ANNEX
 
 Ex Libris 
 C. K. OGDEN 
 
 ' 
 
 S\R DAVID SALOMONS BART. 
 
 BROOMHILL, 
 TUNBRIDGE WELLS. 
 
 " 
 
 
 ilA N ' >.X 
 

 
 j 
 
 . 
 
 
 ^
 
 fc 
 
 o 
 
 CO 
 
 Q co 
 
 W 8 
 
 M O 
 
 < O 
 
 P W T3 
 C0 PH <w 
 
 . 
 
 W Xij- 
 
 a
 
 A RESUME 
 
 OF THE 
 
 EARLIER DAYS 
 
 OF 
 
 BY 
 
 WILLOUGHBY SMITH 
 
 READ AT THE EXTRAORDINARY GENERAL MEETING OF THE SOCIETY OF TELEGRAPH 
 
 ENGINEERS AND OF ELECTRICIANS, HELD IN PARIS DURING THE 
 " EXPOSITION INTERNATIONALE D'ELECTRICITE," SEPTEMBER 21ST, 1881.
 
 BATMAN BROTHERS AND LILLY, 
 
 PRINTERS, 
 
 JUTTON HOUSE, FABRISHJDON ROAD, 
 I OX DON, E.G.
 
 A RESUME OF THE EARLIER DAYS OF 
 ELECTRIC TELEGRAPHY. 
 
 BY WILLOUGHBY SMITH. * . 
 
 Read at the Extraordinary General Meeting of the Society of Telegraph 
 Engineers and of Electricians, held in Paris during ike Exposition 
 Internationale d' Electricite, September 2 is/, 1 88 1. 
 
 WHEN it was finally settled that we should hold an 
 Extraordinary General Meeting in Paris on the occasion 
 of the Exposition Internationale d'Electricite, it occurred 
 to me that it would be a fitting time, whilst in the midst 
 of such a wonderful collection of electrical appliances 
 gathered under one roof, for us to pause, look back, and 
 contemplate, for the mutual instruction of all who chose 
 to avail themselves of the opportunity, the way in which 
 such great and successful results have been achieved. 
 Consequently my purpose in this necessarily short paper 
 is to pass in brief review some of the most important 
 incidents, many of which have come under my own 
 personal experience, in connection with the earlier days 
 of Electric Telegraphy. Of all the modern applications 
 of electricity, one of the most important is its adaptation 
 to telegraphic purposes. 
 
 No account of a practical electric telegraph had been 
 published prior to the date of Messrs. Cook and Wheat- 
 stone's patent of June, 1837. I sa Y practical because 
 Ronalds, as early as 1823, published a description of an 
 electrical telegraph of originality and merit ; but as he 
 proposed to use static electricity in connection with a 
 wire enclosed in a glass tube, his telegraph was not 
 adapted for practical use. 
 
 A 2
 
 4 EARLIER DAYS OF 
 
 In the September following the date of their patent, 
 Messrs. Cook and Wheatstone made their first practical 
 experiment on the London and Birmingham Railway, 
 and demonstrated beyond all doubt that as beacon fires 
 and torches had been superseded in 1792 by the aerial or 
 semaphore telegraph, so now the latter would have to 
 succumb to a far more important rival. But incredible 
 as it may appear at the present time, the contest was 
 long and fierce, and not until 1852 did the final defeat 
 arrive, as far as England was concerned, when the 
 Electrical Telegraph superseded the Semaphore which 
 had so long done duty between Liverpool and Holyhead. 
 On June I2th, 1837, Messrs. Cook and Wheatstone 
 patented "improvements in giving signals and sounding 
 alarums in distant places by means of electrical currents 
 transmitted through metallic circuits." This patent in- 
 cluded the five-needle instrument requiring five-line 
 wires ; a four-needle instrument on the same principle, 
 also requiring five-line wires ; methods of insulating and 
 supporting the line wires by covering them with cotton 
 and varnish and placing them in a resinous cement, in 
 troughs or tubes; and also methods for localizing faults 
 in the same. 
 
 On April 1 8th, i838/JVfr. Cook obtained a patent, the 
 -title of which was precisely the same as the first patent, 
 and contained no additional matter worthy of note. 
 
 On the 2ist January, 1840, Messrs. Wheatstone and 
 Cook obtained a patent, the title of which was similar to 
 the two already mentioned, and in it are described a 
 signal apparatus by which the letters of the alphabet are 
 presented at an opening in a dial-plate by means of an 
 electro-magnet in the circuit, which acts upon the pallets 
 of an escapement put in motion by an independent clock- 
 work ; an electric alarum ; a magneto-electric machine 
 to be used in connection with the apparatus above des- 
 cribed, &c.
 
 ELECTRIC TELEGRAPHY. 5 
 
 On the 8th September, 1842, Mr. Cook obtained a 
 patent for stretching, suspending, and insulating wires 
 on posts erected at suitable distances in the manner now 
 so familiar to all of us both across the country, and 
 parallel with railways. This patent also mentions that 
 " the wires so stretched, suspended, and insulated, may 
 be used for one half the circuit, in conjunction with the 
 earth for the other half." 
 
 On the 6th May, 1845, Messrs. Wheatstone and 
 Cook secured a patent for " Improvements in Electrical 
 Telegraphs, and in apparatus relating thereto, part of 
 which improvements are applicable to other purposes." 
 In this patent it is proposed to give audible signals at 
 the same time that visible signals are made by a needle 
 or pointer, a distinct and different sound being given 
 for each direction or deflection of the pointer. This was 
 to be done by either the pointer itself striking a bell, or 
 by alarum mechanism. Also by using the " derived 
 current," a short branch circuit, or what we should now 
 term a shunt, taken from the main wire without disturbing 
 its continuity, to communicate to a sensitive signal 
 apparatus, the signals passing through the main wire. 
 This patent includes eleven other suggestions or im- 
 provements, which I think we need not stop to investigate. 
 
 Who really invented the Electric Telegraph is a 
 problem as impossible to solve definitely, as who invented 
 Railways or steam navigation. I therefore mention 
 these patents simply because soon after the date of the 
 last named, the first Electric Telegraph Company in 
 England was formed for the purpose of working them, 
 more for the general use of the public than Messrs. 
 Cook and Wheatstone had hitherto done, they having 
 principally confined their attention to applying the tele- 
 graph to the use of Railway Companies as an adjunct 
 to their several systems of signalling. 
 
 Although one could loiter and note many points of
 
 6 EARLIER DAYS OF 
 
 interest which occurred in the history of Electric 
 Telegraphy from its commencement in 1845 to ^50, 
 I fear time will not allow of our doing so ; let us 
 therefore pass to the year 1850, and observe what 
 progress had been made to that date. In England the 
 "Electric Telegraph Company" had extended their 
 wires to 2,215 miles, and were using double and single 
 needle instruments, an important modification of the 
 five and four needle instruments patented by Messrs. 
 Cook and Wheatstone in 1837. ^ n Ireland they had not 
 yet realized the value of the Electric Telegraph, for 
 although they had 500 miles of Railway, they had not 
 5 miles of telegraph wire. But in America they had 
 12,000 miles of wire, and were using the Morse, Bain, 
 and House instruments, all on the permanent recording 
 principle. The only reasons I can assign for the more 
 rapid progress which had been made in America, are their 
 low charges and rivalry resulting from competition, as 
 there were no less than twenty different Companies at 
 work, who did not confine their lines to the Railways, but 
 erected them on public roads or private property as 
 occasion required ; whilst in England one Company had 
 the monopoly, and erected their wires on the property 
 of the Railway Companies only; but as there were 5,447 
 miles of railway open at the time, their powers of ex- 
 tension were not limited in that direction. In France 
 there were only 620 miles of wire in operation, suspended 
 similarly to those in England, the instruments used 
 being what were termed revolving pointers. In Prussia 
 they had 2,468 miles of wire, mostly subterranean, and 
 similar instruments were used to those worked by the 
 French Government. 
 
 In America they were also at this date employing 
 the Electric Telegraph in their Meteorological observa- 
 tions, and had -commenced a system of storm warnings 
 along many parts of their coast, similar to that we
 
 ELECTRIC TELEGRAPHY. 7 
 
 are now so familiar with in England. It has been said 
 that to predict an eclipse is an object merely of 
 curiosity, but to predict an approaching storm would be 
 of inconceivable benefit. It is true we have a con- 
 siderable acquaintance with the nature of heat, water, 
 air and electricity ; but when these four ingredients of 
 nature, in a compounded state, are floating round our 
 globe, and producing all those various agitations and 
 combinations known under the general denomination of 
 " weather," we are puzzled and perplexed, and although 
 Electric Telegraphy has done much to enlighten us, still 
 there *is much more to be learnt. No doubt if the in- 
 formation which is being daily recorded by patient 
 observers in all parts of the civilized world were readily 
 and quickly communicated to one centre, great advance- 
 ment in our Meteorological knowledge would soon follow, 
 and I hope the day is not far distant when such a system 
 will be an accomplished fact. 
 
 Electric Telegraphy having made such satisfactory 
 progress, it became absolutely necessary that England 
 (the island centre of the commerce of the world) should 
 no longer remain in her isolated position, but be put into 
 direct telegraphic communication with other nations. 
 Consequently, in 1850, a Company was formed in con- 
 nection with the concession just obtained from the 
 French Government by Mr. Brett for the sole right to 
 establish telegraphic communication between England 
 and France. Mr. Brett had guaranteed that with his 
 instrument, which was a modification of House's step- 
 by-step motion pointer, one wire, and two persons only, 
 the one stationed in France and the other in England, 
 one hundred messages of fifteen words each, printed in 
 clear Roman type, should be sent ready for delivery in 
 one hundred consecutive minutes. That was a bold 
 undertaking on the part of Mr. Brett, considering the 
 various opinions held in those days as to the practic-
 
 8 EARLIER DAYS OF 
 
 ability of ever being able to either lay or work a sub- 
 marine line. 
 
 It was arranged that 25 miles of copper wire '083 of 
 an inch diameter should be covered with gutta percha 
 (at that time a material of comparatively recent impor- 
 tation, the mechanical and electrical properties of which 
 were just developing themselves) to half-an-inch in 
 diameter. The copper was unannealed, and covered to 
 the full gauge in one covering of gutta percha, in one 
 hundred yard lengths. The copper was joined by what 
 is termed a bell-hanger's twist, and then further secured 
 with soft solder. Gutta percha in a plastic state was 
 then applied and pressed into shape in a wooden mould, 
 the diameter of the finished joint being about 2 inches. 
 Of course in those days the joints were not tested separ- 
 ately, nor was any notice taken of the resistance or 
 conductivity of the copper, or the resistance or electro- 
 static capacity of the gutta percha. The coils, when 
 joined into suitable lengths for carriage, were simply 
 immersed in water, at no specific temperature, and if 
 no serious deflection was obtained on a very sluggish 
 vertical Galvanometer from 24 cells of the old form of 
 sand battery, they were pronounced good, and joined to 
 the length which was being coiled on to a large iron reel 
 constructed to hold the 25 miles in one continuous length. 
 This drum or reel was placed amidships on the deck of 
 a steam tug. Owing to the great number and size of 
 the joints, to place the wire on the reel with the required 
 uniformity the inequalities were filled in with cotton 
 waste, and when thought necessary thin wooden laths 
 were laid longitudinally the whole length of the reel, 
 between the layers of the wire. 
 
 Previous to the submersion of the main wire the 
 shore ends were laid, one from a horse-box, which was 
 to serve as the shore station, in the yard of the Railway 
 Company at Dover, to the far end of the staging erected
 
 ELECTRIC TELEGRAPHY. 9 
 
 in connection with the new harbour which was then being 
 constructed. The other was laid a considerable distance 
 beyond low water mark, over the rocks into the light- 
 house on Cape Grisnez. These shore ends consisted of 
 a copper wire '065 diameter covered only with cotton 
 and a solution of India-rubber, and enclosed in a very 
 thick tube of lead. Early on the morning of the 23rd 
 August, 1850. the tug steamed out of Dover Harbour 
 to the end of the shore length. The end of the main 
 wire was then passed into a boat where it was joined 
 to the shore end and the laying of the main wire 
 was commenced, H.M.S. Widgeon leading the way. 
 As every no yards left the tug a leaden weight was 
 attached to the wire, varying from 14 to 24 Ibs., 
 according to the depth of water or nature of the bottom. 
 At first it was attempted to fix the weights without easing 
 the speed of the tug or the egress of the wire, but as 
 each weight was in two halves, and had to be placed on 
 the wire, and then fastened by hammering the two studs 
 of the bottom half which came through corresponding 
 holes in the top half, it was soon found that the two men 
 so employed struck each other more often than they did 
 the desired object, consequently it was found necessary 
 to stop the vessel while each weight was being attached. 
 The weather was all that could have been desired, and 
 with the exception of the confusion caused at times by 
 the lashing effect of the wooden laths as they became 
 partially released from the drum, all went well, and at 
 six the same evening the anchor was dropped from the 
 steam-tug close under the inhospitable-looking shore of 
 Cape Grisnez, and at the position of the buoy attached 
 to the shore end. The end of the main wire was then 
 passed into the cabin, where it was connected to the 
 printing instrument. But it was soon apparent that 
 either the orthography of the sender at Dover was very 
 far from correct, or that the instrument was not working
 
 ro EARLIER DAYS OF 
 
 as it was contemplated it would. Letters in " bold 
 Roman type" were certainly received, but it required 
 great ingenuity to arrange them in their proper order. 
 As the evening was fast approaching, too much time 
 could not be given to the experiments on board, conse- 
 quently the end of the main wire had to be passed into 
 a boat in which the shore end had previously been 
 secured ; there the splice was made and then passed 
 overboard with the sincere wish that the communication 
 that evening effected between England and France 
 might serve to still further establish the two nations in 
 unity and love. All then landed and made the best of 
 their way to the lighthouse on the top of Cape Grisnez, 
 where a single-needle instrument was placed in circuit 
 with Dover, but although many anxious hours were 
 passed in endeavouring to get signals from Dover, none 
 came. The trials of the next day were equally unsuc- 
 cessful, and then it became apparent that the line must 
 have broken down ; and as all the attempts made to 
 lift the wire at or near the splice failed, the line was 
 eventually abandoned. Had the main wire been used 
 instead of the heavy lead-covered shore end, I believe 
 better results would have followed. But criticism can 
 now only be made through the light of experience, which 
 would not be just. Let us rather award all honour to 
 those gentlemen who devoted their money, time, and 
 energy to a practical demonstration of the feasibility of 
 Submarine Telegraphy. From that moment it became 
 simply a question as to the best way to protect the con- 
 ductor and its insulator during the laying and after 
 submersion. 
 
 After careful consideration, it was, in 1851, finally 
 agreed that instead of one wire of the size used in the 
 experimental line of the previous year, four copper wires 
 065 diameter covered separately with two coverings of 
 gutta percha to '25 inch diameter should be twisted
 
 ELECTRIC TELEGRAPHY. n 
 
 together and well-protected with thick iron wires laid 
 spirally around them. Yarns saturated with tar were 
 laid in the interstices of the twisted conductors, and 
 designated the " worming," and the whole covered with 
 similar yarns laid on at right angles and designated the 
 " serving." Twisting the gutta percha wires, the 
 worming, and the serving, were all done at one operation 
 in one machine, and the wire, thus covered, was de- 
 signated the " core." The whole was then passed 
 through the lay-plate of an ordinary wire rope twisting 
 machine, where ten galvanized iron wires "25 diameter 
 were laid round it, and as it passed from the machine it 
 was coiled in the factory yard. The copper conductors 
 were insulated by the Gutta Percha Co., and forwarded 
 in suitable lengths to a small wire-rope manufactory in 
 High Street, Wapping, where it was twisted, wormed, 
 served, and covered with iron wires as described. Owing 
 to litigation as to patent rights, the manufacture of the 
 cable was seriously delayed and not finally completed 
 until the iyth September, 1851. Preparations were at 
 once made to coil the cable in one length on board the 
 Blazer, a hulk lent by H.M. Government, and on the 
 24th September all was coiled, and the Blazer, in charge 
 of two steam-tugs started for the South Foreland, 
 where on the following day one end of the cable was 
 landed and the laying commenced, H.M.S. fearless 
 showing the course, and the Blazer being towed by 
 two tugs with a third tug in attendance. To enumerate 
 the many exciting incidents which occurred on board 
 the Blazer that day would occupy too much of our 
 time. Owing to the unfortunate delays during the 
 manufacture of the cable, the laying was obliged to be 
 undertaken, to save the concession, during the prevalence 
 of equinoctial gales and the period of the strongest tides. 
 Although the weather was fine at starting, it soon 
 changed, and before noon it was blowing a gale, and
 
 12 EARLIER DAYS OF 
 
 the Blazer^ helpless old hulk as she was, rolled and 
 pitched most unmercifully. The contrivances employed 
 to retard the egress of such a heavy cable were insuffi- 
 cient, consequently during the two stoppages which 
 occurred, one, when from some cause on shore there was 
 a total cessation of signals, and the other from the 
 breaking of the tow-rope, much cable was wasted, and 
 eventually the expedition arrived off Sangatte late in 
 the evening with insufficient cable to reach the shore. 
 Never had there been such a continuous struggle between 
 mind and matter as there had been the whole of that 
 eventful day, and it was to be regretted that matter for 
 the time gained the victory. But nothing daunted, a 
 sufficient length of cable to reach the shore was at once 
 ordered ; and in the meantime three wires were twisted 
 together, and, without any protection, laid to the beach, 
 and there connected to the wires already laid to a room 
 at the railway station at Calais. Thus on the 3oth Sep- 
 tember, 1851 (or within a few days of thirty years ago), 
 was telegraphic communication established between 
 England and France. On the igth October the addi- 
 tional length of new cable replaced the three wires, and 
 on the 1 3th November the cable was first used for the 
 transmission of public messages, and with slight inter- 
 ruptions has continued to do so to the present time. 
 
 The possibility of submarine telegraphy having been 
 so satisfactorily demonstrated, it was but natural that 
 many and various schemes should have been immediately 
 devised for its further development. But, unfortunately, 
 owing to the great haste and total disregard of even 
 ordinary precautions, some of the schemes proved lament- 
 able failures. For instance, in the June of the following 
 year an attempt was made to connect England and 
 Ireland by means of a line from Holyhead to Howth. 
 In this instance one gutta percha covered copper wire, 
 similar to those in the Dover and Calais cable, was
 
 ELECTRIC TELEGRAPH?. 13 
 
 covered with No. 8 iron wire for shore ends, then with 
 twelve No. 14 iron wires for a few miles, but the greater 
 portion of the core was only partially covered, bird-cage 
 fashion, with six No. 14 galvanised iron wires, consequently 
 through the core having no serving or hemp covering, 
 the gutta percha was much exposed. When completed 
 it was coiled on trucks and taken by railway to White- 
 haven ; but, as the steamer which was to receive it could 
 not enter the harbour there, it had to be re-coiled and 
 forwarded to Maryport, where it was coiled on board a 
 paddle steamer and taken to Holyhead. There it was 
 electrically tested for the first time, and found to be 
 faulty. As it had never been immersed in water, and 
 was then perfectly dry in the holds of the steamer, of 
 course only faults of a very serious character could be 
 detected and rectified, which was done at the cost of 
 much time and trouble. Had it been covered with 
 water, its imperfect electrical condition would have been 
 discovered, and most likely it would not have been laid 
 at all. All attempts to repair this line after it was laid 
 proved fruitless, for the simple reason that the necessary 
 strain to lift it drew the iron wires together, cutting 
 through the gutta percha as they forced the core from 
 the centre. 
 
 Later in the same year an attempt was made to lay 
 from Portpatrick to Donaghadee a three-strand hemp 
 rope about two inches in diameter, in the interstices of 
 which was laid a copper wire which had first been covered 
 with india rubber, and then with gutta percha. But, 
 owing to the inadequate means provided for sinking such 
 a rope, the strong currents carried it far from the pro- 
 posed course, and eventually the attempt was abandoned. 
 
 Later still in the same year another attempt was made 
 between the same points, with a cable similar in con- 
 struction to that 'laid in the previous year across the 
 English Channel, but owing to unfavorable weather and
 
 14 EARLIER DAYS OF 
 
 inadequate means for stopping, or even retarding, the 
 egress of such a heavy cable, the length provided proved 
 too short, and eventually this expedition returned also un- 
 successful. Thus had three cables been lost in attempting 
 to electrically connect England and Ireland, but in the 
 following year a heavy multiple cable was successfully 
 laid from Portpatrick to Donaghadee. Four light cables 
 were also laid in that year between England and Holland, 
 and a heavy multiple one between Dover and Ostend. 
 
 The year 1853 was also notable for the large amount 
 of subterranean wires laid in England by several of the 
 then existing Telegraph Companies ; but as I went some- 
 what into detail on that subject in my paper on " Under- 
 ground Wires," which I read before the members of 
 this Society on the 28th March, 1877, an< ^ which was 
 published in Vol. VI. of the Journal of the Society's 
 proceedings, it is unnecessary for us to go over the 
 same ground again. 
 
 From 1851 to 1856 there had been successfully laid, 
 not in connection with England only, but in various 
 parts of the world, about 1 600 miles of cable, and it was 
 thought that sufficient experience had been obtained to 
 bring to a successful issue the long-cherished scheme of 
 laying a cable from Ireland to Newfoundland, and thus 
 bring America into direct telegraphic communication 
 with England. Soundings were taken, and it was 
 asserted that although rough and jagged rocks did pro- 
 ject high above the bottom, still there was a level plateau, 
 or smooth path, between the two islands, as though 
 nature, anticipating the wants of man, had prepared a 
 resting place for telegraph cables, where, if properly 
 placed, they would lie in perfect peace, and where the 
 ooze, which it was alleged was constantly descending 
 through the water, resembling the fall of snow through 
 our atmosphere, would effectually cover and protect the 
 cable to the end of time.
 
 ELECTRIC TELEGRAPHY. 15 
 
 Experiments were made with long lengths of sub- 
 marine and subterranean wires, and the results stated 
 to be highly promising as to a remunerative speed being 
 obtained through an Atlantic cable 2000 miles in length, 
 using a comparatively small conductor. The mechanical 
 properties of various specimens of what were considered 
 suitable cables were fully tested, and finally the choice 
 fell on one in which the conductor consisted of a seven 
 wire strand of copper, weighing 107 Ibs. per knot, trebly 
 covered with gutta percha, the weight of which was 
 261 Ibs. per knot, the external diameter being f of an 
 inch. Around the gutta percha were spirally laid hemp 
 yarns saturated with a composition of tar, pitch, and 
 tallow, over which were also spirally laid eighteen iron 
 wire strands, each strand being composed of seven wires 
 of the same diameter as those used in the conductor. 
 The weight in air of the completed cable was 20 cwt, 
 and in water 13*4 cwt. per knot, its breaking-strain being 
 about 3*25 tons. The final selection of this cable was 
 not made until late in the year 1856, and as it was 
 stipulated that 2500 miles must be shipped and ready 
 for laying in the June of the following year, it allowed 
 only six months for all the work to be done. In that 
 short space of time 116 tons of copper had to be pre- 
 pared and drawn into 17,500 miles of wire, which had 
 to be twisted into 2500 miles of strand; 250 tons of 
 gutta percha had to be what is termed " manufactured," 
 and placed around the conductor in three separate 
 coverings; 1687 tons of charcoal iron had to be specially 
 prepared and drawn into 315,000 miles of wire, from 
 which 45,000 miles of strand had to be twisted, and then 
 laid around the core. Add to all this the fact that ships had 
 to be selected and got ready to receive the cable, machines 
 both for manufacturing and laying had to be constructed, 
 and no surprise need be expressed that experience soon 
 taught that so short a time was totally inadequate for
 
 16 EARLIER DAYS OF 
 
 such an important work to be executed in a proper and 
 efficient manner. 
 
 Lengths of the core varying from 1600 to 2400 yards 
 were immersed in water at no fixed temperature, and 
 there tested for insulation with 504 cells of the old form 
 of sand battery, and a horizontal galvanometer. If the 
 deflections of the coils of each batch were constantly 
 uniform, their electrical condition was certified, and each 
 coil, wound on a wooden reel, was forwarded to the cable 
 factory. 
 
 Soon after the commencement of the manufacture of 
 the cable, frequent want of continuity in the conductor 
 occurred through imperfections in the copper, from im- 
 proper annealing and other causes. Unfortunately 
 several hundred miles of the core were manufactured be- 
 fore the machines were ready to cover the same with the 
 serving and iron strands ; the fatal edict therefore went 
 forth that all the core in stock should be submitted to 
 tensile strain as it passed over a barrel of small 
 diameter, to develop, if possible, any defects in the 
 conductor before it passed the wire-covering machine. 
 Now when a wire or strand of copper covered with gutta 
 percha is stretched, the metal becomes permanently 
 elongated and holds the elastic material in that position 
 until a favourable opportunity occurs for the gutta 
 percha to return to its normal state, distorting the wire 
 in so doing. Unfortunately it was not long before such 
 an opportunity occurred ; the cable as manufactured was 
 coiled, not in tanks where it could be kept under water, 
 but where it was exposed to the powerful effects of the 
 sun's rays. The consequence was that the gutta percha 
 became plastic in parts, and the conductor, being forced 
 out, came into direct contact with the serving, which, being 
 dry, acted the part of an insulator, and thus prevented 
 detection while the cable was in a dry state. It was to 
 be regretted that the manufacture of the cable was too
 
 ELECTRIC TELEGRAPHY. 17 
 
 far advanced for advantage to be taken of Sir William 
 Thomson's discovery of the great variation in the elec- 
 trical conductivity of commercial copper, especially in 
 that then being used for the conductor of the Atlantic 
 cable. The importance of that discovery has since been 
 fully realised, not only in every branch of electrical 
 science, but by also enabling a purer copper to be 
 obtained, for whatever purpose required ; for, as is now 
 generally known, the higher its electrical conductivity 
 the purer the copper. 
 
 It was not until the 5th of August that one end of 
 the cable was landed at Valentia and the laying of the 
 cable commenced ; but soon after starting, difficulties 
 began to manifest themselves, and when 335 miles were 
 laid the cable broke just astern of the paying-out 
 machine, and the expedition was abandoned for that 
 year. The remaining portion of the cable was coiled in 
 the dockyard at Plymouth, where it was overhauled, but 
 not immersed, and passed through a hot insulating mix- 
 ture, then spliced to 700 miles of new cable, making the 
 total length about 3000 miles. 
 
 In the following June, the cable having been re- 
 coiled, half on board the Niagara and half on board 
 the Agamemnon, various experiments were made in the 
 Bay of Biscay, the results of which were considered 
 satisfactory to the proposal of splicing the cable in mid- 
 ocean and both ships paying out cable at the same time, 
 one towards Ireland and the other towards Newfound- 
 land. On the 26th of June the splice was made in mid- 
 ocean, but after paying out three miles from each ship 
 the cable broke at the stern of the Niagara. The 
 ships at once returned to their former positions and re- 
 spliced, but when about forty miles had been laid from 
 each ship a total cessation of signals occurred, and those 
 on board each ship supposing an accident had happened 
 on the other (an incident which shows one of the disad-
 
 j 8 EARLIER DAYS OF 
 
 vantages of this mode of laying cables) they each returned 
 to the position where the splice had been made, and on 
 meeting anxiously inquired one of the other, " How did 
 the cable part ?" The cause of rupture this time proved 
 to be a profound mystery, for all was proceeding satisfac- 
 torily on both ships when the apparent want of continuity 
 occurred. The cable was again spliced, and about 200 
 miles laid when it broke close to the stern of the 
 Agamemnon, and the ships returned to Ireland. 
 
 On the 2Qth July they returned again to mid-ocean 
 and spliced the cable there one 2 more, but t had not 
 proceeded far in paying out before there was reported a 
 want of continuity. Fortunately they continued to pay 
 out slowly from both ships, and after two hours' anxious 
 suspense the signals reappeared as mysteriously as they 
 had previously disappeared. Such incidents frequently 
 occurred again, but no actual rupture took place, and on 
 the 5th August each ship had completed its task, the 
 Niagara having laid 1030, and the Agamemnon 1020 
 knots. It was true the cable was laid, but its electrical 
 state must have been in a most deplorable condition, for 
 nothing but what was analogous to brute force could 
 produce spasmodic, emaciated shadows of the intended 
 signals, and in twenty-four days it was unable to produce 
 those results, the whole being pronounced a failure. It 
 was a failure in one sense of the term, but it should not 
 be forgotten that much valuable experience had been 
 obtained, and that the problem was solved as to the 
 feasibility of laying and working an Atlantic cable. 
 
 About this period, time began to develop the fact 
 that the system of using such a strong solvent of gutta 
 percha as coal-tar naphtha between each separate cover- 
 ing, with a view to causing more perfect adhesion, had 
 very prejudicial effects on the gutta percha, consequently 
 I began to experiment with a view to obtaining an adhesive 
 solution or compound which would in no way injuriously
 
 ELECTRIC TELEGRAPHY. 19 
 
 affect the gutta percha, and soon found that suitable 
 proportions of Stockholm tar, resin, and gutta percha 
 produced a compound suitable for the purpose. Not only 
 did this compound possess good adhesive qualities, but it 
 was also a good insulator, and when applied between 
 each alternate covering of gutta percha, improved the 
 insulation three-fold. This compound has done and is 
 still doing good service, and its qualities are so well known 
 that no further notice need be taken of it here. 
 
 It was about this time a cause for surprise, and regret 
 that the English government should have financially 
 connected itself with such a gigantic scheme as that of 
 laying a cable from Egypt to India while the success or 
 failure of the Atlantic cable was still in the balance. Had 
 they awaited the result of the Atlantic expedition perhaps 
 the lamentable failure which so soon followed might have 
 been avoided. The cable of the Red Sea and Indian 
 Telegraph Co. consisted of a core containing 180 Ibs. 
 of copper, in the form of a seven wire strand doubly 
 covered with 212 Ibs. of gutta percha per mile ; served 
 with tarred yarns covered with eighteen iron wires. The 
 gross weight per knot was about twenty-one cwt. It was 
 laid in six sections, from Suez to Kurrachee. To give 
 the history of the manufacture, laying, and ultimate failure 
 in working would only be to repeat, with unsatisfactory 
 additions, what has already been said concerning the 
 Atlantic cable. Suffice it to say that the 4000 miles were 
 manufactured and laid by the contractor in the short space 
 of fifteen months, and during the same time the same 
 contractor was experimenting in laying hemp-covered 
 cables in the Levant, where about 800 miles were lost. 
 
 These serious failures, including as they did the loss 
 of about 8000 miles of cable, in so short a time, naturally 
 suggested a thorough investigation of the whole matter ; 
 therefore a joint-committee appointed by the Lords of 
 the Committee of Privy Council for Trade and the 
 
 B 2
 
 20 EARLIER DAYS OF 
 
 Atlantic Telegraph Co., was formed to inquire into the 
 construction of submarine cables ; and I believe the 
 arduous labours of that committee did much for the 
 ultimate success of submarine telegraphy. 
 
 In June, "1859, the English government ordered 
 1 200 nautical miles of cable to be laid between Falmouth 
 and Gibraltar. Engineers and electricians were ap- 
 pointed to inspect the manufacture of this cable. The 
 core consisted of equal weights of a seven-wire copper 
 strand trebly- covered with gatta percha and compound, 
 weighing altogether 800 Ibs. per mile. This core was 
 covered with tarred yarns and solid iron wires in the 
 ordinary way. Instruments were practically employed 
 for measuring the electrical qualities of the copper and 
 gutta percha at a uniform temperature, and the results 
 were recorded in electrical units. That is to say, the 
 specific resistance of the copper and gutta percha was 
 so noted that the measurements would be comparable 
 with the results obtained during manufacture and after 
 submersion. The core in suitable lengths was submitted 
 to a pressure of 600 Ibs. to the square inch, and it was 
 invariably found that the resistance of the gutta percha 
 increased when so. circumstanced. The gutta percha on 
 one of the coils having been by accident pierced through 
 to the conductor while placing it in the pressure tank, it 
 was found that the water was forced along the strand 
 and out of the ends which were brought through stuffing- 
 boxes for testing purposes. To obviate this in future, I 
 had the centre wire of the strand passed through the 
 compound before the other six wires were laid around it. 
 This had the desired effect, as water at a pressure of 
 1000 Ibs. per square inch could not be forced through 
 six inches of a core containing a strand so constructed. 
 This cable was also coiled into watertanks as manufac- 
 tured ; but the ships engaged to lay it were not fitted 
 with tanks, consequently soon after shipment of the first
 
 ELECTRIC TELEGRAPHY. 21 
 
 section an increase in the resistance of the conductor 
 denoted an increase in the temperature of the cable, 
 which was found to arise from fermentation, but whether 
 caused by the hemp serving- or the oxidation of the iron, 
 opinions differed. By thermo-electric measurements the 
 temperature varied from 84 to 62 Fahr. in different 
 parts of the same coil, while the extreme temperature 
 kept uniform at 60. It k was thought that the high 
 temperature might have permanently injured the gutta 
 percha, but a length of core taken from the hottest part 
 showed no signs of deterioration. 
 
 Owing to the delay of the ships the season was too 
 far advanced to lay it where first intended ; it was there- 
 fore arranged to lay it between Rangoon and Singapore. 
 But the ship carrying the first portion came to grief in 
 the English Channel, and eventually it was laid in three 
 sections between Malta and Alexandria, where, owing to 
 the shallowness of the water in which parts of it were 
 laid, and the light character of its construction, it was 
 frequently out of order, and had to be superseded by a 
 cable which was laid direct from Malta to Alexandria 
 in 1868. 
 
 In December, 1860, 1 first called attention to the fact 
 that mechanical injuries to the core of a cable, from 
 accident or otherwise, were often concealed by the in- 
 sulating properties of the tarred serving, and that 
 according to the nature of the fault and the amount of 
 tar so was the period of the concealment a fact sufficient 
 to account for the failure of insulation of so many cables 
 after they had been laid in apparently perfect condition. 
 To obviate this I advised that the serving should be 
 saturated with a conducting instead of an insulating 
 preservative fluid, and that the core should be always 
 kept in water and passed direct from the same to the 
 iron-covering machine, where the presence of the wires 
 at the lay plate would compress the serving, and force
 
 22 EARLIER DAYS OF 
 
 the fluid into any defect in the gutta percha, and imme- 
 diately develop the same. The suggestion was readily 
 adopted, and soon proved effective ; since then tarred 
 serving has been the exception, not the rule. 
 
 On the completion of the first Malta and Alexandria 
 cable, the Government ordered one to be laid in the 
 Persian Gulf; but before deciding on the form of cable, 
 a great many experiments -were made, especially on the 
 electrical and mechanical properties of various cores 
 consisting of gutta percha or india-rubber, pure and 
 simple, or compounded with other substances. Eventually 
 the one selected was to consist of a segmental copper 
 wire, weighing 225 Ibs. per knot, insulated by four cover- 
 ings of gutta percha, with the compound already referred 
 to applied in the ordinary way, weighing 275 Ibs. per 
 knot, the total weight of the core being 500 Ibs. per knot. 
 A strand and solid wire, of the same copper and of equal 
 weights, under the same conditions, conduct precisely the 
 same ; but in lateral induction there would be a slight 
 disadvantage in using a strand, and it was therefore to 
 obtain the mechanical advantages of the strand and the 
 electrical advantages of a solid that a segmental wire was 
 proposed, which was constructed of four copper wires, 
 each forming the segment of a circle, drawn into one 
 copper tube ; but the mechanical difficulties in the 
 manufacture were so great that it was abandoned for 
 a solid wire before the completion of the cable. The 
 electrical qualities of the core were tested in suitable 
 lengths at a uniform temperature of 75 Fahr., while 
 under a pressure of twenty atmospheres per square inch, 
 and the electrical resistance of the gutta percha invariably 
 increased about 40% under such conditions ; but returned 
 to its normal resistance when the pressure was removed. 
 
 The now well-known accumulation method was for the 
 first time applied to test the joints in this core. The 
 serving was saturated with tan water instead of tar,
 
 ELECTRIC TELEGRAPHY. 23 
 
 tallow, etc., as heretofore used, and the served core was 
 coiled in watertanks, Twelve iron wires of the required 
 diameter to envelop the serving- when laid helically 
 around it were then covered with two layers of yarns 
 wound in opposite directions. 
 
 r A hot bituminous compound, composed of Stockholm 
 tar, ground silica, and pitch, was applied with each covering 
 of yarn ; the whole passing through powerful compressed 
 grooved rollers, which gave it a smooth black surface. 
 As manufactured it was coiled into water-tight tanks, and 
 under the same conditions the 1450 knots were transported 
 in four sailing vessels and laid in four sections between 
 Kurrachee and Faa, the intermediate stations being 
 Gwadir, Mussendom, and Bushire. The very satisfactory 
 results on the completion of this cable certainly com- 
 pensated for the great amount of care and attention 
 which had been bestowed on its selection, manufacture, 
 and laying. 
 
 During the time occupied in the manufacture and 
 laying of the last mentioned cable two attempts to lay 
 specially constructed cables were made in the Mediter- 
 ranean. It had been proved, by a series of well con- 
 ducted experiments, that if hemp yarns were placed 
 around an iron or steel wire at a certain angle, and in 
 certain proportions, the breaking strain of the two would 
 be greater than the sum of their separate breaking 
 strains; consequently an ordinary gutta percha core was 
 surrounded with wires so covered with hemp to be laid 
 between Toulon and Algiers, but owing to the " springy " 
 character of the cable, and the incompleteness of the 
 paying-out arrangements, the cable broke while being 
 laid, and having to manufacture new cable to replace 
 the portion lost, it took twelve months to complete the 
 undertaking. The other experiment was with 130 knots 
 of an ordinary gutta percha core, covered with yarns 
 saturated with a conducting fluid, put on at a uniform
 
 24 EARLIER DAYS OF 
 
 tension, with a slight twist, and then covered with a con- 
 tinuous tubular metallic casing, which was applied by 
 winding a flat ribbon of copper in a helical direction 
 around it in such a manner that the edge of each con- 
 volution overlapped that of the preceding one. The 
 whole length was coiled round a cone on a revolving 
 disc, or turn-table, fixed in the hold of a steamer, and 
 attempts were made to submerge it between Oran and 
 Carthagena, but its construction proved totally unsuited 
 for the purpose, as did also the arrangements of the 
 revolving disc or table, and the expedition was abandoned 
 after having lost about half the cable. 
 
 Notwithstanding these and other failures, sufficient 
 had been successfully accomplished to warrant another 
 attempt to connect Ireland and Newfoundland, for dur- 
 ing the five years which had elapsed since the last attempt 
 great improvements had been made in every branch of 
 submarine telegraphy. It was accordingly decided to 
 manufacture 2300 knots of cable, consisting of a core 
 containing a seven-wire strand of copper weighing 300 
 Ibs. per knot, covered with four coverings of gutta percha 
 and four of compound applied alternately to a diameter 
 of '464 and weighing 400 Ibs. per knot, the total weight 
 of the core being 700 Ibs. per knot. The serving was 
 ordinary hemp saturated with tan liquor. The external 
 protection consisted of ten solid iron wires, each sur- 
 rounded separately with Manilla yarns covered with a 
 preservative compound of tar, India-rubber, and pitch. 
 The whole cable was coiled in three separate tanks on 
 board that marvellous conception of Brunei, the Great 
 Eastern, and the expedition started in the summer of 
 1865. From the extraordinary care, even to the minutest 
 detail, bestowed on all connected with this cable, it was 
 believed that success would be the reward; but during 
 the manufacture, coiling on board, and the laying, "faults 
 occurred in the insulation which, when found, suggested
 
 ELECTRIC TELEGRAPHY. 25 
 
 the horrible thought that the hands of some evil-disposed 
 person or persons had been at work, as there was the un- 
 deniable fact that some blunt instrument had been forced 
 between the wires of the outer covering, through the 
 gutta percha to the conductor, and then withdrawn. One 
 such fault occurred when seventy-nine knots had been 
 laid, and ten knots had to be picked up to remove the 
 same. Another similar fault developed itself when 750 
 knots had been submerged, and three knots had to be 
 picked up from 2050 fathoms to remove it. Another 
 fault appeared in 2200 fathoms, when 1214 knots had 
 been successfully laid; but this was never found, for after 
 picking up three knots the cable parted, and the end of 
 the sea portion was lost. After various unsuccessful 
 attempts to recover it the expedition returned home, 
 the season being too far advanced to allow of any further 
 attempts being made that year. 
 
 It was generally admitted that had not those 
 mysterious faults occurred in the way and at the time 
 they did, there would have been no difficulty in success- 
 fully laying the cable ; and therefore it was decided that 
 1 600 knots of new cable should be manufactured, and in 
 the following year (1866) another trial be made to connect 
 Valentia and Newfoundland, and that then an attempt 
 should be made to pick up the lost end of the previous 
 year's cable, and complete the laying of that cable also. 
 The 1600 knots of new cable differed somewhat from the 
 former one. The iron wires were more pliable, were 
 galvanised, and no compound used on them or their 
 jute covering. 
 
 The core of these cables was not submitted to pressure 
 during the electrical tests, for in no instance had pressure 
 developed faults ; but actual experiments proved that 
 faults which would be detected under a pressure of one 
 atmosphere, became more or less concealed according 
 to the amount of pressure applied. The results of an
 
 26 EARLIER DAYS OF 
 
 experiment with a faulty coil is given in the following 
 table, from which you will more easily see the proportion 
 in which the resistance increased with an increase of 
 pressure. 
 
 i atmosphere = 49 meg-ohms per knot. 
 
 J 7'5 = 5< 
 
 35 > == 53 
 
 52-5 = 54 
 
 70 = 55 
 
 8'/*5 ,. = 57 
 
 105 = 58 
 
 122-5 = 58 
 
 140 = 63 
 
 i57'5 = 64 
 
 175 = QO 
 
 183-5 . "5 
 
 On removal of the pressure the resistance suddenly 
 increased to 178, then gradually began to fall, and at 
 the end of seventeen hours had returned to its normal 
 condition, or 49 megohms. If the pressure be allowed to 
 decrease very gradually, it is very interesting to note the 
 gradual increase in the resistance until the pressure is 
 entirely removed, and then the resistance begins to 
 decrease ; but if, as stated, the pressure be suddenly 
 removed, the resistance as suddenly increases. 
 
 Owing to the hardness of the iron wire used in the 
 cable of 1865, its adhesive compound, and the long time 
 it had been lying coiled in the tanks on board the Great 
 Eastern, several exciting scenes occurred while sub- 
 merging that portion, from the layers of cable in the 
 tank sticking together and becoming entangled, thus 
 causing what are technically known as " foul -flakes," 
 but with these exceptions nothing could have progressed 
 more favourably than did everything connected with the
 
 ELECTRIC TELEGRAPHY. 27 
 
 work ; and after fourteen days and nights of care and 
 anxiety, all engaged had the satisfaction of knowing that 
 their labours had been crowned with success. After 
 necessary preparations the expedition started to recover 
 the lost end of the 1865 cable, and after twenty days 
 and nights' hard struggle with the elements, the end was 
 recovered. After speaking with Valentia and ascertain- 
 ing the electrical condition of the whole length to be 
 perfect, the end was spliced to the cable on board, and 
 paying out towards Newfoundland commenced. All 
 proceeded well until within a few hours of completing 
 the laying, when at six a.m., while receiving from 
 Valentia a summary of the news from The Times of that 
 morning, a sudden fault in insulation occurred. This 
 was soon found to be on board, so the cable was cut and 
 respliced, and to the credit of all be it recorded that in 
 less than two hours the fault had been removed and 
 paying out recommenced. At five p.m. the same day 
 the laying of this cable was also successfully completed. 
 It was fortunate that the fault of the morning had been 
 detected before it passed out of the ship, for on careful 
 examination it showed that one end of a broken wire in 
 the top coil had by some means got bent at right angles 
 and had entered the coil beneath, and the pressure caused 
 by one of the tank men stepping on it forced the wire 
 through the gutta percha to the conductor. As this 
 fault was precisely of the same character, and occurred 
 in a portion of the same cable, as those of the previous 
 year, it was but fair to assume that they were caused in 
 a similar way, and not maliciously, as at first supposed. 
 The particulars of my system for testing and working 
 these cables, which I employed during the laying and 
 after submersion, I gave in the paper which I read at 
 this society's meeting on the I2th February, 1879, 
 and which is published in the journal of our proceedings, 
 and therefore I need not notice it further here. No
 
 28 EARLIER DAYS OF ELECTRIC TELEGRAPHY. 
 
 doubt the laying of the one cable, and the recovery and 
 completion of the other, were the crowning achievements 
 in submarine telegraphy, and gave to that branch of 
 our science an impetus that knew no bounds until all 
 civilised countries were connected telegraphically as one 
 nation. At the present there are at work in all no less 
 than 70,000 miles of submarine telegraph cables. Thirty- 
 three years' experience has taught us that neither the 
 electrical nor mechanical qualities of either gutta percha 
 or copper deteriorate by long working or submersion, 
 and also that the iron wires, when properly protected 
 from oxidation, retain all their original qualities; con- 
 sequently the best form of a submarine telegraph cable 
 will be that in which these conditions are fulfilled. 
 
 The progress in the perfection of all instruments and 
 appliances connected with electric telegraphy is so well 
 shown in the Exhibition which has brought us together 
 here that I need not refer to them now. Although I 
 have, I fear, encroached somewhat too long upon your 
 time, I feel sure you will in conclusion allow me to 
 embrace this fitting opportunity of acknowledging the 
 uniform kindness I have always experienced in this 
 country, whenever I have visited it, upon either business 
 or pleasure. My first visit was when I landed at Cape 
 Grisnez on the completion of the experimental line in 
 1850, after an exciting day's work and eighteen hours' 
 fasting ; I have at the present time a vivid recollection 
 of the kindness of the lighthouse keeper, who not only 
 shared with me his frugal meal, but rendered all the 
 assistance he could to my personal wants.
 
 RESUME DES 
 
 PREMIERS JOURS 
 
 DE LA 
 
 PAR 
 
 WILLOUGHBY SMITH. 
 
 LU A LA REUNION GENERALE DE LA SOCIETE D'INGENIEURS DE TtfLE'GRAPHE 
 
 ET D'ELECTRICIENS, ftUI SE TENAIT A PARIS PENDANT 
 "L'EXPOSITION INTERNATIONALE D'ELECTRICITE," LE 21 SEPTEMBRE, 1881.
 
 RESUME DES PREMIERS JOURS DE LA 
 TELEGRAPHIE ELECTRIQUE. 
 
 PAR WILLOUGHBY SMITH. 
 
 Lu a la Reunion generate de la Societe, d' Ingenieurs de Telegraphs et 
 d'Electriciens'quise tenait a Paris pendant V Exposition Internationale 
 d' Elcctricite, le 2 1 Septembre, 1881. 
 
 LA decision etant prise de tenir une Reunion Extraor- 
 dinaire a Paris a 1'epoque de 1' Exposition Internationale 
 d'Electricite, j'ai cru devoir profiter de cette occasion 
 favorable, au milieu d'un recueil si merveilleux d'appa- 
 reils electriques sous le meme toit, pour contempler la 
 maniere dont nous avons realise de si beaux resultats, 
 ayant pour but 1' instruction de tous ceux qui s'y inter- 
 essent. Je me suis done propose dans ce traite, neces- 
 sairement court, de passer en revue quelques-uns des 
 evenements les plus remarquables des premiers jours de 
 la telegraphic electrique auxquels j'ai eu plus particu- 
 lierement affaire. Parmi les applications modernes de 
 1'electricite, 1'une des principal es est son adaptation 
 au telegraphe. 
 
 Aucun recit de la pratique du telegraphe electrique 
 n'avait ete public avant 1'epoque du brevet d' invention 
 de MM. Cook et Wheatstone au mois de juin 1837. 
 Je parle de la pratique, car M. Ronalds publia en 1823 
 meme, un expose d'un telegraphe electrique original 
 et meritoire ; mais comme il se proposa de se servir de 
 1'electricite a la statique avec un fil de metal que ren- 
 fermait un tuyau en verre, son telegraphe ne fut pas 
 approprie a 1' usage ordinaire.
 
 32 LES PREMIERS JO URS 
 
 Au mois de septembre 1837, MM. Cook et Wheat- 
 stone firent leur premiere experience pratique au chemin 
 de fer de Londres a Birmingham, et demontrerent d'une 
 maniere frappante que le telegraphe aerien ou sema- 
 phore devrait ceder a un rival plus formidable, ainsi 
 qu'il a lui-meme remplace le fanal et le flambeau 
 d'autrefois. Mais quelque incroyable qu'il semble aujour- 
 d'hui, la lutte fut longue et acharnee, et ce n'etait qu'en 
 1852 que la derniere defaite arriva, du moms a 1'egard 
 dePAngleterre, quandle Telegraphe electrique rempla<?a 
 le semaphore dont on s'etait servi pendant si longtemps 
 entre Liverpool et Holyhead. 
 
 Le 12 juin 1837 MM. Cook et Wheatstone breve- 
 terent " Certains amendements du systeme des signaux 
 et des alarmes electriques en lieux eloignes, moyennant - 
 des courants electriques conduits a travers des circuits 
 metalliques." Ce brevet comprenait 1' instrument a cinq 
 aiguilles qui demandait des fils de metal a cinq lignes ; 
 un instrument a quatre aiguilles, sur le meme principe 
 et demandant aussi des fils a cinq lignes ; des moyens 
 d'isoler et de soutenir les lignes de fil metallique en les 
 couvrant de coton et de vernis, et en les pla^ant dans 
 un ciment collant en auges ou tuyaux; et enfin des 
 moyens d'y localiser des defauts. 
 
 Le 1 8 avril 1838, M, Cook prit un brevet au meme 
 titre que le premier et qui ne contenait point de matiere 
 additionnelle digne de remarque. 
 
 Le 21 Janvier, 1840, MM. Wheatstone et Cook obtin- 
 rent un brevet, le titre duquel ressemblait aux deux deja 
 nommes. Us y decrivent un appareil de signaux par 
 lequel les lettres de 1' alphabet se presentent a une ouver- 
 ture dans un cadran moyennant un electro-aimant dans 
 le circuit, qui agit sur les palettes d'un echappement que 
 Ton met enmouvementa 1'aided'horlogerieindependante; 
 une alarme electrique ; une machine magneto-electrique 
 qu'il faut employer avec T appareil decrit ci-dessus, etc.
 
 DE LA TELEGRAPHIE ELECTRIQUE. 33 
 
 Le 8 septembre, 1842, M. Cook obtint un brevet pour 
 tirer, suspendre et isoler des fils de metal sur des colonnes 
 placees a quelque distance les unes des autres de la 
 maniere si familiere aujourd'hui, tant a travers le pays 
 que paralleles aux chemins de fer. Ce brevet fait aussi 
 observer que " les fils ainsi tires, suspendus et isoles, 
 pourraient s'employer pour la moitie du circuit conjoin- 
 tement avec la terre pour 1'autre moitie." 
 
 Le 6 mai, 1845, MM. Wheatstone et Cook se firent 
 donner un brevet pour "des amendements dans les 
 telegraphes electriques et dans ses appareils, dont une 
 partie pourrait etre appliquee a d' autres usages." On 
 se propose dans ce brevet, de faire entendre des signaux 
 en meme temps qu'on en fait voir, a 1'aide d'une aiguille 
 ou d'un index, chaque direction ou deviation de T index 
 produisant un son different. Cela s'accomplissait en 
 faisant frapper une sonnette par 1'index, ou bien en em- 
 ployant le mecanisme d'alarme electrique. On y arrivait 
 aussi en se servant du " courant derive," qui est un 
 branche-circuit ou ce que Ton appelle de nos jours en 
 anglais "shunt," enleve au fil principal sans deranger 
 sa continuite, pour faire communiquer les signaux qui 
 traversent le fil principal, a un appareil sensitif de 
 signaux. On trouve dans ce brevet onze amendements 
 nouveaux qu'il n'est pas necessaire de rechercher. 
 
 Le vrai inventeur du telegraphe electrique nous est 
 aussi inconnu definitivement que 1'inventeur de chemins 
 de fer, ou de bateaux a . vapeur. Je nomme done ces 
 brevets tout simplement parce que la premiere Com- 
 pagnie du telegraphe electrique en Angleterre fut 
 formee peu de temps apres la date du dernier, ayant pour 
 but de les operer au profit du public plus encore que ne 
 1'avaient fait jusque la MM. Cook et Wheatstone, qui con- 
 centraient leur attention particulierement a P application 
 du telegraphe aux besoins des Compagnies de chemins de 
 fer comme accessoire aux systemes ordinaires de signaux.
 
 3 4 LES PREMIERS JO URS 
 
 Quoiqu'on put s'arreter et observer maints faits in- 
 teressants survenus dans 1'histoire du telegraphe elec- 
 trique depuis son commencement en 1845 jusqu'a 1850, 
 je crains que le temps ne nous manque. Passons done 
 a 1'annee 1850 et remarquons quels progres ont ete faits 
 jusqu'a cette epoque. En Angleterre la Compagnie du 
 telegraphe electrique avait prolonge ses fils jusqu'a 
 2,275 niilles anglais, et employait des instruments a 
 une et a deux aiguilles, qui etaient une modification 
 importante de 1'instrument a cinq et a quatre aiguilles 
 brevete par MM. Cook et Wheatstone en 1837. En 
 Irlande on n' avait pas encore realise la valeur du tele- 
 graphe electrique, car, quoiqu'il y eut 500 milles de 
 chemin de fer, il n'y avait pas 5 milles de fil telegraphique. 
 Mais les Americains possedaient 12,000 milles de fil et 
 employaient les instruments ecrivants de Morse, de Bain 
 et de House. Les seules raisons que je puisse assigner 
 pour les progres rapides qu'avaient faits les Americains, 
 sont leurs bas prix et la rivalite ; puisqu'il y avait non 
 moins de vingt Compagnies differentes qui ne bornaient 
 pas leurs lignes aux chemins de fer, mais les etablissaient 
 sur les grandes routes ou sur une propriete privee quand 
 1' occasion 1'exigeait. En Angleterre une seule Com- 
 pagnie avait le monopole, et n'etablissait ses fils que sur 
 les chemins de fer; mais comme il y avait alors 5,447 milles 
 de chemin de fer en operation, ses moyens d' extension 
 n' etaient pas limites de ce cote-la. En France il n'y avait 
 que 620 milles de fils, suspendus comme les fils anglais, 
 les instruments employes etant ce qu'on appelait en 
 anglais " revolving pointers" (aiguilles tournantes). II 
 y avait en Prusse 2,468 milles de fil, souterrains pour la 
 plupart, et les instruments employes ressemblaient a 
 ceux du gouvernement franais. 
 
 En Amerique on se servait aussi a cette epoque du 
 telegraphe electrique dans les observations meteorolo- 
 giques, et un systeme d'avertissements de tempetes
 
 DE LA TELEGRAPHIE ELECTRIQUE. 35 
 
 etait commence le longd'une grande partie de leur cote, 
 semblable au systeme si familier aujourd'hui en Angle- 
 terre. On a dit que c'est de la curiosite tout simplement 
 de predire une eclipse, mais ce serait un bienfait incon- 
 cevable de predire un orage imminent. Nous avons, il 
 est vrai, une connaissance considerable de la nature de 
 la chaleur, de 1'eau, de 1' atmosphere et de 1'electricite, 
 mais quand ces quatres proprietes de la nature, se 
 reunissent, flottent autour de notre globe et produisent 
 toutes ces agitations, toutes ces combinations variees 
 connues sous le-nom general de " temps," nous sommes 
 embarrasses, et bien que la telegraphic electrique nous 
 ait eclaires considerablement, il reste encore beaucoup 
 a apprendre. Si les renseignements enregistres journel- 
 lement par de patients observateurs partout le monde 
 civilise etaient communiques facilement et promptement 
 a un meme centre, de grands avancements dans notre 
 connaissance meteorologique en seraient probablement 
 le resultat, et j'espere que le jour ne tardera pas a venir 
 ou un tel systeme sera un fait accompli. 
 
 Les progres de la telegraphic electrique etant si 
 satisfaisants, il devenait absolument r necessaire que 
 1'Angleterre, centre du commerce du globe entier, ne 
 restat plus dans la position isolee ou elle se trouvait, 
 mais qu'elle communiquat directement avec les autres 
 nations par le telegraphe. Par consequent, en 1850 une 
 Compagnie se forma, apres une concession obtenue du 
 gouvernement frangais par M. Brett, pour se faire 
 accorder le seul droit d'etablir une communication tele- 
 graphique entre 1'Angleterre et la France. M. Brett 
 avait garanti qu' avec son instrument, qui etait une 
 modification de celui de M. House, un fil, et deux per- 
 sonnes seulement, Tune se tenant en France et 1'autre 
 en Angleterre, cent messages de quinze mots chacun, 
 imprimes en caracteres remains lisibles, seraient envoyes, 
 pretsa distribuer, dans 1'espace decent minutes consecu- 
 
 C 2
 
 36 LES PREMIERS JO URS 
 
 tives. C' etait une entreprise aventureuse" de sa part, 
 vu les opinions diverses sur la possibilite de parvenir 
 jamais a deposer ou a exploiter une ligne sous -marine. 
 
 On se decida a couvrir de gutta percha, (materiel 
 d'une importation recente dont les proprietes mecaniques 
 et electriques commenyaient a se developper), jusqu'a 
 un demi-pouce (127 millimetres) de diametre, 25 milles 
 de fil de cuivre, 0*083 d'un pouce (2*1 millimetres) de 
 diametre. Le cuivre n'etait pas detrempe et etait couvert 
 d'une simple couche de gutta percha en pieces de cent 
 metres chacune. Le cuivre etait uni par. ce que Ton 
 appelle un joint de poseur de sonnettes, et puis garanti 
 en outre de soudure. On y appKqua alors du gutta 
 percha plastique que Ton pressa dans une forme en bois, 
 le diametre du joint complet etant de 2 pouces environ. 
 II va sans dire qu'on ne faisait pas alors 1'epreuve de 
 chaque joint separement ; on ne regardait pas non plus 
 les proprietes de resistance ou de conductivite specifique 
 du cuivre, ni la resistance, ni la capacite electrostatique 
 du gutta percha. Les replis, joints en pieces commodes 
 a porter, etaient plonges dans 1'eau, sans temperature 
 specifique, et si 1'on n'obtenait pas de deviation serieuse 
 sur un galvanometre vertical, d'une sorte tres-lente, de 
 24 piles de 1'ancienne methode de batteries a sable, on 
 les prononyait bons et les joignait a la piece qui etait 
 repliee sur une immense bobine en fer construite expres 
 pour tenir les 25 milles en une seule piece. On posait 
 cette caisse ou bobine par le travers sur le pont d'un 
 remorqueur a vapeur. 
 
 Afin de placer le fil sur la bobine de la maniere 
 uniforme desiree, les inegalites, occasionnees par le 
 grand nombre des jointures, furent bouchees de coton 
 de rebut et quand cela semblait necessaire, on posa de 
 minces lattes longitudinalement tout le long de la bobine 
 entre les couches de fil. 
 
 Avant la submersion du fil principal, les bouts du
 
 DE LA TELEGRAPHIE ELECTRIQUE. 37 
 
 cable qui appartenaient au rivage furent poses, Tun, 
 depuis un wagon pour les chevaux qui allait servir de 
 station dans la cour de la Compagnie de chemin de fer 
 a Douvres, jusqu'a 1'extremite du plancher eleve pour 
 le port neuf qu'on y construisait ; Pautre, a une distance 
 considerable au dela des eaux basses, sur les rochers, 
 jusqu'au phare du Cap Gris-Nez. Ces bouts se com- 
 posaient d'un fil de cuivre 1*65 millimetre de diametre, 
 couvert de coton et d'une solution de caoutchouc, et 
 enveloppe d'un tuyau de plomb tres-epais. 
 
 Le 23 aout, 1850, de bon matin, le remorqueur 
 cingla du port* de Douvres, vers le plancher ou se 
 trouvait le bout de fil. On y joignit alors 1'extremite 
 du fil principal qu'on avait fait passer dans un bateau, et 
 la deposition commen^a dirigee par H.M.S. Widgeon. 
 A chaque piece, de 1 10 metres de longueur, qui quittait 
 le remorqueur, on attacha un poids de plomb pesant de 
 14 a 24 livres selon la profondeur de 1'eau ou la nature 
 du fond. On essaya d'abord de fixer les poids sans 
 alleger ni la vitesse du remorqueur ni la sortie du fil ; 
 mais comme chaque poids etait partage en deux et 
 devait etre pose sur le fil, et puis attache par un 
 marteau, frappant les deux clous de la partie inferieure 
 qui sortait par des trdus correspondants dans la partie 
 superieure, on decouvrit bientot que les deux ouvriers 
 ainsi employes se donnaient des coups a eux plus 
 sou vent qu'a 1'objet desire, ce qui obligea les autorites 
 d'arreter le vaisseau pendant que chaque poids, etait 
 attache. 
 
 II faisait le plus beau temps possible, et a 1' exception 
 de 1'embarras occasionne de temps en temps par les 
 lattes, qui se degageaient partiellement de la caisse, 
 tout alia bien, et a six heures du meme soir on baissa 
 1'ancre du remorqueur tout pres de la plage inhospitaliere 
 du Cap Gris-Nez, a la meme position que la bouee 
 qui tenait le bout attache au rivage fran9ais. On fit
 
 38 LES PREMIERS JOURS 
 
 passer alors le bout du fil principal dans la cabine ou on 
 le lia a 1' instrument ecrivant du telegraphe. Mais il 
 devint bientot evident que 1'orthographe de 1'expediteur 
 a Douvres etait fautive, ou bien que 1' instrument ne 
 fonctionnait pas de la maniere attendue. On regut, il 
 est vrai, des "caracteres remains " tres-distincts, mais 
 il fallait beaucoup d'habilete pour les arranger en ordre 
 lisible. Comme le soir s'avangait on n'avait pas le temps 
 de faire beaucoup d' experiences sur le vaisseau ; on 
 dut faire passer le fil principal dans un bateau ou le 
 bout qui restait au rivage, venait d'etre attache ; puis 
 on fit 1'epissure et la jeta a la mer en souhaitant sincere- 
 ment que la communication effectuee cette soiree-la entre 
 1' Angleterre et la France servirait de reunir les deux nations 
 plus etroitement dans I'amitie. Ensuite, tout le monde se 
 debarqua et s'achemina vers le phare au haut du Cap, 
 ou un instrument a une aiguille se trouva en communi- 
 cation avec Douvres ; cependant, quoiqu'on passat plus- 
 ieurs heures anxieuses a faire jvenir des signaux de 
 Douvres, on n'y reussit pas. 
 
 Les epreuves du jour suivant echouerent egalement, 
 et alors il se trouva que la ligne s' etait cassee, et comme 
 on ne reussit pas a lever le fil dans le voisinage de 
 1'epissure la ligne fut abandonnee. Si le fil principal 
 avait ete employe au lieu du bout lourd et couvert de 
 plomb, je crois que Ton aurait obtenu de meilleurs 
 resultats. Mais la critique ne peut s'exercer que par la 
 lumiere de 1' experience, qui ne serait pas tout a fait juste. 
 Decernons plutot 1'honneur a ceux qui ont sacrifie leur 
 argent, leur temps, et leur energie a demontrer d'une 
 maniere pratique la possibilite de la telegraphic sous- 
 marine. 
 
 Des ce moment on ne s'occupait que de la meilleure 
 methode de proteger le fil et son isolateur pendant la 
 deposition et apres la submersion. Apres un examen 
 attentif il fut convenu (en 1851) qu'au lieu d'un fil
 
 DE LA TELEGRAPHIE ELECTRIQUE. 39 
 
 de la dimension employee dans la ligne experimentale 
 de 1'annee precedente, quatre fils de cuivre, 1-65 milli- 
 metres de diametre, converts separement de deux couches 
 de gutta percha jusqu'a 6*35 millimetres de diametre, 
 seraient tresses, bien proteges, et entoures de fils de fer 
 epais deposes en helice. Fils de chanvre goudronne 
 furent places dans les interstices des conducteurs tordus, 
 et nommes en anglais, le "worming" Le tout etait 
 garni de pareils fils deposes a angles droits et nommes 
 le "serving" Ces trois operations s'accomplissaient 
 tout a la fois par une seule machine, et le conducteur 
 isole, ainsi protege, fut nomme le "core" (ame). Le 
 tout passa alors dans une machine ordinaire pour 
 tordre les cordages metalliques, ou il fut enveloppe de 
 dix fils de fer galvanises, o 63 millimetre de diametre, 
 et pendant qu'il sortait de la machine on le replia.dans 
 la cour de la fabrique. Les conducteurs de cuiVre furent 
 isoles par la Compagnie du Gutta Percha, et expedies 
 en pieces commodes a une petite fabrique de cordages 
 metalliques en High Street, a Wapping, Londres, ou ils 
 furent tordus, etc., et couverts de fils de fer tels que nous 
 1'avons decrit. 
 
 A cause d'un proces au sujet des droits de brevet, 
 la manufacture du cable ne fut terminee que le 17 sep- 
 tembre, 1851. On se prepara immediatement a replier 
 le cable en une piece sur le Blazer, ponton qu'avait 
 prete le gouvernement anglais. Le 24 sept, tout 
 fut replie, et le Blazer, traine par deux remorqueurs 
 a vapeur, partit pour le South Foreland ou, le jour 
 suivant, un bout du cable fut debarque et la deposition 
 commencee sous la direction de H.M.S. Fearless. 
 Le Blazer etait hale par ses deux remorqueurs, un 
 troisieme y assistant. II nous manque de temps pour 
 enumerer les incidents -animes qui survinrent ce jour-Id, 
 sur le vaisseau. 
 
 A cause des retards facheux pendant la fabrication
 
 40 LES PREMIERS JO URS 
 
 du cable, la deposition dut etre entreprise, afin de 
 menager la concession, a 1'epoque des brises de 1'equi- 
 noxe et des plus hautes marees. Quoiqu'il fit beau au 
 commencement, le temps changea bientot, et avant midi 
 il faisait un vent frais, et le Blazer, qui n' etait qu'un 
 vieux ponton, plongea impitoyablement. Les moyens 
 employes pour retarder la sortie d'un cable aussi lourd 
 n'etant pas suffisants, on en perdit une grande quantite 
 pendant les deux arrets, dont Tun etait occasionne par 
 une cessation absolue des signaux a terre, et 1'autre par 
 le cassage de la corde de hallage. Eventuellement 
 P expedition arriva pres de Sangatte, le soir etant deja 
 avance, sans pouvoir parvenir a la plage a cause du 
 manque de cable. Une lutte aussi longue n'avait jamais 
 eu lieu entre 1'esprit etla matiere jusqu'a ce jour memo- 
 rable, et il etait a regretter que pour le moment la 
 matiere remporta la victoire. Mais, commandant tout 
 de suite un morceau de cable qui pourrait arriver a la 
 plage, on torda, en attendant, trois fils, les plaga sans 
 protection, sur la terre, et les lia aux fils deja deposes 
 dans une salle de la gare de Calais. Ainsi s'etablit le 
 30 septembre, 1851, une communication telegraphique 
 entre 1'Angleterre et la France. 
 
 Le 19 octobre le morceau de cable neuf remplaQa 
 les trois fils et le 13 novembre on Pemploya pour la 
 premiere fois a la transmission de messages publics. 
 Malgres certaines interruptions il a fonctionne de cette 
 maniere jusqu'a ce jour. 
 
 Comme on avait demontre si heureusement la possi- 
 bilite de la telegraphic sous-marine, il etait a supposer 
 qu'on ne tarderait pas a suggerer de nombreux projets 
 pour son developpement. Mais, malheureusement, plu- 
 sieurs echouerent par la precipitation et le dedain des 
 precautions ordinaires. Au mois de juin de 1'annee 
 suivante, par exemple, on essaya de relier 1'Angleterre 
 et 1'Irlande par une ligne depuis Holyhead jusqu'a
 
 DE LA TELEGRAPH IE ELECTRIQUE. 41 
 
 Howth. Dans ce cas un fil de cuivre, enveloppe de 
 gutta percha semblable aux fils du cable de Douvres 
 a Calais, fut couvert de fil de fer No. 8 (4*2 millimetres), 
 pour les bouts de rivage, puis - de douze fils de fer 
 No. 14 (2*1 millimetres), pour quelques milles, mais la plu- 
 part du cofcducteur isole n'etait qu'en partie couverte, 
 d'apres un cage d'oiseau, de six fils de fer galvanises 
 No. 14. Par consequent, le conducteur isole n'ayant 
 point d' enveloppe de chanvre, le gutta percha se trouva 
 bien expose. Le cable acheve, on le replia sur des 
 wagons de chemin de fer et 1'amena a Whitehaven ; 
 mais comme le bateau a vapeur qui devait le recevoir 
 ne pouvait entrer au port il fallait 1'expedier a Maryport 
 ou on le replia sur un autre bateau et le porta a Holy- 
 head. On y en fit la premiere epreuve electrique et le 
 trouva fautif. On ne 1'avait jamais immerge, et il se 
 trouva par consequent tout a fait sec a la cale du 
 bateau. II n'etait done possible de decouvrir que les 
 fautes les plus serieuses, travail bien long et ennuyeux. 
 Si on 1'avait immerge, on aurait decouvert son etat im- 
 parfait et renvoye sa deposition. On essaya vainement 
 de reparer cette ligne apres la deposition. La force neces- 
 saire pour la relever reunissait les fils de fer, et coupait 
 le gutta percha en tirant le conducteur isole du centre. 
 
 Un peu plus tard un essai se fit pour deposer, depuis 
 Portpatrick jusqu'a Donaghadee, une corde de chanvre 
 a trois torons, environ deux pouces de diametre, dont 
 les interstices furent bouches d'un fil de cuivre que Ton 
 avait d'abord couvert de caoutchouc puis de gutta 
 percha. Mais a cause des moyens msuffisants dont on 
 se servait pour enfoncer une telle corde, les courants 
 forts 1'emporterent loin de la course proposee, et enfin 
 1' expedition fut abandonnee. 
 
 Plus tard encore on fit un nouvel effort, entre les 
 memes points, avec un cable semblable a celui depose 
 1'annee precedente a travers la Manche, mais le temps
 
 42 LES PREMIERS JO URS 
 
 etait si mauvais et les moyens si insuffisants pour arreter 
 ou meme retarder la sortie d'un cable aussi lourd, que 
 la piece fournie prouva trop courte et cette nouvelle 
 expedition echoua. Ainsi on perdit trois cables en essay- 
 ant de relier 1' Angleterre et 1'Irlande par le telegraphe 
 electrique ; mais 1* annee suivante on deposa vec succes 
 un lourd cable multiple depuis Portpatrick jusqu'a 
 Donaghadee. Quatre cables legers furent deposes cette 
 meme annee entre 1'Angleterre et la Hollande, et un 
 autre, lourd et multiple entre Douvres et Ostende. 
 
 L' annee 1853 fut aussi remarquable pour le grand 
 nombre de fils souterrains deposes en Angleterre par 
 plusieurs Compagnies telegraphiques qui s'y trouvaient 
 alors ; mais comme j'en ai donne quelques details dans 
 mon essai sur les " Fils souterrains" que j'ai lu devant 
 les membres de cette Societe le 28 mars, 1877, et qui 
 etait public dans le sixieme journal des precedes de cette 
 Societe, il n'est pas necessaire que nous revenions sur 
 nos pas. 
 
 De 1851 a 1856, on etait parvenu a deposer, non- 
 seulement en Angleterre, mais a d'autres endroits, 
 quelque 1600 milles de cable, et on se croyait assez 
 experimente pour amener a une fin heureuse le projet si 
 longtemps cheri de deposer un cable entre 1'Irlande et 
 Newfoundland, et ainsi relier 1' Amerique et 1'Angleterre 
 par la telegraphic. On avait sonde, et decouvert 
 certains rochers escarpes ; il se trouva toutefois un 
 plateau ou sentier uni entre les deux iles, comme si la 
 nature, devancant les besoins de l'homme, avait apprete 
 un lieu de repos pour les cables telegraphiques ou, bien 
 places, ils demeureraient paisiblement ; et le limon qui, 
 disait-on, descendait continuellement dans Teau, comme 
 la neige dans 1'atmosphere, couvrirait et protegerait le 
 cable a jamais. 
 
 On fit des experiences avec quelques fils sous-marins 
 et souterrains, et les resultats encourageants firent
 
 DE LA TELEGRAPHIE ELECTRIQUE. 43 
 
 esperer aux autorites d'obtenir une vitesse remuneratoire 
 par un cable transatlantique, d'une longueur de 2000 
 milles, en se servant d'un conducteur comparativement 
 petit. On mit a 1'epreuve les proprietes mecaniques de 
 differents specimens de cables, et enfin on en choisit un 
 dont le conducteur se composait d'une tresse. de fil de 
 cuivre pesant 107 livres (environ 50 kilos.) par mille 
 nautique, couvert trois fois de gutta percha pesant 261 
 livres (environ 120 kilos.) par mille, le diametre 
 exterieur etant 9-5 millimetres. 
 
 On posa en helice autour du gutta percha des fils 
 de chanvre satures d'une composition de goudron, de 
 poix, et de suif, sur lesquels on posa aussi en helice dix- 
 huit tresses enfer.dont chacune se composait de sept fils 
 du meme diametre que ceux du conducteur. Le cable 
 pesa 20 quintaux par mille nautique dans 1'air, et 13-4 
 quintaux dans 1'eau, sa force a rompre etant de 65 
 quintaux environ. 
 
 Le dernier choix de ce cable, ne se fit que vers la 
 fin de 1856, et comme il fut convenuque 1'onembarquat 
 2500 milles, prets a deposer, au mois de juin de 1'annee 
 suivante, il ne restait que six mois pour accomplir la 
 tache. Dans cet espace de temps il fallait appreter 1 1 6 
 tons de cuivre, les tirer en 17,500 milles de fil, et puis 
 tordre le fil en 2500 milles de tresse; 250 tons de 
 gutta percha devaient etre fabriques et places autour 
 du conducteur en trois couches differentes ; 1687 tons 
 de fer devaient etre aussi appretes specialement et tires 
 en 3 15,000 milles de fil dont il fallait tordre 45,000 milles 
 de tresse que Ton pla9ait autour du conducteur. 
 Ajoutons a tout cela le besoin de choisir et d'armer des 
 vaisseaux pour recevoir le cable, de construire des 
 machines a fabriquer et a deposer, et nous ne nous 
 etonnerons pas que ce temps fut entierement insuftisant 
 pour executer d'une maniere convenable une oeuvre 
 aussi importante.
 
 44 LES PREMIERS JO URS 
 
 Pendant la fabrication, les pieces du conducteur 
 isole, d'une longueur de 1600 a 2400 metres, furent 
 plongees dans 1'eau sans temperature determinee, et on 
 eprouva leur isolation moyennant 504 piles del'ancienne 
 forme de batteries a sable et un galvanometre horizonta 1 
 Si les deviations des replis de chaque fournee se trou- 
 vaient egales, on certifia leur bon etat electrique et ils 
 etaient roules sur une bobine pour les expedler a la 
 fabrique de cables. 
 
 Bientot apres le commencement de la fabrication du 
 cable les defauts du cuivre, occasionnes par la trempe 
 peu propre, etc., amenerentla solution de continuite dans 
 le conducteur. Malheureusement la fabrication de plu- 
 sieurs centaines de milles du conducteur isole se fit 
 avant la preparation des machines pour y appltquer la 
 garniture et les tresses en fer. Le decret fatal fut alors 
 public que Ton soumettrait tout le conducteur en magasin 
 a une force extensible pendant qu'il passait sur un baril 
 de peu de' diametre afin de developper, si possible, les 
 defauts du conducteur avant de le faire entrer a la 
 machine dont on se servait pour garnir le fil. Or, lors- 
 qu'un fil ou tresse de cuivre, couvert de gutta percha, 
 s'etend, le metal s' allonge et tient le materiel elastique 
 dans cette position jusqu'a ce qu'une occasion favorable 
 survienne pour que le gutta percha puisse retourner a 
 son etat normal, ainsi deformant le fil. Malheureuse- 
 ment, bientot apres, une telle occasion se presenta. Le 
 cable fabrique ne fut pas replie en reservoirs ou on 
 pourrait le garder sous 1'eau, mais ou il devint 
 expose aux rayons puissants du soleil. Par consequent 
 le gutta percha devint plastique a certains endroits, et 
 le conducteur repousse, communiqua directement avec 
 la garniture de chanvre qui, etant sec, joua le role d'isola- 
 teur et ainsi empecha la decouverte des defauts jusqu'a ce 
 que le cable se pla^at dans 1'eau. . C'etait a regretter que 
 la fabrication du cable etait trop avancee pour que 1'on
 
 DE LA TELEGRAPHIE ELECTRIQUE. 45 
 
 put profiter de la decouverte de Sir William Thomson 
 des grandes varietes de conductivite qui se trouvaient 
 dans le cuivre commercial, surtout en cette espece que 
 Ton employait alors pour le conducteur du cable trans- 
 atlantique. On a realise, depuis, toute 1' importance de 
 cette decouverte, non-seulement en toutes les branches 
 de la science electrique, mais aussi en faisant obtenir 
 un cuivre pur pour quelque objet que ce soit. Tout le 
 monde sait aujourd'hui que le cuivre qui a le plus de 
 conductivite electrique est le plus pur. 
 
 Ce n'etait que le 5 aout que Ton debarqua, a Valentia, 
 un bout du Cctble et en commen9a la deposition ; mais 
 des difficultes s'eleverent bientot apres le depart et 
 lorsqu'on en avait depose 335 milles, le cable se cassa 
 juste a 1'arriere de la machine sur le vaisseau employe 
 pour le filer, et 1' expedition fut abandonnee pour cette 
 annee. On replia le reste du cable dans 1* arsenal de port a 
 Plymouth, ou il fut inspecte mais non immerge, puis passe 
 dans un melange chaud employe pour isoler, et enfin, 
 episse a 700 milles de cable neuf ce qui fixa la longueur 
 totale a 3000 milles. 
 
 Au mois de juin suivant le cable etant replie, la 
 moitie sur le Niagara, 1'autre sur I* Agamemnon, diverses 
 experiences se firent dans la Baie de Biscay, dont les 
 resultats favoriserent la proposition d'episser le cable 
 au milieu de F ocean et de le faire filer des deux vaisseaux 
 a la fois, 1'un vers 1'Irlande et 1'autre vers Newfoundland. 
 Le 26 juin on fit Tepissure au milieu de 1'Atlantique, 
 mais apres avoir file trois milles de chaque vaisseau le 
 cable se cassa a Tarriere du Niagara. Les vaisseaux 
 retournerent tout de suite a leurs anciennes positions, 
 episserent de nouveau et repartirent ; mais lorsque une 
 quarantaine de milles avait ete deposes, tout d'un coup 
 les signaux s'arreterent, et les directeurs de chaque 
 vaisseau ne sachant ce qui etait arrive a 1'autre, 
 (incident qui montre le desavantage de cette methode
 
 46 . LES PREMIERS JOURS 
 
 de deposition) retournerent au point ou ils avaient 
 fait 1'epissure, et en se rencontrant, demanderent 
 anxieusement, " Comment il se passa que le cable fut 
 casse ? " La cause de cette rupture etait enveloppee de 
 mystere. Tout allait bien sur les deux vaisseaux quand 
 la continuite s'arreta. On epissa de nouveau le cable 
 et en deposa quelque 200 milles lorsqu'il se cassa a 
 1'arriere de F Agamemnon, et les vaisseaux revinrent en 
 Irlande. 
 
 Le 29 juillet ils se retrouverent au milieu de 1' ocean 
 et les directeurs y episserent, encore une fois, le cable ; 
 mais ils-n'en avaient pas beaucoup file lorsque le bruit 
 courut que la continuite s' etait rompue. Malgre cette 
 nouvelle effrayante on continua a filer lentement des 
 deux vaisseaux, et au bout de deux heures d'attente 
 penible les signaux reparurent aussi mysterieusement 
 qu'ils etaient disparus. De pareils incidents se 
 repeterent souvent, mais il n'y avait pas de rupture, et le 
 5 aout chaque vaisseau acheva sa tache, le Niagara ayant 
 depose 1030, et I* Agamemnon 1020 milles nautiques. 
 Le cable fut depose, mais .sa condition electrique etait 
 deplorable, car il ne fallait rien moins qu'une force gigan- 
 tesque pour produire 1' ombre spasmodique des signaux 
 desires ; et au bout de 24 jours on ne pouvait plus 
 obtenir meme ces resultats-la, le projet tout entier ayant 
 echoue. C' etait un echec d'un cote, mais on ne devrait 
 pas oublier que beaucoup d' experience precieuse etait 
 acquise et que le probleme au sujet de la possibilite de 
 deposer un cable transatlantique etait resolu. 
 
 Afin d' obtenir une adhesion parfaite, on avait mis 
 jusqu'a cette epoque entre chaque couche de gutta 
 percha de la naphte minerale, qui etait un dissolvant fort 
 de gutta percha, et qui eventuellement produisait des 
 effets nuisibles. Je commensal done a faire des expe- 
 riences en vue de decouvrir un compose adhesif qui ne 
 nuirait nullement au gutta percha, et je trouvai bientot
 
 DE LA TELEGRAPHIE ELECTR2QUE. 47 
 
 qu'une certaine quantite de goudron de Stockholm, de 
 resine, et de gutta percha produirait 1'effet desire. Non- 
 seulement ce compose possedait-il de bonnes qualites 
 adhesives, mais il etait aussi bon isolateur, et en 1'appli- 
 quant a chaque couche alterne de gutta percha, il ameli- 
 orait T isolation de triple. Ce compose s'est montre et se 
 montre encore tres-utile, et on connait si bien ses qualites 
 que je n'ai pas besoin d'en parler davantage. 
 
 Vers cette meme epoque, il etait sujet de surprise et 
 de regret que le gouvernement anglais se fut prete a un 
 projet aussi gigantesque que celui de deposer un cable 
 de 1'Egypte a 1'Inde, alors que le sort du cable transat- 
 lantique balancait encore. Si Ton avait attendu le resul- 
 tat de cette expedition-la, peut-etre aurait-on evite Fechec 
 lamentable qui s'ensuivit. Le cable de la Compagnie 
 telegraphique de la Mer Rouge et de 1'Inde se composa 
 d'un conducteur isole contenant une tresse a sept fils de 
 cuivre, pesant 180 livres, recouverte deux fois de gutta 
 percha de 2 1 2 livres par mille anglais, garni de chanvre 
 goudronne et protege exterieurement de 1 8 fils de fer. 
 Le poids en gros par mille nautique etait de 2 1 quintaux, 
 On le deposa en six sections depuis Suez jusqu'a Kurra- 
 chee. Donner 1'histoire de la fabrication, de la deposi- 
 tion, et enfin de Fechec dans 1' exploitation de ce cable 
 serait repeter, avec des additions peu satisfaisantes ce 
 que Ton a deja dit au sujet du cable transatlantique. Le 
 fournisseur fabriqua et deposa les 4000 milles dans le 
 court espace de quinze mois, et pendant ce temps il 
 deposait aussi dans le Levant des cables garnis de 
 chanvre, dont quelque 800 milles se perdirent. 
 
 Ces echecs serieux, comprenant la perte d' environ 
 8000 milles de cable en si peu de temps, amenerent 
 naturellement une investigation complete de toute 
 1' affaire. Un comite fut done nomme en partie par les 
 Lords du Conseil prive de Commerce, et en partie par 
 la Compagnie du telegraphe transatlantique pour faire
 
 48 LES PREMIERS JO URS 
 
 une enquete sur la construction de cables sous-marins, 
 et je crois que les efforts laborieux de ce comite-la con- 
 tribuerent beaucoup au succes definitif de la telegraphic 
 sous-marine. 
 
 Au mois de juin, 1859, le gouvernement anglais 
 ordonna qu'on deposat 1200 milles de cable entre Fal- 
 mouth et Gibraltar. Des ingenieurs et des electriciens 
 furent charges de surveiller la fabrication de ce cable. 
 Le conducteur isole se composa de poids egaux de tresse 
 de cuivre a sept fils, recouverte trois fois de gutta percha 
 et de compose, pesant tout ensemble 800 livres par mille 
 anglais. Le chanvre goudronne et les fils de fer le 
 couvrirent de la maniere ordinaire. On employa des 
 instruments pour mesurer les qualites electriques du 
 cuivre et du gutta percha a une temperature uniforme, 
 et on en nota les resultats en unites electriques ; c'est- 
 a-dire, on nota la resistance specifique du cuivre et du 
 gutta percha de maniere que la mesure fut comparable 
 aux resultats obtenus pendant la fabrication et apres la 
 submersion. Ou soumit le conducteur isole en pieces 
 commodes a une pression de 600 livres par pouce-carre, 
 et on trouva invariablement que la resistance du gutta 
 percha, ainsi place, s'augmentait. Le gutta percha sur 
 un des replis ayant ete perce, par accident, jusqu'au 
 conducteur, on trouva, en le pla<?ant dans le presseur, 
 que 1'eau s'etait efforcee le long de la tresse et sortait 
 par les bouts que Ton avait fait passer en serre-etoupes 
 pour en faire Tepreuve. Pour eviter un tel accident & 
 Favenir, je plagai le fil du centre de la tresse dans le 
 compose avant de poser les six autres autour. Cela 
 produisit Peffet desire, car Peau, a une pression de 1000 
 livres par pouce-carre, ne pourrait s'efforcer a travers 
 six pouces de conducteur isole contenant une tresse 
 ainsi construite. 
 
 En fabriquant ce cable on le replia en reservoirs ; 
 mais les vaisseaux charges de le deposer ne possederent
 
 DE LA TELEGRAPHIE ELECTRIQUE. 49 
 
 pas de reservoirs. Par consequent peu de temps apres 
 I'embarquement de la premiere section", on decouvrit une 
 augmentation de la resistance du conducteur denotant 
 une augmentation de la temperature du cable, venant 
 de la fermentation occasionnee soit par la garniture de 
 chanvre soit par 1'oxydation du fer. Par des mesures 
 thermo-electriques, la temperature variait de 27 a 16 
 degres centigrade en differents endroits du meme repli, 
 tandis que la temperature extreme restait a 15 degres. 
 On craignit que la haute temperature n'eut gate le 
 gutta percha, mais une piece du conducteur isole prise 
 de la partie la plus chaude, ne montra aucun signe de 
 deterioration. 
 
 A cause du retard des vaisseaux la saison etait trop 
 avancee pour deposer le cable a 1'endroit fixe. On se 
 decida done a le deposer entre Rangoon et Singapore, 
 mais le vaisseau qui en portait la premiere portion fit 
 naufrage dans la Manche et eventuellement on le deposa 
 en trois sections entre Malte et Alexandrie. Mais 1'eau 
 basse dans laquelle se trouverent certaines parties du 
 cable et la legerete de sa construction le detraquerent 
 souvent, et il dut ceder en 1868 a un cable que Ton 
 deposa directement de Malte a Alexandrie. 
 
 Au mois de decembre, 1860, je fis remarquer que 
 les dommages faits au conducteur isole d'un cable 
 par accident ou autrement, se cachaient souvent par 
 les proprietes d' isolation de la garniture goudronnee, et 
 qu'ils resteraient caches selon la nature du defaut et la 
 quantite de goudron qui s'y trouvait fait qui explique 
 P isolation fautive de tant de cables apres etre deposes 
 en ce qui paraissait une condition parfaite. Je conseillai 
 done que le fluide preservatif duquel on saturait le 
 chanvre, fut un fluide conducteur au lieu d'un fluide 
 isolateur, qu'on gardat le conducteur isole toujours dans 
 1'eau, et qu'on le fit passer directement de 1'eau a la 
 machine employee pour le couvrir de fer, ou la garniture 
 
 D
 
 50 LES PREMIERS JO URS 
 
 serait resserree et le fluide force dans les defauts du 
 gutta percha en les developpant tout de suite. On 
 adopta volontiers cette suggestion qui prouva bientot 
 efficace. Depuis lors la garniture goudronnee a ete 
 1' exception et non la regie. 
 
 Le premier cable de Malte & Alexandrie etant 
 termine, le gouvernement anglais ordonna que Ton en 
 deposat dans le Golfe persique, mais avant de decider 
 la forme de cable, on fit de nombreuses experiences 
 plus particulierement sur les proprietes electriques et 
 mecaniques de certaines conducteurs isoles qui se 
 composaient de gutta percha ou de caoutchouc, pur et 
 simple, ou bien- melanges avec d'autres substances. 
 Enfin on en choisit un qui devait se composer d'un fil de 
 cuivre segmentaire pesant 225 livres par mille nautique, 
 et qui etait isole de quatre couches de gutta percha, pesant, 
 avec le compose deja mentionne, que Ton devait appli- 
 quer de la maniere ordinaire, 275 livres par mille. Une 
 tresse et un fil solide du meme cuivre et de poids egaux, 
 aux memes conditions, conduisent 1'electricite d'une 
 maniere tout .a fait pareille, mais quant & T induction 
 laterale, il y a quelques desavantages en employant une 
 tresse. On se proposa done un fil segmentaire afin 
 d'obtenir les avantages mecaniques de la tresse, et les 
 avantages electriques d'un solide. On construisit ce 
 conducteur de quatre fils de cuivre dont chacun formait le 
 segmented' un cercle, tire dans un tuyau de cuivre; mais les 
 difficultes mecaniques d'une fabrication parfaite devinrent 
 si grandes, qu'on 1'abandonna pour un fil solide avant 
 la termination du cable. Les qualites electriques du 
 conducteur isole place sous une pression de vingt 
 atmospheres par pouce-carre furent mises a 1'epreuve, 
 en pieces commodes, a une temperature uniforme, de 24 
 degres centigrade, et la resistance electrique du gutta 
 percha s'augmentait environ 40% a ces conditions, en 
 retournant a sa resistance normale lorsqu'on eloigna la
 
 DE LA TELEGRAPHIE ELECTRIQUE. 51 
 
 pression. On appliqua, pour la premiere fois, a 1'epreuve 
 des jointures de ce conducteur la methode d' accumulation 
 si bien connue aujourd'hui. 
 
 On satura la garniture d'eau a tanner au lieu de 
 goudron, suif, etc., que Ton avait employes jusqu'ici, et 
 le conducteur garni fut replie en reservoirs. On couvrit 
 alors de deux enveloppes de chanvre, enroulees en sens 
 contraires, douze fils de fer du diametre demande pour 
 envelopper la garniture, et les placa en helice. On y 
 appliqua avec chaque couche de chanvre, un compose 
 chaud et bitumineux fait de goudron de Stockholm, de 
 silice broyee, et de poix. On fit passer le tout dans un 
 puissant cylindre resserre et cannele, et cela y donna une 
 surface unie et noire. Apres la fabrication on le replia 
 en reservoirs etanches, et aux memes conditions, on 
 transporta les 1450 milles nautiques en quatre vaisseaux, 
 et les deposa en quatre sections entre Kurrachee et Faa, 
 les stations intermediates etant Guadir r Mussendom, et 
 Bushire. Les resultats satisfaisants apres la termination 
 de ce cable compenserent certainement toute 1' attention 
 et tout le soin que Ton avait consacres au choix, a la 
 fabrication, et a la deposition. 
 
 Pendant la fabrication et la deposition du cable dont 
 nous venons de parler, on essayait deux fois de deposer, 
 dans la Mediterranee, des cables d'une construction 
 particuliere. On avait prouve par une serie d'experiences 
 bien dirigees, que, si Ton pla^ait le chanvre autour d'un 
 fil de fer ou d'acier a un certain angle et en certaines 
 proportions, la force a rompre des deux depasserait la 
 somme de leurs forces individuelles. Un conducteur 
 ordinaire de cuivre et de gutta percha fut done entoure de 
 fils, couverts de chanvre, que Ton allait deposer entre 
 Toulon et Algers, mais a cause de 1'elasticite du cable 
 et Fetat incomplet des arrangements pour filer, le cable 
 se cassa pendant la deposition. Afin de fabriquer un 
 cable neuf propre a remplacer la partie perdue, il fallait 
 
 D 2
 
 5 2 LES PREMIERS JO URS 
 
 une annee pour terminer Pentreprise. On fit une autre 
 experience avec 130 milles de conducteur isole ordinaire, 
 couvert de fils de chanvre satures d'un fluide conducteur 
 mis a une tension uniforme par un leger tordage, et puis 
 recouvert d'une enveloppe continue, tubulaire et metal- 
 lique, que Ton y appliqua en tordant un ruban de cuivre 
 en helice, de telle maniere que le bord de chaque 
 contournement recouvrit celui du precedent. On replia 
 la piece entiere autour d'un cone sur un disque tour- 
 nant, fixe dans la cale d'un bateau a vapeur. On essaya 
 de la submerger entre Oran et Carthagene, mais sa 
 construction se trouva mal-appropriee a un tel dessein, 
 ainsi que les arrangements du disque tournant, etl' expe- 
 dition fut abandonnee apres avoir perdu la moitie du cable. 
 Malgre ces echecs on avait assez fait avec succes 
 pour justifier un autre effort de reunir Newfoundland et 
 1'Irlande, Pendant les cinq annees qui s'etaient ecoulees 
 depuis la derniere expedition, de grands amendements 
 s'etaient accomplis en toutes les branches de la tele- 
 graphic sous-marine. On se decida done a fabriquer 
 2300 milles de cable compose d'un conducteur con- 
 tenant une tresse a sept fils de cuivre pesant 300 livres 
 par mille, recouverte de quatre couches de gutta percha, 
 et quatre du compose, appliquees alternativement a un 
 diametre de 1 1 '8 millimetres et pesant 400 livres par 
 mille. La garniture consista en chanvre ordinaire sature 
 d'une liqueur a tanner. La protection exterieure se com- 
 posa de dix fils de fer solides chacun entoure separement 
 de chanvre de Manille, et recouvert d'un compose 
 preservatif de goudron, caoutchouc, et poix. On replia 
 tout le cable en trois reservoirs distincts, sur ce batiment 
 si merveilleux de Brunei, le Great Easter n> et 1' expedition 
 partit dans 1'ete de 1865. On croyait assurement que le 
 succes recompenserait les soins extraordinaires, dispenses 
 jusqu'aux plus petits details sur tout ce qui se rapportait 
 a ce cable ; mais, pendant qu'on le fabriquait, le repliait
 
 DE LA TELEGRAPHIE ELECTRIQUE. 53 
 
 sur le vaisseau, et le deposait, des defauts se decouvrirent 
 dans 1' isolation qui suggererent la pensee horrible que 
 quelque scelerat avait nui a 1'ouvrage. II semblait 
 qu'on avait force un instrument grossier entre les fils de 
 la couche exterieure a travers le gutta percha jusqu'au 
 cuivre, en le retirant ensuite. Un de ces defauts survint 
 quand on avait depose 79 milles, et il fallait ramasser dix 
 milles pour le corriger. Un autre defaut de la meme 
 nature se developpa apres la submersion de 750 milles, 
 et on ramassa trois milles d'une profondeur de 2050 
 toises pour le corriger. Un autre defaut apparut en 
 2 200 toises, apres la deposition heureuse de 1214 milles, 
 mais on ne trouva pas celui-ci, car, apres avoir ramasse 
 trois milles, le cable se cassa, et le bout qui etait dans la 
 mer fut perdu. On essaya vainement de le recouvrer et 
 puis 1' expedition retourna, la saison etant trop avancee 
 pour faire d'autres efforts. 
 
 On etait d'avis qu'il-n'y aurait eu aucune difficulte 
 en deposant le cable avec succes si ces defauts mysterieux 
 n'etaient pas survenus. On se decida done a fabriquer 
 1600 milles de cable neuf, et a essayer 1'annee suivante, 
 1866, encore une fois de relier Valentia et Newfoundland. 
 On pensait qu'alors on pourrait ramasser le bout perdu 
 du cable de 1'annee precedente afin d'en terminer la 
 deposition. II y avait quelque difference entre le cable 
 neuf (1600 milles de longueur) et le precedent. On 
 n'employa point de compose dans les fils de fer ni dans 
 leur garniture de chanvre ; ils furent galvanises et devin- 
 rent plus flexibles. 
 
 On ne soumit pas le conducteur isole de ces cables 
 a la pression pendant les epreuves electriques, car la 
 pression n' avait jamais developpe les defauts; mais les 
 experiences actuelles prouverent que les defauts, dont on 
 s'apercevrait sous une pression d'une atmosphere, se 
 . cacheraient plus ou moins selon la quantite de pression 
 appliquee.
 
 54 LES PREMIERS JOURS 
 
 Nous avons calcule les resultats d'une experience 
 avec un repli fautif dans le tableau suivant, ou on verra 
 plus facilement la proportion dans laquelle la resistance 
 augmentait avec une augmentation de pression : 
 i atmosphere = 49 megohms par mille. 
 i7'5 = 5 1 
 
 35 = 53 > 
 
 5 2 '5 = 54 
 
 ?o =- 55 
 
 87'5 = 57 , 
 
 TQ5 = 58 
 
 122-5 = 5 8 
 
 140 = 63 ,, 
 
 i57'5 > = 6 4 
 
 175 = QO , , 
 
 183*5 = "5 
 
 Lorsqu'on avait ote la pression, la resistance s'aug- 
 menta subitement jusqu'a 178, puis diminua graduelle- 
 ment. Au bout de dix-sept heures, elle avait regagne sa 
 condition normale, c'est-a-dire, 49 megohms. Si Ton 
 ote peu a peu la pression, c'est interessant a remarquer 
 1'augmentation graduelle de la resistance jusqu'a ce que la 
 pression soit tout a fait enlevee, quand la resistance 
 commence a diminuer. Mais, comme nous 1' avons deja 
 dit, si Ton ote subitement la pression, la resistance 
 s'augmente de meme. 
 
 La durete du fil de fer employe au cable de 1865, son 
 compose adhesif, et le temps pendant lequel on 1' avait 
 garde replie aux reservoirs sur le Great Eastern tout 
 cela occasionna des scenes animees pendant la submer- 
 sion, resultant des couches du cable dans le reservoir 
 qui se collaient et devenaient emmelees, produisant ce 
 qu'on appelle en anglais "foul flakes." Outre cela, 
 tout ce qui se rapportait & cette oeuvre fit des 
 progrds aussi favorables que possible, et au bout de 
 quatorze jours et nuits de soucis, tous les employes eurent
 
 DE LA TELEGRAPH1E ELECTRIQUE. 55 
 
 la satisfaction de voir leurs efforts couronnes de succes. 
 Apres quelques preparatifs necessaires, 1' expedition partit 
 pour recouvrer le bout perdu du cable de 1865, et le 
 trouva a la suite d'une lutte avec les elements qui dura 
 vingt jours et nuits. Ayant parle avec Valentia afin de 
 constater la perfection de 1'etat electrique de la piece, on 
 y epissa le bout du cable qui etait sur le vaisseau, et on 
 commenca a filer vers Newfoundland. Tout allait bien, 
 quand a 6 heures du matin, peu de temps avant la fin 
 de la deposition, pendant qu'on recevait de Valentia un 
 resume des nouvelles du Times de ce jour, tout a coup 
 un defaut d'isolation se decouvrit. On le trouva bientot 
 sur le vaisseau ; on coupa et epissa de nouveau le cable, 
 et, a 1'honneur de tous les ouvriers, en moins de deux 
 heures, ils avaient corrige le defaut et recommence -de 
 filer. 
 
 Vers les cinq heures de la meme journee, on termina 
 avec succes la deposition de ce cable. C' etait par la 
 bonne fortune qu'on avait decouvert le defaut avant de 
 sortir le cable du vaisseau, car apres un examen attentif, 
 on trouva que le bout d'un fil casse du repli superieur 
 etait courbe a angles droits, et etait entre au repli in- 
 ferieur. Un des ouvriers avait march e sur le cable, et 
 le poids de ses pieds forca le fer a travers le gutta percha 
 jusqu'au cuivre. Comme ce defaut ressembla exacte- 
 ment a celui de 1'annee precedente et se decouvrit dans 
 le meme cable, on supposa avec justice que tous les 
 deux arriverent d'une maniere pareille, et non avec 
 malveillance. 
 
 J'ai donne les details du systeme que j'ai employe 
 pour eprouver et exploiter ces cables pendant la deposi- 
 tion et apres la submersion, dans 1'essai lu a la reunion 
 de cette Societe le 13 fevrier, 1879, e ^ public dans le 
 journal de nos precedes ; done, je n'en ferai plus 
 mention. La deposition du cable de 1866, le recouvre- 
 ment et la termination de celui de 1865, ont couronne
 
 56 LA TELEGRAPHIE ELECTRIQUE. 
 
 1'oeuvre de la telegraphic sous-marine, et donne a cette 
 branche de notre science une force impulsive sans bornes 
 jusqu'a ce que tous les pays civilises fussent relies par 
 le telegraphe, comme une seule nation. A present il y 
 a en operation non moms de 70,000 milles de cables 
 telegraphiques sous-marins. L' experience de trente- 
 trois annees nous a appris qu'une longue exploitation ou 
 submersion ne fait deteriorer ni les qualites electriques 
 ou mecaniques du gutta percha ni ceux du cuivre ; et 
 aussi que les fils de fer, bien proteges de 1'oxydation, 
 gardent toutes leurs qualites primitives ; par consequent 
 la meilleure forme d'un cable telegraphique sous-marin 
 sera celle qui embrassera ces conditions. 
 
 Les progres que Ton a fait dans la perfection de tous 
 les instruments et appareils employes a la telegraphic 
 electrique se montrent si bien dans 1' Exposition qui nous 
 a reunis en ce lieu que je n'ai pas besoin d'y faire 
 allusion. Bien que j'aie empiete quelque peu sur votre 
 temps, vous me permettrez en terminant, de reconnaitre 
 la bonte universelle dont on m'a comble dans ce pays 
 toutes les fois que je Pai visite, soit pour les affaires, 
 soit pour le plaisir. Ma premiere visite eut pour 
 destination le Cap Gris-Nez, lors de la termination de la 
 ligne experimental en 1850, apres une journee animee 
 de travail, et dix-huit heures de jeune. J'ai encore un 
 vif souvenir de la bonte du gardien du phare, qui a non- 
 seulement partage avec moi son repas frugal, mais qui 
 s'est occupe de tous mes besoins personnels. 
 
 IIATMAN BEOTHEES AND ULLT, HATTON HOUSE, FARBINGDON R'AD, B.C. 
 
 Vv

 
 N 
 
 ; ^ 
 
 -.