History of the Atlantic Cable & Undersea Communications
Report of the Joint Committee Appointed by the Lords of the Committee of the Privy Council for Trade and the Atlantic Telegraph Company to Inquire into the Construction of Submarine Telegraph Cables together with the Minutes of Evidence and Appendix. (1861)
Evidence of Wildman Whitehouse, 15 December 1859
1713. (Chairman.) What is your profession?—I am a member of the College of Surgeons, but not now practising; lately I have devoted myself to electro-telegraphy and that must be called my profession.
1714. I believe you have given great attention to electrical questions, and especially to deep-sea telegraphy for some years?—Yes, I have.
1715. Since when?—For the last eight or nine years.
1716. I believe you made experiments with Mr. Brett?—I did. One of the Mediterranean cables, for which Mr. Brett was the contractor, was the first one that I experimented upon. The object of my experiments was to determine as many points as I could connected with deep-sea telegraphy; the laws of retardation, and induction, and the most suitable form of instruments to be used; in fact, I spent almost night and day for three years, working constantly, in trying to determine all the points that I could.
1717. What was the longest of the cables that you experimented upon previous to 1856?—160 miles was the length of the cable, as well as I remember, containing six wires.
1718. Was that cable coiled?—Yes, it was coiled in a tank at Messrs. Glass and Elliott's works, at Greenwich.
1719. What were the general results of those experiments?—It is difficult to give them in a few words. The first object was to ascertain, as nearly as I could, the law governing the retardation of the current at different distances, in order that I might discover the limits, if any, beyond which deep sea telegraphy would cease to be practically attainable.
1720. (Mr. Varley.) I think you said that the cables were 160 miles in length upon which you experimented; did not you experiment upon a greater length than that?—At a later period I did, but not at that time.
1721. Was each wire 160 miles in length?—166 miles, I think, was the exact length.
1722. (Chairman.) Did you join the wires together?—Yes, I joined them together. The whole cable was 166 miles in length, and it contained six insulated conductors.
1723. Therefore you got six times that length?--Yes.
1724. What was the nature of your experiments?—The first was with regard to the velocity of the electric current under varying circumstances, trying the length of time that it occupied in travelling the 166 miles, then comparing it with the time required for two lengths of 166 miles, and three, four, five, and six lengths.
1725. Were the results satisfactory?—They were remarkable, certainly, and their invariable uniformity under like conditions was in the highest degree satisfactory. I experimented upon different forms of current, and upon different modes of using the same form. I mean the voltaic current as continuously employed by interruptions or repetitions of similar currents, or otherwise, alternately, a positive and a negative. Then I tried the magneto-electric current from the induction process, and from permanent magnets also. My object was to work out, as far as I could practically, the best form of instrument for use in submarine lines, the best form of current and the best mode of using it.
1726. What result did you come to?—I found that the voltaic current was at first very difficult to use with advantage in so long as length, rapidly. I found that its rate of transmission was slower than the magneto-electric current; I found if used in repetition instead of alternating, that the charge of one current was blended with the charge of the next current, and that it came out almost as a continuous discharge at the other end of the whole length of the cable. I had, in some of the experiments, a seconds pendulum beating, giving the contact every second at one end of the circuit, which gave the voltaic current punctually as the pendulum beat; at the end of the whole distance, through the six wires joined up in a single length, it came out as a continuous discharge, without interruption. There was no variation in the current perceptible on the galvanometer receiving it, so that I then tried reversing, making the pendulum reverse the current at each beat one second positive and one second negative; I could then get distinct signals through the whole distance from that end, the negative current discharged the remains of the preceding positive current, and left the wire comparatively empty and ready for another current, so that you got a small portion of each current; a considerable portion of the current was consumed in discharging the preceding one; but a portion passed through the cable and was perceptible as a signal through the whole distance.
1727. Did you employ weak currents or strong currents?—The widest range which I had at my disposal.
1728. (Mr. Varley.) Can you state the limits that you started from?—I started from a single pair, the size of a sixpence, and I recorded signals through 1,000 miles with that distinctly, beating seconds alternately. I think the highest number of cells I had was 360, of the ordinary zinc and copper plates.
1729. What sized plates?—I used four-inch, the usual size.
1730. How were they charged?— They were charged with acid; it was the ordinary sand battery, such as is used at the gutta-percha works.
1731. (Chairman.) You also tried, you say, the magneto-electro currents?—Yes; I tried the voltaic current, beginning with the very weakest I could send. It required an exceedingly sensitive instrument and very perfect adjustment to show anything from a single pair. Then I gradually increased the number of pairs till I got to 360. I tried as far as I could to get at the ratio of the velocities of those different powers.
1732. With what general result was that experiment attended? Of which current was the velocity the greater?—Up to a certain point there seemed to be a question whether a sufficient amount of current arrived to actuate the instrument. It was rather a matter of the arrival of a portion of the current than the arrival of the maximum current, because beyond that no increase in amount or intensity appeared to augment the velocity at all. With about 18 or 20 cells I always found the slips recorded as high a velocity as with 360 cells; below this the observations were extremely critical and liable to instrumental error.
1733. You required a more delicate instrument:— I required a more delicate instrument. Although I could adjust the instrument to receive a signal from a single cell through 1,000 miles, yet I could not depend upon its accuracy to a minute fraction of a second for measurement of time.
1734. (Mr. Saward.) Were those signals received by relay?—Entirely. I have records of hundreds.
1735. (Chairman.) Will you be good enough to communicate those records?—Yes. In order to ensure accuracy I made the current on leaving the battery before entering the cable pass through a relay to make its mark, and on leaving the cable and entering, the earth it passed thorough another relay, and made its mark; these marks were made by two styles on the same slip of paper.
1736. Was that in order to measure the speed of its travelling?—Yes, the recording mark was the electrochemical decomposition.
1737. (Mr. Varley.) You have not stated the size of the wire nor what form the insulating material was?—It was an ordinary sixteen gauge wire, double covered with gutta-percha of the usual thickness.
1738. (Chairman.) Were the wires joined together at the ends?—I had the command of the whole of the twelve ends; I could either make them go in circuit in the same direction or alternate them in opposite directions; I could test the effect of those different arrangements one against the other.
1739. (Mr. Varley.) What was about the speed that you obtained with the six times 166 miles in one continuous circuit with the battery power?—A retardation of about 2½ seconds.
1740. What was the relative speed of the different battery power used compared with the magneto current?—The magneto currents was usually from 2½ to 3 times as rapid as the voltaic.
1741. If I understood you rightly you stated that you found, after you got up to 20 cells, that increasing the number of cells did not alter the speed of the current?—Not perceptably.
1742. Did it accelerate it at all?—The 20 cells were sufficient to make the distant relay respond readily with moderately careful adjustment, and it required no subsequent adjustment at all. The experiments were perfectly satisfactory, and the velocity seemed to be the same from 20 cells to 360; it is true that I was working then with the sand battery, which has so much resistance; although you multiply the cells you do not increase the current in the same ratio.
1743. I should have anticipated with 300 cells you ought to have noticed a retardation of the speed?—I did not in any instance observe an increased retardation as the result of increased battery force. The slips are all marked with the memoranda taken at the time, showing exactly what was used.
1744. Do you think increasing the surface is objectionable?—I do not; but I think there may be a limit beyond which if you increase it you fail to attain any proportionate results.
1745. Supposing you increase the surface to an infinity of plates whose surface is a mile square; do you think you would do any harm by decreasing your speed if you had proper insulation?—I do not understand the meaning of the question.
1746. Supposing you increase the surface infinitely great, merely keeping the same number for tension, do you think you would cause retardation or weakening of the signals through the cable?—I do not think so; the general results of my experiments have gone in the opposite direction; whenever I have used induction coils with a larger secondary wire than before I have got a more rapid current in every instance, when I have not sacrificed the adequate amount of intensity. You must have a certain amount of intensity to overcome the resistance of the circuit, and if you go beyond that amount of intensity you retard your current; in the first few miles you get a more rapid current, and in the last part of a long line it exhausts itself much more rapidly, the retardation being under those circumstances nearly as the square of the distance.
1747. You are alluding now to small-sized plates?—In this instance I was speaking rather of the induction coil current, because I could not vary the size of the plates adequately in the voltaic battery; but I was able to avail myself of coil currents of various characters, depending upon the sectional area and length of the secondary wire employed in the coil.
1748. In the sand batteries, not when you use the resistance battery?—The results were very remarkable, when I came to use the induction coil current, the secondary wire of which was 40 gauge, the early part of the journey was very rapidly travelled, while the latter part was even at a slower rate than the sand battery. I am speaking of the rate of retardation of induction coil currents; I was going to say of different sectional areas, but having been generated in a secondary wire of a different size, I found that the current generated in a 40 gauge wire travelled very rapidly the early part of its journey, the first 100 or 150 miles, but after it had travelled 1,000 or 1,200 miles, it charged the wire even more slowly than the ordinary voltaic current that I had been using, whereas a current generated in a wire of larger gauge, 28 gauge for the secondary wire for instance, would traverse the cable with equal rapidity for a much greater distance, and did not get exhausted on quickly; in fact, I have found it capable of charging at much greater amount of surface. I was, therefore, led to have the secondary wire for the Atlantic coils unusually large, as you probably know.
1749. What was the site of the wires in those secondary coils?—18 gauge in the secondary.
1750. And in the primary?—The primary wire upon each pair of coils consisted of forty-eight 14 gauge wires, arranged in parallel circuits as one large wire.
1751. Of what length?—About 100 yards each.
1752. (Chairman.) I believe you subsequently experimented upon the gutta-percha covered subterranean wires of the Magnetic Company?—I did; we had but very few experiments upon them, and they were not very satisfactory ones. When the wires of a company are in constant use commercially, and the stations are at a distance from each other, it is difficult to get a satisfactory experiment; you cannot be certain, not being at both ends, that the arrangements at the other end are what you wish. Sometimes an earth wire is left in imperfect connexion with the circuit, which defeats your object entirely.
1753. You formed a circuit of nearly 2,000 miles, did not you?—Yes.
1754. Did you arrive at similar results generally with those in the other case?—We were not able so carefully to test the velocity. We had a galvanometer put in at each junction, so that we could trace the progress of the wave from end to end; we were able then by a relay put in circuit at our end, to record through the whole length of the 2,000 miles.
1755. Do you consider that such a circuit represents faithfully a submarine circuit?—Not rigidly; it represents the nearest approach that we could get to it. Wires buried in the earth are subject to induction of the same character as those submerged in the ocean, though not to the same degree; it was very difficult without comparison to tell exactly how much less the inductive influence would be felt in those wires.
1756. Were those experiments made with battery power, or magneto-electric currents?—Magnetic electricity derived from induction coils.
1758. (Chairman.) What were the induction coils?—They were induction coils of my own; the peculiarity of them consisted in the secondary wire being placed next to the iron instead of outside, and the primary over that, as I had found by actual experiment that that arrangement will give you a much greater quantity of electricity with less intensity, and that is what I wished to get.
1759. (Mr. Varley.) With regard to under-ground wires you stated that you thought when they were joined up at the distant end you were not sure that your request had always been complied with, and that you had the wires joined up as you desired them?—It may have been so; indeed I know by testing that the wires had been joined up as desired, but as I was not on the spot to see that every arrangement was perfect, and that the conditions were carefully maintained during the experiment, it would have been more satisfactory to me had it been repeated with greater care.
1760. Would not your galvanometer show whether they were right or not?—Yes, to a certain extent; but I could not be sure whether something had not happened during our experiment at the other end diminishing our circuit.
1761. The earth would be a total one or a known wire in the office unless the office was extremely slovenly, or a wire being connected with one of the terminals or else the insulation, the testing box, whatever they used at the other end must have been imperfect?—There are other things that might have occurred, there might have been some imperfect insulation between the wires that I know nothing about, it might never have been detected in the ordinary working of their instruments; it would require extreme care and repeated testings to ascertain it, but for such an experiment as this, one ought to have been at both ends; I ought to have gone to the other end to have tested everything there first.
1762. Have you reason to suppose that any such fault existed?—I was not perfectly satisfied with the result, it was too brilliantly successful in point of speed; in truth the experiment was hastily made, it lasted a very short time, and I felt the necessity of repeating it again and again before we relied upon the result; so far as a mere preliminary experiment goes, it was satisfactory.
1763. Can you give any idea of the speed of transmission of the current through that line of 2,000 miles; I mean the length of time each current took; the speed of transmitting words?—I cannot tell you the speed of transmitting words; the induction coils were made to reverse the current alternately; the relay at the distant end was made to reverse the contact, it was merely enough to work a Morse instrument, so that every reversal of current gave us a dot.
1764. (Mr. Saward.) What was the form of signals used on that occasion?—Mere dots.
1765. How did you get your time?—It was counted by the minute; you could make so many dots in a minute. We got a large number of dots in a minute.
1766. (Mr. Varley.) Did you try different lengths?—Yes.
1767. Will you be good enough to give us those results?—We had only two evenings to try the experiments on those wires, and on neither evening had we all the time I could have wished. It was very late before we could get possession of the wires.
1768. You did not transmit any words through the length of 2,000 miles?—No; we did not attempt it; we had not arranged any code for transmitting. We used simply a reversing the key to work the induction coils.
1769. Were the induction coils that you used upon that occasion similar in construction (I do not mean as to the size of the wire) to those subsequently used for working the Atlantic cable, or were they simple Rhumhorff coils?—They were similar in construction to those used for the Atlantic, but smaller in size, and with solid iron bars instead of cylinders.
1770. Were those solid bars joined together in a continuous iron loop rope?—Yes, at each end.
1771. (Mr. Saward.) How many waves of electricity did you succeed in getting through per minute?—We got two, and two and a half a second; but I was not quite certain whether, during the whole of that experiment, the last 400 mile loop of the circuit was perfect. I do not think it was. I think there was during the most apparently successful part of the experiment from some cause, lateral contact, producing short circuit between the ends of that loop.
1772. (Mr. Varley.) That was two and a half reversals, which if converted into dots, would have been one and a quarter dots?—If converted into dots in your usual way. If you require a positive and a negative to make a dot, it would have been a dot and a quarter; but on that occasion we made every reversal print a dot.
1773. Did you attempt dots and dashes?—In truth the instrument would not give dashes; it was merely to record the number of currents. It would not have been able to give dashes; it was not so arranged.
1774. Did you at a subsequent period try experiments over shorter lengths of subterranean or submarine wires?—I tried every length I could get of submarine wires down to four or five miles, as the Atlantic cable was made from time to time.
1775. Were not some experiments tried between Dublin and London with regard to printing?—I spoke on another occasion direct from London to Dublin.
1776. Did you transmit dots, letters, or words?—It was a dial instrument, and every reversal of the current threw it one step forward; we did not try to signal words. I had only about ten minutes or a quarter of an hour, when the time came to put else instrument in circuit.
1777. You merely sent reversals through?—We sent alternately currents received upon a relay to actuate a step-by-step dial instrument. The total length of circuit was 660 miles, all subterranean or submarine.
1778. At what speed did you get your reversals through it?—I can hardly tell you; I have no accurate account. I am quite sure that we got them through as quickly as one in a second; that was the longest length of subterranean or submarine wire at that time laid down.
1779. Do not you think there would be a great deal of difference, between the simple transmission of reversals at equal periods, and the transmission of signals, requiring unequal times; for instance, such as would be required for the transmission of dots, dashes, and spaces in the ordinary Morse Alphabet?--There is the greatest difference the moment you begin to make your signals irregular, either in duration of time or strength of current; you introduce another element; you get your wire discharging itself unequally, and you are very apt indeed to lose a part of the signal; it requires very careful adjustment, and you are obliged to compensate, as it were, for the difference that time has introduced; you send into the wire, currents equal in amount; you send a second current, and if you give them an unequal time to discharge themselves, you are obliged to introduce some system of compensation, in order that the currents may arrive at the other end in equal strength, without interfering with one another.
1780. Do you think the simple reversal of currents, and the formation of dots at the distant end would give any clue, as to the speed of transmission that may be expected through a line of 2000 miles, similar to the one you experimented upon?—Yes, practically the very best clue you can get, but you must make proper allowance. I marked that carefully as to what allowance we had to make, and I found where I could get through a certain number of dots, I could not get through more than a certain number of the usual concerted signals at varying intervals. Having found the number of equal reversals I could get through in equal time, I could deduce from that, in what time I could work off the alphabet; there was about 12 per cent. difference; for instance, if I could get through 100 equal signals at equal times, I must reduce the speed of the instrument to 88, when I came to work the alphabet. I do not think there was a greater difference than that; but it was when the currents were compensated in the way I speak of, which requires extremely careful arrangement. In the Atlantic instruments there was arranged a system of compensation, which would take off from any given signal just so much as would allow for the lapse of time.
1781. (Chairman.) Did you find the results at which you arrived in your experiments upon those subterranean and submarine wires corroborated by the experiments which you subsequently made upon the Atlantic cable?—Most perfectly; we got 120 signals through from Newfoundland by these induction coils in a minute.
1782. Were those signals or words?—Simply reversals; we never got words through at that rate; those were the quickest signals that we ever had.
1783. Those were merely reversals?—Yes; that was their way of calling us; on some occasions we could see that they were trying their instruments, and it gave us the opportunity of adjusting ours, when they came to send words they did not attempt to send them as fast as that; I always explained to the clerks that it required a lower speed, and they knew exactly how much slower to work.
1784. (Mr. Varley.) When you say that you had 120 reversals a minute, do you mean 60 positive and 60 negative?—Yes, the highest speed of words we ever got actually was two words and three quarters per minute. I have an exact record of it, I think it is 41 words in fifteen minutes, it was a message occupying 13 minutes, which came through at the rate of two words and three quarters a minute, with spaces between the letters, and double spaces between the words.
1785. The usual allowance of spaces?—Yes.
1786. Do you think that speed could be maintained throughout the day without interruption?—They could have worked faster than that—the currents came faster than that—with ease, for the instruments when they got 120 reversals in a minute could have sent four words easily by our arrangement with the antecedent compensation that I introduced. The mode adopted was, the induction coils gave us at every reversal of the primary current, of course, an equal amount of secondary current; but when they were to make a dash, this, occupying time, required that there should be the compensation. I had so arranged that the key to the instrument making the dash produced the compensation exactly to the amount that was required; we could arrange or adjust it so that there should be either more or less compensation. I could see on the mirror galvanometer distinctly on one occasion that they had overdone their compensation in Newfoundland, and I sent them word.
1787. Was the over compensation by putting resistance into the circuit or cutting out some wire?—No, cutting out or neutralizing too much of the current. The truth is if you send equal currents into an empty wire, or a wire which is charged with the previous current, they will travel at different speeds; the new current which is to charge the wire will, if it go into an empty wire, tell its tale more quickly, it would tread upon the heels of the previous signal in working too quickly, and wipe out a part of the previous signal. To prevent that, I had to send in a small fraction of the current preceding, and I had to make antecedent compensation in that way, because I found that they travelled at different velocities. These facts came out very curiously in some experiments I was making at Keyham, and they were borne out when the line was laid across the Atlantic. Currents of equal strength from the same apparatus will travel at different speeds, in accordance with the antecedent state of the wire, so that the mere varying lapse of time alters the rate of travelling of different signals.
1788. Suppose the wire be charged with positive electricity, and you are sending a positive signal after that, will that increase or decrease the speed?—I never did that; the construction of these instruments was such that it must always be a negative current to go next.
1789. Would the signal go quicker when the wire was empty or when it was partly negative?—Quicker when empty. In fact, it was to retard the current and prevent it taking off the end of the dash, that I designed the antecedent compensation on purpose to remedy it, and put it into its place again.
1790. Do you think that system of compensation could be depended upon for that purpose?—Yes, most perfectly; and when once arranged, it is the easiest thing in the world. It is a mere piece of mechanical clock work, which acts automatically the instant you put down the key, and releases itself at the proper time.
1791. (Chairman.) It is entirely self-acting?—Entirely, without that arrangement our limitation of speed is much greater.
1792. (Mr. Saward.) Is it attached to the working key?—Yes; it is in fact a part of the key; the difference between simple reversals and working signals is much greater than 12 per cent. unless you can compensate in that way. With great accuracy of compensation you make the working speed nearly approach to the most rapid reversals, so that when we got 120 reversals through in a minute, we could have signalled words at high speed if very accurately compensated in that way; but without such compensation there would have been a reduction of speed to the extent of probably upwards of 40 per cent. required to signal words.
1793. (Mr. Varley.) Did you allow two currents for a dash?—The same time as two currents; we send a positive, and allow a lapse of time for what would have been the next negative.
1794. It was occupying the interval of three currents with one positive and one negative?—Yes.
1795. Did I understand you rightly to say when you were sending a succession of dots, take the letter V which is composed of three dots and a dash, that the wire is empty after making the first dot?—No, because the b1ank and space following that dot has been produced by a current.
1796. It is a longer blank than usual, as if you were beginning a new word?—We will assume that all the currents are short circuited for four or five seconds before commencing a despatch. If then the first signal required were a dot, it is necessary that in this empty state of the wire it should be compensated antecedently, else it would be converted into a dash by marking too soon.
1797. On producing your first dot you send in a current which produces its mark, and then you send in after that dot a short portion of the opposite current?—An equal amount exactly, in order to terminate the dot at the proper time.
1798. An equal amount, and a little more, do not you?—We send in the alternate currents as they come from the induction coils, to make the dots; when you come to the difference of time between the dot and the dash, you must compensate for them.
1799. You do not compensate for the spaces?--We compensate for them by the key exactly as we do for the dash, while the dots produce themselves; you have two keys at work; each key is held with the mechanism exactly the right length of time; you could not put it down before the proper time, at the time when the change took place the key was released; you could only raise it again at the proper time; there was the same mechanism for the dash key, and the blank key. If you printed the end of a word, you kept the blank key down during two intervals, the letter-space being one interval; the word-space two intervals which would have occupied the time of four currents; it is an extremely simple arrangement.
1800. In that way you got two and three-quarter words in a minute?—Yes, and we could have easily sent a good deal more; there was plenty of strength in the signals.
1801. With the battery what was the speed through that cable? —I am almost afraid to say; we had six, eight, and ten seconds for contact with Daniell's battery, before they could receive the shortest signal.
1802. What Daniell's battery was that?—We began with 480 pairs, and then I reduced them, arranging them for quantity, so that we worked from 25 to 50 cells, but never more.
1803. With 50 cells of what surface?— These batteries of Professor Thomson's were about 12 by 9, or something of that sort.
1804. Did you get the dots through in nine seconds?—It required 20 seconds to make a single dot, 10 seconds positive current to indicate the dot and 10 seconds negative current to discharge the wire.
1805. For a dash could you work through it?—It was worked; the Queen's letter went through with the Daniell battery; we did not send a single message from this side with the induction battery; whereas every message from the other side was sent by the induction cods except these four words, “Daniell's now in circuit.” Those were the only words that came from Newfoundland, except by my own induction coils and apparatus.
1806. Have you any idea of the speed of a message through, with 50 pairs of Daniell's battery, the number of words in a given space of time?—I doubt if at any time they exceeded the speed of one word in a minute; there were many stoppages in it, and many parts required repeating; they read off by half a degree of deflection upon their horizontal galvanometer.
1807. (Mr. Saward.) On the Newfoundland side?—On our side they were almost all read upon Thomson's mirror galvanometer. Some were received both by relay and by galvanometer, some came by relay alone; for the greater part of the time afterwards I kept Thomson's in circuit, and threw the relay out of circuit, simply because the other was more convenient and less liable to derangement in its action from the terrestrial currents. I was extremely pleased with it, and I printed some messages with my own hand from Thomson's galvanometer; some of them were recorded by relay with perfect accuracy; there were several recorded by relay and by Thomson's galvanometer on the same slip, so that there can be a strict comparison, and the relay in one or two instances is quite more accurate than the printing from Thomson's instrument. (Several original message slips handed inillustrating these points.)
1808. (Chairman.) Can you give any idea approximately, of the velocity of the messages received from Daniell's. battery?—It was with very great difficulty that we could get anything through, as the signals from this end did not produce a deflection at Newfoundland of more than half a degree; all that was sent from this side we were obliged to work very slowly; if they had worked from that side with a Daniell's battery to us, we should have a fair comparison; it was not fair to compare the working from the lame side of the cable with the working from the perfect side; we did not get anything like half a word a minute through to them at the best time during my stay at Valentia.
1809. (Mr. Varley.) Do you think a word might have been got through in four minutes with Daniell's battery with such a cable as that if it were perfect? —Yes; a word in four minutes could have been easily got if the cable had not been injured. It ought to be done unquestionably.
1810. With regard to the cable at Keyham, did you try any experiments with Daniell's battery there before it went into the sea?—Yes. The Daniell's battery was fitted up and used on board the “Agamemnon.”
1811. Did you record any speeds there?—I tried the velocity. I did not try to record words.
1812. What was the difference of speed that was noticed by the induction coil in the whole Atlantic cable at Keyham, and with Daniell's battery?—Daniell's battery occupied considerably more than double the time of the induction coil; it never was less than five seconds to get a signal adequately.
1813. And the other?—A second and three quarters. I tried it with extreme care; I made a great many experiments, taking various lengths. I took the very first movement produced by the Daniell battery on a large sensitive galvanometer.
1814. It was about as one to three?—Yes.
1815. (Chairman.) Was it the experiment which you made upon the Magnetic Company's wires which induced you to believe that a telegraph could be successfully laid across the Atlantic?—Yes. It was those experiments which conclusively satisfied me by confirming the results of all my previous researches. I can hardly say that I felt satisfied of it beforehand, because that was the first opportunity I had of experimenting on a large scale with wires actually laid down. I relied greatly upon the results obtained in working from London to Dublin 660 miles in a single length, not having the ends brought back again. It is far more satisfactory to work through an actual space than a virtual space, by the wires being brought back again.
1816. What results did you find in these experiments?—Though I was only ten minutes at work, I got signals through rapidly and stronger than I expected. I had sent my assistant to work the instruments at Dublin, and he received my signals at Dublin while I received his in London.
1817. You were one of the early promoters of the Atlantic Telegraph, I believe?—Yes, in conjunction with Sir Charles Bright and Mr. Cyrus Field.
1818. You were concerned in designing the Atlantic Telegraph cable, I believe?—I was very little concerned in designing it; I must disown a good deal of that; I had nothing to do with the designing the outside iron wire at all; I sent the committee a specimen of cable which I thought would have answered the purpose better.
1318. (Mr. Seward.) Did not you approve of the electrical conditions of the cable?—I am responsible for it up to a certain point, but before I gave an opinion at all as to what would be the best kind of insulated conductor, I said, “I want three months for experiment, I do not think that such an important subject ought to be decided off-hand. I do not say that I withhold my opinion, but,” I said, “I do not think it would be fair to ask my opinion suddenly, in that way.”
1820. Did not you think that a small conductor offered the best chance of overcoming retardation?—That is saying rather too much; a large conductor of the size supposed necessary to fulfil the theoretical requirements of the case was deemed impossible of manufacture, and I was asked pointedly this question, “Do you think it can be done with a conductor of the usual size?” I said, “Yes, I am certain it is possible.” The result has proved that I was correct, but I did not select the size of the conductor; it was rather selected on commercial grounds.
1821. You would have preferred a smaller conductor?—Larger, but not nearly up to the supposed theoretical requirements, for I had satisfied myself that the law which regulates the size of conductors in over-ground circuits was utterly untenable in rigid application to submarine wires, as proposed by me; and I still think so from experiments I have recently made, by which I can show the effect of surface induction alone in almost annihilating the current as compared with the effect of resistance on the same current.
1822. What were those experiments that you have made recently?—Some important ones s taking the magneto-electro currents that I purpose to work with, testing their force through a certain amount of resistance with Weber's Dynamometer, or an instrument made on that principle, I got 20° of deflection with a certain amount of force, the needle generating the current revolving at a certain speed. Then without altering the resistance at all, I increased the surface by induction plates and I lose nineteen-twentieths of the whole current by the mere effect of the surface.
1823. Can you explain how you increase the surface?—By plates of tinfoil insulated by gutta-percha, representing a surface equal to about five miles of the Atlantic cable. I found that in a surface of five miles of the Atlantic cable you can show the absorption of nineteen-twentieths of the entire current which was previously traversing through a resistance equal to about 120 miles of the Atlantic cable, showing that the surface is of immense value in consuming the current. This effect is observable only when the currents are alternately in opposite directions.
1824. Will you describe the apparatus you adopted for those experiments?—It it very simple. A magnetic needle is made to revolve rapidly within a helix. The needle and helix are therefore the generator of currents, which are equal in force and of opposite polarities alternately, positive and negative: although the currents are in opposite directions, yet the deflection produced upon the dynamometer is always in the same direction. With the revolution kept up at a certain speed you command uniformly a certain deflection; maintaining that speed you know that you have got an equal amount of current that will stand for any length of time, like a constant battery. I find that with a given deflection, if you apply to any part of the circuit these induction plates, the induction between the two surfaces is so great as to withdraw from the dynamometer nineteen-twentieths of the whole current, which remains upon the surface of those plates, and is there contained by the next current, each current in succession consuming its predecessor to the extent I have mentioned.
1825. (Mr. Varley.) What was the nature of the apparatus?—A needle and helix; a needle about 3 inches in length, and a helix just large enough to take it round with a 36 gauge wire.
1826. What was the length of the electric generator?—Very small indeed; it is about half a mile of40 gauge wire.
1827. That would be equal to how much of the cable tested?—I hope I have explained myself correctly as to the principle of this apparatus. Perhaps the diagrams which I now present will aid my description. They represent the two modes in which I arranged the circuit. There are two ways in which I tried the experiments, and in both the result came out accurately. The, first was passing the current before it came to the dynamometer, through a series of induction plates; it passed through one set of plates on one side of the helix, and another set on the other. It had, therefore, to pass through a condenser, so called, before it got to the dynamometer. In that way, nineteen-twentieths of the current were consumed. It was more remarkable afterwards, is the arrangement represented by the second diagram, when I took the circuit from this generating helix direct to a dynamometer, and arranged the wire so that you could only throw it into the condenser as a sort of cul de sac outside the circuit. When I did so it still went in the same large proportion into the condenser, although the condenser was not as part of the circuit, and only, so to speak, attached.
1828. How did you apply the dynamometer? every current passing round the dynamometer in only one direction?—No; the peculiarity of Weber's dynamometer is that though the currents be passed in opposite directions, yet they produced a deflexion in the same direction. Of course, you change the polarity of both parts at once; therefore, the deflexion remained in the same direction. The remarkable thing to my mind was that though there was perfect continuity in the circuit, from the generator through the dynamometer, yet the current preferred to go into the condenser to the extent of nineteen-twentieths, just as if the condenser had been made an actual part of the circuit. Thus it became evident that surface induction takes precedence of conduction, and exerts a far greater influence on the results than has generally been supposed.
1829. You said that the tinfoil represented 5 miles of the Atlantic cable?—Taking the mean of the internal and external surfaces. There were two sheets of the thinnest gutta-percha between the sheet of tinfoil, so that there is more inductive force; but the surface charged is only that.
1830. It would be something like thirty yards square, would it not, to be a mile of surface?—The outer surface is a square inch to the running inch of the cable, as near as possible; 144 inches in length give you a square foot as nearly as possible.
1831. You do not think that increasing the size of the conducting wire will materially increase the speed of the transmission of signals?—I am quite sure that beyond certain limits, which I cannot yet define, the opposite will be the case, unless the quantity of electricity employed be increased in at least an equal ratio.
1832. A statement was made in a printed publication, shortly after the Atlantic Telegraph was established, to the effect that when the size of the conductor was increased, the speed would actually be diminished?—I think I can tell you exactly how that statement arose. Having the cable with six conducting wires in it, I was able to use the whole of the six conductors as six parallel conductors, increasing the area six-fold, and yet increasing its inductive surface at the same time six-fold. I was conscious that it was not like increasing the sectional area of a single wire six-fold, but increasing its surface only in the proper proportion; yet the results were so remarkable that I could not but be struck by them.
1833. (Chairman.) That produced the retardation?—In every instance, when I used two wires instead of one, the retardation was greater; when I used four it was further increased; and when I used six it was greater still. It is true that allowance must be made for the increase of surface being greater here than the fair proportion due to mere increase of sectional area would be; but the obvious increase of retardation was such that it at once arrested my attention. I had expected, as others did, and as some still do confidently, that by increasing the sectional area I should have got a higher speed; at all events, if it had not resulted in increased speed, I expected that by increasing the conducting power I should have got my signals heavier and stronger. I got them weaker and slower, in opposition to all one would have noticed in overground wires. I got literally less current, and, of course, because more remained on the surface.
1834. From that experiment you think that increasing the size of the wire would not materially increase the speed?—I do not know from what sized wire you propose to start as a minimum, nor am I prepared to say how far you ought to go in that direction; there are certain limits which would put stop to it.
1835. Supposing the Atlantic cable was made with a conductor a foot in diameter?—First, I believe it to be practically impossible to insulate a submarine conductor of such magnitude; and secondly, if made, I believe that such a cable would consume, by its enormous surface for induction, any amount of current which could be thrown into it from an ordinary battery during rapid signalling. Clearly it would convey a continuous current with a facility strictly proportional to its sectional curve and conductivity, but for rapid signalling I think it would prove a failure.
1836. For theoretical consideration do you think you would increase the speed materially by such an increase of the conductor?—I think you would have to increase your battery power to a fearful extent, in order to allow for the great loss produced by surface induction.
1838. You have stated that you had requested to be allowed to make certain experiments before you should be asked to pronounce upon the core of the Atlantic telegraph; will you explain the circumstances which prevented those experiments being tried?—Chiefly the lapse of time; I was then speaking to Mr. Cyrus Field, who was the most active man in the enterprize, and he had so much steam that he could not wait so long as three months, he said, “Pooh, nonsense, why the whole thing will be stopped, the scheme will be put back a twelve month, cannot you say now that you know that will do;” I said, “But I cannot at all tell you that that is the best; I think that we are bound to find out what is the best before we go into it.” It was then determined that if I took the three months it would lose the year; he then pressed me about it and he said, “We hope that you are not going to stop the ship in this way,” I had pretty nearly determined that I would not give an opinion, but he extorted the opinion that I was quite willing to say “that it could be done with that,” but I was determined to withhold anything like an official opinion till I had made the experiments; I had got the permission of two companies who had under-ground wires of different sizes running between the same towns to experiment during the night, and I could have seen the different ratio of retardation upon those two lengths of wire in similar circumstances, when I should be better qualified to give an opinion.
1839. Did you test the core of the Atlantic telegraph?—Yes, the whole of it.
1840. When it was completed?—During its progress, as it was made in portions.
1841. (Mr. Varley.) It was not tested under water at all, I believe?—The core was always tested under water.
1842. After it was covered with iron?—No, there was no opportunity afforded of testing the cable after it was made under water.
1843. (Chairman.) Did you test it at Messrs. Glass and Elliott's works and at Messrs. Newall's?—Yes, constantly at both places.
1844. In your opinion at the date when it was first completed and coiled on board the Agamemnon in 1857 the cable was electrically perfect at that time?—I should hardly like to say that it was electrically perfect; I believe it was as perfect as we could get it, under the pressure of the then existing circumstances.
1845. Was it electrically perfect when it left the gutta-percha works?—Quite so, although there were variations noticed under our extreme tests, yet it was as perfect as could be made, and far more perfect than any that had been previously made.
1846. (Mr. Varley.) You think that the Atlantic cable underwent some deterioration in the course of manufacture?—No question about it.
1847. (Chairman.) In the manufacture of the outer covering?— Yes, chiefly from exposure to heat after it was made; but other sources of injury existed. I will point out one that it was almost impossible to avoid; during the process of manufacture we found a break of continuity—a most serious thing; after it had been covered and spun into cable, and the cable was coiled in a tank we thought this was impossible, and we had done our best to render it impossible by using a strand of seven wires, instead of one solid wire; but on examination we found that there had been a false soldered joint made, not one made in the usual way.
1848. Was that in the copper?—Yes; and it was cemented over. It had been most carelessly done.
1849. (Mr. Saward.) Did you trace the defect to the person who made it?—I showed it to the Gutta-percha Company immediately, and for their sakes Isaid no more about it. They made the most searching inquiry, adopted the most rigorous precautions, and made it impossible that anything of the sort should take place again.
1850. (Chairman.) What control had they over it?—The core was covered in short lengths, and put together with extreme care; but, in spite of all their watchfulness, this occurred, and the copper wire was soldered by some careless person.
1861. (Mr. Saward.) Do you imagine that that joint was made at the gutta-percha works?—It was a false joint; a thing made to save trouble after the first covering with gutta-percha, and the lazy rascal just patched it together before it went through the second covering, there was no real joint made. The gutta-percha went over it, and then it went through a second covering, and it was not discovered till it gave out in this way in the tank. I said, “If any careless rascal has done that once, he may have done it again; what are we to do?” There was a length of several hundred miles lying at the gutta-percha works already completed when I detected this flaw. I said, “I have no confidence in that, unless it is tested?” They said “How is it to be tested?” It was evident that it was no use testing it by the ordinary test for continuity, and I desired it to be tested with a strain, so as to draw these ends asunder. I said, “We must put the whole of this under just strain enough to detect if there is a breach of continuity.”
1852. (Chairman.) Was it done?—Yes.
1853. (Mr. Saward.) What was the strain put upon it?—I tried a great many experiments to determine how much strain would show these bad joints without producing the least permanent stretch; of course we could not produce any permanent stretch without injury. I tried to determine by the friction-brake, or some arrangement of that sort, how much strain could be put on to detect any break of continuity that occurred.
1854. Did you detect any other flaws?—Yes, we had a second.
1855. (Mr. Varley.) Do you remember how much strain you put upon it?—It was estimated at about 40 pounds. Yet, although you know how many pounds weight you put upon the friction-brake, but you cannot always tell what strain that gives; there is a source of accident you could not determine, and could not avoid. During the process of winding under strain (it could not be wound by hand, it was wound by steam) there was a large drum and a swift, from which the core was to run, when the drum began to be filled with the core, the circumference was small, and the winding took place slowly, and the core came easily off the swift; but as the drum filled, and its circumference got larger, each revolution then took off nearly three times as much from the swift as it would before; as you came to the small part of the swift, the velocity of rotation increased rapidly, and the speed of the swift was about as much as four times what it was at the beginning, so that the friction varied; we had to stop sometimes because the swift would be overrun, and there would be jerks, and the inertia then would give the core a stretch, so that it was subjected to a trial that really I should be very sorry to subject a cable to again.
1856. (Mr. Saward.) You think that subsequent experience would lead to very great improvement in respect to the core?—Yes, I do.
1857. (Chairman.) You detected most of these defects, did not you?—Yes. In some instances, however, we could not ascertain when the stretching of the copper wire had taken place, and the contracting power of the gutta-percha was put upon it; as it could not retract suddenly, it would retract gradually. I have ascertained since, that gutta-percha is to a certain degree plastic, even at a low temperature, and that is, to my mind, one of the most serious matters connected with the cable; at the same time, if any such point had escaped, it would ultimately force its way through. I have learned that you cannot be careful enough in winding or rewinding, or handling insulated conductors. If you get the copper conductor stretched beyond what the outside will admit, you are certain to have it find its way, sooner or later, to the outside. The other source of injury to which the cable was exposed, after it left the gutta-percha works, was the heat at Greenwich; the hottest day that we have had for 11 years occurred, while that cable was left exposed to the sun during the whole of a Sunday.
1858. Not placed in tanks?—In dry tanks; if it had been sheltered properly the heat would not have affected it; but the gutta-percha was seen on that day oozing out in drops between the iron wires which it was covered with. I have given the sources of accident before it was covered, but this, which was the most serious of all, occurred afterwards.
1859. From the heat to which it was exposed before it was put on board ship in the first instance?—Yes; it was actually in process of being coiled on board the Agamemnon when this injury from heat occurred.
1860. (Mr. Varley.) At Birkenhead was not it put under water?—Neither at Greenwich nor at Birkenhead was it submerged. In fact the cable was supposed at that time to be so tender that they almost thought it ought to be shut up in a glass case to keep it from rusting; the wire was so thin that there was a great anxiety to keep it from damp or water; the least rust, even half an inch of rust on the cable, it was believed, would have necessarily caused it to have parted at that spot, so that it was most carefully protected from wet.
1861. (Chairman.) Did you accompany the first expedition for laying the cable?—No; I was at one end, at Valentia; keeping up the communication. I did not go out in either the Niagara or the Agamemnon.
1862. You tested the cable after it returned from its unsuccessful voyage in 1857, did not you?—Yes; I should like to deal a little further with the injury at Greenwich, which is, perhaps, the most serious matter of the whole; if it had been tested under water after that the injury would have been detected, I feel certain. It was an intensely hot day; the hottest, Mr. Glasier at the Observatory told me, we had had for 11 years; they had sent for me early in the day to say that there was something they could not understand about the cable. I found on examining the whole of the upper layer that in parts the gutta-percha was, so to speak, exuding or sweating out in large drops, the size of a pea; it was gutta-percha evidently, softened by heat, and blackened with the tar.
1863. Was not there a serving of hemp outside?—Yes; but the gutta-percha had forced its way through the hemp and between the iron wires; it was visible upon the outside. I was in the greatest consternation at this, because they were actually paying it on board the Agamemnon; they had paid some two to three miles before they noticed this and sent for me; some had gone on board the Agamemnon that had been burnt or exposed to the sun in this way. Of course it was all most rigidly examined, and I took many pieces out, in order to see whether the internal structure was affected. Of course, if the gutta-percha had found its way to the outer surface, it was minus below. I did not know to what extent the injury had gone, whether the gutta-percha had come from the surface of the core merely, or whether, in fact, the entire core had become plastic inside; and I cut a great many pieces to examine it; every piece that I examined showed that the conductor remained central; therefore I said, these drops are, so to speak, a film that has been drawn from the outer side of the gutta-percha; the inner part next to the wire has not been rendered fluid. I thought that the thermal conducting power of the iron had drawn these drops equally from every part of the surface; all the worst pieces that I could find I cut out; in every one the conductor remained central, so that my mind was satisfied that although injury had occurred, it was not so bad as it might have been. I could not see anywhere at that time that the cable was so seriously injured as to stop the expedition, because all that was burnt in this way I cut out, and all that was the least suspicious, as far as I could; and I set aside between 40 and 50 miles on that occasion; wherever I could I examined it, and I found the conductor remaining central; that reassured me, however. Since the return of the last expedition and the failure of the cable, I have found other parts of the cable, where the gutta-percha had become plastic, and the conductor, instead of remaining central, had fallen down or been forced laterally through the gutta-percha, and the conductor was in contact with the tarred yarn (specimen handed in).
1863a. (Mr. Varley.) Do you mean after the cable was put on board?—The injury was evidently in part at least due to the repeated handling in being put on board and taken out twice.
1864. (Mr. Saward.) Was not there any precaution against the effects of heat?—A shed had been ordered, but it was not built; the estimates were, I believe, prepared: the only doubt was whether there should be something like a tent hired or a shed built over it. I know the directors had intended a shed to be made, but it was not done. Then there was extreme difficulty in this respect. In those coils of the cable that were exposed to the sun it was not merely the upper layer that was exposed, but the outside of every layer was exposed to be burnt; and if I had cut out every bit upon which the sun's rays had fallen, we should have had the cable cut up into two mile lengths all through--because no flake contained more than two miles, or two miles and a half—so that if I had cut out of the cable completely those pieces that the sun's rays could have touched, namely, the whole of the top layer, and the outside and the inside of each layer, that would have ruined it.
1865. (Mr. Varley.) Did not you notice any great increase of the escape of the current where the wires had got to the outside at any place?—I never did till after the return of the expedition trace it to the outside directly; the gutta-percha itself impregnated the tarred yarn, so that not being tested under water, we did not notice the escape; you know how difficult it is to get accurate testing when the gutta-percha is exposed to heat. In high temperatures the gutta-percha will, instead of being an insulator, become a conductor. This was being done in July and August, and it was impossible to tell how much was attributable to mere temperature, and how much resulted from that sort of accident; I do not believe any kind of research could have told you how much was due to temperature.
1866. (Mr. Saward.) Reverting to Captain Galton's question, if I understand your feeling upon the subject, you had some misgivings as to the perfection of the cable before it went to sea the first time?— Distinctly, but it was impossible at that time that they could be set at rest, in consequence of the temperature; that is a very important point; it had been ascertained a year or two before that in hot weather cables could not be manufactured without a large apparent defect in the insulation. On one occasion at Messrs. Glass and Elliot's they thought that same defect had occurred and that the cable had “gone bad;” they flooded it with cold water, and instead of showing that it was worse it got better; the expression was that it was all right again; the lower temperature improved the insulation directly. A similar appearance showed itself at Mr. Newall's works during the manufacture of the Atlantic cable and was reported by me to the directors; he had a steam boiler and an engine in fact where the cable was stored; it was a most difficult thing to detect whether any deterioration in the insulation had taken place, because the temperature was uniformly high; they were working, I believe, at times night and day with the steam boiler as well as the engine in the same room that the cable was stored; a large, allowance evidently was to be made for the temperature, but no one knew or could determine how much; and it was impossible to say whether, in a given length, the loss of current arose from an absolute fault in the cable, or from the temperature alone; nothing but plunging it into water could have shown that accurately.
1867. (Chairman.) You were not able to do that with the Atlantic cable?—No; at Messrs. Glass and Elliott's it was equally difficult; it varied a good deal, and in one short length of cable the effect of temperature was something remarkable; it showed as much as 64° loss by the detector during the day, and it sometimes went down 2½° at night. The cable that had been heated during the day cooled itself at night; therefore, I saw the utmost effect of the temperature; that piece, about 6 miles in all, only showed 2½° of defect at night.
1868. On board ship was the condition improved? —Little, if indeed at all, improved. It had been coiled on board by gas light, a temporary gas-main having been laid from the factory to the ship. There were 30 or 40 men in the hold, working night and day, and all this was going on in July. There was believed to be some heating in the cable also; at all events the hold was insufferably hot, and it was impossible, unless it had been flooded, to tell whether the defect arose from defective insulation, or was merely the result of temperature.
1869. The temperature in the hold never became improved?—Not to my knowledge.
1870. In fact on board ship you could not even tell that the cable was electrically perfect?—I could not. I never could assure myself of its perfection. There was this question of temperature which, nobody had worked out perfectly, and which I had not an opportunity of putting to the test there. Nothing but submergence could have ascertained it; it was like the cable some years ago in Glass and Elliott's tank, which during a hot day they said was bad, and the moment it was flooded they found it was perfect.
1871. (Professor Wheatstone.) If you refer to Mr. Glass's evidence, you will see that he says the effect of heat was to make the wire get out of the centre?—Of course it would in a greater or less degree arise in the manner which I have described, if it made the gutta-percha plastic enough. But it would injure the insulation even if the wire remained in the centre. Feeling this difficulty with regard to the effect of temperature on insulation, I made some time previously experiments at the gutta-percha works very carefully indeed. I took two miles of a coil of the very best insulated core they had, which at an average temperature was the most perfect we could select. I plunged it into a large tank, to which we gradually added warm water, and brought it up to 100°, and that coil of insulated core which was quite perfect at an average temperature, was utterly bad at 100°. We cooled it again, and it regained its perfection. It was as good a core as ever was made; it was not warmed enough to injure it permanently; but warmth is quite enough during its continuance to destroy the insulation.
1872. (Mr. Saward.) Do you think that warmth would destroy the centricity of the wire?—If the conductor had not been stretched, I do not think it would injure it seriously; it has no natural tendency to get out of the centre.
1873. (Chairman.) You had no opportunity, I believe, before the first expedition sailed, of testing the electrical condition of the cable?—I never had an opportunity of testing the finished cable under water at all, till after its final submersion.
1674. When it was coiled at Keyham; after it returned from its first-voyage did you test it?—I had spoken strongly of the necessity of its being tested under water, and some large tanks were built on purpose; at least, they were supposed to be built on purpose, caulked, and pitched, with the idea of making them water-tight; but when the cable was put into the tank it would not hold six inches of water; the tank being of wood and built on the ground, when it came to have these 2000 tons weight upon it, it gave in every direction; the joists and timbers werestrained and the planks gaped, so that you could put yourfingers in between them.
1875. It was quite impossible to have tested it under water then?—Yes, quite; the tank ought to have been dug out in the earth.
1876. When the cold weather came, did you test the cable?—I was constantly working through it.
1877. Was its insulation more perfect than when it first arrived?—It was much more perfect; the cold weather did, as I anticipated, improve its condition materially.
1878. When it was placed on board the ships for the second expedition was it imperfect?—Do not let me give a false idea; it was, as far as I could ascertain, absolutely free from “faults”; yet the cable ought to have been tested under water, and nothing but such testing as that would have satisfied me that it was absolutely perfect. I beg to hand in a specimen as part of a length which when tested dry showed no defect, nor indeed until after prolonged immersion in water.
1879. (Mr. Saward.) The directors were not aware at that time of the necessity for that testing, were they?—I think so, else why had they spent 3,000l. in building tanks for that purpose.
1880. They were not made aware of it, were they?—They had been made aware of it long before, but my opinions had been overruled or forgotten, I suppose.
1881. The directors were not made aware that the cable was in an imperfect state, or assumed to be in an imperfect state?—The directors knew that I had wished to test it under water. This would have removed any doubts or suspicions which had been entertained.
1882. Those suspicions were not communicated to the directors or to myself, were they?—I most say that it would have been very invidious and a very thankless office to throw out suspicion, and shake confidence in the cable when I had no direct evidence that it was imperfect, and when my wishes as to testing under water had been entirely frustrated. We could work through it, and though we knew that there was a great loss upon it, in spite of all that it was as perfect probably as any cable that had ever been made.
1883. (Chairman.) What class of tests did you use at Keyham?—The usual modes of testing; we were also working through it constantly with our experimental instruments that were to work through the Atlantic; my chief occupation was preparing, these instruments and testing them in every way.
1884. Sometimes it was worked with induction coils and sometimes it was worked with Daniell's battery, was not it?—Professor Thomson worked with Daniell's battery; all the other signals were worked with the induction coils.
1885. (Mr. Varley.) Did you not test the cable before the last expedition?—Yes, very frequently.
1886. Did not you notice anything to indicate that there were faults much nearer to one end than to the other?—There were, I am confident, no “faults” which could be detected without submersion. There were some parts that I was uneasy about, but I could not prove anything about them.
1887. When I saw the tests that had been made at Valentia, there were some distinct tests of a piece of 1,000 miles, a piece of 500 miles, and a piece of 200 miles, and the whole cable Of 2,500 miles, and those tests all pointed to a fault existing about 500 or 600 nobles from one end of the cable?—I know there were some parts not so perfect as others; we could only find that there was a general imperfection spread over a considerable part; there was not at that time, I am quite confident, anything like one definite “fault.” I always said that a length of 200 miles, one particular coil, was about our worst length; we tested it in various ways, and every part tested equally; the length in question was divided into four parts, which were each tested from each end; these were again subdivided and similarly tested with great care, no one part showed more loss than another. It was evident, therefore, that that length was as a whole less perfect than the rest, and it was arranged that it should be the last to come out of the ship in paying out, so as probably not to be used at all. The great loss of cable on the occasion of the first expedition caused this length to be almost entirely payed out; about 20 miles only being brought home in the Agamemnon.
1888. (Chairman.) You tested the cable immediately after it had been submerged, did not you?—I did.
1889. In what state was it then?—It was most difficult to test; they left their current on at the Newfoundland side for many hours, and during that time we could not test it.
1890. What was the reason for leaving the current on?—To assure us that it was all right. They were not at that time ready either to send or to receive from us.
1891. (Mr. Saward.) Was that done for some hours or for some days?—I cannot say at this moment.
1892. Would it be for a few hours or a day or two?—It was almost constantly on with occasional reversals and ship signals till the fourth day. We got a strong deflection by voltaic currents from them until the fourth day, when they began to send currents from the induction coils.
1893. (Chairman.) When the currents first came were they strong?—Yes; it was impossible till the reversal came to say that they were not earth currents; when the reversals came then we knew they were keeping up the ship signals; keeping up the current for ten minutes and then at a particular time reversing. When they lid not send us the ship signals they sent us this constant current while they were unshipping the instruments and getting them into order.
1894. At first the currents were very good and very strong?—Yes.
1895. How long was it before they began to show symptoms of weakness?—As soon as they began to send these coil currents, on the fourth day, we put a relay in circuit, the relay spoke out loud and there did not seem to be any defect at all.
1896. When was the induction coil put into circuit?—As soon as they took off their battery, and ceased to send the ship signals, they began to send the coil signals.
1897. How long did they send strong currents from the coil signals?—For several hours they did that; they simply sent us alternate currents, and now and then attempting e letter. They had not at that time got their apparatus in proper order; they had taken out the experimental one with the coils, and they just kept us alive as it were, counting these signals. The signals were very strong; they made the relay speak out loud, so that you could hear it across the room.
1898. For how many days did that continue?— When we began to find that the relay required adjustment constantly, owing to the variation of the terrestrial currents, then we got Thomson's instrument also into circuit,—as far as I can remember two or three days; and we still got signals on the relay sufficient to print perfectly for us to read from; but we were perplexed in not being able to speak to them.
1899. Could not they receive any messages from you?—They did not for several days afterwards.
1900. How many days?—I have notes of the first day when we established perfect inter-communication with them.
1901. What day was that?—They continued saying to us we do not receive your signals, we cannot read them, send slower, which showed us evidently that they were not receiving properly from us, and we could not communicate with them; they then desired us, if we understood them, to send a battery current for five minutes; of course, as soon as they had stopped, we sent a battery current for five minutes and then a reversal; they could make no mistake, and we opened a communication in that way. I then, from comparison of the results at both ends of the line, felt certain that there was a fault at our end, and that therefore we could not work our induction coils with advantage, and I determined to use the lowest intensity we could work with. I arranged for the signals to be worked with 25 cells; we did work with as many as 50, but never more than 50 by my desire.
1902. What size of induction coils did you use?— Five feet.
1903 Do you consider that the strong currents sent from the induction coils produced any bad effect upon the cable?—Had the cable been sound, it would have been impossible, for the coils to produce any injury; but where injury already exists, there any powerful current will augment the mischief. I do not think that they had previously produced any bad effect, or it would have been manifested at Keyham. The same coils had been used there.
1904. (Mr. Saward.) Would not there be a difference between testing the cable at Keyham and in the water, on account of its being submerged?—Of course, both in consequence of the penetration of the water under pressure of depth, and also more especially from the fearful strain which the cable had undergone in the process of submersion.
1905. (Chairman.) What was the exact construction of the induction coils?—Iron cylinders very carefully insulated with gutta-percha, upon which the secondary wire was wound insulated, and over this the primary wire. The induction coils were laid parallel and connected end to end; the distance of their striking was less than a quarter of an inch. I never got them to strike at a quarter of an inch.
1906. What was the size of the wires?—Eighteen gauge for the secondary wires, and 48 parallel wires of 14 gauge for the primary; it was a large circuit.
1907. (Mr. Varley.) If I remember rightly, when I saw those coils they gave a spark about a quarter of an inch in the air, and that spark lasted for a considerable portion of a second?—It did, it formed an electric arc.
1908. Do you think that applying such a power as 400 cells of Daniell's battery would be likely to injure the cable?—I do; I do not hesitate to say that in my opinion it would injure it far more than a spark from an induction coil; the battery would give you an almost unlimited amount of current in consequence of the duration of the contact. I mean, as long as you continue the contact, so long you would have tension; with the induction coil current it is limited in amount; its very existence is momentary and cannot be maintained.
1909. What is the difference during the time the electricity is flowing out of the secondary wire; is the tension capable of giving a spark through the air at a quarter of an inch, and that flying from a battery of equal tension?—It becomes a mere question of duration and amount. I believe that a far higher tension can be produced in the cable by such a battery than by any single shock or spark from an induction coil.
1910. Do you mean that its tension immediately falls on very much?—Instantly; because in relation to the surface to be charged and the conductor to convey the current, the amount of electricity evolved at a single spark is almost insignificant when compared with that from a Daniell's battery.
1911. I have been informed that those who had the misfortune to touch the cable at the time when the current was discharged from the induction coil received so severe a shock that they have been almost ready to faint from it; is that correct?—Most certainly they would suffer for their carelessness if they touched the conductor or any bare wire; but certainly not if they handled it discreetly by the gutta-percha. If there were a defective place, of course you would get a great deal. We used to get it from the battery at times; the secondary current, in some way, in spite of good insulation, used to find its way into the battery.
1912. You said you thought that there was moredanger arising from Daniell's battery of 400 cells; are you aware whether Daniell's battery of 400 cells will not give a spark that will jump through a greater distance than the 200th of an inch in the air?—I am perfectly aware that you cannot get the voltaic current to take the initiative in striking, as a coil does; but under the conditions which we are considering, I believe that you can excite and sustain a much greater electric tension throughout the whole cable by the battery of 400 cells than by a coil.
1913. (Professor Wheatstone.) The face of your induction coil must have been enormously greater than that of a battery of 400 elements?—Such a battery after contact once made will give you an arc with greater tension effects than the coil, and with far greater striking distance.
1914. With Rhumkorff's coil, a spark is much greater than that produced by a continuous current from a battery?—The spark is longer, but I believe the real power exerted is far less.
1915. (Chairman) If some slight injury had been produced by the use of the induction coils, would the subsequent use of Daniell's battery have increased it?—Suppose a slight injury existed, any strong current would injure it. That was my reason for reducing the use of the Daniell's battery to 50 cells as the maximum. The Queen's message was sent across by 25 cells at my desire. I set aside the use of the induction coils upon the same principle which induced me to limit the number of Daniell's battery to 25 pairs.
1916. To what do you attribute the failure of the cable?—I think the combination of causes I have mentioned, with the subsequent straining and stretching during paying out, is quite sufficient to account for its failure. The risk incurred of stretching the conductor is one upon which you can hardly lay too much stress in the frequent winding and unwinding, coiling and uncoiling, in the way I have mentioned.
1917. You mean stretching the conductor by which the copper would have a tendency to be pushed outside the gutta-percha?—Yes. The result of this commonly manifests itself by a sudden starting out of the conductor, but it may show itself only after a lapse of time.
1918. Had not the cable remained comparatively perfect for several days until the use of the induction coils?—No. I do not think you can interpret it in that way. I am quite willing, if it is so, to let it be so. The first time I was able to test after submersion, I found considerable loss upon the cable at the Irish end, and said, “We will go very cautiously to work about this; we will reduce our intensity; we shall probably not be able to signal at all with these induction coils; we will use Daniell's battery, at the lowest power.”
1919. (Mr. Varley.) If you thought the Daniell's battery more liable to injure the cable than the induction coil, why did not you reduce the power of the induction coil?—I had not the means of reducing the length of the secondary circuit in the coils, nor could I increase the quantity or duration of their current; whereas I could reduce and re-arrange the Daniell as I thought necessary.
1920. (Mr. Saward.) Was not the other side using induction coils?—Entirely, and from the first until Professor Thomson sent out and had the Daniell's put on.
1921. They had no other, had they?—Every message came by the induction coil except the four words I have mentioned, without any injury to the cable at that end during the time they were used. The first careful testing made at Newfoundland after the use of the 480 Daniell's cells showed that injury had commenced at that end also.
1922. (Mr. Varley.) You have stated that an induction current would be more likely to produce injury at Keyham than under water?—No; I think you must have misunderstood me. I said that the immersion under water would not increase its liability be injury from the use of the coils at all; that the iron would produce induction more than the water, and the iron was equal, at Keyham, as when it was submerged.
1923. The power of the induction coil would be wholly unable to jump through such a thickness of gutta-percha; it appears to me that it arises from the passage of the current, rather through any moisture that may be in the fault, and not from its power of perforating the gutta-percha?—No; I quite feel with you, and therefore if any fault exist, it is desirable, in order to avoid increasing it, to work with the lowest power you can signal with. I did not put any limit to that; that was the principle I adopted, and suspecting a fault to exist as they did not receive our signals, I at once disused the induction coil, and reduced the battery power in that way.
1924. You do not suppose that there is any difference between an induction coil or a battery with respect to the electric tension?—I regard an induction coil somewhat in the light of a unit jar, by which you can throw in a definite amount to the larger jar (the cable), and which being alternately of opposite polarities can never accumulate therein. Whereas a prolonged contact with a powerful battery may surcharge it even to overflowing. I think it cannot excite in the whole wire anything like the amount of tension that the continued action of the Daniell battery would.
1925. In the case of a dot made with an induction coil, that dot would last for a shorter period than the complete discharge of that coil. You say it lasted for a second, and then made the dot?—The duration of the secondary current depended upon the revolutions of the commutatae, which made the primary contacts; therefore, if working the induction coil at the rate of 120 currents in a minute, it could only be of half as second's duration, because it would be cut short.
1926. Therefore, when you are making a contact for the dot with the induction coil, you have the current on the whole of the duration of the dot?.—Yes, you have half a second contact.
1927. In making a dash you have only half a second contact, instead of the longer contact you would have if you had used Daniell's battery?—True; half a second's contact and a pause of a second, during which no current was sent into the wire. We could not make a signal in less than five seconds with Daniell's battery. I think that the lapse of time and the continued contact of the battery charges the whole cable much more highly than a single blow of the induction coil; at all events, this is proved by actual experiment; that the very Daniell battery we were using would increase the size of any aperture made by the induction coils fourfold. Where the induction current had produced a certain amount of injury, in a faulty bit of cable purposely subjected to this trial, then the Daniell battery was immediately after applied, and the injury was found to be increased fourfold; that was the result of the experiment.
1928. (Professor Wheatstone.) If you had not employed your induction shocks previously, do you think the Daniell battery would have been sufficient to have produced the effect itself?—I do; we had made a hole in the gutta-percha on purpose; in fact we had pricked it; then, when the induction coil had been put on and a number of currents had passed, and it had done its worst, we put on the Daniell battery and increased it four-fold.
1929. Would putting on the Daniell battery originally have produced the same effect?—Neither induction current nor the Daniell would produce the aperture; but when once made, either alone was sufficient to enlarge it. We made the aperture on purpose, and the current of the Daniell battery without the coils produced this effect three or four times.
1930. (Mr. Varley.) I have noticed this effect in testing gutta-percha under water, passing the current from a battery of excessive tension, the passage of the current softens the gutta-percha and it falls away; that is the way in which a powerful current will rapidly increase the aperture. I want to understand why, as a general impression seems to exist abroad that an induction current would not produce this effect, the induction current lasting for half a second, a current of much higher tension should not produce the same effect as Daniell's battery lasting the same time?—I believe the
1931. 700 cells were certain to injure a wet surface?—480 cells in series were used by Professor Thomson’s orders upon the Atlantic cable.
1932. Was not that used after else cable failed?—It was intended to be used when his instruments were ordered before-hand.
1933. I think you have stated that it required five seconds to get a signal through from Daniell's battery, five seconds positive and five seconds negative?—It required at Keyham five seconds before I got a movement on the galvanometer, if you charged it, the cable, with either current, it took five seconds before the first deflection was produced by the reversal.
1934. Therefore you could not get dots in less than five seconds?—There was five seconds retardation, and it required considerably more than that to make a dot and its space.
1935. I asked that question with a view to ascertain the speed obtained by Daniell's battery?—Much depends upon circumstances, and it is difficult to answer that question; we have no accurate data to proceed upon. It would not be doing justice to the battery current to estimate it, except in the way of comparison of the velocity with regard to the working speed attainable in either way, inasmuch as it would require special instruments to do full justice to it, and these we did not possess.
1936. Have you absolute proof of that fact?—In a variety of instances I have tried it, and always found five seconds at Keyham; of course, in the Atlantic cable when laid down we could not take signals in that way, as you could not be at both ends at once; but I have proof that we received 120 currents in a minute from the induction coil; and at that rate, with proper antecedent compensation, we could have sent words at the rate of 100 currents a minute, I am certain of it. I am taking merely the two operations.
1937. Did you try any experiments with the Atlantic cable to see the speed at different lengths?— Yes, and they corresponded very nearly with those I had made out before at Messrs. Glass and Elliott's, they are as nearly identical as possible. The retardation is very nearly in the simple ratio as the distance, not as the square of the distance, and the working speed bears a strict relation to the amount of retardation.
1938. Can you give the result of those experiments? —They corresponded so closely with my previous experience, that I did not think it necessary to multiply minute observations and records. I was satisfied with the information derived from the previous experiments. I had made hundreds of other experiments on shorter lengths, on the lengths that Messrs. Glass and Elliott made the year before.
1939. (Chairman.) I believe you made some experiments upon Messrs. Silver's mode of insulation?—Yes, I have.
1940. What is your opinion of their mode of insulation?—At first it seemed to be quite perfect, but the specimens which I have had from them have not stood the test of time. From some cause the rubber was softened and became tarry.
1941. Next the wire?—Yes, apparently beginning at the wire.
1942. (Mr. Varley.) Have you found its insulating powers decrease in consequence?—It is so altered in its consistence that there is danger of the wires coming through completely.
1943. (Mr. Saward.) Have you tested Messrs. Silver's plan by keeping the current on continuously for a long time. I have been given to understand that a current kept on continuously for a lengthened period through wires insulated by their method injures the India-rubber?—I have not made any experiments of that sort. I know if any injury has begun, you are, so to speak, forcing it; by keeping it constantly charged in that way you can increase the injury to any amount.
1944: I am not speaking of an intense current or of a weak current, but as current of electricity constantly on?—I have not tried that, I have tried india-rubber under very considerable pressure for a certain length of time. I think for above a week I kept up the pressure of very nearly a thousand atmospheres upon a ten feet length of Messrs. Silver's wire, and it seemed to stand it perfectly. Maintained a pressure of never less than 12,000 lbs. to the square inch.
1945. (Chairman.) What thickness was the india-rubber?—A 16 gauge wire double covered to gauge ¼ inch.
1946. (Mr. Varley.) Did you attempt in any of those experiments to put windows in the apparatus to see what was going on inside?—I should have blown the windows out.
1947. That is 3,000 or 4,000 lbs. upon the square inch?—I know these would have burst. I did not expect to see anything, therefore it did not occur to me to put windows. I have made some other experiments with india-rubber, which have surprised me very much; I have noticed that india-rubber on continued immersion in water undergoes a change in its external appearance. I find that india-rubber absorbs water to a very large extent; I have specimens of native that have absorbed from 12 to 16 per cent. I have other specimens of cold vulcanized india-rubber that have been in water 12 weeks, and one specimen has absorbed 12½ per cent. of water in that time, but then it was an exceedingly fine film, so that I should have the maximum of absorption saturation in the shortest possible time.
1949. (Mr. Saward.) Was that by mere immersion?—By mere immersion I found that the other larger and thicker sheets absorbed just as much but more slowly in proportion to their thickness; the greater the surface you have, of course the more rapidly it will absorb.
1949. (Chairman.) Do you find when it was submitted to pressure that it would imbibe much water?—I did not try that. I believe the effect to be produced by a chemical rather than a mechanical force. It seems to become hydrated.
1950. (Professor Wheatstone.) How did you determine the quantity of absorption?—By having every piece carefully weighed.
1951. Are you quite sure that the water went through the inner substance of the india-rubber and was not merely adhering to the surface?—First of all I weighed the india-rubber after having dipped it in water for five or ten minutes, and dried it on a soft Turkish towel and then on blotting paper. I got the whole of the surface perfectly dried, and then I weighed it again; this was the standard weight to which I subsequently referred each piece. I then immersed it, and day after day I dried it in exactly the same way and weighed it, noting the increase.
1952. (Chairman.) Have you experimented upon gutta-percha in the same manner?—Yes, and it does not absorb anything; I cannot get 250 square inches to absorb a single grain of water it may be porous, but it is not absorbent; india-rubber is not porous, but is absorbent.
1953. (Professor Wheatstone.) Do you find the insulating power of india-rubber affected by this absorption of water?—That will require a considerable time to be tried. The insulating power is not apparently lessened until the hydration has penetrated completely through the entire substance of the rubber. It does then appear to be a less perfect insulator, but I am not at present prepared to say to what extent.
1954. (Mr. Saward.) Would any insulating medium which can be worked in such a way as not to be passed through tanks of water have that advantage?—Most probably.
1955. (Professor Wheatstone.) What do you consider to be the best insulating material?—Vulcanite, though heretofore there have been difficulties in the practical use of it for wires. I cannot get it to absorb water; the ordinary cold cured vulcanized rubber does absorb water, but I have not found anything equal to vulcanite for insulation.
1956. (Mr. Saward.) Are you at all familiar with the article made under Mr. Wray's patent?—It is a compound of rubber.
1957. It is a compound of rubber and shellac?—I have some now under examination by hydraulic pressure and otherwise, and am not yet prepared to offer an opinion upon it.
1958. (Mr. Varley.) Have you any idea as to the ratio of vulcanite as an insulator when compared with gutta-percha?—I believe it to be equal to pure and perfectly dry gutta-percha or glass.
1939. (Mr. Saward.) What are the component parts of vulcanite?—India-rubber is the basis; then there is some sulphur compound; they may vary it. By stopping the baking process early you get the soft vulcanized india-rubber, and by baking it further you have what is called the hard rubber or vulcanite.
1960. Do not you think that the sulphur would be an objection?—Its action upon the copper can be avoided entirely by tinning the copper wire.
1961. (Mr. Varley.) Do not you think the sulphur would ultimately be taken up by the sea-water, and there would be a danger of leaving the compound porous. In the atmosphere you would have the smell of sulphur very strong?—Not from the vulcanite. You may have a perfectly polished surface. I have had some for years, and there is not the slightest appearance of sulphur in the form of bloom upon it, as there is upon the soft vulcanized rubber generally.
1962. My reason for asking the question is because I have used a vulcanite comb fur some time, and after the varnished surface seemed to be worn off the smell from it was very strong?—I am not aware that it is ever varnished. The constant contact with animal oils or grease may, I think, possibly have produced the effect you mention. I have had vulcanite on purpose to examine carefully into it, and I never saw the least bloom upon the surface.
1963. (Professor Wheatstone.) Is vulcanite flexible enough for coiling?—It is very resilient; it is not so flexible as vulcanized rubber. I have seen one specimen of wire beautifully covered, which, I think, is quite as flexible as it need be.
1964. (Chairman.) Who covered it?—Mr. Hooper.
1963. (Mr. Varley.) Would it get into a kink without danger of breaking?—I think you might snap it possibly by sudden violence.
1966. (Mr. Saward.) Would not it be a bad material to pay out?—I do not think it would be difficult to pay out safely; it is just a question that can be determined practically; it is so good an insulator that it is worth while to try.
1967. Do you know anything of its expense?—I do not. I think the objection to the action of sulphur upon the wire can easily be got over by tinning the wire or coating it with zinc. I have seen recently another specimen in which vulcanized soft rubber had been actually made adherent to the metal as perfectly as if it were incorporated with it. You cannot detach it. Mr. Hooper showed me some taps the other day that he used years ago for his ordinary hydraulic beds, and the material had incorporated itself perfectly with the metal.
1968. (Mr. Saward.) What is your opinion as to the external form of a cable for an Atlantic telegraph?—Distinctly that there ought to be hemp in it; it should be of light specific gravity; it should neither stretch nor shrink, and should be flexible.
1969. Would a cable of this description meet your view (No. 9), that has steel inside; would you consider that a suitable cable for the Atlantic?—Yes; it would be a very good one; it approaches very nearly the specimen I sent for the Atlantic before it was decided upon. I sent in one with iron wires coated with hemp spun into a strand.
1970. Your ground of objection to the old Atlantic cable was on account of its want of buoyancy?—Chiefly on account of its specific gravity; you may add any unlimited amount of iron, but you do not increase its strength, because you load it exactly in the same proportion; for the Atlantic I should prefer a little more hemp than that. I do not know what specific gravity that is.
Return to the Wildman Whitehouse main page
Last revised: 22 April, 2010