed in an unit of time by an unit of current at a certain temperature and tension. ' His expression was : " The unit of current is that current which in one minute, at a temperature of C., and under 760 mm pressure, develops one cubic centimetre of explosive gas." The value of the deflections of magnetic needles suspended in multipliers of wire, inserted in the circuit of a battery and voltameter have been compared with the volumes of gas developed in certain times, and the laws of their deflection thus ascertained. Ampere observed that when a positive current moved in a wire from south to north, over and parallel to a magnetic needle, the latter was deflected, with a tendency to place itself at right angles to the wire, the north pole pointing westward. Subsequently, Biot and Savart have occupied themselves with the task of establishing the relation between the deflection of the needle and the distance of the galvanic current moving in a straight conductor of infinite length, and have found that : " The total effect of an infinitely long and straight con- ductor upon any magnetic element is in inverse proportion to the distance of the element from the nearest point of the wire." For the effect of a circular current upon a magnetic element, "Weber has given a mathematical development, from which he proves that when the distance of the magnetic element from the centre point of the current circle is x, its diameter y, the intensity of the current g, and the magnetic intensity of the element which is deflected /*, the force J with which the deflection takes place is expressed by which, translated into words, is, that a magnetic element in the axis of a circular current is attracted or repelled from the SCIENCE AND PRACTICE. 237 centre with a force which is directly proportional to the superficial contents of the circle, and inversely to third power of the distance of the element from the periphery. 16. Tangent Galvanometer. When a magnetic needle is deflected by a ring in which a current is circulating, the force of the deflection is the resultant of the attraction and repul- sion of the ring upon all the magnetic elements composing the needle ; but as these attractions and repulsions depend upon the position of the needle with respect to the ring, and alter with it, when the former is turned from its normal position the resultant is evidently no longer the same. To reduce the difference, however, to an amount so minute as to be neglected, the diameter of the ring must be made so great in comparison with the length of the needle that the distance of each magnetic element from the periphery, in whatever position the needle may stand to the axis of the ring, may be said to be the same. When the needle n s, Fig. 122, is at rest, and the ring surrounding it is in the plane of the magnetic meri- dian, any current in the ring which tends to deflect it from its position of rest acts at right angles to the direc- tion of the horizontal component of the earth's magnetism, and its force may be represented by a line through the pole of the needle at right angles to magnetic north and south. Such a current will deflect the needle, which will take up a position, n s', making an angle, a, with its former position. Suppose the whole magnetism of the needle to be collected in the pole n, and this to be acted upon, in the direction n s by the magnetism of the earth, and in a direction at right angles to n s by the magnetism of the ring ; when the pole n takes its place at n', these forces are balanced and equivalent to a force acting in the direction n' s 1 . Let this force be decomposed into the forces a n and an' at right angles to each other ; each of 238 THE ELECTRIC TELEGRAPH. these forces may be further decomposed into forces acting at right angles to the direction of the needle and parallel to it. Thus, the force a n, the horizontal force of the earth's mag- netism upon the pole, is equivalent to c n at right angles to the needle, combined with b n r parallel to it, while, in the same way, a n is equivalent to the forces e n and d n'. But as the forces b n and e n act parallel to the needle, they can produce no deflection, whilst c n and d n', acting m opposite directions between which the needle is balanced, must be equal to each other. The angle a = n n' = a n b = a' n d] by plane trigonometry, n d = n' d. cos. a. n' c = n' a, sin. a. And since rid = ric, n' a' cos. a = n' a, sin. a, whence -? = n' a, tan. a. cos. a Putting, for the sake of clearness, the horizontal component of the magnetic force of the earth, n #=:M, and the strength, or magnetic force of the current, n a =. S, we have S = M, tan. a. By the same reasoning, when the strength of the current is altered to S', the angle of deflection will be altered to a'. M remains constant, and S:S'=M tan. a : M tan. a'; or, S : S' tan. a : tan. a'. Therefore the magnetic effects or strengths of two currents are proportional to the tangents of the angles through which the magnetic needle is deflected from the magnetic meridian. But to compare the values of S and S', measured with dif- ferent rings and needles, we have the proportion S : S' = C tan. a : 0' tan. ', SCIENCE AND PRACTICE. 239 C and C' being constants of sensibility of the two arrangements. They are functions of the diameters of the rings, and of the distances of the magnetic poles. If n s (Fig. 123) is the direction of the magnetic meridian .p and the plane of the deflecting ring, and n s the direction taken up by the deflected needle, by plane trigono- metry a b tan. a = an Following Weber's equation, however, and, since a n is equal to p M the Fig- l 23 - attraction between the magnetic pole and the earth we get 27TV 2 tan. a = M in which g (or S), the strength of the current, is 2 tan. a. Therefore and by the same reasoning C' =M Pouillet's tangent galvanometer is constructed on these principles. It consists of a copper ring of large diameter, erected in the plane of the magnetic meridian, and of a short magnetic needle in its centre. A circular copper band, or ring, is bent outwards at the ends to form parallel connections, which are properly insulated from each other and attached to the terminal screws for re- ceiving the wires of the galvanic circuit. In the lower half of 240 THE ELECTRIC TELEGRAPH. the ring a wooden frame is supported, which keeps it in form, and carries a compass-box, containing the magnetic-needle. The latter is short and is cemented to the middle of a long glass fibre, which serves as a pointer, and allows the divisions of the scale over which it moves to be of a considerable size. " The ring, with its continuations, is supported upon a tripod with levelling screws, in which it is turnable for facility of placing it in the plane of the magnetic meridian. The distance of the ring from the needle renders the latter per- fectly insensible to weak currents, and a multiplier becomes necessary. Messrs. Siemens have constructed a tangent galvanometer in which the copper ring is replaced by four separate, thick, well-insulated copper wires, bent in form of a circle of about the same diameter as the ring in Pouillet's instrument, and terminating round the pedestal on which they are supported in four pairs of brass terminal screws. With this arrange- ment the galvanometer can be made twice, thrice, or four times as sensitive by letting the current pass as many times round the needle. It may also be used as a differential gal- vanometer by letting the current pass in reverse directions through the convolutions. The magnetic needle with its pointer of glass or aluminium is suspended at the end of a fibre of unspun silk hung from an adjusting screw, on the top of a glass tube, and is lowered on to the card by turning the screw when the instrument is not in use. The support turns upon a vertical axis, by which the coil may be placed north and south. Gagain constructed a galvanometer in which he professes to have succeeded in reducing the error arising from the altered position of the magnet, by removing the plane of the ring to a distance of half its radius from the centre of the needle, by which, when the latter is deflected, the one half is just so much more as the other half is less strongly acted upon. Bravais, who undertook the mathematical demonstra- tion of the correctness of Gagain's theory, has proved that when a magnetic needle is subjected to the action of a circular current in the magnetic meridian, the centre of the SCIENCE AND PRACTICE. 241 needle occupying the apex of a cone having the circular current for a base, the tangents of the angles of deflection are almost strictly proportional to the intensities, when the height of the cone is equal to a quarter of the diameter of the base. 17. Sine Galvanometers. When the needle is deflected from the magnetic meridian by the action of the ring or coil, we have seen that the force n c with which the earth's mag- netism strives to bring the needle back to the line n s is equal to the product of the directive force, M, of the earth's magnetism on the needle and the sine of the angle of deflection, or n c M sin. a, and n' c=n d. M sin. a is therefore the value of each of the forces which pull in opposite directions and between which the needle comes to rest. If the convolutions of the ring, which have been hitherto supposed to remain in the same plane, be turned round the vertical axis of the galvano- meter in the direction of the needle, the latter will be deflected still farther from the meridian, but always through a less angle than that through which the coil is turned after it. Hence, in time, a point is reached where the plane of the coil coincides with the direction of the needle, 8i or they are parallel to each other. Let this be now the position of things in Fig. 124, the needle in the position n parallel to it having been turned through the angle a ; the force with which the coil deflects the needle is now not only at right angles to its own plane, but also to the direc- tion of the needle, and is represented, directly, by the line n d, which is also the expression of the current moving in the coil, whilst that part of the earth's magnetism which balances this force is n c, as before. R Fig. 124. and the coil 242 THE ELECTRIC TELEGRAPH. But ri c a ri sin. a = M sin. a; therefore also S = M. sin. a. Any other intensity of current, S', moving in the same ring, will require the instrument to be turned through another angle, a, in order to bring the needle to the zero point of the scale ; and we get another equation, by the same reasoning S' = M sin. '. The relation of the two currents is S : S = sin. a : sin. a'. That is to say, the currents are proportional to the sines of the angles through which the galvanometer is turned to make the coil and needle parallel. Instruments constructed on this principle are called Sine Galvanometers . Multiplier Sine Galvanometer. The galvanometer which Professor Du Bois Rey- mond uses in his beautiful experimental researches on animal electricity consists of a multiplier of from twenty to thirty thousand turns of fine insulated cop- per wire, which act upon a Nobili's astatic pair of magnets. The two mag- netic needles are placed upon a common centre with reversed poles, and are of nearly equal direc- tive forces, so that the difference between them, which determines the di- rective force of the system, is very small, whilst, being Fig. 125. placed one in the centre and the other over the coil, they are deflected in the same sense. SCIENCE AND PRACTICE. 243 This galvanometer is shown in perspective in Fig. 125. The bed of the instrument supporting the coils, c c, is turn- able about a centre in the tripod levelling stand, d d\ its circumference is divided into degrees of arc. The silk fibre by which the needle system is suspended is attached to an adjusting screw in the middle of the cross beam b b, resting upon the upright pillars, a a. A glass cylindrical case and glass top protect the instrument from dust and the fibre and needle from currents of air. The lower needle swings in the centre of the coils of wire ; the upper one acts as a pointer to and is suspended over a graduated card; its position being observed through the telescope, e. The multiplier is wound on two bobbins, which are placed side by side, with the needle system between them. They are of about the same length and magnetic action, and may be used either separately or together. These instruments were used for testing the Malta- Alex- andria cable and others, whose electrical conditions have been under the surveillance of Messrs. Siemens. In their instruments, each of the bobbins have about 3,500 units' resistance and about 12,000 turns, making a total resist- ance of 7,000 units, and, in all, 24,000 turns round the needle. The card inside is graduated from the line, 0, parallel to the direction of the coils, to 90 on each side. These galvanometers may be made of almost any required sensitiveness for weak currents, by making the needle system sufficiently astatic. 18. The astatic condition of a pair of needles is measured by the time which it occupies in making an oscillation across the magnetic meridian. Matteucci had a pair which took seventy seconds to make a single oscillation ; but from five to ten seconds is a very convenient degree of directive force to obtain for the measurement of high resistances by weak currents, otherwise the needle system is liable to change its zero by trifling disturbances over which the operator has no control. B 2 244 THE ELECTRIC TELEGRAPH. An astatic pair of needles never takes the direction of the magnetic meridian, but assumes a position at an angle which increases as the difference between the force of the needles is diminished, or as they become more astatic. Dubois calls this the arbitrary deflection of the needle pair. The cause of this arbitrary deflection has been ascertained by Nobili to be that the two needles are never suspended absolutely in the same vertical plane, but that the vertical plane which coincides with one needle makes always an angle with that which coincides with the other. This will be easily understood from Fig. 126, where n s and n' s' represent the horizontal projection of two needles of equal size and magnetic moment, and N s the line of the magnetic north and south. Suppose the force of all the magnetic elements of the needles which tend to turn them to the poles N and s to be collected in the ends n, ri, s, and s', and to act in lines parallel to the line N s, then, if / represent the force acting on the end n t the product / I will be the static moment with which this force tends to turn the north end of this needle in the direction of the arrow, and 2 f I that exerted upon the whole needle, the south pole being attracted in the reverse direction, but in the same sense with regard* to the point of suspension o. The other needle, n' tf, having an equal amount of magnetism, but being at an angle a with n s, the total force with which it is drawn round in the other direction is 2/ 1'. It is evident, therefore, that the needle system can only come to rest when the opposing forces or when 2 f 1=2 ft, 1= I SCIENCE AND PRACTICE. 245 and this can only occur when the line bisecting the angle a stands at right angles to the line N s. When the magnetism of the two needles is of different intensity, let n s have the magnetism f t and n s' an amount equal to /' ; then, in order that the needles take a certain direction, the forces must be balanced, or I __ 2/' f ~ 2/ Now the limits of the possible values of _ . t are evidently I 1 and oo on the same side of the line N s. The value of __ is 1 when the line bisecting the angle a is at right angles to N s, and it will be infinite when n' s and N s coincide ; but when this takes place it follows that the proportion iZ. between the magnetism of the needles must also be infinite, that 2 /, the magnetism of the needle n s, must be very small infinitely so in comparison with 2/', the magnetism of the needle n' s. Therefore, when the disproportion between the magnetisms of the two needles of an astatic pair is very great, the stronger magnet points magnetic north and south. Between these extremes, as the relation -Z. becomes finite, the needle pair places itself at various angles between and 90. 19. Magnetism of the Coils. Copper is not a magnetic metal ; nevertheless, Dubois, TyndaLL, Melloni, and others, have found large multipliers of insulated copper wire mag- netic to a degree great enough to cause a permanent deflec- tion of the astatic pair to 30 where no current passed through the circuit. Suspended within such magnetic coils, the needles usually show a disinclination to come to rest on the zero line, but take up with equal facility a position on either side of it. The magnetism of the copper coil has been variously attributed to the mixture of magnetic metals in the copper, to the iron which adheres to the wire as it leaves 246 THE ELECTRIC TELEGRAPH. the drawing plate, and to magnetic matter in the insulating covering. The impurity of the metal may be avoided by taking galvanic copper, or, as this is rather brittle and requires to be melted over and over again before it can be drawn, by taking copper which, when tested by being held in the neighbourhood of a delicate magnet, affords no trace of any magnetic metal. The adhesion of iron to the wire may be prevented by drawing it through holes in agate plates, or, if these are not to be had, by letting the wire drawn through ordinary steel dies be placed for a few hours in a bath of cold muriatic acid before being covered with silk. Professor Tyndall traced the magnetism of his coil to the silk, and believed that the green dye used in colouring contains some magnetic substance. When he substituted bleached silk, he found the disturbance vanish. For our part we have always found white silk injurious to the eyes of the workmen employed in winding the coils, and prefer green on that account. Mr. Vogel, of Berlin, whose wires are perhaps the most uniformly drawn and the best covered of any we have yet met with, has introduced the use of aniline for dyeing the silk with which he covers his wires, thus satisfactorily removing the last difficulty, as aniline is totally free from any perceptible magnetic influence upon the most delicate needle system. 20. Sine and Tangent Galvanometer. A combination of both principles in one instrument has been made by Siemens and Halske. It is furnished with two separate coils of wire on the same ring one of a few turns of thick wire, the other of many turns of thin wire. Two magnetic needles are also used with this instrument : that for tangent readings is short, and attached to a long brass or aluminium pointer ; that for sine readings is longer, and attached to a similar pointer. The ring round which the wire is coiled is supported by a circular plate, carrying in its centre the compass-box; and is turnable in a graduated metal ring, for the purpose of reading off the angles through which the coil is turned for sine readings. When the instrument is used as tangent SCIENCE AND PRACTICE. 247 galvanometer, the angle of deflection is read off on a card, inside the compass-box, with the shorter needle ; when used as a sine instrument the other needle is used, and the circular plate, with the coils, turned in the same direction as the needle is deflected, until they coincide. The thicker coil, whose resistance is about one-tenth of an unit, consists of sixteen convolutions of copper wire 1*34 millimetres diameter ; the other coil has above a hundred and fifty units resistance, and consists of over a thousand convolutions of insulated copper wire of 0'25 millimetres diameter. Should the deflection due to a current be too great to be read off, an arrangement is adopted by which a known fraction only of the current goes through the galvanometer, the other going through a shunt. This is done by inserting the shunting circuit parallel to the galvanometer coil. 21. Weber's Mirror Galvanometer. In most of his experi- mental researches in galvanism, Professor Weber has em- ployed a galvanometer, the magnetic needle of which is a circular steel mirror reflecting the divisions of an ill mm* - nated scale placed at some distance from it into a telescope Fig. 127. through which the observer reads off the deflections of the mirror. Its principle is precisely that of the receiving in- strument used in Gauss and "Weber's telegraph already explained. 248 THE ELECTKIC TELEGRAPH. Fig. 127 shows a plan of this arrangement, c c is a paper scale divided into 1,000 equal parts, usually millimetres ; a b, a telescope ; g g, the multiplier ; and m a magnetised steel mirror, about one-eighth of an inch thick and 1 inch diameter, polished on the side facing the telescope, and sus- pended by a long fibre of unspun silk. When undisturbed by the passage of a current through the coil, or from other causes, the mirror takes the direction of magnetic north and south, and reflects the central division of the scale into the telescope ; but when a current passes through the coil the mirror is deflected, making an angle, a, with the line n s, and reflecting some division d of the scale into the telescope, or that point in the line (d m) which makes with the line a o m the angle 2 a. Within 5, the values of sines and tangents are, within a very small fraction, equal to their angles ; so that, when the angles do not exceed this limit, they may be taken without further reduction as proportional to the cur- rents producing them. With Weber's instrument this limit is never exceeded : the length o c being 0'5 metre, and o m, the distance of the scale from the mirror, usually more than 5 metres. Besides, where the diameter of the mirror is small in comparison with the diameter of the coil, we have seen that the currents are proportional to the tangents of the angles a. With this instrument = tan. 2 a ; but as, for very small angles, we may put tan. 20 = 2 tan. a without appreciable error, we can accept also the values of o d as being proportional to the tangents, and therefore to the intensities also. Fig. 128 shows a vertical section of such a galvanometer. a a and d a! are two circular coils of fine copper wire insu- lated with a covering of silk, forming the multiplier. In some of these instruments the coils a a and d d are divided into a number of coils of different lengths and gauges of wire, terminating in a series of binding screws, b b, outside the frame of the coil. The frame on which the wire is wound is of vulcanite. The mirror m is suspended by a fibre,/, from a little windlass, r, on the cap c of the vertical SCIENCE AND PRACTICE. 249 glass tube g g. The mirror is raised or lowered in the coil by turning the milled head of the reel r, and may be removed entirely from the galvanometer after taking off the glass tube g, through a slit in the frame between the coils a a and a' d. e is a glass plate to guard the mirror from currents of air, and d a solid cylindrical block of copper put behind the mirror for the purpose of retarding the freedom of its oscillation, and bringing it quickly to repose. The checking action of a solid mass of non-magnetic metal in the presence of a moving magnet has already been alluded to. Arago believed this action to bo due to the attraction and repuk sion of currents of magneto- elec- tricity set up in the mass by the moving magnet, and which have the effect of opposing its motion. 22. Thomson's Mirror Galvanometer is a modification of "Weber's, differing from it in many important regards. Weber's instrument is admirable in fact, necessary for a certain class of measurements ; but for others the mass and sluggishness of the heavy steel mirror are objectionable, as well as the distance of the galvanometer from the observer, entailing as it does a length of connection wires which in fine measurements may be found to be inconvenient. In addition to this, the instruments, with all their adjuncts of telescope, scale, illuminators, &c., are expensive and cum- bersome. Professor Thomson has avoided all this in taking a mirror whose weight does not exceed a few grains, and whose momentum is therefore very small, and in dispensing with the telescope by throwing a spot of light directly upon Fig. 128. 250 THE ELECTRIC TELEGRAPH. the scale, and lessening its distance. For all measurements in which the instrument serves as a galvanoscope, as in Wheatstone's bridge, PoggendorfFs compensation method of comparing electro-motive forces, &c., and when the readings are not very different in value, it must be confessed that this galvanometer is much to be preferred. Mr. Becker has given it a very convenient form, by fixing the coil in the back of a brass barrel or cylinder, in the front of which a glass plate enables the interior to be seen from before, and prevents dust and currents of air getting to the needle. In the centre of the coil is suspended, by a fine cocoon fibre, in a frame, a small silvered mirror* of microscope glass, between one- eighth and one- fourth of an inch diameter. A little magnet, made of a piece of thin watch-spring, is fastened to the back or silvered side of the glass, and, being magne- tised, operates as the needle of the system. Above the brass barrel a vertical rod carries a curved permanent adjusting magnet, and a rack and pinion enables the latter to be turned round horizontally to bring the point of light to any part of the scale which may be desired. The adjusting magnet is elevated or depressed on the vertical rod for the purpose of increasing or decreasing the directing force upon the magnet needle. When the ends of the adjusting magnet coincide with the poles of the earth's magnetism, it adds to the directive force of the latter, and the instrument becomes proportionally unsensitive. The magnet may, however, be turned round so as to oppose the directive force of the earth, and in this position be lowered towards the mirror, until it very nearly neutralises the earth's * The process of depositing metallic silver upon glass is as follows : (A) Dissolve 10 parts of nitrate of silver in 50 parts of water, and neutralise with (about) 6 parts of liquor ammonia ; add to this a solution (B) of 1 part of tartaric acid in 4 parts of water, and dilute the whole (A -f- B) with 500 parts of water. The things to be silvered should be placed conveniently in a vessel, the solution poured in, and then put away in a quiet place for a few hours, at a temperature of from 40 to 50 C. When silvered they may be washed by a gentle stream of water, dried, and varnished with a solution of amber in chloroform. I f SCIENCE AND PRACTICE. 251 directive force. The instrument so placed has its maximum sensibility. The scale, divided from the middle towards the ends into equal parts, is fixed upon a wooden stand at a distance of two or three feet from the mirror. Behind the scale is a parafiin lamp, whose light falls through an adjustable slit underneath on to the mirror, which reflects it back upon the scale ; and in order that the point of light shall be as well defined as possible, a small plano-convex lens is placed before the mirror, through which the rays converge into a focus, throwing a sharp image of the slit upon the scale. Mr. Yarley has made some of these instruments for the measurements of the Atlantic cable, and has substituted a plano-convex lens, silvered on the curved side, for the mirror in Professor Thomson's instrument, dispensing of course with the lens in front. 23. Rheostats. In the early experimental investigation of the laws of the galvanic current, the comparison of resist- ances was made by lengths of metal wire, which becoming sometimes rather great, an inconvenience was very soon felt in handling them. Wheatstone first overcame this by rolling the wire round a cylinder of dry boxwood, on which a worm was cut just deep enough to receive it comfortably, and to facilitate the variation of its length ; the other end of the wire was coiled upon a cylinder of brass in such a way that the point where the wire touched the cylinder as a tangent to its circumference should be the point of contact, and from this point the length of the wire on the non-conducting roller to the end was measured. The cylinders of boxwood and brass were fixed in bearings parallel to each other upon a wooden board. The worm on the wooden roller was cut from end to end, comprising about forty turns to the inch. The wire, whose thickness did not exceed the one-hundredth of an inch, was connected at one end to a metal cap which covered the nearer end of the wooden roller, round which it followed the course of the worm until it left it to be wound upon the metal cylinder, to the further 252 THE ELECTRIC TELEGRAPH. end of which the other end of the wire was connected. The metal cap was in permanent connection, through a spring pressing upon its periphery, with a terminal screw ; while a similar spring-contact kept the brass cylinder con- nected with another terminal screw forming the ends of the system. The axis of the wooden roller was furnished with a pointer or index which turned with it over a circular dial, and indicated the fractions of turns, whilst a straight bar between the roller and cylinder, graduated correspondingly with the worm of the former, showed the number of whole turns upon it. If, in the point where the wire met the brass roller, perfect contact had been made, the length indicated by the rule and index would have represented the resistance ; but this was never strictly the case : there was always a resistance to passage at the point in question, which, being nearly con- stant, had more effect when the length of the wire in circuit was small than when it was great. 24. Jacobi's .Zs^osfo^. Jacobi, of St. Petersburg, also in- vented a Rheostat, whose purpose, like that of Wheatstone's, was to render the handling of lengths of wire for resistances convenient to the operator. It consisted of a roller of dry wood, in the worm of which, from end to end, a long German- silver wire was wound tightly. One end of the roller was furnished with a metal cap, to which that end of the wire was permanently attached ; the other end of the wire was insu- lated. In front of the roller was fixed a straight round bar of brass, on which a metal jockey wheel, with a groove in its rim, rode over the German-silver wire, pressing upon it suf- ficiently to make a tolerably good contact. The Rheostat was put into a galvanic circuit by means of terminals : the one in connection, through the metal bearing and cap, with the wire, and the other forming one of the supports of the guide-rod. The current passed through the support and cap and through the convolutions of the wire, until it reached the jockey wheel, by which it left the wire. By turning the handle on the axis of the roller, the jockey wheel travelled along the guide-rod, and more or less resistance was intro- SCIENCE AND PRACTICE. 253 duced by bringing, in this way, the jockey wheel towards one end or the other of the roller. 25. Poggendorff's Rheostat. Professor Poggendorff called the instrument which he arranged for the same purpose a Rheocord. Four parallel wires were stretched on a board between terminal screws, the two middle ones being connected permanently together. Between the two on the one side and the two on the other, sliding contacts were introduced, which could be brought to the extremes at each end.- The current from the terminal went therefore through No. 1, crossed over the sliding contact, went down No. 2, crossed to 3, traversed 3 as far as the other sliding contact, crossed to 4, and left 4 at the terminal on that side. When the sliding contacts were brought down to the bottom, the current passed from terminal to terminal over the contacts without going through any length of the four wires ; whereas, when the slides were at the top, the current had to pass through the whole length. The places of the contacts were read off by their distances from the bottom. These distances being I and I', the resist- ance R, between the terminals, expressed in length, was, therefore B=2(Z -M') We have spoken of these arrangements in the past tense, as we believe they are one and all superseded by those which follow. 26. Siemens' Resistance Boxes. A much more handy method of varying the length of the interposed resistance wire is by means of a succession of short circuits between different points of its length, the wire being stationary in- stead of being continually wound and unwound or touched by contact rollers, as is the case with the Rheostats, by which the wire may easily become elongated and hardened, and is always liable to be damaged. The method we speak of is best understood by supposing a length of wire (Fig. 129) between the terminals a and b, so arranged that a point c, at the distance a c from the end a, equal to one unit, can be put into direct communication, by means of a short circuit, with the terminal b. A current 254 THE ELECTRIC TELEGRAPH. passing between a and b will encounter only one unit of resistance. In the same way, if the point g y midway between a and b, be put in short circuit with either a or b, the current will meet on its way between a and b with only half the total resistance of the wire. In the same way the intermediate Fig. 129. points, g and h, being connected by a shunt, the resistance between the ends will be equal to the sum of that between a and g and that between h and b. The resistance of the shunt in each case being infinitely small, its resistance does not appear in the result. By this means a considerably greater length of wire can be made use of, and the body of the wire be protected by a case or otherwise, the points c, d, g, h, &c., only being neces- sarily at the command of the operator. A highly useful arrangement, for this purpose, is shown in Fig. 130, where the various points in question are con- Fig. 130, nected to a series of brass terminals, e, /, g, &c., so placed that, by inserting a metallic wedge or plug between any two of them, the length of wire contained between them is short- circuited. If a plug be thus inserted in the hole between e and /, for instance, the resistance d will disappear the current passing through the plug, the resistance of which is infinitely small. Between the two first terminals, a and e, a length of wire, whose resistance is equal to one unit upon the reel b, is inserted; between e and /upon the reel^, twice that length ; the same between / and g, and so on, in the following order : SCIENCE AND PRACTICE. 255 1, 2, 2, 5, 10, 10, 20, 50, 100, 100, 200, 500, 1000, 1000, 2000 and 5000 ; making a sum total, when all the plugs are out and the current passes from terminal to terminal through all the intervening lengths, of 10,000 units. This simple arrangement enables the operator to make the resistance of his apparatus infinitely small by inserting all the contact plugs, or to open any resistance in whole numbers between one unit and ten thousand of them at pleasure. The lengths of wire b d, &c., are of German- silver, insulated with two coatings of silk, sometimes further guarded from the air by a protecting varnish, and wound double upon bobbins of dry wood or vulcanite. The purpose of this method of double winding is to avoid the effect of induction currents in the bobbins on making or breaking contact with the battery ; induction takes place, of course, but the currents circulating everywhere in opposite directions, the effect is eliminated. The coils or bobbins are arranged in a mahogany case and the terminals put upon a thick slab of vulcanite. There is a difficulty which must not be overlooked in using these resistance scales, which is not met with in using the Rheostat. It is the spring which the current makes on changing the contact pegs. When measuring resistances of insulation where the capacity of a jar for charge is present, this evil is principally felt, and it becomes necessary to put the galvanometer out of circuit before making any change, in testing both the insulation and resistance of conductors of long cables, in order to avoid the strong charge and discharge currents which would otherwise alter the magnetism of the needle. 27. Eisenlohr's Resistance Column. Another form of resistance scale is that arranged by Professor Eisenlohr. Seven brass rings are fixed upon a cylinder of dry wood, at equal distances from end to end. In the space between each two of these rings is coiled a length of well-insulated wire, one end being soldered to the upper and the other end to the lower ring. The length in the first space has a resist- ance equal to one German mile of telegraph wire, that is to 256 THE ELECTRIC TELEGRAPH. say, about 64 of Siemens' units ; that in the second space has a resistance representing two such miles, and so on to the sixth space, the wire of which has a resistance of six miles. Any two neighbouring rings can be brought into short-circuit by means of the brass contact pieces which turn on pins in the upper rings. In this way either or all of the coils may be short-circuited the current passing only through those which are open. When all the brass contact pieces are closed the resistances are all short-circuited ; when all are open the resistance between the top and bottom rings is equal to 1 + 2 + 3 + 4 + 5 + 6=21 miles. 28. Ohm's Law. Until the end of the first quarter of the present century physicists were still in darkness as to the mode and laws of the propagation of the galvanic current. The immense velocity with which the galvanic impulse is transmitted led to the seeking an analogy between it and light ; and on this wrong scent much time and labour were lost, when Ohm, a German physicist, conceived the happy idea that a juster analogy was to be found in the propagation of heat, and proceeded to apply to galvanic electricity the formulae of Fourrier and Poisson. He expressed the intensity of an electric current as directly proportional to the electro- motive force, and inversely to the resistance of the circuit. Algebraically, if E is the electro-motive force, R the resistance, and I the intensity, Of these magnitudes R is made up of two resistances that interior and that exterior to the element. The internal resistance, or resistance of the element, is, again, the sum of the several resistances due to the passage of the current from one plate to the liquid, to its passage through the liquid, and to its passage from the liquid to the other plate. We will call this resistance of the element, r. The remaining com- ponent the external resistance is that due to the passage of the current through the interiors of the plates, the wire connecting them, and through whatever conductor may be SCIENCE AND PRACTICE. 257 otherwise inserted between them. Let this be p. Substi- tuting these values for B, in (I., i = .JL_ ..... (ii. The truth of this equation may be proved experimentally, as follows : Evidence of the direct proportion of the intensity to the electro-motive force is obtained by comparing the known function of the deflections of a magnetic needle of a galvano- meter due to the current in a circuit in which r and p the circuit resistances remain constant while the number of pairs is charged. The resistance r of a pair of plates of equal surface, at the same distance, diminishes as their surface is increased, and vice versa ; but the resistance of more pairs joined up in series, increases proportionally to their number. Therefore, we take a single pair of plates of known surface and connect them in the circuit of a galvanometer, and of a length of wire determined by a Rheocord, or other adjustable resistance, and note the deflection, ^>. Then we double the electro-motive force, E, by inserting, in the place of these , two pairs of plates of each double the surface of the former , by which the resistance r remains unchanged ; the wire p remains also the same, but we have another deflection, ^> 1 . For the intensity I, with the single pair, we have the expression p + r and by the second reading, with two pairs, ^ T _-C.^oN_. 2E 4J . . . Ij _C ( that we can, without appreciable error, neglect it, the intensity of the whole battery becomes I H = .... .(V. P that is to say, that when the resistance of the battery is very small in comparison with the resistance of the circuit exterior to the battery, the strength of the current is increased in direct proportion to the number of elements added to it. s 2 260 THE ELECTRIC TELEGRAPH. Dividing both numerator and denominator of the above fraction, (IV., by the number of elements, n, we get E which becomes, if we set p = 0, affording us light upon another relation of the galvanic current, viz., that when the resistance exterior to the battery is so small that it may be neglected, the current of a number of elements will do no more work than that of a single pair. The first of these laws applies to a battery used for working a long line of telegraph, whose resistance with the coils of the apparatus is very great in comparison with that of the elements, and where it is evident a large battery is necessary. The second law applies to a local circuit, where the resist- ance of the circuit is small and a few elements do as well as a great number. Secondly, let n elements be so combined that all the copper poles are connected together to form a common positive pole, and all the zincs to form a common negative pole. In this case we have still a single element, but of n times larger surface. Theory and experiment prove alike that the electro- motive force of the system is exactly that of a single element, and, according to Ohm's law, the intensity is expressed by I.-- 1- "* . . (YIL p+ np + r Here the external resistance p remains the same, but that of the battery is reduced to . And now by setting, in turn, the resistances p and r as very small in comparison with each other, we find mathematically what good the combination can do us. SCIENCE AND PRACTICE. 261 When p = 0, I n = ...... (VIII. T or, when the circuit resistance external to the battery is in- appreciably small, the intensity increases as the number of parallel plates increases, and in working with such circuits it proves that we do well to take elements of large surface. Whenr=0, . wE E .\ n - - = - .... (X-A.. n P 9 a very important result, which says that when the external resistance p is very great in comparison with that of the element, no greater intensity is obtained by increasing the surface of its plates. Two other cases belong under the same head, but seldom occur, viz., that of combining similar elements of different sizes, and of combining elements of different electro-motive forces in series and parallel. The resistance of similar elements of different sizes will, of course, be different let them be r^, r 2 , r 3 . . . r n ; but the electro-motive forces will be equal, and the intensity of the current of each element joined up in series will be T* n f\ 1= _ 5 __ .(X. p + ^i + r 2 -f r 3 . . . . r n and that of any section in the circuit, the product of this with the number of elements, or, l n = _ !L5 _ : _ . . (XI. 9 + ^ -\- r 2 -{- r s . . . . r n But where these elements are connected up parallel, the intensity of the circuit becomes r 2 T * _ T _n (X +_L + _L + ...._L' r z r s r n In the other case, where the electro-motive forces of each 262 THE ELECTRIC TELEGRAPH. of the elements, E p E 2 , E 3 . . . E n , are different, and resist- ances, r 19 r 2 , r 3 ... r n , also, the intensity of the current is P + r, + r 2 + r 3 + . . . r n the elements being connected together in series, and, TT-+ + ~+ + :r I - r i r * ^ r n . . (XII. ' / 1 1 1 1\ hp V~+77+7T '7-J when they are connected together parallel. Lastly, in addition to these methods, it is sometimes neces- sary to determine the best combination and number of elements for a battery, under given circumstances, in order to produce a given effect. For this, some of the elements may be connected parallel, and then these combinations con- nected together in series. This problem of finding the most advantageous combinations is solved nearly as follows by Eisenlohr : If we call the ^surface of the exciting plates = 1, and connect the same in equal elements, the surface of each element will be - of the whole surface ; and if the resistance X of the whole parallel = 1, that of each of the elements sepa- rately = x, and that of all the elements, in series, = x 2 ; then, n being the external resistance, 1-1 T71 (XIII. The denominator x + - of this fraction evidently attains its minimum, the value of x being variable, when p = a? 2 ; but when the denominator of a fraction is minimum, the fraction itself has its maximum value ; therefore, the strength of the current of a battery of given surface of plates is at its maximum SCIENCE AND PRACTICE. 263 when the external resistance of the circuit is equal to that of the battery* Now, let the given battery consist of N" elements, the resistance of each of which is r units, and the external resist- ance p as before ; to determine the manner in which we must couple them in order to get their maximum intensity we must arrange x rows of each - elements parallel. The resistance x of each row must therefore be P _ rx x ' N whence the number of rows, As a numerical example : We have a galvanometer, the resistance (p) of whose coil is 16 units, and wish to arrange CN"=) 100 elements, each of which has (r=) 4 units resist- ance, in such a way that the galvanometer needle is deflected to the maximum. Inserting these numbers in (XI V., we find for the number of rows, and by dividing the total number of elements by the rows, JL=- 10 * " ~20~~ we have the number of elements in each row. The practical limits are evidently when #=N and when a?ssl In the first case the elements are all connected up in series, in the other they are all parallel. It very frequently * If p represents the surface of a rectangle whose sides are a and t>, p = a b ; and, if one side of another rectangle of the same superficial area = x, the remaining side must be , because x = a b. But the sum of the sides of x x a rectangle of given surface are least when the sides are equal, as in this case, when x = , which can only occur when a b #2 = p. (Dub. p. 63.) 264 THE ELECTRIC TELEGRAPH. occurs, however, that the value of x comes out greater than JS", which is no absurdity ; as it proves only that then N is not great enough to give us the maximum which x expresses, and in this case we must take all the elements in series. III. CONDUCTING POWERS or MATERIALS. 29. Specific Conducting Powers. The conducting power of a material is independent of the form and dimensions of the body measured. AVe have already seen that the resistance of a geometrical body of any material is directly proportional to its length and inversely to its sectional area ; it is also inversely proportional to its conducting power. By length is understood the distance between the points where a current enters and where it leaves the body ; by sectional area, the section at right angles to the direction of the current through the body, or to the line joining these two points ; and by its conducting power, the ability which the material has to communicate the electricity from atom to atom along its length. Algebraically expressed, therefore, the resistance r of any body is I 80 I being its length, s its section, and c its conducting power. From this we have the value of y> &c., alternately positive and negative, and to regard those beyond |3 as = 0. Hence the resistance, B^ of a pure metal wire, at the temperature t, whose resistance at C. is R , is __ E / " ^ 100 1000,37647 t + 0,000834 Two exceptions are found to this rule, in the metals iron and thallium : the per-centage variation of the conducting power of pure iron between and 100 is 39 '2, and that of pure thallium between the same limits, 31*4, while the other pure metals vary only 29-3 per cent. 32. Alloys. The conducting powers of alloys of lead, tin, cadmium, zinc, and some other metals, with each other, are proportional to the volumes of the metals entering into their compositions. These metals form a class separate from the others, and are of limited number. For the most part the metals belong to the other class, those which, when alloyed with each other or with one of the metals above mentioned, have smaller conducting powers than are propor- tional to their respective volumes. To this class belong bismuth, antimony, platinum, palladium, iron, aluminium, sodium, gold, copper, silver, and so on This difference is the result of the different natures of the alloys, and depends upon the nature of "the combination of the metals forming them. Many alloys are unquestionably chemical combinations, others are solutions of one metal in another, others, perhaps, only mechanical moistures, and others, again, solutions of one of these in an excess of one of the metals. Of the alloys which enter most largely into matters con- nected with telegraphy, German-silver deserves to be espe- SCIENCE AND PRACTICE. 269 cially mentioned. It is from this alloy that, at present, almost exclusively resistance-coils are manufactured. The conducting power of a specimen when hard-drawn, as deter- mined by Dr. Arndsten, is only 10,532 times as great as that of pure mercury at C., whilst that of a specimen when annealed, according to Dr. Siemens, is still less. Another advantage which this alloy has in common with most of the others is that temperature exerts a comparatively small effect upon its resistance. The conducting power c at a temperature i, according to Arndsten, of this alloy, whose conducting power at C. is 100, is expressed by c = 100 0,0387 t + 0,0000557 P which is very little over a tenth part of the change found for the pure metals. 33. Metals annealed. The degree of hardness or softness of a metal or alloy affects materially its conducting power. That of a hard- drawn wire is not the same as when the wire has been made hot and let cool again ; and to the fa,ct that not sufficient importance was attached to this property has been justly attributed the differences between the results of different observers. Dr. Matthiessen has found it necessary, in order to obtain comparable results, not only to heat wires to 100 C. before measuring their resistances, but even to keep them during several days at that temperature before their resistances became constant below that point. The conducting powers of the metals and alloys are increased by annealing. Peltier first pointed out this phenomenon in the behaviour of copper ; and Matthiessen repeated his experiments with copper and silver, with the following results : Metals. Temp. Conducting Power. Hard. Annealed. CoDDGr . , 11-0 14-6 95-31 95-36 97-83 103-33 Silver 270 THE ELECTRIC TELEGRAPH. The comparison being made with pure silver at 100 C., from which it appears that the conducting power of copper in- creases 2-5 per cent., and that of silver nearly 8 per cent., by annealing. According to Dr. Siemens, the conducting powers of copper, silver, and brass, hard and annealed, are as follows, compared with pure mercury at C. : Metal. Hard. Annealed. Copper 52-207 55-253 Silver 56-252 64-380 Erass 11-439 13-502 This property of copper is especially of advantage in the manufacture of telegraph cables, galvanometers, &c., where a great length of conductor is required with little resistance, and where the metal must be as soft and as little liable to change its molecular condition as possible. 34. Metals fused. Although not strictly within the domain of telegraphy, the conducting power of fused metals is interesting, and brings us to a material most im- portant in the reproduction of standards of resistance mercury. Of the metals which have been subjected to electrical measurement when in a state of fusion, all, as far as we know, with the exception of bismuth, lose their high con- ducting power, and at the point of solidification regain it very rapidly. Tin may be taken as a fair example of this behaviour of the metals. A determination, which we made in 1862, with some pure tin, melted in a glass spiral surrounded by hot oil, gave the following results as compared with the conducting power of pure mercury at C. : SCIENCE AND PRACTICE. 271 Temperature, C. Conducting Powers. 280 1-879 x 273 1-880 f > fused. 263 J 227 1-990 ' 220 4-211 i 70? 6-631 | solid. 24 7-892 ) This is shown graphically by the curve a, b, c, d, Fig. 131, the curve ef showing the corresponding conducting powers of mercury at the different temperatures. Mattecci first observed that bismuth behaved differently at its point of solidification to the other metals. At the freez- ing point of water the conducting power of this metal is about 0-74 times that of mercury ; between this point and 250 C. the melting point its conducting power follows the law common to pure metals ; at its fusing point its conducting power increases suddenly until it equals that of pure mer- cury at the same temperature, very nearly; from which point it decreases again in conducting power as the tem- perature increases. This behaviour is probably due to its crystalline structure. Under the head of fused metals we come to mercury, which is always in this state at ordinary temperatures. Dr. Werner Siemens proposed to employ a body of this metal as unit of resistance. He considered it of the greatest importance to have the unit of resistance expressed as a geometrical body of that material, which is commonly referred to as unit when speaking of conducting powers, by which all practical problems are facilitated. As an instance, if it be required to know the resistance of a cer- tain length and section of any metal, it is only necessary to calculate what it would be in mercury, knowing the length and section of a body of the latter representing 272 THE ELECTRIC TELEGRAPH. the unit, and to multiply the result with the conducting power of the metal required. A great point in favour of mercury is the ease with which it may be procured in a chemically pure state. M-H-H-H rrrn MM LMLLL LI M M M U_ ~~rrr MTFFFffl LUJJJJJJ i M \ JJ-VU MINI Fig. 131. There are two ways which are highly recommended : one followed by Dr. Werner Siemens ; the other followed by Dr. A. Matthiessen with equally good results. By the former method the mercury is placed in an evaporating dish, with about three-quarters of an inch thickness of con- centrated sulphuric acid over it. It is then carefully boiled SCIENCE AND PRACTICE. 273 for some hours, adding sulphuric acid when necessary, and, from time to time, a few drops of nitric acid. Dr. Mat- thiessen allows his mercury to stand some weeks under a cover of dilute nitric acid, with which it is frequently agitated. Another recommendation in favour of mercury is the fact that, being fluid, its molecular condition cannot be subject to changes, as may be the case with the solid metals, especially in the process of drawing, and that, so long as it is pure, it must always have the same conducting power at the same temperature. Lastly, the variation of the conducting power of mercury with difference of temperature is considerably less than that of the pure metals in a solid state. The amount of this variation, as determined by Becquerel, for all temperature degrees between and 100 C., may be taken, without sensible error, as directly proportional to the difference of temperature, being equal to the 0,00104th part of the con- ducting power for each degree of the centigrade scale. A more recent determination by Dr. Matthiessen, which has been equated by the method of least squares, with two members, taking the conducting power of mercury at C. as = 1, gives the conducting power C t , at the temperature t C. C, = 1 0,00074432* 0,0000008261 P. The conducting power of mercury alloyed with even a very minute quantity of foreign metal is greater than that of pure mercury; hence great care is necessary, in using this metal, to procure it as free from impurities as possible. 35. Electric Permanency of Metals. The conducting powers of some of the metals in a solid state at the same tempera- ture seem to be pretty permanent; others, again, appear to alter their conducting powers materially in the course of a few months. As a rule, the hard-drawn wires are the least permanent, by reason of their becoming gradually annealed by exposure to variations of temperature. Dr. Mat- T 274 THE ELECTRIC TELEGRAPH. thiessen's experiments, undertaken for the Unit Committee of the British Association, show that annealed German- silver increased its conducting power in the course of a year at the rate of nearly 0*2 per cent. ; some specimens of gold, silver, and copper, both annealed and hard-drawn, also altered their conducting powers. The change in the conducting power of hard-drawn silver was the most con- siderable, the two specimens experimented upon having been found to have increased at the rate of over 3 '9 and 2*8 per cent, respectively. That some of the specimens of metals experimented with should have shown no change is no argument that they would not do so if allowed to get older ; and errors of observation may sometimes, especially in measuring such minute differences, as well be taken to account for agree- ment as for disagreement. At present the question must be considered an open one, whether metals do change their conducting powers by age ; and if so, if the change in the same wire is always in the same sense ; and if so, what becomes of the conducting power at last, has still to be determined. Another vexed question to be set at rest is, whether the passing of electric currents through a wire is able to alter its conducting power ? Professor Kirchoff says that the conducting power of any wire, at a given temperature, certainly undergoes changes if electric currents are transmitted through it and it is exposed to fluctuations of temperature. Schroder van der Kolk also says that the conducting power of a copper wire undergoes a change whenever weak currents are allowed to pass through it. If this were the case, of what use would be our resistance- scales? Dr. Matthiessen has happily found this to be a fallacy. He allowed a current from two Bunsen's cells to pass through a series of wires of different metals for six days, at the end of which no change in their conducting powers had taken place a result which some experiments of our own, undertaken in the winter of 1862-3, in Germany, SCIENCE AND PRACTICE. 275 with, the view of determining the same question, completely corroborate. We connected a finely -adjusted annealed German-silver (wire) resistance to a self-acting make-and- break apparatus, or " Wippe," which sent reversed currents from a battery of large DanielFs elements, at an immense speed through it both night and day. In addition to this, the wire was kept in a recess in an iron stove in the laboratory, so that, without any interference on our part, its temperature was raised by day at least to the tempera- ture of boiling water, and during the night descended to within a few degrees of the freezing point. The battery was varied at intervals from one cell to twenty, during about six weeks, but the conducting power of the metal did not vary in the least. Suspecting that this constancy might be due to the reversals, we repeated the experi- ment with a zinc current made and interrupted with great rapidity, with the same result. The belief in the inconstancy of metals may have its origin in the discovery that all the old resistance- scales and rheostats are no longer exact. This may have its origin in three causes : 1st, the greater perfection of our systems of measurement enabling us to detect differences which may have existed before, but which we were unable to appreciate ; 2nd, the process of annealing, which, when commencing with a hard- drawn wire, may extend over a very long time if the wire is only exposed to variations of the temperature of the atmosphere ; and 3rd, the oxidation of the surface when the air has access to it. There can be no question that much has still to be learned in this branch of science. Wires of German-silver and some other alloys become, when exposed freely to the air for some years, so brittle as to be incapable of being wound up on reels without danger of occasional ruptures of continuity. That this brittleness is accompanied by a change in conducting power is probable. It remains, how- ever, to find whether the exclusion of air by means of some such material as paraffine will prevent the brittleness in ques- tion, or if it is due to molecular changes of the alloy. T 2 276 THE ELECTRIC TELEGRAPH. 36. Conducting Powers of Fluids. In constructing a battery for any given work, it is necessary frequently to consider the resistance of the fluids used. Wheatstone, Horseford, and others, have invented apparatus for the determination of the conducting powers of solutions, &c. That of Horseford is the best and simplest. It consists of an oblong wooden trough, varnished inside with shellac. On the top are two cross-bars of wood, with guides, over- lapping the sides, to keep them straight. To each of these cross-bars is attached a plate of platinum, only so much smaller than the interior section of the trough as to allow it to be moved freely, with its cross-bar, from end to end. Usually one of the bars, with its platinum plate, is fixed, and the other movable. A divided scale on the upper edge of the trough facilitates the observation of the distance between the plates. Copper wires are soldered to the pla- tinum plates, and serve as connections with the measuring apparatus. When the apparatus is to be used, it is placed as level as may be upon a table, and filled to a convenient height with the solution whose resistance is to be measured. If the relation of the resistance to the distance between the plates is to be determined, the solution is poured in, and the resistances measured with various distances. By this measurement it becomes evident that the resistance of a conductor is directly proportional to its length. By keeping the distance between the plates unaltered, and varying the height of the solution in the trough, the experimenter may convince himself of the truth of another law, namely, that the resistance of a conductor is inversely proportional to its transverse section. But the experiments which are most important are the conducting powers of solutions of the various salts and of the acids. For these measurements, of course, the distance between the plates and the height of the solution in the trough must be constant through the whole series. Becquerel has determined the conducting powers of some of the concentrated solutions, and also of the same diluted with water. When the conducting power of pure silver is SCIENCE AND PRACTICE. 277 taken as 100 millions, those of some of the concentrated solutions are as follows : 1. Concentrated solution of sulphate copper, specific gravity = 1-1707 at 9 C 2. Concentrated solution of sulphate zinc, specific gravity = 1-441, at 14-4 C ; 3. Concentrated solution of chloride sodium, at 9 '5 C 4. Concentrated solution of chloride copper, specific gravity = 1-4308, at 9-25C , 10*35 5. Concentrated solution of nitrate copper, specific gravity, =1-5790, at 10 C 8-40 A striking property of sulphuric acid is that when diluted to a certain point it attains its maximum conducting power ; this point is when the solution has a specific gravity of about 1-215, or when 100 parts, by weight, of the solution contain 29 -6 parts of acid, after which its conducting power again diminishes, as appears by the following table of some of Saweljev's determinations given by Wiedemann in his elaborate treatise : 5-42 5-77 31-52 Specific gravity. 80s HO, in 100 parts by weight. Temperature. C. Kesistance. 1,003 0-5 16-1 16-01 1-018 2-2 15-2 5-47 1-053 7-9 13-7 1-884 1-080 12-0 12-8 1-368 1-147 20-8 13-6 0-960 1-190 26-4 13-0 0-871 1-215 29-6 12-3 0-830 1-225 30-9 13-6 0-862 1-252 34-3 13-5 0-874 1-277 37-3 0-930 1-348 45-4 17-9 0-973 1 393 50-5 14-5 1-086 1-492 60*6 138 1-549 1-638 73-7 14-3 2-786 1-726 81-2 16-3 4-337 1-827 92-7 14-3 5-320 278 THE ELECTRIC TELEGRAPH. 37. Determination of Gfatranic Polarisation. Tlie deter- mination of the resistances of fluids cannot always be made by the direct substitution of a metallic resistance giving the same deflection of the needle of the measuring instrument, because the electro-motive force in the circuit of a fluid resistance is not always that of the cell or battery by which the resistance is measured, but generally this minus the electro -motive force of a polarising layer of gas forming on the plates or electrodes immersed in the fluid. In a circuit, therefore, containing a metal resistance, R, and the resistance of a fluid column, r, the current, I, measured by the galvanometer, is < E being the electro-motive force of the measuring battery, and e that of the polarisation of the plates with the reverse sign. The truth of this employment of the law of Ohm may be proved by varying the two members of the denomi- nator : R, by varying the length of wire, and r by varying the distance between the electrodes. Whereas, when the fluid conductor is removed, and a metallic resistance (=>) of the same value introduced in its stead, the deflection indi- cates another current, say I', and we have " To find now the value of e, we put so much extra resistance, r, into circuit Number 1, that the needle's deflection is sensibly diminished, indicating a current of some inferior strength, I L , expressed by I = E "" * (II R + r + r' Equations (I. and (II. combined give the value of the difference, E e, between the two opposite electro-motive forces, by an expression from which the unknown resistances, II and r, of the wire and fluid columns are eliminated : - E-^r ..... (III. SCIENCE AND PRACTICE. 279 To arrive at e y we must eliminate E, in order to do which two other observations are necessary. First, the constant cell, E, is connected in the circuit of the galvanometer, and a resistance, =W, is added by degrees until the needle indicates again the intensity I, whence -*' and still more resistance, p, until the intensity is further reduced to I,, the same as in equation (II., by which From these two equations the value of E is obtained n - which, subtracted from (III., leaves the value of e The polarisation is, therefore, equal to the product of the two indications of the galvanometer divided by their dif- ference, and multiplied by the difference of the resistances added to the circuits, to reduce the deflections each time from I to I P Lenz and Saweljev found the polarisation of plates of different metals in different fluids to be as follows : 1. Platinum plates in diluted sulphuric acid (6 parts acid to 100 parts water) 1185 2. Platinum plates in nitric acid 538 3. Copper plates in sulphuric acid 466 4. Zinc plates in sulphuric acid 315 5. Graphite plates in nitric acid 273 6. Amalgamated zinc plates in sulphuric acid 217 7. Iron plates in sulphuric acid 72 When, however, the fluids are of such a nature that no gas is formed on either of them, that is to say, when the gases are recombined in the moment of their formation, little or no polarisation is observed. 280 THE ELECTRIC TELEGRAPH. The following three determinations in the same unit will illustrate this : 1 . Copper plates in sulphate copper solution 15 2. Amalgamated zinc plates in nitric acid 6 3. Copper plates in nitric acid 2 The electro-motive forces observed in these instances were probably due to differences between the metal plates them- selves, and not to polarisation by gas. In the same unit in which the above values are expressed, the electro-motive force of an ordinary DanielFs element is only 470, or less than that due to "polarisation by gas of the platinum plates of an ordinary voltameter. It can be no matter of wonder, therefore, that an element with an electro- motive force so small as that of Daniell's should be found incompetent to effect the decomposition of large volumes of water, and that for this purpose we are obliged to employ elements of greater force, such as Grove's or Bunsen's. "With the continuation of the current the polarisation in- creases, and its amount depends, within a certain limit, upon the strength of the decomposing battery ; but it attains a maximum. Increase of temperature of the fluid which is decomposed is followed by a decrease of the polarisation. 38. Insulating Substances. The conducting powers of in- sulating materials have been determined by various observers qualitatively. The only quantitative measurements to be relied upon are those made with the telegraph cables where a great and uniform surface is at the command of the experi- menter. Necessarily, therefore, the information which these determinations afford us comprehends only a limited number of materials. At a temperature of 72 F. the conducting power of gutta-percha has been found to be about forty-five times that of india-rubber, while at 92 F. the relation between their conducting powers is almost double this. To calculate the conducting powers of insulating materials from their resistances as dielectrics of cables, we must sup- pose the propagation of the electric current from the central conductor through the insulating covering to take place in SCIENCE AND PRACTICE. 81 concentric cylinders. It is evident then, from what has gone before, that the resistance through such a cylinder concentric with the conductor will be directly proportional to its thick- ness and inversely proportional to its surface, that is to say, of its length and circumference, and to the conducting power of the material. If we have a metallic conductor insulated with a material whose conducting power is c, the diameter of the conductor, 2 r, and the outer diameter of the insulating covering, 2 R, the resistance, d W, of a differential cylinder, whose thickness is dx, diameter 2#, and length /, will be dx and by integration between the limits of x = r and x = R, the sum of all the differential cylinders which make up the space occupied by the insulator, or, in other words, the resistance of insulation will be ../ dx ' r (I. W=l Zxirlc ~2 r whence the conducting power, c, is R : 2*1 W I3y measuring the value of W by any of the known methods of determinating great resistances (which will be treated of further on), and being in possession of the dimensions of the cable, we can calculate the conducting power. Having another cable, insulated with a different material whose conducting power is c', length /', resistance W, and R' ratio of diameters , the conducting power of these two cables will obviously stand in the relation log-. log. e ;r 282 THE ELECTRIC TELEGRAPH. Or, if the two cables have the same length, these conducting powers will be as E E' tan -7 l S- e TT -ir- -IT- E E' And if, further, the relation of their diameters be - = c : c = W : W, inversely, therefore, as the resistances. With the aid of (I., it is easy to calculate the insulation resistance of a wire covered with any insulating material whose conducting power is known in comparison with that of some other material ; or, when the material is the same, as, for instance, when gutta-percha is used to insulate both cables, for any unit of length (a knot, for example), the resistances will be as -W : W = log. nat. -?- : log. nat.-^- T> log. nat. W=W LT . : . . . (III. log. nat. 39. Variation of the Conducting Power of Gutta-Percha with Temperature. The per-centage variations of the conducting powers of insulating materials are considerably more than those of the metals. The conducting power of gutta-percha, for example, at a temperature of 20 C., is about twelve times as great as at the freezing point of water. The conditions under which the cores of submarine cables had been tested were too uncertain, until the date of the Persian Gulf cable, to justify any dependence upon the quantitative results obtained in measuring under various temperatures. A tolerable idea of the curve which the con- ducting power of gutta-percha made when, in a graphic representation of the measurements, the temperatures were taken as abscissae and the conducting powers as ordinates, SCIENCE AND PRACTICE. was arrived at by us from tests of the Malta- Alexandria cable during the process of sheathing. The difficulty of ascertaining accurately the length of the cable and tem- perature of its interior at any moment, however, precluded the possibility of any mathematical expression with con- fidence. The first, and so far as we believe, the only good results hitherto obtained, are those published by Sir Charles Bright in a paper read recently before the Institution of Civil Engineers. Experiments were made by Messrs. Bright and Clark upon four coils of the insulated core destined for the Persian Grulf cable. Each coil had a length of one nautical mile. They were placed in a felted iron tank holding about 1,200 gallons. At starting, the coils were maintained for three days in water kept in motion, containing a large quantity of melting ice, and may therefore, at the end of that time, when their resistances were measured, be presumed to have taken throughout the temperature of the water. After the first measurements were made, the water was allowed to increase in temperature very gradually up to 38 C., the gutta-percha resistance being measured at regular intervals. These experiments occupied thirty-three days, during which time nineteen series of observations were made, the mean results of which, reducing the observed resistances at C. uniformly to 100, are as follows : Temperature. Resistance. Temperature. Resistance. C. 100-00 20 C. 8-45 2 84-14 22 6-82 4 64-66 24 5-51 6 47-65 26 4-47 8 37-15 28 3-51 10 28-97 30 2-99 12 23-18 32 2-48 14 16-89 34 1-92 16 14-37 36 1-68 18 11-05 38 1-43 284 THE ELECTRIC TELEGRAPH. From these results Messrs. Bright and Clark find that the curve between the resistance and temperature is a logarithmic one, and have arrived at the empirical formula, E^ = R, (0,8944) <1- ; expressing the resistance R^ , at a temperature ^, as equal to the product of the resistance R^ at some lower tempera- ture, t, and the (^ t) power of a constant hase, 0,8944. That this empirical formula is only an approximation to Tumperatures. Fig. 132. the true expression is obvious from the curves shown graphically in Fig. 132, the black line representing the values obtained from the foregoing table, and the dotted line those given by the above formula. The agreement is never- theless sufficient to enable us to calculate, by means of the given formula, the resistance of a gutta-percha covered wire for any temperature between 0C. and 38 0., with very slight error, from the resistance at any given temperature. Hence it becomes needless to test the cores of cables at the gutta- percha works at an uniform temperature of 24 C., as has been customary with cables hitherto made, since the resistances may be reduced to a standard temperature, and thus save the expense and trouble of keeping up warming tanks. The difference of temperature, in combination with the SCIENCE AND PRACTICE. 285 influence of pressure, is strikingly observed in submerging a cable. As the cable reaches the bottom the copper resistance becomes gradually less, showing a lower temperature ; the insulation resistance increases at the same time, owing to the pressure, to a marked extent, and afterwards further increases as the gutta-percha becomes electrically sensible of its altered temperature. IV. METHODS OF MEASUREMENT. 40. Kirchhoff's Laws. Hitherto we have regarded the current as traversing simple closed circuits. Problems often occur in practice in which it is necessary to consider the circuit as made up of several parallel branches, or shunt circuits. The question, for example, whether a single battery could be used for telegraphing at the same time to Bristol, to Hull, and to Paris, belongs to this branch of the subject, as do also the mathematical solutions of the Wheatstone's bridge, PoggendorfFs, and other methods indispensable in electrical measurements. Kirchhoff * has provided for the solution of such questions two propositions, which he has proved mathematically and experimentally. The first is, that " The sum of the intensities in all those wires ivhich meet in a point is equal to nothing." o (Fig. 133) is the point in which seven wires meet ; the currents i lt I 2 , and i 3 , ap- proach, and ?!, 2 , %, and i, recede from it. If we give the plus sign to those currents which ap- proach and the negative to those which recede from the point, the sum of all the intensities is, * Fogg. Ann. 64, p. 513. 286 THE ELECTRIC TELEGRAPH. In other words, the sum of those currents which approach the point is equal to the sum of those which recede from it. The truth of this is evident at the first glance ; for, other- wise, the point must be a reservoir, which is contrary to all our notions of electricity. The second proposition is that " The sum of all the products of the intensities and resistances in all the wires which form an enclosed figure is equal to the sum of all the electro-motive forces in the same circuit." A circuit by which this law is illustrated is shown in plan in Fig. 134. E is a galvanic battery, whose circuit divides itself, in the points a and b, into the parallel ways r and r 2 respectively. Let the intensities in the three sections of the conductor be I, i lf and i 2 ; according to the law just expressed, the sum Fig. 134. of the product of the intensity of the current in each of the branch circuits between a and b, multiplied by its resistance, will equal nothing, since no electro-motive force is found in this cir- cuit, or, in *'**= . . . (1. whence, that is, the currents in these circuits are inversely proportional to the resistances. Further, by the same law, I R + I K + = E and by the first proposition, (2. (3. By knowing the electro-motive force, E, and the three - SCIENCE AND PRACTICE. 287 resistances, we are now in a position to find the value of the current in any part of the circuit. Substituting in (4. the value of . E-IE ' 1= ~^~~ ; and of . E IE we have E-EI E-EI , 2 "Whence the current in the lower circuit, between a and b, is I = E _ r i + r s (5 ' K which, inserted in the equations (2. and (3., gives us the currents in the branch circuits : and If, further, we desire to find the resistance E,' of the whole circuit, we must insert the value of I, (5. in the fundamental p formula, I = , from which E' = E + r \ r * - .... (8. r +r 2 Of this, E, is the resistance of the undivided circuit between a and b, and the fraction - 1 expresses the resistance of r l +r 2 the shunt circuit between the same points ; therefore, the resistance of two parallel conductors is equal to the product divided by the sum of their resistances. This must be true for every value of r 2 between and oo ; 288 THE ELECTRIC TELEGRAPH. so that, if we take away the shunt, the resistance r 2 becomes infinite ; and giving this value to it in the above equation, which may be written also R' =- R + - !J _ the whole circuit resistance becomes, R' = R + r r The circuits and resistances of three or more parallel circuits are calculated in the same way. It happens sometimes that several lines leave the same station, and the question has been raised under what con- ditions a single battery suffices to work them all at the same time. Let the resistances of three lines, for instance, be r lt r 2 , and 7*3, (Fig. 135), and the inten- sities in them i l9 i 2 > aD ^ 4> respec- tively, when the current of a battery, w h se electro-motive force is E and resistance R, passes parallel through "* them. We must now find the values of these intensities, and com- pare them with those which would Fig. 135. b e obtained if only one of the lines were inserted at a time. By KirchofFs second law, IR + *>, = , I R + t, r 2 = E, I B + t , r a E whence the intensities in the three lines are, E- IR E - IR to === and E IR SCIENCE AND PRACTICE. 289 for which expressions, if we can consider B, = 0, we obtain the same values for the intensities in the three branches, which would be due were only one branch inserted at a time, or, , and E ' a = ~, Therefore, we can accept as a law that when the resistance (E) of the battery is inappreciably small in comparison with that of the lines, or other circuits, the current in each of the latter is of the same intensity as it would be were the battery in circuit with that line alone, and that when the resistances of the lines are equal, the currents circulating in them will be equal also. It more seldom happens, however, that the resistances of several lines are equal than that the currents traversing them from a common battery are required to be so. In the latter case it is necessary to distribute the battery in such a way as to make the intensity in each branch as nearly the samo as possible. To find how this distribution is to be made, let us suppose the three lines t\, r 2 , and r 3 , (Fig. 136), in which we have to put Fi &* 136> three parts, E 1? E 2 , and E 3 , of the whole battery ; E x being common to all the lines, E 2 common to r 2 and r& and E 3 serving r 3 only. Retaining the same designations, we have the fol- lowing equations for the three lines : _ I, E, + t, r, = E' T 1 E 1 + I 2 B 2 + ; 2 r 2 = E 8 + E r I, E, + I, E 2 + 7 3 (R 8 + r s ) = E 3 + E 2 + E, in which, when the resistances of the lines are considerable 290 THE ELECTRIC TELEGRAPH. in proportion to the resistances of the elements, we can set RJ = R 2 = R 3 = 0, and the expressions become, E, Dividing the two last equations by the first, E, whence E 2 : E, = r 2 - r, : r, and E 3 : E, = r 3 r, : r, proportions which imply that the number of elements E x in the circuit of smallest resistance being given, the numbers E 2 , E 3 , &c., to be added to each of the other circuits must bear the same proportion to E L which the difference between their resistances bears to the smallest resistance. More generally if the resistances of the lines were equal, the single battery would produce a like current in each of them ; but if not, the battery power which must be added to the lines of superior resistances to produce the same current as in the smallest line, must be exactly proportioned to the superiority of their resistances. 41. WTieatstone' s Balance. On the same laws depend the mathematical proof of the truth of the beautiful and useful system of resistance measurement invented by Professor Wheatstone, a description of which appeared in the Philo- sophical Transactions of 1843. The poles of a battery, E, Fig. 137, are connected to the points of union c and d of parallel circuits, r lt r 3 , and r 2 , r 4 , and between some points, a and b, in these two conductors a wire, r 5 , is inserted. The current takes the course indicated by the arrows, and several complete circuits are formed, for SCIENCE AND PRACTICE. 291 which Kirchhoff's laws provide expressions. I, i i9 3 , i& 4 , and i 5 are the currents, &, r lt r 5 the resist- ances in the several cir- cuits, and E the electro- motive force of the bat- tery. For the currents ap- proaching and receding from the points a and b, and 1) - *. * 8 'i = 2) , * 4 ''. for the circuit r 3 , r 4 , and r 5 , 3) '*. v* + V5- and, lastly, for the two parallel circuits, The principle of Wheatstone's balance is based upon the relation which must exist between the resistance r lt r 2 , r s , and r 4 , when the position of the points a and b is so arranged that their electrical tensions are the same, or that no cur- rent passes between them when they are joined by a con- ductor. Supposing this to be the case, the value of fl' 5 in the above equation becomes = 0, and from 1, 2, and 3 we obtain the values of the intensities in three of the sides, in terms of the remaining one, '2 '4 These values we set in 4, and have ^,,+,,^3-^-^=0, and, dividing each side by ?' 4 and clearing away the fractions, obtain the expression, 292 THE ELECTRIC TELEGRAPH. , r 2 r 3 = 0. r, r a This is, therefore, the relation which must be established between the four sides r l9 r 2 , r 3 , and r 4 before the condition can be fulfilled that the intensity of the current in the bridge a b is null. The way in which this system is applied for the measure- ment of resistances will be explained directly. Before we come to that, however, there is a case worth considering which often occurs in testing cables or conductors whose ends have different temperatures that of a foreign electro- motive force in one of the branches of the system. With cables, this arises mostly from earth currents or from electro- motive force between the earth plates ; in other instances, from thermo-currents. Let the four resistances A, B, C, and D (Fig. 138) be adjusted so that the current in the balance a b is inappre- ciable, while the inten- sities of the currents due to the battery E, inserted between the points c and d in the several branches, are increased or dimi- nished by the current of the element, E t , set up in Fig. 138. the side C. If the inten- sities of the currents in the circuits under these conditions are a, |3, 7, , and c, and R the resistance of the battery between c and d, a y = ft I = y 3 = Aa 3(3=0 C y D ft = E' 2) 3) 4) 5) 6) By eliminating (3, S, and c with the aid of 1), 2), 3), and 4), SCIENCE AND PRACTICE. 293 we obtain the value of the intensity, y, in the side c, con- taining the strange element, in terms of both the electro- motive forces E' and E. =".-*- E' - C _AD *y ~ B E A (D + K) + B (A + K) And from these, the relation between the two electro-motive forces h JT = BC AD ~ E A (D -f E) + B (A + R) which supplies a method of comparing the electro-motive forces of two batteries. The value of the resistance C, when TSf the relation -f- is known, is found by the formula AJD _j_ E^ A (D + R) + B ( A + E) = ~B~~ " E B or, when the resistances A, B, C, and D are very great in pro- portion to R (the battery resistance), the latter may be neg- lected, and C becomes C:rz AD_ E; A (D + B) Lastly, if we suppose E' ^= 0, or that there is no electro- motive force in the side C, the member JS' A (D + R) + B (A + R) Ei. 4ft of the above falls away, and we get the common expression of relations of Wheatstone's balance B Four corollaries from the two laws developed by Kirchhoif have been published by Bosscha, which in many calculations 294 THE ELECTRIC TELEGRAPH. of galvanic circuits are found of great value, frequently saving time in developments. These corollaries are : 1. If, in any system of circuits, containing any electro- motive forces, a conductor exists in which the current = 0, the currents in the remaining circuits are not altered in the least degree if the circuit of the conductor in question is divided, or it is removed, together with whatever electro- motive force it may contain, from the system. 2. If the conductor in question contains no electro-motive force, the currents will not be altered, if, after its removal the points between which it previously existed are connected directly with each other. If, on the other hand, it contain an electro-motive force, the points can only be joined again by inserting between them an equivalent electro-motive force. 3. In a system of linear conductors containing electro- motive forces, the current set up in any conductor, a, by an electro-motive force contained in any other conductor, b, will be identically the same as that which would be set up in b by an equal electro-motive force in a. 4. If in a svstem of linear conductors there are two of / them, a and >, in which the electro-motive force in a occasions no current in b, whatever current may be circulating in b will not be altered if a is divided or removed, nor will the current in a be altered if b is divided or removed, however the electro- motive forces in the remaining circuits may be arranged. 42. Siemens's Apparatus for Testing Cab/es. The testing apparatus, as at present used in the testing- room of Messrs. Siemens, is a modification of their old plan. It was rearranged and endowed with its present form in order to render it available, not only for measurement by the bridge method, but also for measurements of great resistance by deflection of the galvanometer needle, of sensibility of the instrument, of the electro-motive force of the battery, and of charge and discharge of a cable. For the construction of such an apparatus, the following parts are necessary : An adjustable resistance coil of 1 to SCIENCE AND PRACTICE. 295 10,000 units ; two branch or proportion resistance coils, of each 10, 100, and 1000 units ; a coil of 10,000 units ; a galva- nometer, two four-sided commutators, a battery commutator, a contact key, a constant element (Daniell's), a battery of n elements, and the necessary terminal screws for connecting earth, galvanometer, and cables. The testing-board, fitted up with the more portable of these articles, is shown in plan, in Fig. 139, with the connections between the various members in dotted lines. Fig. 139. These connections are made of thick copper wire, well in- sulated with gutta-percha; they are led from one point to another, underneath, and their directions indicated by strips of ebony or other wood, let into the top of the board, and polished with it. In the laboratory, however, the operator does better to make these connections in the air, because, in the event of a leakage anywhere, he has them more immediately under his inspection, and can eradicate the disturbing cause with greater ease than if the wires were fixtures. Fig. 140 shows a general plan of the same, arranged in a more theoretical way. The resistance r (= 10,000 units) 296 THE ELECTRIC TELEGRAPH. is connected in the same circuit as the upper branch, r, of the proportion resistances r and p, by which the resistances in these branches may be r = 10,100, 1000, or 10000, in the upper, and ftZ p = 10, 100, or 1000, in the lower side. The upper circuit includes, also, a triangular commutator, U 3 , for com- pleting this circuit, with or without the inclusion of a Daniell's element. The triangular commutator consists of three brass slabs i, 11, and HI. The Daniell's cell is connected between n and in, while i and n are in circuit of the bridge. Be- tween i ii and i in are holes Fig. 140. for a contact plug, which, when put in the hole i IT, completes the circuit without the element ; and when in i m, introduces the element into the circuit. The point of contact, c, of the upper branch, r, with the lower, p, is connected with the beam of a key, k, the back or reposing contact of which is to earth ; the front or working contact to one corner of the four- sided current director, Uj. The ends of the galvanometer coil are not connected imme- diately to the corners of the parallelogram, but are inter- cepted by a current-director, u 2 , for enabling the operator to observe the deflection of the needle on either side of the Zero line. Three metal slabs, L, L L , and E, with holes between them, for the insertion of contact plugs, form the upper right-hand side of the bridge. The earth-plate is connected by a wire to E ; L X is connected with the upper corner of the bridge ; L with the conductor of the cable, at one end, and should both ends be at the disposition of the operator, he connects the other end to E, when he wishes to measure the copper resistance. The purpose of these three slabs is SCIENCE AND PRACTICE. 297 for removing the cable end from the bridge for a certain time, and being able to replace it for the observation of charge and discharge currents. The bar terminals T allow either one-third, two-thirds, or the whole of the battery to be brought into circuit. It remains now to arrange the board for various cable measurements. 43. Copper Resistance, when both ends of the Cable are on the Board. One end of the cable to be measured is attached to L, the other end to E, and a plug is put in the hole L L X ; a plug is also put into the hole z l of the battery commutator T, by which one-third of the battery is put into circuit. The plug in hole I n of the tri- angular commutator, U 3 , completes the upper branch, r r lt without the constant element. The plugs of t^ are put into the holes, so that either the zinc or the copper current shall enter the cable, as may be required. In u 2 , also, two plugs are inserted. In measuring the resistance of a cable conductor, the lever of the key is kept down on its front contact the whole time, and, whenever the resistance, R, is altered, the plug between the slabs, G and G 1 (forming a short circuit across the galvanometer coils), is inserted to prevent the charge and discharge currents passing through and dis- turbing the needle of the galvanometer. The current diverges into two circuits from the point c, converging again in the point d into the main line, d, u l9 c. The upper circuit includes so much of the resistance, r r 1 , as is left unstoppered, the triangular commutator, u 3 , the point of junction, a, the terminals L 1 , L, cable, E, and d. The lower circuit is from c, through p, b, R, to c. Between c and d the circuit is made up of TJJ, T battery, u x , K, &c. Having obtained a balance of the galvanometer needle, or when the current in the circuit a, U 2 , G, U 2 , b is null, the pro- portion r x ~ 9 IT or, X--RL - 298 THE ELECTRIC TELEGRAPH. is established. The values which this proportion may have are, obviously 1. When r p r 10 100 1,000 may be = , , or p 10 100 1,000 2. When 10 , , or . p J 100 1,000' 10,000 The limits within which these proportions are confined are therefore, ^r = 1,000, and J-QQQ = 0,01 ; and, since the greatest value of the adjustable resistance, R, is 10,000 units, and the least value one unit, the limits of the measureable value of x with the bridge system so arranged, are Maximum : x = 10,000 X 1,000 = 10,000,000 Minimum : x = 1 X 0,001 =: 0,001. Whenever it is possible to do so, it is to be preferred to make the ratio = 1, and r as nearly equal to R as can be, by which the greatest sensibility of the system is obtained. 44. Copper Resistance, when only one end of the Cable is on the Board. The connections of the board remain the same, except that, instead of the second end of the cable being con- nected to E, it is connected to earth. The upper circuit is then from c through r, U 3 , a, LJ, L, cable and earth ; and the lower from c through p, b, R, d, and earth. In measuring the copper resistance of a cable with earth - plates in the circuit, it is necessary (1) to have the plates, both of the apparatus and cable-end, sufficiently large, and buried in moist ground, or sunk in water to avoid the introduction of resistance into the cable circuit ; and (2) to have both the plates of the same metal, and of the same quality of metal, at the same temperature, and, as nearly as possible, in water of SCIENCE AND PRACTICE. 299 the same kind, or a current will be set up between them which will make the observed resistance smaller or greater, according as the current due to the plates is in the same, or in the opposite direction to the current of the measuring battery. 45. Insulation Resistance by Bridge. As the resistance of insulation is generally very great, in comparison with that of T the conductor of a cable, the proportion has to be made as large as possible. The limit to which we may go, in this direction, we have seen is when r_ = 10,000 p 10 and, as the greatest value of R is 1 0,000, cable resistances under ten millions of units may be measured by this method. The whole force of the battery is introduced by inserting the contact plug in hole z 3 of T. One end of the cable is connected to L, the other insulated. The direction of the current is determined by the position of the two plugs in Ui. The circuits of the currents are obvious. When the cables are long, or badly insulated, this method of measuring the resistance of their dielectrics is the best ; but for short or well- insulated cables it would not be found sufficiently delicate, and the method by comparison of the deflections of the galvanometer needle is to be pre- ferred. 46. Insulation-resistance by Deflection of Galvanometer Needle. The bridge system is interrupted for this measure- ment, and the current of the whole battery allowed to go in a simple circuit through the galvanometer coils, and into the cable. The resistances r and R are made infinite, and o null, as in Fig. 142, by which the current of the battery goes from the key k, through c, p, U 2 , coil of galva- nometer, the other side of u 2 , a, L' L, cable, through the in- sulating material to earth, and from earth through u x , battery, u ly to key. During the whole time of measuring insulation resistance, the beam of the key is depressed upon the work- ing contact. The plugs of u l are placed so as to put the 300 THE ELECTRIC TELEGRAPH. copper or zinc pole of the battery to earth at pleasure, and those of U 2 so as to determine to which side of the zero line of the galvanometer the needle shall be deflected. The galvanometers used by Messrs. Siemens with these bridges are either Dubois' sine multipliers or Weber's reflecting galvanometers. Professor Thompson's reflecting galvanometer is well adapted for such work ; better, perhaps, than either of the others, as its sensibility is greater than that of the sine instrument, and it is less unwieldy than Weber's. With a very sensitive galvanometer, and a given battery power, it is sometimes found that the needle is deflected beyond the range of observation. In such an event the operator inserts a shunt by completing the circuit of the upper branch r r , making it, instead of in- finite, equal to some resistance . which will take as much of the T~t current from the galvanometer as is necessary. The current goes then (Fig. 141) from the key to c ; here it is split into two parts ; one goes through p, b, L T 2 , and galvanometer to a, and the other through r to a. At a these parts combine again, and the whole current goes over L L into the cable. The resistances r r' consist of four coils of 10 1 , 10 2 , 10 3 , and 10 4 units, which can be introduced separately or combined, and take off the -- part of the current from 9+r the galvanometer (g being the galvanometer resistance). With very rare exceptions these powers of 10 will be found sufficiently comprehensive for supplying the required shunt, and the function of the deflection of the needle will have Fig. 141. to be multiplied with in order to arrive at the value SCIENCE AND PRACTICE. oOl which would be obtained were no shunt used and the range of the instrument wider. 47. Charge. This is a method which has two objects : the first is to ascertain the inductive capacity of the cable, and to obtain data for judging of the concentricity of the conductor in the insulating medium ; the second is by observ- ing the loss of static charge by recombination of the elec- tricities through the dielectric, in a given time, to conclude upon the degree of its insulation. The connections of the board for measuring charge are precisely the same as those for insulation by deflection. When the cables are long, it is requisite to employ the shunt resistance r r in order to keep the needle within readable bounds. The plug y g is left out when the throw of the needle is to be observed, and the key pressed down. At the moment of completing the circuit the electro-static charge passes through the galvanometer and enters the cable ; the needle is impelled from its position of rest with a sudden jerk to one side, and afterwards continues to oscillate over the zero point until it comes again to rest. The first throw or swing is noted. If the discharge is also to be observed, in order to compare it with the charge, before letting go the key the operator removes the plug from between L' and L, by which the cable end is insulated from the board. 48. Discharge. As the coils of the galvanometer are seldom entirely free from magnetism, and the needle seldom so exactly centered that the same strength of current gives an equal deflection on each side of the zero line, it is pre- ferable to obtain the first swing of the needle due to the discharge current on the same side as that which was observed for charge. To this end it is necessary only to alter the position of the contact plugs of the galvanometer commu- tator u 2 before the discharge current passes through, by which the galvanometer coils are reversed in relation to the points a and b ; and, as the discharge is in the contrary direction to the charge, the deflection of the needle will be, for both, on the same side of zero. After the cable has been left insulated at both ends during 302 THE ELECTRIC TELEGRAPH. one minute, or whatever time may be fixed upon since the removal of the plug L L', and the plugs of U 2 rearranged, the plug L L' is suddenly replaced, and the throw of the needle observed. If the time which has elapsed is only a few seconds, and the cable is well insulated, the deflection of the discharge current is almost equal to that due to the charge ; but when the interval is long, or the cable indifferently in- sulated, the greater part of the charge is recombined through the dielectric. The methods of comparing these two indications was sug- gested by Dr. Siemens, and may be considered one of the most important test methods in cable work. Dr. Siemens assumes the strength of the instantaneous current which produces the swing to be proportional to the sine of half the angle of deflection, or * = C sin. J C being a constant of sensibility of the instrument, and i the current producing the deflection . For discharging, there- fore, *\ being the returning current and \^ the swing which it produces, we have '. = C sin - J- The difference of the two : i = i lt expresses the loss during the interval between the two observations, or that portion which has recombined through the insulating coating. The charge being taken as unit, the loss, L, is, therefore, (h ' \L \L sin. H __ sin. I sm. JL. i t , 2 2 , 2 sm. J-. sin. L 2 2 Observations of two following charge currents may also be taken as data for calculating the loss, instead of those of a charge and a discharge. The plug L L' is removed from its place and the key held down ; the plug is then replaced for an instant, and again removed. On completing the circuit SCIENCE AND PRACTICE. 303 the needle is observed to swing through an angle of < degrees. After a lapse of one minute, or other given interval, the stopper L L' is replaced again, and the swing ^J due to this second charge current observed. The quantity of electricity which is measured by the second observation is obviously not that which remains in, but that which has escaped from the cable ; in other words, we make good the loss which the charge has sustained in the time between the observations, and this loss, L, is in terms of the whole charge, 0, sin. ' 2 L = "1 5 sin. -L- 49. Constant of Sensibility of the Galvanometer. Galvano- meters with single magnets do not vary their constants of sensibility considerably unless it is very soon after the mag- netising of the needle, but those with astatic systems are very inconstant, altering sometimes during an observation. For all calculations from measurements by deflection the constant of sensibility must be known. This constant is the deflection produced when an unit of electro -motive force is in circuit with an unit of resistance. The unit of resistance used in measuring insulation of cables is one million, or 10 6 times the small unit used in measuring metallic resistances. This small unit is the resistance of a column of mercury a meter long and a square millimeter transverse section ; the great multiple of it is not called an unit, but simply a " million," and the insulation of a cable is expressed as having so-and-so many millions, meaning so- and-so many million times the little column of mercury. A million metres of mercury, or its equivalent resistance in any other metal, would be very difficult to employ, and we are happily prevented the necessity of employing it to obtain the same deflection of sensibility by using a shunt and a smaller resistance. In the circuit represented in Fig. 142 the current of the 304 THE ELECTRIC TELEGRAPH. element E has to pass through a resistance, r 1 , and through the parallel circuit g (galvanometer) and R (shunt). The whole resistance * R s , ,, , &g The intensity, I, of the current in r is the sum of the intensity i in the shunt, and that indicated by the deflection a . Let the latter intensity be F (a), then and the current F (o) in the galvanometer is Now, with a given value of E, p, and b } 2) c a G b, and 3) c K R b. u x is stoppered so that the currents of the battery and constant cell oppose each other in the gal- vanometer circuit ; the key is kept down ; of the resistance scales, r = 0, x Fig. 144. = 10, or 100, 306 THE ELECTRIC TELEGRAPH. or 1,000, according to the strength of the battery, and II is varied until the needle of the galvanometer rests upon the zero-line. When the balance is obtained with, say p = 100, we alter this side to 110 and get another reading for R^rRi. p The relation of the electro-motive forces x is expressed by the equation E __ (E, R) + (110 - 100) E' ~ (110 100) as will be explained afterwards. These are the principal applications of the beautifully arranged testing apparatus invented by Messrs. Siemens. With some trifling modifications it is equally applicable for measurements by other methods. 51. British Association Bridge. An ingenious electrical balance has been arranged by the sub-committee appointed by the British Association in 1861. The purpose of this balance is for copying standard resistances with great exactness. Instead of employing proportion resistances of some unalterable value, the ends of the two branches A and c, Fig. 145, enclose a wire, w x, of sensible resistance, contact being made with it by means of a travelling point, u. Ac- cording as u is moved to the one side or the other, therefore, re- sistance is added to one and subtracted from the other branch; and the resistance of w x being small, a balance of great exact- ness may be obtained between the .proportion resistances A and c, which is otherwise liable to be temporarily deranged by inequality in the temperature of these coils. Fig. 146 shows a special plan of the board. The two branch coils A and c, of equal, or nearly equal, resistances, are SCIENCE AND PRACTICE. 307 wound upon a wooden reel, c A ; their ends terminate in thick copper connections, which dip into the mercury cups a a 1} c and c v The standard resistance s is wound upon a similar reel, and its ends connected to the amalgamated copper wires, which terminate in mercury cups s s t . n is the resistance whose length is to be adjusted. D is a commutator consisting of two parallel arms of copper connecting the mercury cups d with d lt and /with/u or d with/ and d t with/. E is a gra- ,V>f -<^v| ~o o-. Fig. U6. duated scale, underneath which the adjusting- wire w x is stretched, and with which the sliding brass piece H is in contact. The contact-key is of original construction ; its duties are to close the battery circuit first and then the galvano- meter circuit. It is made of three brass springs, 1, 2, and 3, Fig. 147, each insulated from the other at K, and connected by screws with the opposite corners of the bridge ; 1 and 2 x2 308 THE ELECTRIC TELEGRAPH. being in the battery- circuit, 3 at a bottom contact-point, 4 in the circuit of the galvanometer. T is an ebonite button, on which the finger is placed to depress it, and Q a piece of ebonite intended to prevent 2 and 3 making contact with each other, and to push 3 down upon 4. The resistance- wire R being approximately adjusted is placed opposite to the standard in the bridge, and the point Fig. 147. u moved along the wire w x until, on depressing the key, the galvanometer indicates no current. The commutator D is then reversed, by which R and s exchange places, and the contact- point u moved again until the balance is obtained as before. If the balance is obtained without moving u, it is evident that R and s are equal to each other, and that the resistances A -|- x u and c -|- w u are also equal, indicating, at the same time, a small inequality between the values of A and c. If the balance is only obtained by moving u, the 'direction in which this movement takes place shows whether the wire R is too long or too short, and if R is a known length of the same wire as w x, the distance through which the contact is moved gives a measure of the excess or deficiency of the length of R. When the length of R has been adjusted till the balance is attained,' the proportion coils, A and c, are removed and replaced by others, A' and C', whose resistances are ten times as great. A second adjustment is made with these coils, which are afterwards substituted by two others of still greater resistance ; and in this way any required degree of accuracy may be attained. The connections (shown by dotted lines) between the mercury-cups, &c., are made with stout copper wires. The SCIENCE AND PRACTICE. 309 bridge may be used for other than copying purposes, by in- serting proportion resistances between a a and c c , the com- parison resistance or set of adjustable coils between s and s' and the resistance to be measured between r and r. As the resistance of the wire, w x, might be embarrassing in the general employment of the bridge, it may be short-circuited by a wire between the mercury cups, e and e. 52. Balance formed by a Bisected Wire. Balances, the two proportion resistances of which are made variable by moving the point of bisection of a wire, are the most delicate and best adapted for measuring very small resistances ; they are, of course, of more use in the physical laboratory than in the testing-room. Fig. 148 represents a perspective view of the wire bridge constructed by Dr. Siemens for use in his laboratory at Berlin, and with which the mercury unit of resistance was determined, and most of the elegant experiments made by that able physicist carried out. Upon three thick slabs of vulcanite, a a, a a', and a" a t which rest upon a table, is supported the brass guide, A A, on the top of which is a rack, D D, whose teeth engage with those of a horizontal pinion underneath the carriage B B, with which it travels when the milled head c is turned to the right or left. The same slabs of vulcanite support a brass scale, m m, a metre long, graduated in millimetres ; in front of the scale is stretched, between insulated metal clamps, a fine platinum wire, w w, of exactly a metre in length between the clamps, passing between two platinum contact rollers, G, carried by B B. The clamps, between which the ends of the wire are held, are in metallic connection with the bolts and couplings, E E, and from these through thick copper bars to the opposite corners, 1 and 2, of a commutator, s. The other corners, 3 and 4, of s are connected by similar bars with the clamps K K in front. In the clamps K K are sliding connecting rods, L L, between which and the front contact H of a key, J, the standard re- sistance w, and the resistance x, which is to be measured, are connected* 310 THE ELECTRIC TELEGRAPH. The wire, bisected by the contact rollers G, therefore forms two sides of the bridge, and between its ends the galvano- SCIENCE AND PRACTICE. 311 meter N is inserted. A Daniell's element is inserted between the lever of the key and the brass stage A. The galvanometer used with this bridge is one of Weber's construction, with a polished steel mirror. With a single Daniell's element, when the currents of the system are balanced, if the contact rollers be moved 0,1 millimetre out of their place, the galvanometer mirror shows a deflection, represented by five divisions of the reflected scale passing before the fibre of the telescope. The resistance of the platinum wire is about twenty units. It is seldom procurable absolutely cylindrical, although drawn with great care through stone ; the conicalness, how- ever, rarely exceeds an amount which throws the resistance middle-point above 0,2 millimetre from the middle-point of the length. The resistance of passage from the wire to the clamps E E is a source of some little annoyance in using this really beautiful apparatus, and renders it inapplicable, with the same degree of accuracy, when the contact rollers are far from the middle. The position of the roller is read off by a nonims carried by the waggon, B, along the metre scale, m m. One reading only is necessary for giving the value of x : 1,000 a a being the length read off by means of the nonims from the index to one end of the bridge. A second reading is, however, usually made, by inverting the wire by means of the commutator s, and reading the length d with the rollers on the other side of the middle point of the wire, by which we get If no appreciable resistance exists in the junction of the wire with its clamps, and the electrical and geometrical middle points of the wire coincide, a + a = 1000 ; but as 312 THE ELECTRIC TELEGRAPH. this is rarely the case, the value of x is given with the nearest approximation to the truth by the formula x = W 1,000 +0 a' 1,000 + d a When double readings are made that is, when in each experiment the comparison and measured resistances are in- verted there is no need for having the bisected wire so rigidly in contact at both ends, provided the contact at the end from which the lengths are measured is without sensible resistance, and the resistance of the other end does not change during the double observation. On this principle we have constructed a bridge-balance of this kind, in which the bisected wire is soldered between good contact clamps at one end, and at the other is held by a metallic block running upon an adjusting screw, by which the wire can be strained or slackened at pleasure, the total length not appearing, the resistance x being given by Tir & X = W r a A method of increasing very materially the sensibility of the system was used by us in Dr. Siemens' laboratory in 1861. It consisted in inserting between the ends of the platinum wire and the points branching to the galvanometer and resistances W and x, an equal resistance coil, r, which amounts, in fact, simply to lengthening the bridge-wire. With this arrangement, the contact rollers can be moved' considerably farther from the middle point with less error. The resistance a? by a single reading, in this way, therefore, is x = W 1,000 a +r This method of testing with continuations was not prac- tised to any great extent, however, and was, we believe, never published until Dr. Matthiessen constructed a similar arrangement with his apparatus, in connection with the Unit SCIENCE AND PRACTICE. 313 Committee, which, probably suggested the idea of the ad- justing wire in the British Association balance. 53. Determination of the Constants of Galvanic Elements. We have already seen that, according to Ohm's law, the intensity of the current in any galvanic circuit is a function of the electro-motive force, and resistance in that circuit ; or that I- E K" R, being the sum of all the resistances, and E the sum of all i;he electromotive forces in the circuit ; and have considered the value of B, as the sum of resistances interior and exterior to the battery, by which the same value of I is expressed by r being the resistance due to the passage of the current from the platea of the elements to the fluids, and vice versa ; and r l that which is exterior to the element interposed resist- ance. Resistances we can compare directly, or calculate from given dimensions and conducting powers of materials ; in- tensities we can also compare directly with each other, or with some given amount of work done in the decomposition of water or salt solutions, or in the deflection of a magnetic needle, or in heat developed ; and electro-motive forces we can compare with each other by their known relations to these two combined. 54. Determination of the Resistances of Galvanic Elements. There are different ways in which this may be done. The readiest is with the aid of a tangent galvanometer. The element whose resistance is to be measured, is put alone in the circuit of the galvanometer, the deflection of whose needle is observed ; a resistance is then inserted in the cir- cuit, which lessens the intensity of the current and diminishes the deflection of the needle to $' degrees. The two inten- sities are C. tan. = 314 THE ELECTRIC TELEGRAPH. and C.tan.0' = + r + r" ^ of which x is the resistance of the element, r that of the galvanometer and connections, and r that added to reduce the deflection. By combining these two equations, C and E are eliminated, and the resistance x of the element obtained in the same measure as that in which r and r are expressed in _ r tan. (/ -f r") tan. 0' tan. 0' tan. If we use the galvanometer as a sine instrument, instead of reading off the angles of deflection < and $', the coils being turned through the angles a and a ', the formula becomes _ r' sin. a (r + r") sin. a' sin. a' sin. a When neither a sine nor tangent galvanometer, but only" a galvanoscope, the functions of whose needle- deflections are unknown, is at the command of the operator, he can deter- mine exactly the resistance of the element in the following way: The element to be measured is connected in the circuit of a rheostat or other wire, whose length is adjustable, and that of the coil of the galvanoscope. If the resistance of the galvanoscope coil be g, that of the rheostat wire r, and that of the element x 9 the needle being deflected, say a degrees from the magnetic meridian, the intensity, according to Ohm, is - E being the electro-motive force, as before, F the unknown function of the angle, and therefore F (a ) the intensity. A resistance equal to r + g being connected between the poles of the element, the current will be split into two equal parts, one part going through the galvanometer and rheostat, SCIENCE AND PRACTICE. 315 the other through the shunt resistance, which is equal to their sum. The consequence is that the intensity of the current in the galvanometer branch is decreased, while the intensity in the whole circuit is increased to 2 |3 being the new deflection due to this altered state of things. The shunt is taken away, the needle returns again to a and the current has its original intensity, F (a). The resistance r of the rheostat is then increased by r, until the needle descends from ct to j3, with the corresponding intensity. Dividing (I by (II, the value of x is obtained. 55. Determination of the Electro-motive Forces of Galvanic Elements. In calculating the insulation resistances, and other electrical conditions of submarine cables from the deflections of a needle, it is indispensably necessary to know the electro- motive force of the battery used in the measurement. The methods of comparing the electro-motive force of a battery with that of some constant element, taken as unit, are very various. Sometimes the electro-motive forces of elements are referred to the amount of work which, with an unit of resistance in the circuit, they are capable of performing in an unit of time ; as, for instance, the measurement by the voltameter. The intensity I of the current in any closed circuit is directly proportional to the volume of water which it decom- poses in a given time, and when the resistance R is constant, the volume decomposed is also proportional to the electro- motive force, for when the unit of resistance is determined upon, or R = 1 is set in the expression of Ohm's fundamental E equation, I = , we get the equation I = E, that is to say, 316 THE ELECTRIC TELEG11APH. if the whole resistance of the circuit be = 1, the electro- motive force will be equal to the intensity, which is also equal to the number of cubic centimeters of gas developed in a minute. The unit of resistance in this case has been taken as that of a prism of copper, 1 millimeter section, and 1 meter long. According to this method of determination, the electro- motive forces of the following elements are : Zinc Carbon element Deleuil =839 Ditto ditto Stohrer = 777 Zinc Platinum ditto Grove =829 Zinc Copper ditto Daniell = 470 Or, were the resistance of the circuit that of a copper prism of the above dimensions, each of these elements would decompose so many cubic centimetres of gas in one minute, as is indicated by the number set opposite to it in the list. 56. Another method of measuring the electro-motive force is by means of the unit adopted by Eegnault, namely, the electro-motive force of a thermo-electric pair of copper- bismuth wires, whose soldered ends are kept at the constant temperatures of C., and 100 C. To measure the electro- motive force of any element, Eegnault combines a number of these copper-bismuth elements together, until, when the currents of the two batteries are opposed, they exactly com- pensate each other. The number of thermo- elements required to do this is, therefore, the measure of the force of the element under inquiry ; and, as the force of each individual thermo- element is very small, the number brought to balance a galvanic element is great enough to allow of very nice adjustment, and would give very good results if the electro- motive forces of 'the individual thermo -elements were practi- cally equal, which unhappily is not the case. The electro-motive force of a DanielFs element, whose copper-plate is immersed in a concentrated solution of sul- phate of copper, and whose zinc is immersed in dilute sulphuric acid of the strength of 1 part, by weight, of acid, SCIEXCE AND PRACTICE. 317 to 4 parts of water, is now usually adopted as the unit of electro-motive force. 57. Fechner's method of comparing the electro-motive forces of two elements consists in measuring the intensities of the two currents, when the resistances are equal. He prefers to employ a galvanometer with a long thin wire, making many turns round the needle, besides a considerable resistance in order to be able to neglect the resistance of the element itself. 58. Another and preferable method is mentioned by Wiedemann. The two elements are connected up in the same circuit with a tangent galvanometer, or other apparatus for measuring intensity ; first, so that their currents go in the same direction, and secondly, in contrary directions. Let The electro-motive forces be =* E and E', the resistances of the elements = R and R', the interposed resistance = r, and the intensities of the sum and difference = I, and I tt then E 4- E' I. = R + R' + r I E ~ E ' = R + R' + r whence E' = E . * "" Id 59. Another method, requiring, like the last, two observa- tions for obtaining the necessary data for comparison, is that commonly resorted to when no galvanometer is at hand ; a galvanoscope will then fulfil all that is required, in con- junction with a sufficiently well arranged adjustable resist- ance scale or rheostat. One of the elements, E, whose resistance is r, is first con- nected in the circuit of the galvanoscope (resistance g) and of the adjustable resistance R. A deflection of, say degrees, is obtained and noted. The other element, E', whose resistance 318 THE ELECTRIC TELEGRAPH. is r, is then put into the circuit, in the place of E, and the resistance altered to B^, until the needle is again deflected < degrees. The intensity with the first element is E B + 9 + r and that when the second element is used, E' These expressions equalled give the value of E 1? the electro- motive force of the second element in terms of that of the first, ' R + r The resistances, E + 9 anc *- RI + 9i ma j be made so great in comparison with r and r lt that the latter may be neglected, and the calculation simplified by the disappear- ance of these magnitudes from the numerator and denomi- nator of the fraction, or E' = E. R + ^ It sometimes happens that a large electro-motive force has to be measured by a comparatively small one, as in measur- ing the battery used for testing the insulation resistances of telegraph cables by the force of a Daniell's cell. In this case it is better to use a shunt for obtaining the common deflection, degrees. If both the batteries are large, a shunt may be used in both measurements with advantage. The battery E (Fig. 149) has a re- sistance r units, the galvanometer g units, the shunt s units, and the in- terposed resistance E units ; we get a deflection of the needle through degrees. The battery is substituted by E 1 with r 1 resist- SCIENCE AND PRACTICE. 319 ance, the shunt being changed to s 1 till the needle is again deflected degrees, while H and g remain unaltered. The e f |ual intensities in the galvanometer circuits are and ( + O (i + -f) + g from which E' = E A large battery, E 1 , being compared with a small one, E, the shunt is only used in the case of the large one, and s = oo may be inserted in the above formula, by which p. 60. Professor Wheatstone's method is as follows : he connects the element, E, to be measured, in circuit with a galvanometer, while the whole resistance of the circuit is R. The resulting deflection is 0. He then adds r units to the circuit until the deflection is reduced to 0^ These deflec- tions being noted, he puts the constant element E x in the place of E, adjusts the new resistance Rj of the circuit until the needle is deflected degrees, and adds to this r units, until this deflection falls to X degrees as before. From these observations he has four expressions F (*)=-|- F (0) =-J 320 THE ELECTRIC TELEGRAPH. F (0), F (0!), E, and E x eliminated, E = E! r \ 61. Ohm's method, although it has rendered good service to the science in being the means of measurements of great value by Professor Poggendorff, is less to be recommended, as it depends upon the correctness of the functions of the galvanometer deflections ; whereas all those methods from which these functions are eliminated are to be preferred. By the method invented by Professor Ohm, a galvanometer and set of resistance coils are connected in circuit with the elements to be measured, and the intensities observed with two different values of the interposed resistance. E being the electro-motive force, 1^ the resistance of the element, r t and r u the two interposed resistances, and I ; and I y/ the observed intensities, *' = BT+T and l " = R + r whence f*. N T T E !-* Thus E, which is expressed in arbitrary units of resist- ance and intensity, may be compared with the E measured in the same way, of some constant or unit element, or I y and L may be expressed in cubic centimetres of water decomposed in a minute, and r and r in the units of copper prism mentioned before. 62. Compensation Method of Poggendorff. This is the most elegant of all the methods yet introduced. It has, in addi- tion to the advantage of comparing the electro-motive forces of two elements or batteries by a single observation, that of being independent of any detrimental polarisation. This is one of the few null-methods which are applicable in elec- tricity ; that is to say, one of those methods of measurement in which we balance the currents either of two batteries or of two circuits, so that we have a circuit whose electrical conditions are such that, if we insert a galvanometer, no traces of current can be perceived. SCIENCE AND PRACTICE. 321 Two batteries, E and Ej (Fig. 150), are connected parallel between the points b and c, so that their currents are opposed to each other, and between the same c points a conductor whose resistance is R! is inserted. Let the resistance of the circuit c E b be R, that of c E L b be E 2 , and let the intensities of these circuits be i, i lt and ?' 2 , as in the figure, then 1) . . i *, + f t =0 2) . . E i + E! f , = E 3) . . E 2 t, + E, , = E 4 Dividing 2) by 3), R* Fig. 150. For the determination of which it would be needful, in addi- tion to the values of R, R D and R 2 , to know the values of ^ i lt and i 2 . The same is done, however, by reducing the intensity in one of these circuits to nothing, by which the remaining two become equal to each other. By adjusting one of the resistances, R or R x , we arrive at a point where the intensity a in the circuit c E x b is reduced to nothing, by the currents of its proper battery Ej and of E being balanced. Where this is the case, t-, =; o and equation 1) becomes i i l =Q or i = i v These values set in (4, we obtain the relation of the two electro-motive forces E, E, from which we see that with this method it is impossible to compare the electro-motive forces of two batteries when they are equal, and R L a measurable quantity ; because if 322 THE ELECTRIC TELEGRAPH. E = E!, it follows, from the above equation, that R = R L + R, which could only occur if either R, the resistance of the element and circuit c E b, were nothing, which is impossible, although it may be made comparatively small ; or if the resistance R x were infinite, in which case - = 0, and the ixi E above expression would become -== 1, that which is equiva- lent to setting the batteries in a single circuit in opposite direction through the galvanometer. In practice, it is necessary to consider the resistance of the battery E, and the resistance inserted in the same circuit between the points c b, as separate magnitudes ; because it is not always that we can make the resistance exterior to the battery so great as to be able to neglect entirely the internal resistance without error. The circuit as arranged for comparing the electro-motive force of a battery with that of a single cell is as follows : The battery is inserted in the part circuit c E b (Fig. 151), with a coil of wire of known resistance, r ; in the f ex. opposite part-circuit, c E t b, a galvano- / \ scope, G, is inserted, with the con- -L El SJ R| _LE s tant cell E x ; and, from the points of junction, c b, of the two part-circuits, is the adjustable resistance scale R-j. If the resistance of the battery is a?, the resistance of c E b is -^ Fig. 151. and the currents being balanced, the relation of is E i JL - ' + * + R . E, R, The resistance of an element is generally known approxi- mately, and if the battery is composed of n similar elements, of which each has p units resistance, it is near enough to set j = np, in the above formula ; but it is better, where great SCIENCE AND PRACTICE. 323 exactness is desired, to make two observations with the same batteries, E and E', with other resistances, r x and E^, in order to eliminate x altogether. Thus, by the first measurement with R x and r, we have JE_ r + x + E,, i; = ~RT and by another measurement, with E' x and r lf _E_ r,+ x + B'i E, ~~ ~ R', which, being combined* and x eliminated, give _E_ (r-QCR.-R',) E, - (R, - R',) a formula in which differences of resistances only appear. The measurement may be made without knowing the values of either r or K^ ; as it is only necessary to note the difference, A> the addition to or subtraction from the side c E b, which compensates a difference, A', in the same sense, in the circuit c E, x 5, and we have Mr. Yarley has pointed out to us that which he considers a source of objection to this method. It is that the tension of the constant cell is different in the circuit in which its current is 0, to that which it would be if the circuit were closed without this condition, while we measure the force of the battery E, whose elements have only a tension due to an ordinary closed circuit. This objection does not, however, interfere with the correctness of the system ; for as the unit cell * These two measurements enable the operator to ascertain the resistance -p of the battery E. If, instead^'of eliminating x, we eliminate , and seek the value of x, we get _ Ri O'4-R'.Q R'l And by dividing this expression by n, the number of elements in the battery E, we get the average resistance of each of the elements composing it. Y 2 324 THE ELECTRIC TELEGBAPH. is always in the same condition at the moment the observa- tion is made, it follows that we need only alter our phraseology and term the unit of electro-motive force, that of a certain element in those precise conditions of tension, and as the difference between the tensions of the poles of the same kind of element under two given conditions must be always the same, it is easy to deduce from experiment the constant with which the values found must be multiplied, in order to give the first comparison. In measuring the electro-motive force of a battery of n similar elements, in series, it will always be found that the observed value is less than n times the observed electro- motive force of each single element. This is explained by the fact that the decompositions and recompositions which take place in the single element are due only to the forma- tion of its current ; whereas, in each of the n elements under consideration, these decompositions and recompositions are due to two causes first, to the formation of its proper cur- rent ; secondly, to the conduction of n 1 times this current that of the remaining elements. This action gives rise to an impoverished solution in the compartment of the negative metal and to a contrary polarisation. There are other excellent methods ; but the foregoing are the most important, and the last of them can be recommended as the best. 63. The following comparisons between the electro-motive forces of pairs of metal plates in a single fluid medium were made by Professor Poggendorff with his method of compensation. The alternate relations of three metals were always measured at the same time, all three being placed in the same fluid. It has been explained that the electro-motive force of a pair of metal plates is the difference between the electro-positiveness of the two metals, when immersed in the same fluid. If three different metals are taken in the order of this electro-positiveness, and the differ- ence of the first and second be measured, and then the difference between the second and third, it is evident that the difference between the first and third must be equal to I* SCIENCE AND PRACTICE. 325 the sum of the other two differences. Imagine a scale on which the distances from a common point, o (Fig. 152), to certain heights, a b c, represent the electro-positive polarisa- tions of three metals, A, B, C, when plunged into water ; then the distance a b y which is the difference between their electric conditions, will be the expres- sion of the electro-motive force of a galvanic pair composed of the metals A and B, and the distance b c will be similarly the electro-motive force of a pair of B and C. A pair of plates of the metals A and C will evidently have an electro -motive force, a c, equal to the sum of the two electro-motive forces a b and b c. This will also be observed in the following table ; 152. the sum of the two first values being very nearly equal to the third of the same series. A. Two metals and one fluid. (The unit in which the electro-motive forces are expressed is that of a Daniell's Element.) I. In sulphuric acid (sp. gr. = 1'838) 1 part, water 49 parts. 1. Zinc Tin ............................................. 0-409 Tin Copper .......................................... 0-410 Zinc Copper. ......................................... 0'824 2. Zinc Copper .......................................... 0-837 Copper Silver ................................. ..... 0-214 Zinc Silver ..... ....... . ...... .... ...... '..' ....... .... 1-053 3. Copper Mercury ................ .... ................ 0*356 Mercury Platinum ................................ 0*231 Copper Platinum ...................... . ............. 0-604 II. In nitric acid (sp. gr. = 1'222) 1 part, water 9 parts. 4. Zinc (amalgamated) Copper .................... 0*882 Copper Platinum.... ................................ 0*616 Zinc (amalgamated) Platinum . ................. 1-495 III. In muriatic acid (sp. gr. = 1-113) 1 part, water 9 parts. 5. Zinc (amalgamated) Copper ..... . ........... 0*788 Copper platinum ........................... ......... 0*743 Zinc (amalgamated) platinum ; ................ 1*537 326 THE ELECTRIC TELEGRAPH. 6. Copper Silver 0-152 Silver Platinum 0-620 Copper Platinum 0'771 IV. In caustic potash 1 part, water 6 parts. 7. Zinc Iron 1'003 Iron Silver 0-201 Zinc Silver 1-198 8. Zinc Antimony 0*541 Antimony Platinum 0-709 Zinc Platinum , 1'257 Y. In carbonate of soda water concentrated solution. 9. Zinc Iron 0-832 Iron Copper 0-072 Zinc Copper 0-909 lO.Zinc Tin.... 0*235 Tin Platinum 0-842 Zinc Platinum 1-078 VI. In chloride of soda water concentrated solution. 11. Zinc (amalgamated) Iron 0-476 Iron Copper O'i60 Zinc (amalgamated) Copper 0-743 12. Zinc Copper 0*672 Copper Platinum 0*673 Zinc Platinum 1-346 VII. In bromide of potassium 1 part, water 6 parts. 13. Zinc Copper 0-650 Copper Platinum 0-452 Zinc Platinum 1-102 14. Zinc Iron 0*280 Iron Silver 0*439 Zinc Silver 0-726 VIII. In iodide of potassium 1 part, water 4 parts. 15. Zinc Iron 0*447 Iron Platinum 0*427 Zinc Platinum 0*864 16. Zinc Tin 0-439 Tin Copper 0-051 Zinc Copper 0'499 SCIENCE AND PRACTICE. 327 IX. In cyanide of potassium 1 part, water 6 parts. 17. Zinc Silver 0*545 Silver Iron 0-420 Zinc Iron 0'967 18. Zinc Copper 0'052 Copper Bismuth 0-818 Zinc Bismuth 0'874 B. Two metals and two fluids. a) Iron in diluted sulphuric acid 1 part acid, and 49 parts water ^0*461 Copper in concentrated solution of sulphate of copper b) Copper in concentrated solution of sulphate of \ copper ! 0*711 Platinum in nitric acid (sp. gr. 1'34) ) c) Iron in sulphuric acid ....... } Platinum in nitric acid j The following determinations were made by the same physicist with Ohm's method : I. GEOVE'S ELEMENT. Zinc in diluted Sulphuric acid (1 : 4) ; Platinum in nitric acid (fuming) =1-8 12 Ditto (1: 4); Ditto (sp. gr. 1-33) =1-678 Ditto (1:12); Ditto ( ditto 1-33)= -603 Ditto (1: 4); Ditto ( ditto 1-19) = -558 Ditto (1 : 12) ; Ditto ( ditto M9) = -512 Zinc in sulphate of zinc solution ; Ditto ( ditto 1-33) = 1-550 Zinc in solution of common salt ; Ditto ( ditto l - 33)= -765 II. DANIELL'S ELEMEXT. Zinc in sulphuric acid, diluted (1 : 4) ; Copper insulphate copper solution I'OOO Ditto (1:12); Ditto 0-906 Ditto ( 1 : 12); Copper in nitrate of copper solution 0'926 The foregoing are a few of the numerous valuable com- parisons with which Professor Poggendorff has enriched the science of galvanic electricity. Others of equal value have been made by Joule, Wheatstone, Svanberg, and various other physicists. They all agree pretty well amongst each other, considering that different methods were employed in the comparisons. As an instance, the electro -motive force 328 THE ELECTRIC TELEGRAPH. of Grove's element, measured by one of Daniell's, is, accord- ing to Poggendorff 1-812 to 1-670 Joule 1-870 Buff 1-787 Levy and Saweljew 1-920 Beetz 1-708 Regnault 1*732 The differences are so slight as to be amply accounted for by small inequalities in the degree of concentration of the fluids employed. V. UNITS or RESISTANCE. 64. Siemens' Mercury Unit. The best of all the arbitrary units of electrical resistance is that of a prism of mercury defined and determined by Dr. Werner Siemens.* The want of agreement between the resistance etalons distributed by Jacobi, induced Dr. Siemens to direct his attention to the in- troduction of some method of constructing an unit whose reproduction, with small chance of error, would be a matter of comparative facility. The metal, mercury, adapts itself best to this purpose ; its conducting power, as we have already said, is less than those of the other pure metals, while its molecular structure at the same temperature is always the same. The difficulty first experienced was to get a vessel in which the mercury could be contained, and glass tubes were selected as giving an unalterable form to the body. The unit of resistance which Dr. Siemens defined is that of a prism of pure mercury 1 millimetre section and 1 metre long, at C. Glass tubes are generally irregularly conical, very seldom, for any length, cylindrical, so that, in the selection of uniformly conical tubes consisted the principal difficulty. From a great number of glass tubes of different calibres, cut to the length of a meter, a few were selected as being most uniform, and the amount of their conicalness * Fogg. Ann., Bd. ex., Seite 1. SCIENCE AND PRACTICE. 329 quantitively determined with the aid of a drop of mercury sucked into one end, and its length measured at intervals along the whole length. The effect which conicalness has upon the resistance was calculated as follows : The tube is regarded as a truncated cone, A, B, c, D (Fig. 153), whose parallel bottom and top have the radii n and r, and whose length is L At the distance, x, from the top, suppose a disc, M N, of the thickness d x, and radius z, to be drawn ; then if W is the re- c F* 153 sistance of the cone, in the direction of its axis, and d W the resistance of the disc M N, in the same direction, and This value of z differentiated gives R r which we must insert in the first equation to obtain the resistance of the differential disc. And this integrated gives W, the resistance of the cone. f I L - r) TT s 2 (R-r) TT \ r R t/r or W = . . .' . . (L R r TT v 330 THE ELECTRIC TELEGRAPH. To ascertain the values of the radii, B, and r, for each tube to be used as a measure, the only way is to fill the tube with pure mercury at a known temperature and weigh it, and to accept the reciprocal-lengths of the mercury drop, with which the tube is calibrated, as the relations of its sectional area, at different points. The weight, Gr, of the mercury is arrived at by repeatedly filling and weighing, making the necessary reductions for temperature and atmospheric pressure, and taking the mean of several observations. If a is the specific gravity of mercury at C, the volume V is For the value of V we have also the geometrical expression for the contents of a truncated cone. which, divided by R r, becomes IfrV" " E/S And if we make -5- > Whence _ Y ^ "~T^r n . . 1 + and a are constants : a is the length of a knot in centimetres =185200, ]3 is the weight of a Ib. in grammes =453,6. a is the specific gravity of drawn copper =8,899. A A 2 356 THE ELECTRIC TELEGRAPH. The highest value, probably, yet found for the conduct- ing power of pure copper is sixty times that of pure mercury. Commercial copper may be considered of good quality when its conducting power is over fifty. Dr. Matthiessen's results on this subject are valuable. He has found the conducting powers of annealed wires of the following sorts of commercial coppers, when the conducting power of pure copper at 15'5C. is taken as 100, to be : 1. Lake Superior, native, not fused 98-8, at 15-5 2. Ditto, fused, as it comes in commerce ... 92*6, at 15 -0 3. BurraBurra 88-7, at 14*0 4. Best selected' 81-3, at 14*2 5. Bright copper wire 72*2, at 15*7 6. Tough Copper 71-0, at 17'3 7. Demidoff 59-3, at 12-7 8. RioTinto 14-2, at 14-8 76. Measuring the Insulation Resistance of the Core. The first measurements of the electrical conditions of submarine wires were made with a simple galvanometer, inserted between one end of the cable to be measured and one of the poles of a galvanic battery, the other pole of which was to earth. When the farther end of the cable was put in con- nection with the earth the current of the battery passed through both cable and galvanometer, deflecting the needle of the latter to a greater or less degree as the cable was shorter or longer, or the section of the conductor greater or smaller, than that of some piece of cable taken as a standard. This test served to show at the same time that the conductor of the cable was entire, and gave a remote idea of its length. When the end of the cable was insulated, however, or dry and free in air, the current of the battery passed through the gal- vanometer as before, but, having to traverse the insulating medium surrounding the conductor, its intensity was materi- ally lessened, depending upon the degree of insulation of the cable. If no deflection was observed, it was presumed that the cable was good ; and if the current was strong enough to deflect the needle, the magnitude of its deflection gave an SCIENCE AND PRACTICE. 357 idea of the magnitude of the fault through which the current found a complete circuit. The instruments used for this purpose were of a rough nature, badly insulated, and insensible to currents of small intensity. But notwithstanding the obvious insufficiency of this method of testing, it did not entirely give place to a more scientific way until the date when the Malta- Alexandria cable was begun. The Government tests applied to this core began, however, the work of civilisation; the insulation resistances and the copper resistances were expressed in one and the same unit, and were thus made directly comparable with each other ; the dependence of the resistance of the con- ductor upon the conducting power of the metal used and upon its dimensions, and the dependence of the resistance of the in- sulating covering upon the specific resistance of the material and its dimensions being known, these resistances were calcu- lated and the results compared with those found by actual experiment; and thus the electrical conditions of the cable were judged by the agreement of the results found with those expected. Dr. Werner Siemens was the pioneer who began this very serviceable work, and carried it through ; and, after the first prejudice at innovation had been got over, electricians, one after the other, fell into the same way of thinking and of measuring. In manufacturing the core of a submarine cable, as much care is devoted to the selection of gutta-percha of high specific insulation as of copper of high conducting power. Both are only to be attained by freeing the commercial materials from impurities. The Gutta-percha Company, of Wharf Road, have suc- ceeded signally in the production of first-rate insulation. The way they secure this is by selecting the best gum, and, after the process of cutting the imported blocks into small shavings and masticating it at the temperature of boiling water, of straining the plastic material through sieves of fine wire gauze. By this operation, almost all the natural im- purities of the gum are removed, and the substance rendered homogeneous and of low conducting power. 358 THE ELECTRIC TELEGRAPH. An idea of the gradual perfection to which the covering of submarine wires with gutta-percha has attained, may be gleaned from the relative insulations of the principal cables which have been made. The core of the Atlantic cable, made in 1856, had an insu- lation resistance of 12 millions per knot, at a temperature of 24Cels. ; the Red Sea and Indian cable, made in 1859, had 30 millions per knot ; the Toulon- Algiers cable, made in 1860, 60 millions per knot; the Malta-Alexandria cable, made the same year, 100 millions ; the Oran-Carthagena cable, made in 1863, 350 millions ; the Persian Gulf cable the same ; and, lastly, the Atlantic cable, nearly 500 millions per knot. The insulation resistance of a cable is measured by one of the following methods : (1.) Wheatstone's bridge. (2.) Differential, method. (3.) Deflection. 77. The method of measuring a cable by means of Wheat- stone's bridge was given when treating of Messrs. Siemens' testing-board. With the customary arrangements, this method is limited to the measurement of resistances of 10 millions. When greater resistances are sought to be measured, it would be necessary, if this method were to be used, to increase the proportion of the two branches of the bridge, or the value of the adjustable side, but which would proportionably reduce the sensibility of the galvanometer. 78. To avoid this, the system has been introduced of mea- suring the resistance by mean of a differential galvanometer, the magnetic effects of whose coils upon its needle are very different. When this is done, and a single battery used, the intensity in the smaller circuit is obliged to be so great in order to balance the needle, that it endangers warming the coil and increasing its resistance. To remedy this incon- venience, Messrs. Siemens, in their measurements, have sometimes employed separate batteries in the two circuits of the galvanometer. Such an arrangement is shown in Fig. 164, in which x is SCIENCE AND PRACTICE. 359 the cable resistance ; a, one of the coils of the galvanometer, whose resistance is r ; B, the battery, whose electro-motive force is E, the intensity in the circuit on the right hand being I ; R is the resistance i. inserted ; b, the galvanometer coil ; B', the battery, whose electro-motive force is E', and I 7 the current in the left-hand circuit. Let m and m f be the opposite magnetic effects of the coils a and b upon the needle Fig. 164. suspended between them, when the intensities in the circuits are equal. The relation is first ascertained by direct experiment, by inserting in x a known resistance, and by measuring also the relation , when e and e r are the electro-motive forces of 6 the batteries. This is called determining the constant of the apparatus. A known resistance, W, is inserted, instead of x, in the circuit of a battery, B on the other side the amount of the adjustable resistance with a single element, e r , is varied to R 7 , until the needle rests over zero. Then the intensities in the two currents are I= W^ C 1 - dl '-FT7 i ' ' (2 That the needle may take up this position, however, it is necessary that the currents in the circuits be inversely pro- portional to the magnetic effects of the coils. J?L - - . (3 m' ~ ' I Inserting the values of I and I', given by (1 and (2 in this equation, m B' And if e = e that is to say, a similar electro-motive force be taken on each side, in measuring the constant, 360 THE ELECTRIC TELEGRAPH. m' ' E' + / Sometimes the proportion between m and m 1 is so great that it would be inconvenient to obtain the resistance W large enough to establish the balance in measuring the constant of the apparatus. In this case the operator employs a shunt across the ends of the larger coil of the galvanometer, and introduces its value into the calculation by the con- stant K. This shunt, r,,, Fig. 165, does not exert any de- flecting force upon the needle, its duty being simply to take off part of the current from r, that the needle may not be acted upon Fig. 165. by the full intensity, i. The currents in r and r being i and (,, respectively, by Kirchhoff's laws, I __ i ._ { ti - o and i r i ti r t = whence And since m I m' r By Ohm's law, however, W-f / SCIENCE AND PRACTICE. 361 and which, divided by each other, give rr" I ~ e R' + r' (5 and (6 combined m e W(r + m, e (E' + r 1 ) r" The cable is now inserted in its place, as in Fig. 164 ; E, is adjusted to obtain a balance between the active magnetic effects of the currents. I and I' being the currents, . . .(9 And their relation to each other is L-5--K I ~~ m~ therefore E (E + /) and The resistance of the cable, according to this, is equal to the product of the resistance of the circuit R + r, the relation p of the electro-motive forces ^, and the constant K of the Sit' initial magnetic forces of the coils of the galvanometer. Fig. 166 represents a plan of a convenient method of arranging a board by which not only this, but also the foregoing systems may be introduced, by simply changing the positions of a few contact plugs. * In this equation, as in some of those which precede it, the resistance x is regarded as the whole resistance of the circuit. The galvanometer resist- ance is neglected, as the value of x is very great in comparison with it. 362 THE ELECTRIC TELEGRAPH. A similar arrangement, with the exception of the form of commutator u, which we have altered, is used by Messrs. Siemens on board ship while paying out submarine cables, and also at the land stations. The key acts between the lever 3 and back contact 2, as a short circuit across the galvano- -*^= c&a| meter coil a , the longer of + * J 3 -. *l the two, and between the lever and front contact 1, as contact key in the cir- cuit of the galvanometer coil ft. This application of the key is useful in insu- lation tests, when the cur- rent is allowed to run, for a certain time, into the cable before the observa- tion is made, and the battery circuit on the other side closed, at the same instant that the insulation current is allowed to go through the galvanometer. The other connections and pieces are obvious. The constant of the instrument is measured by allowing the current of the smaller battery b to go, at the same time, parallel through both the coils a and /}. This is done by means of contact plugs in the holes i in of u x , in the holes 1 3, 2 4, and 45 of u, and in all the holes of the pro- portion resistance coils r and p ; all other stoppers are left out. The current of b goes through the circuit b, z, I, IIT of u,, K (1, 3) (the key being pressed down) divides into ' r'q ^ i\ -o ' 'fK "n ! of battery 6. The resistance ( a, r, p, TJ (3, 1), B, u (5, 4), c j of the coil ft and its circuit remains unaltered ; R is adjusted until the deflecting forces of the currents in and ft are balanced, in which case SCIENCE AND PRACTICE. 363 R being the amount of unstoppered resistance in the set of c oils between and u v r f the resistance of a, and r" that of /3. An advantage is derived from the measurement of the magnetic constant of the coils by the galvanometer with a single battery, as no error can occur from inequality in the electro-motive forces. For the measurement of insulation resistance, the cable is connected with the terminal u 2) in which the contact plug is inserted, so that the cable end is, in fact, connected with the beam K 3 of the key, or that which is the same thing, with the upper end of the coil a. Contact plugs are inserted in all the holes of r and p ; diagonally in u 2 , so as to put the copper pole or the zinc pole of the battery B to earth as occa- sion may require ; in u 3 , so as to introduce a sufficiently great battery power to make the observation sensible, when 1 per cent. of the electro-motive force is added or subtracted; in u, in holes 1 2 and 4 5 ; in U! in hole i in ; in the hole of u 2 , when the moment arrives for sending the current of B into the cable ; and lastly, after the expiration of a certain time, the key is pressed down, by which the short circuit is broken (letting the insulation current go through , and the opposite current of b through ft), and R is closed at the same instant. In the changed position of the stoppers of u, the resistance coils R, which were in the circuit of a, in measuring the con- stant, are now shifted into the circuit opposite. The current of B goes through earth, u 2 , u 3 , u 2 , r, a, u 2) cable, dielectric, earth. The current of b goes through the circuit, u t (in, i), key (1, 3), ft /, u (2, 1), R, u (5, 4), b, &c. The resistance of R is adjusted until the needle is not de- flected to either side of zero. The resistance of the cable is then expressed by r 1 the resistance of a is so small, in comparison with #, that it may, without sensible error, be neglected, and the resistance of the cable be called 364 THE ELECTRIC TELEGRAPH. One of the fundamental laws of magnetism is that the magnetic force decreases as the square of the distance between the points acting upon each other increases. The galva- nometer coils, in which currents are circulating, being mag- nets, are subject to the same law. It, therefore, becomes con- venient to so arrange the coil (3 that its distance from the needle may be varied, by which the constant relation K may be obtained of any value most convenient in the calculation of x, as, for example, some power of 10, or the resistance x may be calculated by the distance ; which is, in that case, varied by means of a michrometer screw, and observed with a nonius. The latter method, where time and trouble are not of importance, is less to be recommended, however, as any error occurring in the observation of the distance, obviously comes into x in the square. Nothing is easier than to make TT ~R -= = 10 6 , and then the resistance of the cable is x = (E -f r) millions. The apparent complication of the board is caused by the arrangements of the pieces for measuring by the other methods as well as by this one. 79. Beyond the limits within which the differential _ 2_ method just explained may be em- ployed, we use the method of deflection. For this purpose Professor Thomson's reflecting galvanometer, and the sine- multipliers, are the best measuring in- istruments. One end of the cable, a b, Fig. 167, is connected with the galvano- meter coil, the other end of the coil being in connection with a contact key, K, and, through this, with one pole of Fig. 167. a battery, B, whose other pole is to earth. When in this position the key is pressed down, the charge current passes through the galvanometer. This is avoided by a short circuit or shunt of no appreciable resist- SCIENCE AND PRACTICE. 365 ance between K and a, which is removed when the steady deflection due to leakage, or conduction of the insulator, is to be observed. The details of these arrangements were given in the description of Messrs. Siemens' testing-board. 80. Resistance of Insulating Materials under Pressure. The arrangements made by Mr. Reid at the gutta-percha works enabled Mr. C. W. Siemens to have the core of his cable tested under a hydraulic pressure of 280 atmospheres. He took advantage of this to make some interesting experi- ments on the electrical behaviour of gutta-percha, india- rubber, and a combination of both, under high pressure. The results of these experiments, which the writer carried out under Mr. Siemens' direction, were read before the British Association at their recent meeting at Newcastle. The core of the Malta- Alexandria cable was tested under a pressure of 600 Ibs. per square inch, and from these tests it was observed that the resistance of insulation increased, under this pressure, to the amount of 14 per cent. ; or, more generally, that the resistance, R^, of a coil of this cable, under the pressure of p Ibs. per square inch, whose resistance under atmospheric pressure at the same temperature, was E, could be very nearly calculated by the formula. R p = R (1 -f- 0,00023^) The stronger tank since erected by Mr. Reid enabled the tests of part of the Carthagena-Oran core to be carried to 300 atmospheres. The resistance of the insulation was observed at different stages of the pressure, between vacuum and 300 atmospheres, advancing each time 75 atmospheres. The results of the tests with gutta-percha covered wire showed that the resistance increases with the pressure, and that the curve is not approximately a straight line, as the above formula expresses it, but that it is slightly convex to the axis of x, when the ordinates of a graphic system (Fig. 168) represent the resistances, and the abscissae the pressures, in atmospheres. Wires insulated with india-rubber gave quite opposite results, the resistance decreasing as the pressure was increased, and the curve being somewhat concave to the base line. A wire insulated first with india-rubber and then, 366 THE ELECTRIC TELEGRAPH. over this, with a coat of gutta-percha, gave mean results between those found with the two materials separately. From all the observations made in carrying out these ex- Gutta-percha. _u i_ Pressure. boo atmospheres. Fig. 168. periments the means of the co-efficient a, from these three modes of insulation, are for Gutta-percha ........................ + 0,0041 India-rubber ..................... 0,0009 Combined ........................... -|- 0,0016 to be inserted in the formula aj p) . by which as good an approximation to the true resistance is obtained, as the expressions of the curves as straight lines allow ; and the observations were too rough and too few to make it worth while to calculate the formula with more members. The values obtained for the Carthagena-Oran core agree SCIENCE AND PRACTICE. 367 sufficiently well with the corresponding observations with the core of the Malta-Alexandria cahle, to render it highly pro- bable that the per-centage improvement in the resistance under pressure is independent of the thickness of the in- sulator, as well as of improvement in the quality of the material. 81. Electrification. In the earliest days of cable-making attention was called to the fact that, when a current is kept on an insulated wire, the insulation resistance increases with the time, but not proportionably to it. The late Dr. Esselbach made some elaborate experimental investigations to determine empirically the nature of the curve, during the time he officiated as government electrician to the Malta- Alexandria cable, and had prepared himself to prosecute his researches on this subject with the Persian Gulf line, of which he held the position of general- super- intendent. It is much to be regretted, however, that his untimely death has deprived the science of the most valuable part of that which he had already accomplished in illustra- tion and explanation of the phenomenon. If we take a wire, insulated with gutta-percha, and con- nect the pole of a battery to one end of it, we find that the observations of the current, after stated periods of time, will give us a curve represented by the line a b, Fig. 169, in which the strengths of current are the ordinates, and the times the abscissae. It is evident that the decreasing current observed is due to two causes the one to leakage through the material, and which is the proper insulation current, and the other to electrification. A glance at the line which we obtain will suffice to make it evident that the curve is assymptotical, nearing the axes of the system in both direc- tions. From this it is obvious that the current which we measure is never the true insulation- current 1, but always 1 plus some function of the time, although the curve after an hour or so approaches very near to the line of the true insulation- current. There is but one conclusion to be drawn from this phe- nomenon : it is, that a cable takes an infinitely long time to 368 THE ELECTRIC TELEGRAPH. become completely charged, and that the quantity of elec- tricity which goes into it, to contribute to the charge, at any moment, after closing the circuit, is represented by the differ- ence between the true insulation and the measured current. Time. 40 min. Fig. 169. 60 min. This difference is found experimentally to be inversely pro- portional to the time, or nearly so. 82. Joints in the Core. After the tests of the single knots of core, these are joined up in lengths of six knots, more or less, to be transferred to the machines in the sheathing works. Joining the core is an operation requiring manual skill, and, above all, scrupulous cleanliness. No joint is ad- mitted into a submarine line unless made by a workman who has had a considerable practice. This is a branch brought to perfection by the Gutta-percha Company. The jointer commences by cutting off the two ends of the core, so that the gutta-percha and copper- wire are " flush ; " he then SCIENCE AND PRACTICE. 369 warms the percha, for a distance of about three inches from each of the ends, with a spirit-flame, and, when sufficiently soft, pushes it back until it forms an enlargement as at A and A t , Fig. 170. The wires of the copper strand are then cut off at different distances between the bulbs and the ends. There are two ways in which joints are made in the con- ductors of cables. The first is called the " scarf-joint." It Fig. 170. is made by filing off the two ends of the strand at a corre- sponding angle ; fitting and soldering the slanting faces. When this is done, the whole is wrapped round with two coats of fine copper wire, the first wrappering being soldered all the way, the second only on each side of the joint ; by which, should the conductor part at this point, the outer wrapper-wire will preserve electric continuity. Such an accident might easily occur ; and, in that case, the outer wire would only be extended like a spiral spring. The other method, which is sometimes used, is less convenient for strands of seven than for those of three wires. The wires of the opposite ends of the strand are opened out, and each are joined separately. Each separate joint is made with slanting faces, soldered like the scarf-joint, but not wrappered, and a distance of at least half an inch left between each sepa- rate soldering. When the soldering is finished, the work- man endeavours to bring the same spiral twist into the soldered wires as in the rest of the conductor. This is best done by having previously coiled a length upon the floor, and put a couple of reverse twists in it, which are afterwards concentrated in the soldered place, to give it the right spiral. The soldering completed, the operator washes his soldered B B 370 THE ELECTRIC TELEGRAPH. wire with naphtha, warms it with his spirit- flame, and smears it with some of the compound of resin, tar, and gutta-percha, invented by "Willoughby Smith. The joint is then screwed into a holder (Fig. 170), where one of the knobs of gutta- percha, A AI, is gently warmed until the gum is soft. It is then drawn carefully up to the other knob, leaving, on its way, a perfect tube of gutta-percha upon the wire. The superfluous percha is. removed, the other knob warmed and drawn in the same way over the tube already formed, which is at the same time heated sufficiently to make the two adhere. A thin coating of Chatterton's compound is put over this ; and, finally, after kneading the plastic tubes well together between the finger and thumb, an oblong piece of sheet gutta-percha is warmed in the lamp, and the whole joint clothed with it. When he has kneaded the whole into a homogeneous mass, warming it gently from time to time as it gets consolidated, the operator works the joint as cylin- drical as possible with his hands, and finishes it off on the outside by burnishing or ironing with a small polished steel tool, heated to a degree sufficiently high to smooth the percha. The Gutta-percha Company do not make two tubes of the knobs AAj, but simply draw them together over the joint so as to form a single tube, which they cover up in a piece of sheet gutta-percha. The methods are, perhaps, equally good. Sometimes also joints are made with alternate coverings of gutta-percha and compound, so as to resemble exactly the insulator. 83. Testing the Joints. Joints are the weak points of submarine cables. It is therefore essential to test those which occur between the lengths of core with the utmost precision. The old method of testing joints was to get a steady deflection by the insulation current after the battery had been on the core or cable some time, during which the joint was held in the air, dry ; it was then suddenly plunged into water in electrical connection with the earth ; and any movement of the galvanometer needle, however slight, was taken as an indication of an inferior joint. SCIENCE AND PRACTICE. 371 Mi*. Whitehouse first sought to refine this crude process by measuring only the current which actually went through the joint, and expressing it in terms of the insulation resist- ance of lengths by the cable. The joint between the lengths a and /3 (Fig. 171) was immersed in acidulated water con- tained in an insulated vessel, v ; a plate of metal, also im- mersed in the water, was connected to one pole of a powerful, well-in- sulated battery, B, the other pole being in connection with one end of a galvanometer coil; the other end of the galvanometer coil, and both ends of the conductor of the cable, were to earth. Whatever current was indicated by the deflection of the galvanometer needle must neces- sarily pass through the water in the insulated vessel, and through the joint. This method was employed for some time, but has been succeeded by a modification of it, introduced by Messrs. Bright and Clark, which consists in accumulating the elec^ tricity which goes through the joint upon the plates of a condenser, and sending the discharge suddenly through the galvanometer, instead of letting the current go gradually through, as Whitehouse did. This modification is decidedly superior to its original. The two parts, A and B (Fig. 172), of the cable are placed con- veniently, that the joint between them may be immersed in water contained in the insulated vessel, v, from which a con- nection, I, leads to the plate, 2, of the condenser, c. The other side of the condenser is connected with the lever, 1, of a switch, s, the two anvils, 2, 3, of which are respectively in the circuits of a battery, E, and of a galvanometer, G. The other end of the galvanometer coil is connected to side 2 of the condenser, and the other pole of the battery with tho BB 2 372 THE ELECTRIC TELEGRAPH. conductor of the cable. When all is ready for the test, the lever of the switch is turned upon the contact 2, closing the circuit of the battery from one pole through E, t, B, joint, water in Y, P, /, and plate 2 of condenser ; andfromthe other pole, through s, 2, 1, I 1 , and plate 1 of the condenser. What- ever leakage occurs through the joint carries with it an equi- valent quantity of electricity, which is accumulated upon the plates 1 and 2 of c. After the lapse of a certain time, the lever of the switch is taken from the anvil, 2, and put upon 3, complet ing the circuit between the plates of the condenser with the galvano- meter (c, 2, G, 3, s, 1, /', c, 1). The discharge from the condenser plates passes thereupon through the galvanometer whose magnet needle is deflected, the deflection depending of course upon the leakage of the joint, the length of time during which the condensa- tion continues, the degree of insulation of the condenser itself, and the force of the battery. In the Gutta-percha Works, Mr. Smith uses a length of cable instead of an ordinary condenser. In this case the conductor of the condenser cable is connected with I, instead of c, 2 ; and as the outside (which represents c,) is to earth, the lever s t of the switch must also be put to earth. Mr. Yarley prefers the use of condensers made of alter- nate leaves of tinfoil and paper saturated with paraffin. At first sight, it might seem doubtful that any satisfactory result could be obtained with this method, in which the con- denser conducts, perhaps, ten thousand times as well as the joint which is tested, The objection, however, is easily satis- fied. The proportion between the current which goes through the condenser and the static quantity which is retained as Fig. 172. SCIENCE AND PRACTICE. 373 charge at the moment of discharging, remains unchanged, whatever the current may be. If the current is small, as it must be if the joint is good, the charge will be proportion- ally small ; and it will increase as the current increases. In shifting the condenser by an instantaneous movement from the battery and joint to the galvanometer, no time is given for any material part of the charge, at the moment in the condenser, to neutralise itself through the dielectric ; and, therefore, whatever is present shows itself upon the instru- ment. The core, when joined up in convenient lengths, is coiled carefully upon the drums and transported to the sheathing works. Arrived there, if not wanted immediately to be sheathed, it is stored away in water-tanks, and tested at in- tervals, to ascertain if it has become injured, and to accu- mulate data for subsequent calculations. 84. Self-heating of Cables. During the manufacture of the cable, also, its electrical conditions are ascertained at regular intervals usually twice a day to make sure that no injury has happened to the insulator ; or, in the event of the cable being coiled in a dry tank, that it has not heated spontaneously, and increased the temperature of the core. This is seen by the increased copper resistance and lower resistance of insulation. The cable destined in 1860 for submersion between Ran- goon and Singapore, and subsequently laid in the Mediter- ranean, was coiled in tanks in the yard at Morden Wharf, Greenwich. While there, some of the tanks became leaky, the weight of the cable bearing too heavily upon their foundations ; and it was decided not to pump water over the cables, because the iron rusted very quickly, and the oxide, being washed off by every new supply of water, would soon have reduced the sections of the wires. After some days' exposure to the air in this half- wet condition, we found that some of the cables showed signs of deterioration and a higher copper resistance than was due to the temperature of the surrounding air. The cable which appeared the most de- 374 THE ELECTRIC TELEGRAPH. cidedly affected- a length of 162 knots was cut into three parts, and coiled into a dry tank, where its electrical con- ditions were from time to time narrowly examined. Its copper resistance had risen to a degree which indicated a temperature of 80 Fah., while the air and water in the neighbourhood were not above 57 Fah. Considerable anxiety was, of course, manifested on the occasion of this unlooked-for mishap, and Dr. Miller of King's College undertook for the Board of Trade to inquire into and report upon the subject. In his Report he says that, " These heating effects appear to have been due, not to any permanent chemical change, either in the composition of the insulating coating of the gutta-percha, nor in that of the serving of hemp and tow ; but were owing simply to the effects of oxidation upon the iron at the ordinary summer tem- perature of the air, produced by the moistening of the cable with the water of the river, the slightly brackish nature of which increased the effect. " In support of this opinion a quantity of iron filings was placed in a wooden box, a foot deep, and water at 40 F. poured over them. In a few hours the temperature of the mass had gone up con- siderably, and, after a day and a half, reached 100 F. The rate at which this increase of temperature went on was found to depend upon the frequency with which the filings were stirred and watered. In order to measure the degree of self-heating, and to ascertain at all times the temperature of a cable when coiled in the tanks, both in the yard and on board the ship, Mr. "William Siemens constructed a resistance- thermometer. This " resistance^ thermometer" consists of a coil of fine copper or other pure metal wire, whose resistance at Cels. is .1 00 or 1,000 units. The per-centage variation of the resistance of pure metals between certain limits of tempera- ture being known, by measuring the -resistance of the coil at any moment, its temperature can be calculated, and that of the surrounding medium concluded. A more useful and unerring measure of temperature than this resistance-ther- mometer does not exist. One of the chief advantages which SCIENCE AND PRACTICE. 375 it possesses over the expansion-thermometers is that its indications may be read off at almost any distance ; and by a little ingenious contrivance the unknown temperature of the leading wires will not cause the least error. This is attained by connecting one end of the thermometer wire with the outside casing of the instrument (or earth), the other end with one of the leading wires, and the remaining leading wire also with the casing. The leading wires are made of the same metal, of the same length of course, and are adjusted to be of the same resistance. They are bound up together so that they each must have the same tempera- ture in any point. Thus arranged, the thermometer is put in its place of rest, and the ends of the leading wires con- nected with the measuring apparatus, which consists of a "Wheatstone's balance. The resistances r and / (Fig. 173) are equal to each other, of the same metal, and coiled together upon a common reel. In the circuit of the leading wire I, connected with the casing of the thermometer, an adjustable resistance, R, is inserted. The whole then forms a balance, giving us in the proportion of equilibrium, r' - I + B, or, as r =/ and / = V, Knowing beforehand the resistance of R' at C., its tern- 376 THE ELECTRIC TELEGRAPH. perature giving any other resistance is easily calculated by means of the coefficients given in the table at p. 267. A more handy arrangement still is by substituting a resistance, R, equal in every respect to R', in fact, by a second resistance-thermometer. This is immersed in a vessel of water whose temperature is changed until the electric equilibrium is established ; the temperature of the water gives then, without any reduction whatever, by means of a mercury thermometer, the temperature of the distant coil R'. Only that portion of the apparatus on the left-hand side of the dotted line a b is in the testing-room, that on the right hand is outside. Such an arrangement might be used with good results for ascertaining the temperature of the sea at different depths. In this case all on the right-hand side ofab would be submerged, that on the left on board. With a number of these instruments, placed between the layers of cable on board the ship Queen Victoria it was observed, after the cable had been stowed a few days, that the upper part increased in temperature at the rate of about 3 F. daily, until 86 F. was reached, while the lower part retained the temperature of the hold and water. The obstinate doubts which were urged at the time against the truth of these results were silenced signally when water of 42 F. was pumped upon the cable, and flowed out at the bottom at a temperature of 72 F. This shows the necessity of keeping an iron-covered cable, which has once been under water, always under water, even during tjie transport, although it may be attended with some inconvenience. The tests which are made at the sheathing works are of the same kind as those made at the gutta-percha works. Faults sometimes occur ; but these are easy to determine, as the operator is in possession of both ends of the cable. In order to facilitate th,e discovery and location of faults in sheathed cables, Mr. "Willoughby Smith has invented and patented the idea of serving the core with tanned hemp instead of tarred hemp, which he professes to have found has an inclination to temporarily mend small faults which might develop themselves when submerged. SCIENCE AND PRACTICE. 377 85. Finding the Place of a Fault in the Insulating Covering when both ends of the cable are at hand. The advantage of having both ends occurs before the cable is submerged, and sometimes, but in rare instances, after it is submerged, when the stations at the ends of the line are joined by another wire, whose insulation is good, and which enables the electrician to employ the " loop-method." In this event the ends of the two cables are connected together at the distant station, say at A, Fig. 174, forming a single line, E A d. Fig. 174, Two proportion resistances, r and f>, are connected together in a point, a, with one pole of a battery whose opposite pole is to earth ; their further ends b and d are severally connected to the end of cable and the end of an adjustable resistance, R, the remaining ends of the cable and resistance- coil meet- ing in E. The resistance R is inserted in the cable circuit at that end of it towards which the fault lies. A galvano- scope is also inserted between b and d. The system thus arranged forms the circuits of an ordinary Wheatstone's balance, the sides of which are, respectively, 1) r, 2) p, 3) y, 4) R -|- x. The magnitude x is the resistance of the con- ductor from the fault to E, y the resistance from the fault to d ; therefore, the resistance of the whole conductor / is 378 THE ELECTRIC TELEGRAPH. and, when the currents of the bridge system are balanced, 7" = K + * From these two equations the distances are expressed in units of resistance. These values are reducible to units of length by dividing them by the average resist- ance n of one knot ; and they become L x and L y knots respectively, d L , E + ~ (p+r)n Generally, the branch resistances r and p may be made equal to each other, facilitating the calculation with the above formulae, which then become Z-K l + K, *==__ and y =--- in resistance, or ,.-=* an, *-*+??;' 1 in knots. Another way of doing the same thing is by making the resistance between a and b some power of 10, and inserting in the side a d a set of resistance-coils, R. The ends of the cable are connected immediately to the points b and d y which are joined also by the ends of the galvanometer-coil G, as before ; x and y being the two ends of the cable from the fault in opposite directions, and no current deflecting the needle of Gr on closing the battery circuit, we have the bridge equation JL = = x ^y. / SCIENCE AND PRACTICE. 379 in resistance, or in knots. 86. Rupture of the Conductor while the insulation remains good. Another kind of fault, and one of a serious nature, when it occurs, which is happily seldom, is when the copper conductor, from its inability to withstand the elongation to which some point of the cable is exposed, snaps asunder, and the electric continuity is lost. This fault is of more danger while paying out than before. The only way to determine the distance of the rupture, in such a case, is by comparing the static capacity of the cable from the end to the fault with the average capacity of the core, per knot. There are two very excellent methods of doing this ; the one is by Mr. De Sauty, the other by Mr. Varley. 87. De Sauty' 's Method of comparing the Capacities of Leyden jars ly Bridge System. This is one of the most elegant of the Fig. 175. many applications of the null-methods, its purpose being to compare the capacity of a cable with that of a jar of unit surface ; in determining the distance of a rupture of the con- ductor it is invaluable. The method in question depends upon the same principles as Wheat stone's bridge, the only difference being that capacities are dealt with in one half of the bridge instead of resistances. At the point of junction of two resistances, r and R (Fig. 380 THE ELECTRIC TELEGRAPH. 175), one of which is adjustable between limits, is connected the beam of a contact key, K, between the point-contact of which and earth a battery, E, is inserted. The farther ends of the resistances r and R are connected respectively to the interiors of the unit jar u, and of the cable L, and to the two ends of the galvanometer coil G. The external coatings of both the jar and cable are to earth. On pressing down the key, the charge current flows in r and R, and into u and L, either directly or partly through G, into one of them. The charge current or that current which flows into the jars at the instant of closing the circuit is proportional to, and may be taken as expressing, the quantity of electricity conveyed by the currents to commence their charge, and therefore to their capacities for this charge. c and D being the intensities of the currents flowing into the jar and cable at the moment of closing the circuit, C : D = Ktr : KL K v and ~K L being the two capacities, or . . . (L C K/7 D KL By KirchhofF's law of branch circuits, 1) . . Ar BR G^ = 0, 2) . . A G C = 0, and 3) . . B G D =* 0. When the intensity G in the galvanometer circuit = 0, which is the condition upon which the method rests, and which is obtained by adjusting the value of K, these equations become A r B E = or ^- = 5 , r A C = or A = C and B D = o or B = D from which .= . (II B r ' SCIENCE AND PRACTICE. 381 and as according to (I. -g- = we have also (III. The jar U may be formed by a length /^of the same cable as L, in which case the capacities K^/ and E^ of the two lengths are to each other as the lengths. which, being substituted in III., R r IL= lu (IV. when the length Ig-is some unit used to express the length of the cable as a knot, 2026 yards, for instance, the dis- -p tance l v of the rupture will be knots from the end. In arranging a board for this measurement, a condenser made of alternate plates of gutta-percha and metal, having the same capacity as an unit of length of the cable to be tested, is found more convenient than a piece of cable, which is liable to accident. 88. Varley's Method of comparing Capacities of Jars. An equally good method of comparing the charge of one cable with that of another, or with that of a condenser of known capacity, has been used with good results by Mr. Varley. He em- ploys a differential galvanometer, whose coils have the initial mag- netic effects m and m, and the resistances g and g . The ends of these coils, joined up in the Fig. 176, 382 THE ELECTRIC TELEGRAPH. point a (Fig. 176), are connected by means of a contact key, K, with a battery, E. The other end of the coil g',. goes to one side of a condenser, (/, or to the interior of the standard piece of cable; the other end of g to the interior of the cable whose charge is to be measured. Part of the current of the coil g is shunted by means of an adjustable resistance, r. When the key is pressed down, the current divides itself at #, part i', passing through the coil g, to the unit jar c', and the remainder i, passing through the parallel circuit Q/j* a and r, whose combined resistance is - , into the g + r cable c. Let the intensities in g and r be expressed by i and i lt and at the moment of closing the circuit, let the needle be unaffected by the currents, then I' m' = i m m From the law of branch circuits, therefore m' I r I m g -\- r m g + r V m r and since the intensities of the charge currents at the mo- ment of closing the battery circuit are proportional to the capacities of the jars, or JL i' " a m For simplicity, to save measuring the constant of the m galvanometer, Mr. Yarley makes the coils g and g> of his f SCIENCE AND PRACTICE. 383 galvanometer equal in every respect, by which, therefore, m = m't and To obtain a balance with this arrangement, therefore, it is necessary that the ratio of the capacities of the two jars > should be equal to that of the sum of the resistance of the (j galvanometer coil, and its shunt divided by the resistance of the shunt ; and for attaining this it suffices to alter the value of r. A detailed description of Mr. Yarley's apparatus for this method may be found in Mr. Culley's Handbook. 89. Final Tests of a Complete Cable. To ascertain whether the electrical conditions of a finished cable correspond with the known conditions of the separate lengths of which it is composed, the conditions of the whole cable must be calcu- lated from the results of the separate tests, and compared with the final test of the complete cable. Let the lengths of the coils of the core, tested separately, be 1^ l# l# l n , and their sum L ; the insulation resistance per knot of the same, after the same time and at the same temperature, r lt r 2) r s , r n , and the resist- ance, per knot, of the whole cable x, then we have the equation, whence x = With a cable carefully made and handled, the measured will agree within a trifle with the calculated value of x. Beyond the errors of observation, the leakage of the joints, which is always great in comparison with that of similar lengths of perfect core, should be the only source of difference. The mean value of the copper resistance, per knot, measured 384 THE ELECTRIC TELEGRAPH. with a finished cable, agrees still better with the calculated value. The respective lengths / / 2 , l^ l n , having been found to have the average resistance p l5 p 2 , > 3 , . . . . p n per knot respectively, the mean resistance R, per knot, of the whole cable conductor should be L being, as before, the total length. Lastly, the mean inductive capacity C, per knot, from the single reduced average capacities c lt c 9 , _ Q * ... .(3 In this expression Q and / are constants and known magni- tudes ; therefore, by setting the values 1, 2, 3, &c., in suc- cession for K in (3, we obtain a series indicating the height hj answering to the load Q, when the cable is under the different tensions 1, 2, 3, &c. 100. The electrical operations during the paying out are of importance. The end of thick cable put ashore is taken into the land-station, and there given in charge of the elec- trician, whose duty it is to see it insulated, to speak through to the ship or to measure. Messrs. Siemens, in their expeditions, place a clock at the land station, which puts the end of the cable in position for insulation and continuity tests, and for correspondence, at regular intervals. This arrangement removes the chances of misunderstanding in the event of the insulation becoming bad during the paying out or after it is completed, as the electrician on board the ship knows at what time he may expect to receive currents, and when the farther end is insulated and put to earth. The commutator constructed for this purpose consists of a small disc of metal, projections upon the periphery of which come into contact at certain moments in each hour with three metal springs, connected severally to earth, telegraph instru- ment, and measuring board. The moment each full hour is completed, the end of the cable, represented by the metal disc just mentioned, makes contact with the Morse-instrument of the telegraphing board, and intelligence can be communicated between the ship and shore through the cable. This contact lasts four minutes, 400 THE ELECTRIC TELEGRAPH. then the point of the disc (which makes one revolution per hour) gets beyond the contact-spring,, and the disc turns without meeting with another contact for twenty-four minutes, during which time, therefore, the cable end is in- sulated and the ship measures the dielectric resistance. At twenty- eight minutes past a point of the disc touches a metal spring connected directly with earth, and the ship is enabled to measure copper resistance. This contact is also of only four minutes' duration ; and then succeeds an insulation of the cable end, as before, for another space of twenty-four minutes, giving the ship, in all, forty-eight minutes in each hour for insulation measurements. At fifty-six minutes past, the disc or land end of the cable makes contact with a spring, leading to the measuring-board at the station ; and, during the succeeding four minutes, the operator there measures in- sulation resistance, those on board taking care to insulate the ship end in time to let the land-station make the necessary observations, and, at the full hour, to put the end to the Morse apparatus on board, that the land -station can com- municate his result in the four minutes during which he is allowed to speak. Very recently, Mr* Willoughty Smith, Electrician to the' Gutta-percha "Works, suggested a method by which, during the submersion of a cable, its insulation resistance could be measured on board the ship, and simultaneously an indication of insulation be given at the land end. In addition to this, without entirely disturbing the continuance of these tests, messages can be passed backwards and forwards. In the event of a fault occurring in a cable, it is always of great importance to be able to communicate the results of measure- ments made on shore, where the instruments are steadier and more delicate, *and of necessity the results more to be de- pended upon, to the electrician on board the ship, who uses the data for calculating the distance of the fault, in order that he may judge what steps are expedient to be taken for recovering it ; and such a communication may be delayed for a considerable time, when every moment is precious, when a system of clockwork such as that just described is us ad. SCIENCE AND PRACTICE, 401 Mr. Smith proposes to receive signals upon very delicate reflecting galvanometers, constructed for the purpose upon Professor Thomson's principle, by which the currents trans- mitted may be very weak, and yet sufficient to give sensible deflections. Two such instruments^ of different degrees of sensibility, are to be inserted in the line the less sen- sitive one on board the ship, the other at the land- station and are to be employed, at the same time, for the insulation tests. The only battery to be used is that on board. One pole of this is to be put to earth, the other through the ship's galvanometer to the cable end. On shore, the other end of the cable is to be connected with one side of a galvanometer coil, the other side being put to earth ; and the earth wire having a very great resistance. This will be easily understood by Fig. 182. The cable A B is supposed to be partly submerged. The end A is con- Fig. 182. nected with the galvanometer G, and with the lever of the key K. Between the galvanometer and earth, resistances, R, equal to about a hundred millions of units, and between the anvil-contact of the key and earth resistances to a much less amount, are inserted. On board ship the cable conductor is connected with the marine- galvanometer G f , and the bat- tery B. As this system was used in submerging the Atlantic line, we may suppose A B to represent the 2,000 knots D D 402 THE ELECTRIC TELEGRAPH. of cable partly on board the Great Easteen and partly in the sea. The resistance of the cable in such a position would be, perhaps, 0,25 millions of units (representing an average of 500 millions per knot length). The current deflecting the needle of the ship's galvanometer G' with the battery B, due alone to this resistance of the cable insu- lator, would be 0,000004 B ; whilst the current due to the cable and resistance R would be 0,00000401 B, neglecting the resistances of battery and galvanometers. The addition which the introduction of the resistance or shunt R makes, therefore, to the indication of the ship's galvanometer, will only be about 0,25 per cent, of the entire deflection a dif- ference far too minute to be appreciated. The current from the ship passes from the battery through G' ; after which it is divided between the cable leakage and circuit B, A, G, R, and earth. That portion of the current which takes this course only the 0,0025 of the whole is nevertheless amply sufficient to give a readable deflection of the needle of the galvanometer G on shore ; and this deflection will continue practically unchanged so long as there occurs no alteration in the battery on board and in the cable no fault, the resist- ance of which is so small as to prevent neglecting the resist- ances G' and B in calculation. While it can, however, be said with approximate truth that G -f- B = 0, a small fault in the cable will not make itself known at the shore end ; because it follows from Kirchhoff's laws that when the battery resistance in any circuit is infinitely small, the addition of shunts between the poles makes no difference in the currents circulating in those circuits which were there before. Yery different, however, will be the behaviour of the instrument on board ship, which will indicate immediately the magni- tude of the fault. The electrician will then determine its distance by means of the method which we shall describe directly. In order to do this, however, he will have to measure the resistance of the copper conductor through the fault, which he can do without sensible error by regarding the end on shore as insulated. He will then have to tele- graph to the clerk on shore to put the end of the cable SCIENCE AND PRACTICE. 403 directly to earth. ; for which purpose a contact, c, will have to be provided. The transmission from ship to shore will be done by simply changing the electro-motive force of the battery either by increasing or by lessening it by which the current in the instrument G will be correspondingly altered, and the signals be indicated by the movements of the needle. Each time this occurs, of course, a new charge and discharge will take place which will prevent the rate of speaking ever exceeding that which, with the same receiving apparatus, wculd be attainable were the resistance R not in circuit. The shore telegraphs to the ship, by pressing down the key K, substituting thereby the resistance R' for R, altering therefore the total resistance in the battery circuit. These alterations are responded to by the needle of G'. A great benefit of this system is that, should a fault occur in the cable even while a correspondence is being carried on, it will become at once evident to the ship ; and, if a con- siderable fault, to both ship and shore. Unfortunately, how- ever, beyond the mere qualitative knowledge that the cable is good and as a means of correspondence, the deflections of the galvanometer on shore are totally useless. As data for calculating the distance of a fault they are valueless, because faults can only be determined by help of the resistance of the copper conductor ; and this is so overpowered by the resistance R, which is in the same circuit, that were the fault to occur anywhere at A or at B, or midway between the two, the result, so far as the galvanometer G is concerned, would be absolutely the same. The method is unquestionably one of the most ingenious that has been suggested, and we hope to see it generally employed. Its great merit consists in enabling the insula- tion of the cable to be observed continually on board, and at the same time, without diselectrifying the cable, to corre- spond. We do not, however, believe that a measuring appa- ratus on shore, where the facilities are so much greater for obtaining exact results, can safely be dispensed with. In order that the electrician on board might not always DD2 404 THE ELECTRIC TELEGRAPH. be required to keep his eye upon the galvanometer, which in laying long cables is very fatiguing, we proposed to use in the paying out of one of the Mediterranean cables an arrangement by which, if the insulation fell below a certain value, the needle of the galvanometer would be deflected and make contact with a fixed point, closing the circuit of a delicate relay and small battery. The relay was in turn to close the circuit of an alarm, which was to give instant notice of the occurrence of a fault, and prevent such accidents as allowing one to leave the ship for two or three hours before the electrician is a.ware of it, and when its recovery may be attended with the entire loss of the submerged part of the cable. 101. Determination of the Position of a Fault in the Insula- tion. When a cable has been sheathed with wet hemp, tested daily during the manufacture under water, and kept wet on board the ship, it is almost an impossibility that a fault can leave the ship before being discovered by the electrical engineer. Formerly, when a cable was made up dry, coiled dry in the hold of the ship, and first wetted in entering the sea on being submerged, it was to be expected .that, at this moment, when too late to repair it with facility, the first in- dications should be given of faulty insulation. In this way, many a cable that appeared moderately well insulated when on board the ship, proved to be very bad on leaving it. Although such occurrences are now impossible, faults still occur in cables by abrasion against the bottom, by the line being torn by anchors, and in other ways, after submersion. In such cases, the determination of the exact position of the fault is not an easy matter, on account of its varying resist- ance and polarisation, and the changing direction of earth currents. When the cable is laid, and therefore one end only at the disposition of the operator, a fault may be determined by the separate measurements made at both ends, or by those made at either. It is preferable, however, to have measurements made at each end to compare them, where it is possible, by which the chances of errors are reduced. SCIENCE AND PRACTICE, 405 The cable is represented in Figs. 183, 184, by the line between the points or stations A and B, the fault being at some intermediate point F. From each end two measure- ments are possible first, that of insulation, when the oppo- site end is insulated ; and, secondly that of continuity, when the opposite end is put to earth. In Fig. 183 the end B is insulated, and A sends a current from a battery into a cable. With the exception of a very small proportion, which escapes through the insulating covering of the cable between the V JFbzzZft ) Fig. 183. Fig. 184, station A and the end B, the current goes to earth at r, as is shown by the arrows. Fig. 184 shows the farther end B to earth. The current then divides itself, part going to earth at r, and the remainder at B. Station B can make similar measurements when A insulates, and puts to earth his end of the line. Let the resistance of the conductor between A and F be x ; of that portion between r and B, y ; and the resistance of the fault, s ; then we have five equations for the calculation of three unknown magnitudes. 1) . . B, = x + z 2) . . R' = y + * R being the resistance measured by A, and B/ that measured by B, when the opposite end is insulated, y + * 406 THE ELECTRIC TELEGRAPH. r being the measurement by A, and r 1 that by B, when the opposite end is to earth ; and 5) . . l=x+y I being the resistance of the copper of the complete cable before the fault occurred. With the equations 1, 2, and 5, the values of x, y> and z are as follows, in units of resistance : _ E B' I 2 h 2 E' - K I or, when a knot length of the conductor has an average resist- ance of n units, these equations expressed in knots L x and L are L R/ ~ R + l With the aid of equations 3, 4, and 5 we obtain the values of x, y, and z, as follows : rg-fOp /r'(l-r}-\ r-r 1 L 1 - V r(l-r')J _r'(l-r)r /r(l-r y- r >- r L 1 ~V 7(1 -^T and in knots, . = !^=!iri- A A2HZ'n y n(r'-r)L V 7(1-^0 J SCIENCE AND PRACTICE. 407 The value of x, divided by y, according to the above, which takes a form, -, very convenient in application, is j_ f~r~ T := V v Each of these methods depends upon a similar measurement from each end of the cable, and this requires, of course, the communication of the results of the observations from one station to the other. It is, however, possible to calculate the position of the fault with the data given, by the determina- tions made at either one of the stations alone. The two determinations made by station A are expressed in the equations 1) and 3), while 5) is an equation which follows from knowing the value of I beforehand. From these three equations we obtain x = r and 8 = (B - r) + >/(R-r)(f r) in units of resistance. In knots they are and From the two measurements made by B equations 2) and 4) with the general equation 5), we have in units x = (I /) H y = r' ^(E r )(l r ') and The value in knots, L x and Ly, being, as before, these divided by w, the number of units resistance in a knot at the same temperature. 408 THE ELECTRIC TELEGRAPH. 102. As the ship approaches the shore, where the sound- ings give a rapidly decreasing depth, the deep-sea cable is cut and the end buoyed. The ship then goes towards the land, and, after sending the end of the thick cable ashore, pays it out from the landing-place to the end of the cut cable, where the shore end is put upon one of the ship's boats, which proceeds to the buoy. The deep-sea cable is hauled up, and the jointer, who is in the boat, makes the permanent con- nection between the two cables. Testing joints on the sea is not done with the same pre- cision as on shore. If the joint is made on board the ship, which is sometimes the case in very calm weather, the joint may be tested with the aid of a condenser. Made in the boat, however, it is only possible to test the cable from the shore for insulation, when the gutta-percha joint is dry and when it is wet. The signal of approval being given by hoisting a flag or otherwise, the joint is covered well up in hemp, sheathed, and dropped overboard. Observations of the bearings of the spot are taken, and any striking con- figurations of the coast noted, in order to be able to return to the spot and fish up the joint if it should become bad. 103. Sealing up Faults; Hipp' s Method. A short cable, insulated with gutta percha, laid between Bauen and Fluelen along the "VValdstattersee, as part of the line between Lucerne and Altorf, became so faulty as to allow the escape of nearly all the current sent into it at either end. Either through bad manufacture or exposure to the air for a time before sub- mersion, the gutta-percha of this cable was found, on inquiry, to be brittle, and it was therefore probable that the fault was occasioned by cracks in the material, similar to those observed in the short length examined. To take up the cable was found to be impossible, as it had become deeply imbedded in the mud of Lake Lucerne, and had not strength enough to resist the force necessary to extricate it. Under these circumstances the repair of the faulty cable appeared very doubtful, and it would have been impossible, had not M. Hipp, the ingenious director of the Swiss tele- graphs, luckily hit upon a method by which, if not perma- SCIENCE AND PRACTICE. 409 nently repaired, the cable was at any rate endowed with a new lease of life, and one renewable to a certain limit, from time to time, as its weak condition interfered with its working. Hipp grounded his operation upon the physical facts that, when the the two poles of a battery are plunged into water, the latter is decomposed oxygen being developed at the positive, and hydrogen at the negative pole ; that when the positive electrode consists of a base metal, it is oxidised by the gas formed on its surface, and that the oxide thus formed is a very bad conductor of electricity. Hipp, there* fore, put the positive pole of a battery of seventy-two elements to the Lucerne end of his faulty cable, and insulated the end at Fleuellen ; the loss of current expressed by the deflection of his galvanometer needle amounted at the time to 32, and this lasted during all the first day it was kept on ; the next day the deflection fell to 20 P ; the day following this to 12 Q ; and on the fourth day the loss had decreased to 3^, the battery being kept all the time of the same strength. After working three weeks with positive currents, and keeping the battery on in the intervals between work, the loss fell to 1, and the cable was worked through as if np fault existed. This method, which might be safely followed with a cable laid in fresh water, would be dangerous with one submerged in the sea. The positive current, in this case, finding a way through the insulator, would facilitate the formation of muriate of copper in the fault. This would partly insulate it ; but being a soluble salt, it would very soon become dissi" pated in the surrounding water, and open the wound again greater than before. Besides, the copper of the conductor would be continually wasted, until continuity were entirely interrupted. To avoid this evil Mr. Yarley suggested laying a pure platinum wire along the strand, which would not be liable to dissolution by the current, and therefore preserve the continuity. Such a proceeding, with the present high sensibility of the receiving apparatus used upon long lines, would be no doubt of immense value, and we may some day 410 THE ELECTRIC TELEGRAPH. have to regret that it was not employed in the Atlantic cable. 104. Charge and Distribution along the Wire. The measure- ment of the distribution and charge in an insulated wire is of as much importance as the measurement of its insulating power. The distribution is independent of local faults in the insulating covering, and depends intrinsically upon the geometrical form of the insulator. By measuring the charge of a given length of cable, and comparing the result with the mean charge of the material employed, the electrician has a means of determining, with the greatest certainty, if the insulating material has the same thickness at every point along and around the conductor. A knowledge of the capa- city for charge of a cable is also of the greatest importance in the event of a rupture of the conductor while the insulation remains good. Dr. "Werner Siemens first explained the phenomenon of static charge in a cable, having had his attention directed to it by the retardation of signals on a subterranean line between Berlin and Frankfort-on-the-Maine. He says, in his paper on this subject, that when the resist- ance of the battery is very small in comparison with the Fig. 185. resistance of an uniform cylindrical cable, the tension of the pole of the battery connected with one end of it remain un- altered, even when the other end of the line is put to earth, and that the tensions of different points along its length are in proportion to their distances from the end which is to earth. To explain this, let a c (Fig. 185) represent the line wire, a b the tension of the electricity of the battery inserted SCIENCE AND PRACTICE. 411 between the end a and earth, and c the end which is put to earth ; then a perpendicular raised upon any point of the base a c to the point of intersection with the hypothenuse b c will express the electric tension or the charge at that point. The superfices of the triangle a b c represents, therefore, the whole charge held in the cable under these conditions. As a second case, suppose a battery of equal force to be connected between the end c and earth, so that its current circulates, in the same sense, in the line, as the battery at a. The negative ordinate c d will then represent the ten- sion of this battery and of the end c of the line ; and the vertical distances between the straight line joining b and d and the different points of the base a c will represent the tensions in those points. The covered wire is, according to this, from the end a to the middle e charged with positive electricity, and from the middle to the end c with negative. The whole charge of the cable is equal to the sum of the opposite charges in the two equal triangles b a e and c d e, or null. When the batteries have not equal tensions the sum of the charges is positive or negative, according as the stronger battery is connected with its positive or negative pole to the cable. A third case of charge is when the end c of the cable is insulated. The tension of the electricity in the point c will then be the same as a b, and the charge of the whole cable will be the contents of the rectangle a b X a c, which is equal to twice the contents of the triangle a b c ; in other words, the charge of every point of the cable will be the same. 105. Inductive Capacity of Materials. The amount of this static charge depends, other things being equal, upon the specific induction of the material with which the cylinder is coated. Faraday says that the inductive action is communicated from one coating of a jar to the other, from atom to atom, through the dielectric. When the jar is a cylinder, there- fore, the inductive action must be propagated, from atom, to atom, in rings concentric with the conductor. 412 THE ELECTRIC TELEGRAPH. Dr. Siemens has applied the laws of heat in conductors to electric induction, expressing the capacity K of an insulated wire, with the help of the formula developed for the resist- ance of cylinders, in which I, the inductive capacity of the material, replaces C, the conducting power in the resistance formula, and |3 a constant factor. In Mr. de Sauty's method, as well as in that of Mr. Yarley, the quantity K is measured by comparison with the capacity of a jar or condenser of unit surface, or of known value in the capacity of an unit of length of some good cable. Dr. Siemens' experiments were made with a galvanometer, the needle of which was deflected by the charge- current entering a cable. He supposes that when the deflection is caused by a current of very short duration, the quantity of electricity K passing through the galvanometer is propor- tional to half the sine of the angle of deflection, which is true with the single needles of tangent galvanometers when the angle is not too great ; or, a being the angle of the thread, and t half the time of a complete oscillation of the needle under the influence of the earth's magnetism, 1) . . K = t sin. 2 To compare the charges K and K L of two different wires, having obtained, with the aid of a tangent galvanometer, the angles a and a lt we have, therefore, the proportion, 2) . . : K, : : sin. ~ : sin. -^- while the comparison of the charges by calculation with the dimensions, &c., would give <0 IT IT' n ^ W L_JL_ ) . Jv . IY . . . - -r UK-! log. e - log., - SCIENCE AND PRACTICE. 413 I and /! representing the lengths of the wire ; B, and 1^ the corresponding radii of the insulators ; r and r L the correspond- ing radii of the conductors ; and n and % the electro -motive forces or numbers of similar elements. From these three equations some very important conse- T> quences are deduced. Supposing that - of one cable is equal to - of another, a special instance of which would be 1 if different lengths of the same conductor were covered to the' same thickness with different insulating materials, 4) . . K : K, = n I : n, I, and, further, the same number of elements in action in both cables, that is, n = n l9 5) . . K : K, = I i I, If the batteries are not the same, but the lengths are l=l lt 6) . . K : Xj = n : n t For the relation of the times^ t and t lt which the charge- current takes to appear at the farther ends of the two cables, Professor Thomson and Dr< Siemens have arrived at the proportion 7) In case If, besides, I = ^ the lengths are the same 9) . . f.t^rfir* or, if, instead of the lengths being the same, the conductors have the same diameters, r =. r 19 10) . . #:# 1== p : / 414 THE ELECTRIC TELEGRAPH. From tine above equations the following important laws are deduced : (1.) The charges are in the same proportion as the sines of half the angles of deflection of the needle of a tangent galvanometer. (2.) When the relations between the diameters of the wires and those of the insulators are respectively equal, the charges only depend upon the lengths of the lines and upon the number of cells, and, according to 4), will be in direct pro- portion to the products of the lengths of the cables by the number of cells. (3.) If the proportion between the diameters of the con- ductors and those of the gutta-percha coverings are the same, and the electro-motive forces of the batteries are also equal, while the lines are of different lengths, the charges, accord- ing to formula 5), are directly proportioned to the lengths of the cables. (4.) If the proportion between the diameters of the con- ductors and those of the gutta-percha coverings is the same, and the lengths of the lines are also equal, while the electro- motive forces of the batteries are unequal, the charges, according to formula 6), are in the same proportion as the number of cells. (5.) The times of charge are independent of the batteries. Electro-motive forces do not appear at all in the formulae 7) to 10). (6.) When the proportion between the diameters of the conductors and of the gutta-percha coverings is the same, the times of charge only depend upon the lengths of the lines and the diameters of the conductors, and, according to 8), are directly proportional to the squares of the lengths, and inversely proportional to the transverse sectional areas of the conductors. (7.) If the proportions between the diameters of the con- ductors and insulators are the same, and the lengths of the cables also equal, the times of charge, according to 9), are in the inverse ratio of the squares of the radii of the conductors. SCIENCE AND PRACTICE. 415 (8.) The conductors and coverings being equal in both cables, the times of charge, according to 10), are as the squares of the lengths. With the help of these formulao it is easy to compare the quantity and time of charge in any submarine line with those of a given line. When calculating the speed of telegraph signals, another condition, discovered by Faraday, steps in, viz., the formation of electric- charge waves in the cable. As may be deduced from what has gone before, the charge precedes the current in going to the extremity of the cable. If the communication between the battery and the line be interrupted before the current has reached the end, the electricity diffuses itself and causes a deflection of the gal- vanometer needle, although the battery at that moment may not be in circuit. By reversing the" pole of the battery, instead of cutting it off, that part of the line nearest to the battery will be charged with the opposite electricity, while the electricity already in the more distant part of the line will discharge itself towards both ends of the cable, that is, one part will pass through the instrument and the other combine with, and be neutralised by, the opposite current. An electric wave is therefore formed at the same time, which is neutralised by degrees by the opposite one which follows it, but which, nevertheless, reaches the end. In this manner a succession of signals may be made to pass through a line by rapidly changing the batteries, which will set in motion the receiving instrument at the farther end, if they are of sufficient power on arriving there. If the currents which succeed each other are of the same strength and duration, these waves will all be useful in long lines ; but if long and short signals alternate, the latter will be more or less neutral- ' ised by the former, and very short signals will be absorbed to an extent which will render them powerless to move the armature of the receiving instrument. 106. The Wippe or Self-acting Make and Break. In order to overcome the inconvenience and difficulty attending the observation of angles by the sudden throw of the needle, Dr. 416 THE ELECTRIC TELEGRAPH. Siemens constructed an instrument, called, in German, a " Wippe," which is, in fact, a self-acting commutator capa- ble of interrupting or reversing the currents with great rapidity. If the end c of a cable were connected by two parallel cir- cuits, a and b (Fig. 186), with the two anvils a' and b r , in the reach of two movable contact springs, c and d, c being Fig. 187.- connected with one pole of a battery, the other pole of which is to earth, and d being connected to earth direct, then, upon pressing the spring c for an instant upon , a charge would flow into the cable from the battery, through a, in the direction represented by the arrow ; and if, c and a' being interrupted, the lower spring d were pressed upon the contact b, a quan- tity of the electricity held in charge in the cable would rush out and pass through the side b as shown by the arrow. Thus, by connecting a' and b f with c and d alternately, so that the one contact is always broken when the other is made, a succession of currents will circulate in the arm a towards the cable, and in the arm b from it. By inserting a measuring instrument in either of these SCIENCE AND PRACTICE. 417 circuits, therefore, the quantity flowing into the cable or out of it may be measured, if the contacts follow each other with sufficient rapidity, by a constant deflection. The "Wippe" consists of two permanent steel bar-magnets supported at right angles to each other upon a vertical shaft, turning on points in stone bearings. Around these magnets are two coils of wire at right angles to each other, and whose ends are connected with a commutator, turned by the shaft in such a way that when the magnet system is deflected a certain distance by the current, the latter is then reversed and the deflection continued in the same direction till the magnets have made half a revolution, when the first current direction obtains again, keeping the magnets rotating with an immense velocity. For this a driving battery, separate from the experimental battery, is required. The shaft carries a small metallic eccentric, e (Fig. 187), which, turning between the prongs of a fork, //, rocks the beam b b, lifting alternately the springs s and s / from the contact anvils c and c r . The springs, anvils, and beam are insulated from each other and connected by leading wires with convenient terminal screws. The terminals of the Wippe may be connected in various ways to fulfil the same duties. If both the ends of the cable are at /^~7OOV7wVX?j^\G hand, they may be connected between [ CD the springs s and s f , as in Fig. 188, the galvanometer G being inserted in the same circuit, while a battery, }, is connected between earth and c' ; und c is put to earth. When the beam b is in the posi- Mon shown in the figure, whatever blectricity is in the cable will be discharged, by way of s, c, and earth. When the beam places itself in the Fl - 188 - opposite way, as shown by the dotted lines, the end of the cable which was before on earth will now, in turn, be insu- lated, and the galvanometer end put, through / and V lg.e whence R, S, = 8 in which S and S, are the speeds attainable, r and r,, the radii of the copper conductors, R and R, the radii of the gutta-percha covering (therefore R r, th e -p T> thickness of the covering), log. e and log. e - 1 , the Naperean or natural logarithms of the ratios of the diameters of the gutta-percha to the copper, and / and l lt the lengths. Certain conditions of equality in the dimensions of the two cables simplify the above formula. 1. For instance, if two lengths I and /, of the same cable be taken, the remaining dimensions are equal, and B,= or the speeds, in this case, are to each other severally as the 420 THE ELECTRIC TELEGRAPH. squares of the length. 2. If the lengths are the same, and the cables different, then P r.'log., or the speeds are related directly as the transverse sections of the conductors, and as the logarithms of the proportion between the diameters of gutta-percha and copper. With aid of the formula S = r 2 log.- r the diameter 2 R, of the gutta-percha being given, the thick- ness of conductor which would give the greatest speed is found by the ordinary method of maxima and minima, by the differential calculus. The value of r which, set in the above formula, gives the greatest possible value of S is R R ~ v/T* 1-649 .... When the conductor is a solid or a sectional wire, the value -o is best deduced from the weights of the conductor and core. The weight w of a knot of the copper conductor, and that of a knot of finished core, W, being known, the weight of the insulating coat (W w) will, be also known. Knowing the specific gravity s of the insulating material, and that of p the conductor p, we can find the value of from the r equations 1) . . . W w = (R'- and 2) . . . w = r* I TT ff which, divided by each other, give W 8 e= 2-7183 . . . . etc., the base of the Naperean system of logarithms. SCIENCE AND PRACTICE. 421 From several determinations with specimens of gutta-percha, kindly given us by Mr. Willoughby Smith, we have found the mean value of s, the specific gravity of this material at 4 C. to be 0,9693 .... The value of a-, the specific gravity of telegraph copper may be taken at 8,899 When the conductor is a strand of seven wires, the value of r, which enters into the formula for calculating the speed, is obtained by taking five per cent, from the value of 1*5 times the diameter of a single wire. The following table, extracted from a more detailed one by Professor Thomson, who was the first to put it in this form, gives the relative speeds of working in similar lengths of cables, with different values of . r R r *&K.^ 0,1 10 12-52 0,2 5 35-00 0,3 3,33 . . 58-91 0,4 2,5 79-71 0,5 2 94-21 0,6 1,66 . . . 99-96 1 0,6065=^ l,649=v/ 100-00 0,7 1,429 . . 95-54 0,8 1,25 77-64 0,9 1,111 . . 46-84 " It is easy to understand/' says Professor Thomson, " why there should be a particular diameter of copper which will give a maximum rate of signalling with a stated outer diameter of gutta-percha, since if the copper wire is too small there is a loss of speed owing to a too large resistance not compensated by the smallness of the electro- static capacity; while on the other hand, if the diameter of the conductor be 422 THE ELECTRIC TELEGRAPH. too large as for instance if it fall but little short of the outer diameter of the gutta-percha the thinness of the gutta- percha coat gives rise to a greater loss of speed by increased capacity than is compensated by the gain, owing to dimin- ished resistance." According to this, a cable insulated with a coating one- third the thickness of the diameter of its conductor is in the most favourable condition to transmit with speed. Of all the long cables yet made, the new Malta- Alexandria comes nearest to this theoretical proportion ; the thickness of the insulating covering of this cable being, within five per cent., equal to the diameter of the conductor. From a mechanical point of view, a cable which has a proportionally thinner covering than this would be too much exposed to faults, therefore no engineer would ever recommend it. It is never- theless highly interesting to find in practice that the theory is corroborated. Calculated by the above formula, taking the speed of 475 words, actually obtained on 1320 knots of the Malta- Alexandria cable, as data, the speeds with which some other cables might be worked through, are given in the following table : No. Cables and dates of submersion. Length Knots. r If r log. c^L r Speed: Words per minute. 1 Atlantic, 1858 2,500 4-8 1-00 1 570 0-51 2 Red Sea 1860 .. 1 358 3-4 1-27 1-224 2-15 3 4 Suakin-Aden section .... Malta- Alexandria, 1861 Malta- Tripoli section .... Tripoli- Bengazi section Bengazi- Alexandria section Persian Gulf, 1863 . 629 1,320 230 507 597 1,150 2-95 > 3-48 1-95 > T31 1-082 T249 12-5 4-75 156-5 31-8 23-2 3-20 6 Eurra^hee-G-wader section . Gwader-Mussendom section . Mussendom-Bussire section . Bussire-Fao section Atlantic, 1865 246 357 393 154 2,500 > 3.28 j> ?> 1-73 1-190 71-3 33-8 27-2 189-0 1-15 The calculated speed of working through the Suakin-Aden section of the Red Sea cable, 12 J words per minute, agrees with that which was really obtained, viz., 11 words, allowance being made for the great improvements which had been effected in the relays and transmitting apparatus. With SCIENCE AND PRACTICE. 423 regard to the value obtained for the speed of speaking through the new Atlantic cable =1-15 words per minute, this might seem a direct contradiction to the assertion of Professor William Thomson and Mr. Yarley, that they can telegraph through the whole line at a rate of 8 words per minute. The reason of this very augmented rate of working, is due to the employment of a new transmitting apparatus, invented by these two celebrated electricians. This apparatus consists in a key so arranged as to send for each signal a series of five separate waves of different lengths into the cable. The first wave is positive, and lasts long enough to reach the other end of the cable and move the receiving instrument, immediately upon this succeeds a wave of shorter duration from the negative pole, which wipes out the positive charge and leaves it negative ; this wave is not intended to reach the other end or give a signal. The third wave is again posi- tive, it is cut off still shorter, and " wipes out " the negative charge, and leaves the cable, towards the home end, positive. The fourth wave is still shorter, negative, and so on. By this means, by the time the complete signal or five decreasing waves are given, each of which, except the first, is absorbed in the next succeeding it, the cable is left in its normal condition, and without the necessity of adopting any of the contrivances explained in the first part, the corre- spondence can be regularly proceeded with. INDEX. Absorption of water, 385. Alarm, Cavallo's, 10; Kramer's, 72; Reusser's, 10 ; Siemens', 80 ; Soin- mering's, 19. Alloy unit, 338. Alloys, conducting powers of, 268. Alphabet, Gauss and Weber's, 31; Morse' s,87; Stoehrer's, 175; Wheat- stone's, 49, 50. Alum battery, 228. Amalgamation, 332. Amalgamated zinc, 231. Amber, 1. Ampere, deflection, 21 ; telegraph, 23. Annealed metals, 269. Apparatus for cable tests, 294. Arago, 249. Astatic needles, 243 ; faults of, 244. Atlantic cable, 243, 348. Atmospheric electricity, 202. Attraction and repulsion, 2. Automatic printing, 166. Bain's telegraph, 176. Bakewell's telegraph, 179. Balance, Wheatstone's, 290, 292 ; Bri- tish Association, 306 ; bisected wire, 309. Battery, alum, 228. Battery, commutator, 104 ; Daniell's, 221 ; earth, 230 ; galvanic, 220 ; effects of galvanic, 218 ; Grove's, 327, 328; Kramer's, 224; Meid- inger's, 224; Siemens', 226; sul- phate-mercury, 229 ; Varley's, 227, Beaudoin and Digney, 122. Betancourt's telegraph, 10. Bianchi's lightning-guard, 213. Bisected wire, 309. Bonelli's telegraph, 182. Boucherie's process, 185. Borggreve's commutator, 100. Break, 396. Breguet's telegraph, 65 ; lightning- guard, 209, 211. Brett's cable, 241. Bright, insulator, 189 ; preserving posts, 187; testing joints, 371. B. A. balance, 306 ; unit, 338. Cable, Atlantic, old, 341 ; Atlantic, new, 348; Carthagena-Oran, 346; Dover-Calais, 341 ; Dover-Ostend, 341 ; in the ship, 388 ; Malta- Alexandria, 344 ; Persian Gulf, 345 ; Red Sea, 343; self- heating of, 373; tests, 350 ; joints, 368. Calibrating tubes, 330. Carlisle, decomposition, 18. Cavallo's telegraph, 10. Charge in overland wires, 200 ; in sub- terranean wires, 111. Charge, measurement of, 301 ; along the line, 410. Chatterton's compound, 352. Checking action, 31, 249. Chemical theory, 17. Chemical telegraph, Bain's, 176; Bake- well's, 179; Bonelli's, 17; Davy's, 42; Gintl's, 156, 180; Maison- neuve's, 179 ; Morse's, 40. Chloride of zinc, 186. Circuit, simple Morse, 84 ; for two stations, 93. Circuit, electric, 206 ; galvanic, 216 ; local, 92 ; shunt, 286. Classification, 1. Clarke's insulator, 191 ; joint testing, 371. Clocks for paying out tests, 399. Closed circuit, 148. C. M., 6 Combination of G. P. and I. R., 387. Commutator, battery, 104; Borg- greve's, 100; for three lines, 102; intermediate, 95 ; Nottebohms, 96 ; Siemens', 98. Compensation method, 320. Conducting power, 264 ; alloys, 268 ; annealed metals, 269 ; cable con- INDEX. 425 ductor, 355 ; copper, 356 ; fluids, 276 ; fused metals, 270; gutta-percha, 282 ; India-rubber, 384 ; materials, 264 ; pure metals, 266 ; and heat, 266. Conduction of earth, 35. Conductor, straight, 337 ; circular, 337. Conicity of glass tubes, 329. Constant of sensibility, 301. Constadt's telegraph, 25. Contact theory, 16. Copper resistance, 296, 298. Copying telegraph, 179. Corrections for tubes, 331 ; for ends, 332. Code, Gauss and Weber's, 31 ; Morse's, 87; Stohrer's, 175; Wheatstone's, 49, 50. Coxe's telegraph, 20. Currents, atmospheric, 202. Daniell's battery, 221, 327. Davy's telegraph, 42. Decomposition of water, 18. Depth of ocean, 292. De Sauty's key, 115. De Sauty's charge method, 379. Descent of bodies in water, 394. . 1% Diagrams of bottom, 352. Dial telegraph, Brequet's, 65 ; Kra- mer's, 70 ; Siemens', 75 ; Wheat- stone's, 45. Differential galvanometer, 353. Differential method, 358. Difficulties of solid normals, 339. Direct working instrument, 120 ; Dig- ney's, 122; Siemens', 125; com- plete, 127. Discovery of Leyden jar, 3 ; electro- magnet, 24 ; earth conduction, 35. Discharge, 301. Distribution of battery, 289. Don Silva's telegraph, 10. Double-needle telegraph, 50. Double telegraphing, 154. Dover- Calais cable, 341. Dover-Ostend cable, 341. Dubois' galvanometer, 242. Durability of posts, 186; of cables, 388. Dynamometre, 398. Early observations, 1. Earth conduction, 35, Earth element, 230. Earth plate, 200. Eisenlohr's resistance, 255. Elastic line, 390. Electrics and non-electrics, 2. Electric permanency, 273; circuits, 206; action of pile, 17; measure- ments, 285. Electrical magnitudes, 333. Electricity, static, 1 ; voltaic, 215 ; ne- gative and positive, 3 ; distribution of, 410. Electricity and lightning, 202. Electrification, 267. Electro -magnet, 24 ; laws of, 24. Electroscope, 8. Electromotive force, 333 ; measure of, 315 ; Fechner's method, 317 ; Ohm's method, 320 ; Poggendorff's method, 320 ; Wheatstone's method, 319. Electromotive force of element com- pared, 325. Elements, alum, 228 ; Bunsen's, 316 ; Daniell's, 221 ; earth, 230 ; electro- motive forces, 315 ; Grove's, 327, 328; Kramer's, 224; Meidinger's, 224 ; Siemens', 226 ; sulphate mer- cury, 229 ; Varley's, 227. Embossing telegraph, 86. Erection of overland lines, 198. Escape of current, 391. Experiments of Guillemin, 125. Fabroni, 17. Failure of old Atlantic, 378. Faraday, induction, 25 ; magneto - electricity, conduction, 25. Fardley's lightning- guard, 210. Faults, Hipps' method, 408. Faults, in insulation, 405 ; locality of, 307, 405 ; in conductor, 379, 381. Fechner's method, 319. Final tests of cable, 383. Fluids, conductivity of, 276. Forms of line insulators, 189. Formulae, astatic needle, 245 ; bisected wire, 312; circular current, 236; conducting powers, 264, 267 ; pure metals, 267 ; alloys, 269 ; mer- cury, 273 ; constant of sensibility, 307 ; correction for ends, 333 ; cor- rection forconicity, 331 ; copper re- sistance, 355 ; distribution, 263, 289 ; differential method, 359 ; differential galvanometer, 353 ; dynamometer, 398 ; effects of pressure, 365 ; elec- tromotive force, 318; faults in cables, 405 ; final tests, 383 ; galvanic polarization, 279; gutta-percha,' 284; inductive capacity, 411 ; insulation, 299, 282 ; , loss of charge, 302 ; Ohm's law, 256 ; place of fault, 378, 426 INDEX. 379, 381 ; resistance of element, 314 ; resistance of tubes, 319 ; rate of working, 41 9; sine galvanometer, 24 ; shunt circuits, 286 ; tangent gal- vanometer, 238 ; voltameter, 234 ; Wheatstone's balance, 290. Franklin's theory, 3. French railway telegraphs, 65. French telegraph bell, 68. Fused metals, 270. G-alvani, 15 ; hypothesis of, 15. Galvanism, discovery of, 13. Galvanic current, origin, 215 ; effects, 218 ; measurement, 232. Galvanic element, resistance, 313 ; con- stants, 313 ; force of, 315. Galvanic batteries, 220 ; polarization, 278. Galvanometer, constants of, 303 ; dif- ferential, 353*; Gaugain's, 240 ; mag- netism of coils, 245 ; sine, 241 ; sine and tangent, 247 ; tangent, 237 ; Thomson's, 249 ; Weber's, 247. Galvanised wire, 195. Gauss and Weber's telegraph, 27. German lightning-guard, 212. German silver wire, 268. Glass tubes, 328. Great Eastern, 349. Grey's experiments, 2. Grotthuss' theory, 215. Grove's element, 327. Gutta-percha, 357 ; conducting power, 280 ; inductive capacity, 384 ; under pressure, 365 ; absorption of water by, 386. Heating of cables, 373. Heat produced by current, 218. Heat and resistance, 266. High pressure tests, 365. Hooper's I. R. cable, 388. House's telegraph, 53. Hughes' telegraph, 55. Indian and Bed Sea cable, 343. India-rubber, 384 ; conducting power, 280 ; inductive capacity, 384 ; in- sulation, 385 ; as insulator, 384 ; Hooper's, 388 , absorption of water by, 386 ; under pressure, 365. Induction in cables, 111. Inductive capacity, 411. Induced magnetism, 24. Injection of poles, 185. Insulation resistance, 299 ; under pres- sure, 365. Insulation of G. P., 357 ; of I. R., 384 ; of Wray's mixture, 388; of core, 356. Insulating materials, 280, 384. Insulators, line, 189 ; Bright' s, 189 ; Chauvin's, 192; Clark's, 191; Sie- mens', 190. Insulators, stretching, 193 ; Kohl's, 194 ; Siemens', 194 ; Varley's, 192. Intensity of current, 256. Intermediate commutator, 95. Jacobi's etalons, 328. Jar, Ley den, 3. John's telegraph, 120. Joints in line wires, 196 ; in core, 363 ; testing, 370 ; in cable, 371. Joule, Dr., 331, 336. Kerckhoff's lightning-guard, 213. Kirchhoff's laws, 285. Kramer's telegraph, 70 ; alarm, 72 ; battery, 224. Law, Ohm's, 256 ; Kirchhoff's, 285 ; magnetic deflection, 21 ; electro- magnetism, 24. Laying cable, 391. Lesage's telegraph, 8. Lever key, Morse, 42. . Leyden jar, discovery of, 3 ; construc- tion of, 4. Light, 218. Lightning-guard, 205 ; Bianchi's, 213; Breguet's, 209, 211 ; Fardley's, 210 ; German, 212; Kerckhoff's, 213; Meissner's, 207 ; Nottebohm's, 210; Siemens', 208, 211. Lightning and electricity, 204. Lightning, stroke of, 204. Line insiilators, 189. Local circuits, Moore's, 42 ; Wheat- stone's, 40. Lomond's telegraph, 9. Magnet, electro, 24; construction of, 24 ; laws of, 24; deflection of, 21. Magnetic effects of current, 220. Magnetism of coils, 245. Magneto -electricity, 25. Magneto-electric telegraphs, 21 ; Gauss and Weber's, 28 ; Siemens', 75, 136, 166 ; Steinheil's, 33 ; Wheat- stone's, 45. Magneto induction key, 126. Magnitudes, electrical, 333. Mallet's buckled plates, 189. Malta- Alexandria cable, 344. INDEX. 427 Marie Davy's battery, 229. MattMessen, correction formula, 331 ; alloy unit, 338 ; copper resistance, 356 ; effects of temperature, 267. Measurements of conductor, 353 ; elec- tro-inotive force, 305; Fechner's method, 317 ; Wheatstone's method, 319 ; Ohm's method, 320 ; Poggen- dorff's, 320 ; of galvanic current, 232 ; of charge, 301 ; of discharge, 301 ; of galvanic constants, 303. Mechanical effect, 333. Meidinger's element, 224. Meissner's lightning-guard, 207. Mercury unit, 328. Metals, conducting power, 266 ; elec- tric permanency, 373. Morse's, chemical telegraph, 41 ; elec- tro-magnetic telegraph, 41 ; trans- mitting key, 41 ; simple circuit, 84 ; transmitting plate, 90 ; apparatus with relay, 92 ; for two stations, 93 ; with induction currents, 131 ; with closed circuit, 148. Multiplier, 22. Musschenbrceck, 3. Needle instruments, 46 ; magnetic, 21. Negative electricity, 3. Nottebohn\'s commutator, 96 ; light- ning-guard, 210. Ocean telegraphy, 341. Oerstedt, 21. Ohm's law, 256 ; method, 330. Operations during paying out, 391. Origin of galvanic currents, 215. Ostend-Dover cable, 341. Overland lines, 185 ; erection of, 198 ; charge in, 200. Parallel lines, 288. Paying out cables, 391. Physiological effects, 219; telegraph, 219. Pile, voltaic, 17. Polarized relay, 133; with recorders, 138. Positive electricity, 3. Posts, erection of, 198 ; impregnation of, 185; iron, 188; stone, 187; wooden, 185. Pouillet's, tangent galvanometer, 239. Pressure tests, 351. Printing telegraphs, 165 ; Bain's, 176 ; Bakewell's, 179 ; House's, 53 ; Sie- mens', 50 ; Hughes', 55. Progress in cable tests, 357 ; insula- tion, 358. Quantity, electrical, 333. Railway telegraphs, 65. Rate of working, 419. Red Sea cable, 345. Reid's pressure tanks, 351. Relay, apparatus with, 92 ; Morse's, 42; polarized, 133; Wheatstone's, 40. Repulsion, 2. Resinous electricity, 3. Resistance, boxes, 253 ; column, 255 ; galvanic elements, 313; absolute, 336 ; of conductors, 353, 355 ; of core, 356 ; of cables, 353 ; under pressure, 365, 366 ; thermometer, 374. Reusser's telegraph, 9 ; alarm, 10. Rhepstat, 251 ; Jacobi's, 252 ; Poggen- dorff's, 253. Ritchie's telegraph, 23. Ronald's telegraph, 10. Rupture of conductors, 379 ; De Sauty's method, 379 ; Varley's method, 381. Salva's telegraph, 10. Sand batteries, 228. Schilling's telegraph, 27. Schweiger's telegraph, 20 ; multiplier, 22. Sections.^ Cable, new Atlantic, 349 ; Dover- Calais, 342 ; Dover- Ostend, 342 ; old Atlantic, 343 ; shore end, 343 ; Red Sea, 343 ; shore end, 343 ; Malta, 345 ; Carthagena, 347 ; sec- tions of bottom, 292. Self-acting make and break, 51. Self-heating of cables, 373. Ship's galvanometer, 390. Shunt circuits, 286. Siemens' self-acting make and break, 51 ; pointer telegraph, 76 ; commu- tator, 98 ; submarine key, 113 ; direct working instrument, 125 ; polarized recorder, 138; type tele- graph, 166 ; lightning discharger, 208 ; "iron posts, 188 ; apparatus for testing, 294 ; resistance scales, 243 ; unit, 328 ; wippe, 415. Sine galvanometer, 241 ; tangent gal- vanometer, 246. Single needle telegraph, 46. Slack, 395. Smith, Willoughby, cable testing, 400. Scemmering's telegraph, 18. Sounder, 146. Soundings, 392. Spark, 2. Specific conducting powers, 264. 428 I^vDEX. Specimen of Hughes' printing, 65 ; Bonelli's printing, 185. Speed of telegraphing, 419. SteinheiTs telegraph, 32; lightning- guard, 206 ; earth circuit, 35. Stoehrer's telegraph, 172 ; code, 175., Sturgeon's electro-magnet, 24. Submarine key, De Sauty, 115; Sie- mens', 113. Submarine board, 143. Sulzer, 14. Switch, submarine, 117. Tables. Co-efficients, 267; conducting powers, 266 ; tin, 271 ; sulphuric acid, 277 ; polarization, 279 ; gutta- percha, 283 ; electro-motive forces, 325 ; copper conductivity, 356 ; water absorption, 386. Tangent galvanometer, 237. Telegraphs, by friction electricity, 6 ; galvanic electricity, 13 ; magneto - electric, 21 ; now in use, 46 ; C. M.'s, 6 ; Lesage's, 8 ; Lomond's, 9 ; Keusser's, 9 ; Don Salva's, 10 ; Betancourt's, 10 ; Cavallo's, 10 ; Konald's, 11; Soemmering's, 18; Schweiger's, 20; Coxe's, 20; Am- pere's, 23 ; Ritchie's, 23 ; Constadt's, 25 ; G-auss and Weber's 27 ; Stein- heir s, 32 ; Wheatstone's, 36, 46, 50, 82 ; Morse's, 41, 131 ; Davy's, 42 ; Siemens', 50, 75, 125 ; House's, 53 ; Hughes', 55; Breguet's, 65; Kra- mer's, 70 ; John's, 120 ; Digney's, 122; Bain's, 176; BakeweU's, 179; BoneUi's, 182; Physiological, 219; by sound, 146 ; double,*! 54, 160. Tell-tale, paying-out, 389* Temperature, 266, 282. Terra-voltaism, 230. - Testing joints, 370. Tests of core, 350 ; complete cable, 383. Theory, contact, 16; chemical, 17 ; of Grotthuss, 215; of two fluids, 2; of single fluid, 3. Thomson's galvanoscope, 249. Tourmaline, 1. Translation, 104; apparatus, 107; spring, 141, Transmitting key, 41 ; plate, 90. Transmission, Grey, 2 ; Dufay, 3 ; Winckler, 3 ; Lemmonier, 3 ; Wat- son, 5 ; Franklin, 5. Type telegraph, 166. Unit of electro-motive force," 316 ; Eeg- nault's, 316; DanieU's, 316. Unit of mercury, 318; absolute, 333 ; alloy, 338, Varley's translating apparatus, 107 ; switch, 117; element, 227. Vitreous electricity, 3. Volta, Alexander, 16 ; theory of, 16 ; pile, 17. Voltameter, 232. Vorselmann de Heer, 219. Water decomposed, 18 ; absorption of, 386. Watson's experiments, 5. Weber's galvanometer, 247 ; absolute units, 333 ; telegraph, 27. Wheatstone's alarm, 39 ; relay sys- tem, 40 ; telegraph, 43, 46, 50, 82 ; balance, 290, 292; method, 319. Whitehouse's joint testing, 371. Wippe, 415. Wire, 195 ; preservation of, 195 ; joint- ing, 196 ; suspension of, 198. Wooden posts, 185. Working, rate of, 419. Zinc amalgamated, 231 ; coated wire, 195; chloride, 186. THE END. VIRTUE AND CO., PRINTERS, CITY ROAD, LONDON. NOVEMBER, 1867. GENERAL CATALOGUE OF WORKS PUBLISHED BY VIRTUE ^NI3 CO., 26, IVY LANE, LONDON. Monthly, Price One Shilling, SAINT PAULS. A NEW MONTHLY MAGAZINE OF FICTION, ART, AND LITERATURE. Edited by ANTHONY TROLLOPE, AND ILLUSTRATED BY J. E. MILLAIS, K.A, NEW BOOKS AND NEW EDITIONS. THE GREAT SCHOOLS OF ENGLAND : an Account of the Foundation, Endowments, and Discipline of the Chief Seminaries of Learn- ing in England, including Eton, Winchester, Westminster, St. Paul's, Charterhouse, Merchant Taylors*, Harrow, Rugby, Shrewsbury, &c. &c. By HOWARD STAUNTON. New and Cheaper Edition. With Appendix containing an Account of all the Endowed Schools in England. [In the Press. MILLS AND FACTORIES FOR GINNING, SPINNING, AND WEAVING COTTON: Jute and Silk Manufactures; Bieachiug, Dyeing, and Calico Printing Works : Sugar, Paper, Oil, and Oil-Gas Manufactures; Iron and Timber Workshops, Corn-mills, &c. Israel in Bondage. By the Rev. J. H. INGBAHAM, LL.D., Author of- The Prince of the House of David." Illus- trated with Eight Engravings. Fcap. cloth, 5s. PRINCE OF THE HOUSE OF DAVID ; or, Three Years in the Holy City. By the Rev. J. H. INGBAHAM, LL.D. With Illustrations. New Edition. Fcap. cloth, 3s. 6d. ' RHYMES FOR THE NURSERY. Illustrated by GILBERT. 16mo. cloth, 2s. 6d. RIPPON'S (DR.) SELECTION OF HYMNS FROM THE BEST AUTHORS, including a great number of Originals, intended a.-, an Appendix to Dr. Watts's Psalms and Hymns. New Editions. Nonpareil 32mo. I Long Primer Umo. \ Large Type. s. d. t. d. s. d. Roan ... 1 6 i Roan .... 2 6 I Sheep ... 5 , gilt edges 2 , gilt edges 3 Roan, gilt edges 6 Morocco ... 6 | Morocco ... 6 | Morocco . .90 ROBINSON CRUSOE. With Illustrations on Steel. 18mo. cloth, 2s. ROLLIN'S ANCIENT HISTORY. 2 Vols. 8vo. cloth, 21s. PUBLISHED BY VIKTUE AND CO. 11 ROWBOTHAM'S (J., F.R.S.A.) DEEIVATIVE SPELLING- BOOK. 12mo. liiup cloth, Is. 6d. GUIDE TO FRENCH CON- VERSATION ; consisting of Modern French Dialogues, with the Pro- nunciation. New Edition, by DE LA VOYE. 18mo. bound, 2s. 6d HOWE'S (C. G., Author of "What Put my Pipe out") GOING TO THE DOGS ; or, the Adventures of Frank. Showing how he was brought up to follow neither Trade nor Profession, and what his very genteel bringing up brought him to. Fcap. boards, 2s. ; cloth, 3s. 6d. ROYAL GEMS FROM THE GALLERIES OF EUROPE; 90 large Steel Fngravings of Pictures of ihe great Masters, with Notices Biographical, Historical, and Descriptive, by S. C. HALL, F.S.A. 2 Vols. Elephant folio, cloth, gilt edges, JE5. RURAL SCENES ; or, a Peep into the Country. A New and Revised Edition, with 88 Cuts. 18mo. cloth, 2s. RUTTER (REV. H.), THE LIFE, DOCTRINE, AND SUF- FERINGS OF OUR BLESSED LORD AND SAVIOUR JESUS CH R1ST. With Introduction by the Very Rev. F. C. HCSENBETH, D.D. Illustrated with 50 Steel Engravings. 4to. cloth, gilt edges, 1 8s. SABINE'S (ROBERT, F.S.A.) THE ELECTRIC TELEGRAPH. With 200 initiations, hvo. cloth, 12s. 6d. SANDFORD AND MERTON. With Illustrations. 18mo. cloth, 2s. SAYER'S (Tnos. A.) AIDS TO MEMORY. A Practical System ot Mnemonics, l^mo., is. SCOTLAND (THE PICTORIAL HISTORY OP). Edited by the Rev. J. TAYLOR, D.D. Illustrated with 79 Engravings on Steel, from Drawings by W. H. BARTLETT, and other Artists. 2 Vols., imperial 8vo., 2 5s. SCULPTURE GALLERY : a Series of 80 Steel Engravings. With Descriptions by JAMES DAFFOKNK, T. K. HARVEY, &c. Folio, cloth, gilt, 3 3s. SHAKSPERE'S COMPLETE WORKS. With Notes and Introduction by H. STEBBING, and Inquiry into the Authenticity of the Shaksptre Portraits. With 51 Steel Engravings. Royal 8vo. cloth, 25s. SHARPS' S NEW TESTAMENT. Translated from GRIESBACH'S Text. Fourth Edition. Fcap. cloth, Is. 6d. SHELMERDINE'S (W.) SELECTION OF THE PSALMS and other Portions of Scripture, arranged and marked for Chanting. Fcap. cloth, Is. ONE HUNDRED AND EIGHTY CHANTS, Ancient and Modern. Selected from the most famous composers, arranged for Four Voices, with Organ and Piaioforte Accompaniment. Crown 8vo. cloth, 2s. 6d. SHO WELL'S HOUSEKEEPER'S ACCOUNT BOOK. 4to., interleaved with blotting paper, 2s. SMEDLEY'S (M. B.) TWICE LOST, LINNET'S TRIAL, and NINA. New and Cheaper Edition. Crown 8vo. boards, 3s. ; cloth, 4s. ' By PRANK E. SMEDLEY. GATHERED LEAVES. A Collection of the Poetical Writings of the late FRANK E. SMEDLEy, Author of " Frank Fairlegh." With a Memorial Preface by EDMUND YATES. Portrait, and numerous humorous designs. Printed on superior paper, with Borders. Imperial 16mo. imitation half-morocco, gilt edges, 8s. 6d. FRANK FAIRLEGH ; or, Scenes from the Life of a Private Pupil. Crown 8vo. boards, 2s. 6d. ; cloth, 3s. 6d. ; or with 30 Illustrations by GKORGK CRUIKSHANK, 8vo. cloth, 10s. 6d. 12 CATALOGUE OF WORKS BY FRANK E. SMEBLEY Continued. HARRY COVERDALE'S COURTSHIP, and What Came of it. Crown 8vo. boards. 2s. 6d.; cloth, 3s. 6d. ; or, with Illustrations by H. K. BROWNE, 8vo. cloth, 10s. 6d. LEWIS ARUNDEL ; or, the Railroad of Life. Crown 8vo. boards, 3s. ; cloth, 4s.; or with Illustrations by H. K. BROWNE (Pfliz), 8vo. cloth, 12s. 6d. THE COLVILLE FAMILY ; with SEVEN TALES BY SEVEN AUTHORS. Edited by the Author of "Frank Fairlegh." Crown 8vo. board <, 2s. 6d. ; or with Frontispiece and Vignette Title by PHIZ, cloth, 3s. 6d. SMYTH'S (WARINGTON W., M.A., F.R.S.) TREATISE ON COAL AND COAL-MINING. Illustrated. Post 8vo. cloth, 7s. 6d. SPURGEON'S (REV. C. H.) THE SAINT AND HIS SAVIOUR ; or, The Progress of the Soul in the Knowledge of Jesus. Second Edition. Fcap. cloth gilt, 3s. 6d. STUDENTS' SERIES OF HANDBOOKS, for School and College Use. 12mo. Uniformly bound in cloth, with red edges. 1. ENGLISH HISTORY. Outlines of the History of Eng- land, more especially with reference to the Origin a-id Progress of the English Constitution. By W. DOUGLAS HAMILTON, of H.M. State Paper Office. Price 6s. 2. GRECIAN HISTORY. Outlines of the History of Greece, in connection with the Rise of the Arts and Civilisation in Europe. By W. D. HAMILTON ami EDWARD LEVIEN, M.A. Price 3s 6d. 3. ROMAN HISTORY. Outlines of the History of Rome, from the Earliest Period to the Christian Era, and the Decline of the Empire. By EDWARD LEVIEV, M.A. With Map. Price 3s. 6d. 4. ENGLISH DICTIONARY, containing more than 100,000 Words, or 50,000 more than any existing Work. By HYDE CLAKKE, D.C.L Price 4s. 6d. 5. FRENCH DICTIONARY. A French-English and Eng- lish -French Dictionary, including Technical Terms. By ALFRED ELWBS. Price 3s. 6d. 6. GERMAN DICTIONARY. A Triglot English -Germ an - French Dictionary. By N. E. 8. A. HAMILTON. Price 4s. 7. ITALIAN DICTIONARY. An Italian-English-French Dictionary, )or Students and Men of Business. By ALFRED EI.WKS. Price Is. 6d. 8. SPANISH DICTIONARY. A Spanish -English and Eng- lish-Spanish Dictionary, including Technical Terms. By ALFKED JELWES. Price 6s. 9. LATIN DICTIONARY. A Latin-English and English- Latin Dictionary. With Appendix of Fiench and Italian Words derived from the Latin. By Rev. THOMAS GOODWIN, M.A. Price 4s. 6d. 10. GREEK LEXICON. A Greek-English and English-Greek Lexicon. By HENRY R. HAMILTON. Price 6s. 11. HEBREW LEXICON. A Hebrew-English and English- Hebrew Lexicon. By Professor M. H. BRESLAC. 2 Vols., price 14s. 12. CHRONOLOGY. Chronology of Civil and Ecclesiastical History, Arts, Literature, and Civilisation, from the Earliest Period to the Peace of Paris, 1856. Price 3s. 6d. 13. STUDIES IN ENGLISH POETRY. With Biographical Sketches, arid Notes Explanatory and Critical. By JOSEPH PAY.NK. Price 6s. PUBLISHED- BY VIRTUE AND CO. 13 STUDENTS' SERIES or HANDBOOKS continued. 14. STUDIES IN ENGLISH PROSE. With Biographical Sketches, and Critical and Explanatory Notes. By JOSEPH PAYNE. Price 5s. 15. OLD TESTAMENT HISTORY. Analysis and Summary of Old Testament History. By J. T \LBO*S WHEELKR, F.R.G.S. Price 5?. 6d. 16. NEW TESTAMENT HISTORY. Analysis and Summary of New Testament History. By J. TALBOY* WHEELER, F.R.G.S. Pi ice 5s. 6d. 17. DICTIONARY OF TECHNICAL TERMS. Price 5s. SWAIN'S (CHARLES, Author of "English Melodies," &c.) ART AND FASHION ; with other Songs, and Poems. Post 8vo. cloth, 7s. 6d. TATE'S (W., Author of "The Modern Cambist") ELEMENTS OF COMMERCIAL ARITHMETIC. New Edition. 12mo. cloth, 2s. 6d. N.B. A Key to the above, price 3s. 6d. TAYLOR'S (JAMES) PICTORIAL HISTORY OF SCOTLAND, from the Roman Invasion to the Close of the Jacobite Rebellion, A.D. 79 1746. By JAM-:S TAYLOR, D.D., assisted by Professors LiNDSAYand EAD;B, Dr. JOHN ANDERSON. G. MACDONALD, &u. With numerous Engravings on Steel. 2 Vols. royal 8vo. cloth, 2 5s. THOMSON'S (A.) CONSOLATION FOR CHRISTIAN MOURNERS. 8vo. cloth, 12s. ; 12mo. cloth, 5s. TOMLINSON'S (CHAS., F.C.S., Lecturer at King's College) EXPE- RIMENTAL ESSAYS 1. On the Motions of Camphor on the Surface of Water. 2. On the Motion of Camphor towards the Light. 3. History of the Modern Theory of Dew. With Illustrations. 12mo. limp cloth, Is. CYCLOPAEDIA OF USEFUL ARTS, Mechanical and Chemical, Manufactures, Mining, and Engineering. Illus- trated by more than 3,000 Engravings on Steel and Wood. New Edition. 3 Vols Royal 8vo. cloth, 3 15s. N.B. The Third Volume of this Work, forming an Appendix to previous Editions, is sold separately, price 21s. TRAILL'S (MRS.) CANADIAN CRUSOES; a Tale of the Rice Lake Plains. Edited by AGNKS STRICKLAND. With numerous Illustrations by HARVEY. Fcap. cloth, 3s. 6d. TREASURE-SEEKER'S DAUGHTER : a Tale of the Days of James the First. By HAHNAH LAWRANCIC. 12mo. boards, Is. TREDGOLD ON THE STEAM ENGINE. In Two Sections : 1. MARINE ENGINES. 2. LOCOMOTIVE AND STATIONARY ENGINKS. 1000 pages of Text, and upwards of 220 Engravings. Also 160 Woodcuts and Diagrams. 3 Vols. royal 4to. cloth, 4 14s. 6d. TURKEYS, GEESE, AND DUCKS, both for Pleasure and Profit. By the Author of "Eggs and Poultry." Fcap., Is. TYTLER'S (M. FRASER) TALES OF MANY LANDS. With Eight Illustrations. New Edition. Fcap. cloth, 3s. 6d. VERNON GALLERY OF BRITISH ART. Edited by S. C. HALL, ESQ. F.S.A. Comprising 152 Engravings of the choicest Pictures in the Collection of Robert Veruon, Esq. 4 Vols. folio, cloth extra, with g'lt edges, 8 8s. VIRTUE'S ILLUSTRATED WORKS-Demy 4to. cloth gilt and gilt edges. AMERICAN LAND, LAKE, AND EIVER SCENERY. 120 Steel Engravings, after Sketches by W. H. BARTLETT. Witli De scriptions by N. P. WILLIS, Author of "Pencillings by the Way." 2 Vols., 1 15s. 14 CATALOGUE OF WORKS VIRTUE'S ILLUSTRATED WORKS. continued. THE BOSPHORUS AND THE DANUBE. The Bos- PHORCS by Miss PARDOE. The DANUBE by W. BKATTIE, M.D. Illus- trated with 168 Engravings, from Drawings by W. H. BARTLETT. 2 Vols., 1 5s. each. CALEDONIA ILLUSTRATED. 170 Steel Engravings, from Drawings by W. H. BARTLETT, T. ALLOM, &c. With Descriptions by WILLIAM BEATTIB, M.D. 2 Vols., 2 10s. CANADIAN SCENERY. 118 Steel Engravings, after Drawings by W. H. BARTLETT. With Descriptions by N. P. WILLIS, Author of" Pencillings by the Way." 2 Vols., 1 15s. IRELAND, ITS SCENERY AND ANTIQUITIES. 120 Engravings by W. H. BARTLETT. With Descriptive Text by J. STIR- LING COYNE, N. P. WILLIS, &c. 2 Vols., 1 15s. PALESTINE (THE CHRISTIAN IN); or, Scenes of Sacred History, Historical and Descriptive, by HENRY STEBBING, D.D., F.R.S. feO steel Engravings, from Drawings taken on the spot by W. H. BART- LETT. 1 6s. PIEDMONT AND ITALY, from the Alps to the Tiber. With 138 Illustrations after Designs by DDDLEY COSTELLO, HARDING, PVNE, BARTLETT, BROCKEDON, &c. 2 Vols., 2 2s. PORTS, HARBOURS, WATERING PLACES, AND COAST SCENERY OF GREAT BRITAIN. A Series of 125 Steel Engravings, from Drawings by HARDING, CRKSWICK, COOKE, BARTLEIT, &c. Edited by W. BEATTIM, M.D. 2 Vols., 2. SCOTLAND. By W. BEATTIE, M.D. Illustrated in a Series of 120 Steel Engravings, alter Drawings by ALLOM, BARTLETT, and M'CuLi.ocH. 2 Vols. 2. SWITZERLAND, Illustrated in 108 SteelEngravings, after Drawings by W. H. BARTLETT. With Descriptions" by W. BEATTIE, M.D. 2 Vols., 2. VIRTUE'S ILLUSTRATED JUVENILE SERIES, fcap. 8vo. extra cloth, price 3s. 6d. per volume. 1. WINTER EVENINGS ; or, Tales of Travellers. By MARIA HACK. With Eiglit Illustrations by GILBERT and HAHVEY. 2. FOXHOLME HALL : a Legend of Christmas ; and other Amusing Tales for Boys. By W. H. G. KINGSTON. Illustrated. 3. THE PIRATE'S TREASURE : a Legend of Panama ; and other Tales for Boys. By W. H. G. KINGSTON. Illustrated. 4. GRECIAN STORIES. By MARIA HACK. New Edition, with Eight illustrations. 5. TALES OF MANY LANDS. By M. ERASER TYTLER. Illustrated. 6. THE CANADIAN CRUSOES : a Tale of the Rice Lake Plains. By Mrs. TRAILL, Edited by Miss STRICKLAND. Illustrated. WATTS' AND RIPPON'S HYMNS. Bound in One Volume, 32mo. roan, embossed, sprinkled edges, 3s., gilt edges, 3s. 6d. WEBB'S (MRS. J. B.) NAOMI; or, the Last Days of Jerusalem. With View, &c. of Jerusalem, and numerous Illustrations by GILBERT and BARTLETT. Nineteenth Edition. Fcap. cloth, Is. 6d. WEDGWOODS (THE) : being a Life of JOSIAH WEDGWOOD, with Notices of his Works and their Productions, Memoirs of the Wedg- wood and other Families, and a History of the Early Potteries of Stafford- shire. By LLEWELLYNN JEWITT, F.S.A., &c. With a Portrait and numerous Illustrations. Printed on toned paper. 8vo. extra cloth, 18s. PUBLISHED BY VIETFE AND CO. 15 WHEELER'S (J. TALBOTS, F.R.G.S.) HISTORICAL GEO- (iRAt'HY OF THE U1.D AND NEW TESTAMENTS. Second Edition. Folio, cloth, 7s. 6d. ANALYSIS AND SUMMARY OF OLD TES- TAMENT HISTORY AMD THE LAWS OF MOSES. Tenth Edition. Post 8vo. cloth, red edges, 5s. 6d. ANALYSIS AND SUMMARY OF NEW TES- TAMENT HISTORY. Seventh Edition, revised. Post 8vo. cloth, red edges, 5s. 6d. POPULAR ABRIDGMENT OF OLD AND NEW TESTAMENT HISTORY. New Edition. 2 Vols. Ihmo. cloth, 2s. each. WILKIE GALLERY. A Collection of 37 Engravings from the best Pictures of the late Sir DAVID WILKIE, R.A., including his SPANISH and ORIENTAL SKETCHES, with Notices Biographical and Critical. Folio, cloth, gilt edges, '6 10s. WILLEMENT (E. E.), FAMILIAR THINGS : their History, &c. Third Edition. Fcap. cloth, 2s. 6d. CONVERSATIONS ON BIRDS AND ANI- MALS. Illustrations. 12mo. cloth, 2s. WOODWARD'S (B. B.) THE HISTORY OF WALES, from the Earliest Times to its Final Incorporation with England. Illustrated with 75 Steel Plates, representing Views of Remarkable Places, Antiquities, and Scenery. 2 Vols., super-royal 8vo. cloth gilt, 1 8s. WORBOISE'S (E. J.) SIR JULIAN'S WIFE. A Tale. Fcap. 5s. ; with gilt edges, 5s. d. THE WIFE'S TRIALS. A Tale. Fourth Edition. Fcap. cloth, 3s. 6d. ; with gilt edges, 4s. THE LIFE OF THOMAS ARNOLD, D.D. Second Edition. Fcap. cloth, 3s. 6d. ; with gilt edges, 4s. CAMPION COURT. A Tale of the Days of the Ejectment Two Hundred Years Ago. Third Edition. Fcap. cloth, 5s. ; with gilt edges, 6s. 6d. THE LILLINGSTONES OF LILLINGSTONE. Fcap. cloth, 5s.; with gilt edges, 5s. 6d. LOTTIE LONSDALE; or, the Chain and its Links. Second Edition. Fcap. cloth, 5s. ; with gilt edges, 5s. 6d. EVELYN'S STORY ; or, Labour and Wait. Fcap. cloth, 6s. ; with gilt edges, 5s. 6d. WRIGHT'S (THOMAS, ESQ., M.A., F.S.A.) A HISTORY OF CARICATURE AND GROTESQUE IN LITERATURE AND ART. With Illustrations by F. W. FAIRHOLT, F.S.A. 4to. cloth, red edges, 21s. -DOMESTIC MANNERS AND SENTIMENTS IN ENGLAND DURING THE MIDDLE AGES. With numerous Illustrations by F. W. FAIRHOLT, F.S.A. Fcap. 4to. cloth, red edges, 2ls. THE CELT, THE ROMAN, AND THE SAXON. A History of the Early Inhabitants of Britain. Illustrated by the Ancient Remains brought to light by recent research. Numerous Engravings. Second Edition, with Additions. Post 8vo. cloth, 12s. YOUNG MAN'S COMPANION. A Compendium of Useful Knowledge, including Geography. Astronomy, History, Biography, Na- tural Philosophy, Commercial Affairs, &c. By J. MAVOR, F.A.S. New Edition, brought down to the Present Day. 9 Engravings on Steel. 8vo. cloth, 15s. YOUNG WOMAN'S COMPANION; or, Female Instructor. A Summary of Useful Knowledge, calculated to form the Intellectual, the Moral, and the Domestic Character. Including Biographies of Illustrious Women, Hints on Education, Domestic Management, Receipts, &c. pp \ T^T?T* P. 1 t 1 il 1 ^ u ' ~ -tv-EjAjj tit, rrom best Authors 41 TTCTPTOT 1 "nTPTTATcr A T?V t. j._\j_ujjvy_L JJ1U-LJ.UJM A.KJL, by N. E. S. A. Hamilton. Vol. 1. English ' German French 1 AO \7"^*l O Ci T~I ^ t , * * 1 y: -'. vol. ^5. German English French 43 Vnl L Tfi-Aiunli T? T i n 1 voi. o. xrencn Jinglish German : Complete in 1 vol. . 3 4 4 6 witn trrammar 44. HEBEEW DICTIONAEY, by Dr. Bresslau. Vol. 1. Hebrew English 6 5 5 6 7 3 v 01. a. xungiisn Jtlebrew 4 Complete, with Grammar, in 2 vols. 120 14 * " (^J\i_A.-M.j\x-AJRr bv Dr BrGSsl^n 1 48. COMPOSITION AND PUNCTUATION by J. Brenan . 49. DERIVATIVE SPELLING BOOK, by J Eowbotham 1 1 1 6 GREEK AND LATIN CLASSICS, With Explanatory Notes in English. * LATIN SERIES. A NoteTb^V DELECi , with Vocabularies and JNotes, by H. Young . . C2ESAE. DeBelloGaUico; Notes by H/Young ' CORNELIUS NEPOS; Notes by H Young . g VIEGIL The Georgics, Bucolics, and Doubtful Poems ; Notes by W. Rushton, M.A., and H. Young i , VIRGIL. JEneid ; Notes by H. Young . Epodes ' and Carm en Seculare, by H.' ' M A " Is. LONDON VIRTUE & CO., 26, IVY LANE. 24 GREEK AND LATIN CLASSICS. 9. TEEENCE. Andria and Heautontimorumenos ; Notes by the Eev. J. Davies, M.A. . . . . . Is. Qd. 10. TERENCE. Adelphi, Hecyra, and Phormio; Notes by the Eev. J. Davies, M.A 2s. TEEENCE. Eunurchus; Notes by J. Davies, M.A. . Is. 14. CICEEO. De Amicitia, de Senectute, and Brutus ; Notes by the Eev. W. B. Smith, M.A. . ... 25. 16. LIYY. Books i., ii., by H. Young . Is. Qd. 16*.LIVY. Books iii., iv., v., by H. Young . Is. 6d. 17. L1VY. Books xxi., xxii., by W. B. Smith, M.A. . Is. Qd. 19. CATULLUS, TIBULLUS, OVID, and PEOPEETIUS, Selections from, by W. Bodham Donne . . . . 2s. 20. SUETONIUS and the later Latin Writers, Selections from, by W. Bodham Donne 2s. GREEK SERIES. 1. A NEW GEEEK DELECTUS, by H. Young . . Is. 2. XENOPHON. Anabasis, i. ii. iii., by H. Young . . Is. 3. XENOPHON. Anabasis, iv. v. vi. vii., by H. Young . Is.' 4. LUCIAN. Select Dialogues, by H. Young . . .Is. 5. HOMEE. Iliad, i. to vi., by T. H. L. Leary, M.A. . 1*. 6d. 6. HOMEE. Iliad, vii. to xii., by T. H. I/. Leary, M.A. . Is. Qd. 7. HOMEE. Iliad, xiii. to xviii., by T. H. L. Leary, M.A. Is. Qd. 8. HOMEE. Iliad, xix. to xxiv., by T. H. L. Leary, M.A. Is. Qd. 9. HOMEE. Odyssey, i. to vi., by T. H. L. Leary, M.A. Is. Qd. 10. HOMEE. Odyssey, vii. to xii., by T. H. L. Leary, M.A. Is. Qd. 11 HOMEE. Odyssey, xiii. to xviii., by T. H. L. Leary, M.A Is. 6rf. 12 HOMEE. Odyssey, xix. to xxiv. ; and Hymns, by T. H. L. Leary, M.A 2s. 13 PLATO. Apologia, Crito, and Phaedo, by J. Davies, M.A. 2s. 14. HEEODOTUS, Books i. ii., by T. H. L. Leary, M.A. Is. Qd. 15. HEEODOTUS, Books iii. iv., by T. H. L. Leary, M.A. Is. Qd. 16. HEEODOTUS, Books v. vi. vii., by T. H. L. Leary, M.A. . . . . '. Is. 6d. 17. HEEODOTUS, Books viii. ix., and Index, by T. H. L. Leary, M.A ! 6& 18. SOPHOCLES. (Edipus Tyrannus, by H. Young . . Is. 20. SOPHOCLES. Antigone, by J. Milner, B.A. . . .2s. 23. EUEIPIDES. Hecuba and Medea, by W. B. Smith, M.A. Is. Qd. 26. EUEIPIDES. Alcestis, by J. Milner, B.A. . . .Is. 30. -ZESCHYLUS. Prometheus Vinctus, by J. Davies, M.A. Is. 32. AESCHYLUS. Septem contra Thebas, by J. Davies, M.A. Is. 40. AEISTOPHAJSES. Acharrienses, by C. S. D. Townshend, M.A l8.6d. 40*. AEISTOPHANES. Eana? . . , . . 41. THUCYDIDES, Book i., by H. Young . . . .Is. LONDON: VIRTUE & CO., 26, IVY LANE. > -^hlatZL. 14 DAY USE RETURN TO DESK FROM WHICH BORROWED LOAN DEPT. This book is due on the last date stamped below, or on the date to which renewed. Renewed books are subject to immediate recall. LIBRARY USB REC'D LD TOR? SEP 9 197$ EEC. CIR. HA * T T r o-i A Kr 11 o General Library LiD 21A 50m-ll, 62 TTfwr;r nt California (D3279slO)476B Berkeley YC 10455 1 M51G967