The Cable Under The Sea
It took more than a decade to lay a telegraph wire across the Atlantic, but for moderm international communication, that was just the beginning
ON JANUARY 8, 1815, AMERICAN troops under Andrew Jackson won a great victory against an attacking British force at the Battle of New Orleans. Sharpshooters and cannon behind well-defended positions brought down the staid redcoats by the hundreds, until the famed British discipline broke. The only problem with this success was that representatives of the two belligerents had signed a peace treaty in Belgium two weeks earlier, and only formalities remained before ratification.
The extracurricular battle wasn’t surprising, because information between continents never moved any faster than a ship at sea. Tramp freighters cruised from port to port hunting cargoes, because they had no way of knowing where freight had piled up. American newspapers got their news from Europe by two-week-old mail or from arriving passengers.
In 1854 a small number of visionaries undertook to change all this with a telegraph cable bridging two thousand miles of the North Atlantic. They started manufacturing and laying it before they knew what the difficulties were and what heroic solutions would be required. For years nearly everything that could go wrong did. Cable developed maddening intermittent faults being laid. It broke of its own weight in midocean. A cable laid in 1858 worked for one month and then died. After the Civil War the promoters strengthened their cable, chartered the largest ship in the world to carry it, and tried again. Success finally came in 1866. Many more long-distance submarine cables followed, eventually boosting carrying capacity a hundredfold and more. The first transatlantic telephone cable arrived in 1956. And another historic cable project is going across the Atlantic now: the first fiber-optic cable, with a capacity of nearly forty thousand simultaneous conversations. Fiber optics promises to give new life to submarine cable communications, which have been losing ground to satellite networks.
To no single person or country belongs the credit for the telegraph. Electric telegraphy began with the invention of practical batteries in 1800. Early approaches used a multitude of wires. Some used twenty-six, one for each letter, with tiny sparks, needles, or cork balls nudged by a charge for indicators. The greatest advance, about 1830, was the application of electromagnetic theory to the apparatus. In America Samuel F. B. Morse seized upon magnets to devise a system using a simple open-or-closed key to transmit and a receiver that printed the dots and dashes of his simple code on a strip of paper. He used only one wire, with the ground serving to return the current.
In 1842, just two years after he got his patent, Morse laid an experimental underwater line across New York Harbor to Governors Island. It wasn’t the very first water crossing, but his one-wire design was well suited for submarine telegraphs, which required an absolute minimum of wires. Landlines could use crudely insulated or even bare metal wires, but submerged cables required excellent and expensive insulation.
Eleven years after Morse’s patent the English succeeded in laying a cable across the English Channel. That cable used another invention fresh out of the patent office: a German cable-coating machine invented in 1847 and first applied to electric cables used for detonating harbor mines. The inventor drew on a new type of insulation, gutta-percha.
Gutta-percha, a distant relative of rubber, had just reached the West. It came from an obscure tree that grew only in the jungles of Malaya, Borneo, and Sumatra. Condensation and coagulation of the milky fluid of the tree yielded a rigid, strong material. Water didn’t hurt it. Its insulating qualities even improved in cold, deep water. In fact, its greatest enemies would prove to be sun and air. For the next seventy years nothing could match it for submarine cables.
With this basic technology available, people began pondering the possibility of sending messages across the Atlantic. The first to act was a British engineer, F. N. Gisborne, who started with the modest plan of faster communication between the United States and Newfoundland. By getting the latest information to and from European packets stopping at Newfoundland, his company would slash several days off the information transit time. An America-bound message, just off the boat from London, would start south on a telegraph across Newfoundland. A carrier pigeon would cross the Cabot Strait to relay it to a telegraph station on northern Nova Scotia, which would pass it on to New York.
But surveying a route four hundred miles long across the wilds of Newfoundland and laying the first forty miles of wire proved so expensive that Gisborne had to go to New York for more venture capital. A fellow engineer introduced Gisborne to the wealthy businessman Cyrus Field. They met on a winter night in 1854.
At thirty-four, slim and thoughtful-looking, Cyrus Field resembled a poet more than the exhausted executive he really was. The son of a Massachusetts preacher, he had left school at fifteen and taken a job as an errand boy in New York. Three years later he had entered the wholesale paper business. At twenty-one he made junior partner in a New York paper firm, but it failed shortly afterward. In twelve more years of six-day workweeks, he paid off all his debts dollar for dollar and accumulated a fortune besides. Health failing from the pace, he retired to travel and reacquaint himself with his family.
Field’s greatest adventure began that night. After Gisborne described the Newfoundland project, Field returned home and consulted a large globe in his study. He realized that the first transatlantic cable would have to cross Newfoundland. Why not stretch Gisborne’s cable on to Europe? He wrote to the leading experts on telegraphy and oceanography. The replies were encouraging. Samuel Morse came to New York to see Field and explained why he believed an electric impulse could get through such a cable.
The Navy’s océanographie authority, Lt. M. F. Maury, replied to Field after consulting a recently completed survey of the North Atlantic between Newfoundland and Ireland. The sea bottom was mostly level, he found, not too deep, and covered with a sediment that would be easy on cable housings. He couldn’t be much more specific because the soundings had been taken fifty miles apart. But it was enough for Field, who immediately started gathering fellow investors and effectively assumed control of the company. Over the next thirteen trying years, he would cross the Atlantic more than forty times, suffering from seasickness on each trip.
In 1856, after two years, Field’s company finished laying its British-made Newfoundland cable. Realizing that from this point American support would not be enough, he crossed the Atlantic to make contact with financiers, British cable manufacturers, and the crown. He returned with generous help. After a brisk effort in Congress, the company had the federal guarantees it needed. It was ready to make cable.
To the early Victorian mind the project was enormous. Leading electrical scientists disagreed about how a submerged twenty-two-hundred-mile wire would behave and how much power would be needed. Morse himself vastly overestimated the carrying capacity of the cable and therefore its profitability. But he was much closer to the truth than the astronomer royal, Sir George Airy, who believed that the cable would neither sink to the bottom nor conduct messages.
Field’s Atlantic Telegraph Company gave its British cable manufacturers—cable wasn’t being made in America yet—a paltry four months to manufacture the line, because American backers were anxious to do all the laying during the relatively calm summer months of 1857. Specifications for construction and materials were loose. The two companies working on the outer coverings forgot to check with each other about how to spiral the metal strands and discovered when it was too late that they had laid them in opposite directions. The mistake required an elaborate splice to keep the cable from unwinding underwater.
The first Atlantic cable, which took three attempts in 1857 and 1858 to lay, consumed 367,000 miles of iron wire and 300,000 miles of tarred hemp. At the heart was a single conductor, seven strands of copper. (The Atlantic Ocean, a much better conductor than the impure copper used, would return the electrical current.) William Thomson (later Lord Kelvin), a director of the company and a renowned physicist, strongly urged using a heavier copper conductor, but expediency won out. And the cost cutters had the support of Morse and Michael Faraday, who wrongly believed that a thicker conductor would require more current. Surrounding the seven-strand copper core were three layers of guttapercha, covered in turn with tarred hemp. The outermost layer was made of iron strands, for tensile strength.
One hundred and twenty crewmen, working for three weeks, loaded twenty-five hundred tons of cable aboard two steam-driven ships. One was the frigate Niagara , the U.S. Navy’s biggest ship. The other was the Royal Navy’s Agamemnon . The plan was for the Niagara to begin laying its cable off Ireland; half-way across it would splice to the British vessel, which would continue to Newfoundland. They kept in constant touch with the station at Ireland, checking the cable for continuity. Thomson and his group were able to locate faults in the thin cable (about the diameter of a garden hose) by using a Wheatstone bridge, a recent British electrical invention, to measure the resistance of the cable between the ship and the break, and converting that figure into miles.
There was something hypnotic, no doubt, in the sight of the black cable slipping into the ocean and connecting the ship with land far over the horizon. Here was the first human track upon the ocean. But three hundred miles out of Ireland, the brakes on the Niagara ’s paying-out mechanism grabbed too firmly, and the thin black line broke and vanished beneath the waves. The ships returned to England and discharged their cargo onto the docks of Plymouth. The company ordered seven hundred more miles of cable and rebuilt the cable-handling machinery. Further, it was decided that next time cable laying would begin in mid-ocean, to allow splicing under controlled conditions.
In June 1858 the flotilla met for a rehearsal session in the Bay of Biscay, and then set out once more. A storm along the way nearly sank the Agamemnon —which had 250 tons of cable unwisely stored on the main deck, repeatedly causing it to roll forty-five degrees off the vertical—but the ships made the rendezvous. Before they could complete the difficult splice, the cable broke twice, first at the Niagara ’s machinery, where a second splice was made, and then on the sea floor, where a third splice followed. Finally, three hundred miles later, it broke again at the Agamemnon ’s stern, and the ships returned once more to England. Field and Thomson immediately reported to the company’s directors in London. The board chairman suggested the company sell the cable and quit. A vice-chairman resigned. But under Field and Thomson’s influence, the board voted to try again immediately, while the crews and cable were still available.
The third expedition was a temporary success. The gentlest possible handling of the cable kept it intact across the ocean. On August 10, 1858, for the first time ever, the British Empire had instantaneous contact with its lost colonies. The celebrations started almost immediately and reached such a height in late August that New Yorkers accidentally burned the cupola off their city hall with bonfires and fireworks. Speakers hailed Cyrus Field as “Cyrus the Great” and compared the cable to the discovery of America and the invention of the printing press. The company’s chief engineer was knighted.
On September 1, at the peak of excitement, the cable stopped working. Measurements showed that about 270 miles off Ireland the copper core was now open to the ocean. It had worked for one day short of a month. It had passed about four hundred messages, most of which were between telegraph operators trying to straighten out garbled telegrams. Queen Victoria and President Buchanan had exchanged greetings.
The public and press immediately heaped abuse on the cable company. Some charged fraud. But the experts knew a great hurdle had been passed. An Atlantic cable definitely could be made to work. It would save vast amounts of money; a single message to Canada canceling the movement of two British army regiments to England had already saved the crown a quarter of a million dollars.
Commissions met to examine the failure. The reason was a combination of manufacturing defects, handling damage during the three expeditions, and injury to the insulation from excess voltages. The latter had been the unauthorized inspiration of Dr. Edward Whitehouse, the project’s first electrician, who wired together batteries to put two thousand volts across the cable (he needed the high currents to operate a printing receiver of his own design). The company banished him from the project. Not only did high voltages damage the cable, they charged it up like a huge capacitor, and this charge interfered with the passage of the pulses. The queen’s ninety-eight-word message to President Buchanan took sixteen and one-half hours to transmit with Whitehouse’s recorder in use.
In fact, just a few volts were sufficient to get a clear signal through the cable. What made the weak cur- rents practical was a supersensitive receiving instrument that Thomson had invented in early 1858 after noticing how slight movements of his monocle bounced light around. The device’s only moving part weighed less than one five-hundredth of an ounce.
Thomson’s “marine galvanometer” was small enough to sit on a tabletop. Current from a telegraph wire passed through a coil of fine copper wire. The coil’s magnetic field applied a twisting motion to a tiny iron magnet suspended inside from a silk thread. A mirror attached to the magnet reflected a thin beam of light from an oil lamp across the table. The light, flashing to the left and right of a zero point, indicated the dots and dashes of Morse’s code. It took two operators to work the receiver: one to watch the light and call out the letters and another to write down the message. It raised the transmission speed by about tenfold over old methods.
But all the lessons gleaned from the failure weren’t enough to keep the project going during the next several difficult years. Cyrus Field’s paper company went bankrupt during the depression of 1860, the Civil War began and stopped all refinancing, and for several years the entire project languished. One competing group started work on a telegraph route to Europe going the other way: by land from British Columbia, Canada, to St. Petersburg, Russia, with a short submerged section across the Bering Strait. The group started surveying in 1865 and soon began planting poles.
But by then the Atlantic Telegraph Company had reawakened. After two years of war the British and American governments decided the cable was more necessary than ever. Field returned to England and rounded up old acquaintances and financiers. In 1864 contracts went out for a stronger cable, a little more than an inch in diameter, with a conductor three times heavier. The outer iron wires would be galvanized against corrosion. It would be built in one piece twenty-seven hundred miles long.
And five thousand tons heavy. There was only one ship on earth to transport and lay such a cargo: the Great Eastern . It was a product of an era when a single engineer could design and build with equal success tunnels, bridges, railroads, and a whole new class of ship. In this case the engineer was Isambard Kingdom Brunei, already famous for his Great Western Railway. He had designed the Great Eastern to transport four thousand passengers to the East Indies without recoaling. The 22,500-ton iron vessel first saw water in 1858 amid extravagant press notices. Everything about it was grandiose, from its sixty-foot paddle wheels to the six great masts. It was five times bigger than the next largest ship of the era, and nothing would outweigh it until the Lusitania in 1906. But the Great Eastern began life under a cloud that only turned darker with the years. It took three months to launch and broke Brunei’s health; he suffered a stroke aboard ship shortly before the first voyage. Two days later a steam explosion in a funnel killed several crew members. Even more unfortunate was the fact that the owners tried to operate the ship on competitive North Atlantic routes rather than on the long-distance routes for which it had been built. By the time the cable company struck a deal with the ship’s representative, the Great Eastern had already bankrupted three other corporate owners.
Modified with cable tanks and handling machinery and loaded with the new cable, the Great Eastern left England in June 1865. Three times the electricians detected faults in the cable just after it had gone overboard. Each time the crew recovered the cable by laboriously transferring an end to the bow and winching it up. On the last recovery, six hundred miles from completion, the cable chafed and broke. It came to rest two and a half miles down. The captain started dragging for the cable end by steaming over it at right angles with a grapnel attached to five miles of wire rope. Four times the crew hooked the cable, and four times the rope or its shackles broke under the weight.
The next year the company started again with nineteen hundred miles of cable, better handling machinery, and a stronger grappling rope. In late July 1866 the cable was complete, to be immediately joined by a second cable when the Great Eastern recovered the broken 1865 cable and spliced it to a new section.
“The world changed that day,” says Dr. Vary Coates, of Congress’s Office of Technology Assessments and co-editor of The Transatlantic Cable of 1866 . “Some major assumptions changed forever. Before the cable the State Department closed down at 5:00 P.M. every Friday and opened again Monday morning. After the cable they had to have someone available twenty-four hours a day.”
The public waited a few weeks to see if the cable would hold up—it did—and then the honors and banquets began, although never at the frenzied pitch of 1858. Congress voted a medal of thanks to Field. The queen knighted four more of the British personnel.
For businesses, the cable represented a long stride toward a global economy. Even at a cost of five to ten dollars per word, customers queued up immediately. One of the first was Lloyd’s of London, which used the cable to report the movements of ships in harbor. Owners of tramp steamers could now locate potential cargoes by telegraphing harbor agents and directing their ships to ports. Commodity and stock speculators used the cable to smooth out price differences between the continents. Banks used it to transfer money instantaneously for international business purchases, reducing the risk of nonpayment and making credit available to smaller firms. Businesses cut their inventories, knowing they could quickly order more goods from overseas. The editor of the Times of London had called the cable “a great bore,” but the major news organizations “simply couldn’t compete without it, once it became available,” says Coates. During the Franco-Prussian War of 1870, a team of reporters from the New York Tribune kept their readers in daily contact with the fighting.
Their feasibility and profitability proven, submarine cables multiplied rapidly. A third transatlantic cable arrived in 1869; a fourth, in 1873. By 1926 twenty-four cables had bridged the Atlantic (and the 1873 cable was still working, with forty more years of service remaining).
The intervening years saw several leaps in making cable and terminal equipment do more work. The net effect was to raise a cable’s capacity from four words per minute in 1866 to fifteen hundred words sixty years later. Meanwhile, the cost per word dropped to a few cents.
The first major advance was duplexing. It allowed a cable to carry messages in both directions at once by making the receiver at each end deaf to messages going out from the transmitter alongside. The effect was to raise the capacity of the cable by 70 percent. The next advance was an automatic transmitter capable of multiplexing. It could transmit several messages recorded on paper tape, sending each a piece at a time to ensure that the cable was making money every second. The final achievement, a response to the arrival of transatlantic radiotelegraphy in the twenties, was the Permalloy cable, which raised the message capacity another fivefold. Designers accomplished this by wrapping a thin tape of iron-nickel alloy around the copper conductor and heating it to boost its magnetic properties. The effect of this inductive loading was to counteract the tendency of different frequencies of the impulse to travel at different speeds and obliterate the signal.
The beneficial effect of inductive loading on submarine cables had been understood for decades, ever since the work of Oliver Heaviside, but it couldn’t be exploited until the invention of the strongly magnetic Permalloy metal. (A self-taught English engineer and physicist, Heaviside was a major thinker in nineteenth-century electricity, but his papers were so obscure, and his lifestyle so hermitlike, that he got little public attention then and is even more obscure today. Besides giving a big boost to telephone and telegraph cable designs with his proposal of inductive loading, he recognized that an outer atmospheric layer was ionized and could reflect radio waves. He also realized that the mass of an electric charge would rise with its velocity—a discovery that contributed to Einstein’s relativity theory.)
Inductive loading kept cable telegraphy alive for several more decades, but it couldn’t survive the next competitor, transatlantic telephone. No transatlantic telegraph cables were laid after 1928. Radio-telephone service began on the shortwave band in 1927 between stations at Lawrenceville, New Jersey, and London. Signals curved around the earth by bouncing off Heaviside’s ionosphere, which meant that on some days the only way to be understood was to bellow. Sunspot activity could shut down the system altogether. And limits on the capacity of the system meant a person could wait hours for a call to go through.
At about this time Bell Telephone considered laying a telephone cable with the capacity of exactly one conversation but decided to wait for more powerful technology. The wait paid off. In 1956 AT&T, the British Post Office, and the Canadian Overseas Telecommunication Corporation successfully laid the first transatlantic telephone cable. It could carry thirty-six voices one way; a second cable was required for the other direction. This was a coaxial cable that worked in essentially the same way as the coaxial cable between the wall and your television set, dividing the thirty-six voices across a bandwidth 240 kilohertz wide. It required several innovations. One was the repeater, a vacuum-tube amplifier with as many components as a radio set, spaced at twenty-mile intervals across the Atlantic. Because it is difficult and expensive to raise a submarine cable from two-mile depths, the tubes inside had to be extremely reliable. (AT&T used a twenty-year-old tube design.) Also, each repeater could only be three inches in diameter and had to be flexible enough to pass through the cable-laying machinery. Designers divided the components into separate sausagelike tubes, reinforcing the eight-foot string with steel rings. The outer covering was polyethylene, an extremely durable white polymer discovered in Britain shortly before World War II. One of its many advantages was that it was distasteful to the teredo worm, a marine borer responsible for injuries to many gutta-percha cables.
This cable, TAT-I, was retired in 1978. Today all twenty-five transatlantic telegraph cables, as well as three old telephone cables, have been abandoned to the deep. (The last telegraph cable expired in 1965.) There are currently nine working transatlantic telephone cables. These cables split the message traffic with satellites hovering 22,300 miles over the equator. Cables and satellites together moved 212,000,000 phone calls, plus telex and telegraph messages, between America and Europe in 1985.
Now an entirely new type of transatlantic cable, labeled TAT-8, is under construction. It will be the first transatlantic cable to use optical fibers, piping pulses of laser light. (Alexander Graham Bell himself first studied the transmission of telephone conversations via beams of light. He even invented a working model, using a selenium photo detector, and was so enthusiastic about it that in 1880 he wanted to name his infant daughter Photophone. He considered the photophone his greatest invention.)
From Tuckerton, New Jersey, TAT-8 will cross the ocean to branch into two lines serving both Great Britain and France. Even with a third of its transmitting capacity held in reserve for backup, the $335,000,000 cable will transmit the equivalent of 37,800 voice conversations. “That kind of capacity in one facility is really eye-opening,” says Herbert Schenck, of Undersea Cable Engineers, a consulting firm in Washington, D.C. The most sophisticated copper-core transatlantic telephone cable can carry less than a fourth as much traffic.
At the heart of TAT-8, which is four-fifths of an inch thick in midocean, are three tiny pairs of glass fibers, two pairs for use and one in reserve. Each hair-thin fiber has two kinds of glass. At the center is ultratransparent silica glass eight microns in diameter. Surrounding it is a coating of more glass with a different refractive index, which encourages the light to stay inside the thin core.
Even with the superclear glass, after a laser pulse leaves the terminal at Tuckerton, it will immediately start dimming. Fixing that is the job of the regenerators, devices spaced every thirty-one miles and protected by beryllium-copper shells. These shells hold a regenerator for each fiber, plus back-ups that controllers can switch on by remote control. Regenerators are not amplifiers, explains Ken Roberts, a sales manager for the cable manufacturer Simplex Wire and Cable. The amplifiers in the earlier copper-core cables took the signal and boosted it, inevitably passing along white noise. But in TAT-8, “they’re re-creating the original digital signal, retiming it and recycling it.” That enables a signal to emerge out of a fiber of unlimited length without any deterioration.
Each regenerator contains a light detector, electronics to reconstruct the pulse stream, and a miniature laser to shoot the information to the next regenerator. Everything is powered by a high-voltage, low-amperage current transmitted along a copper tube built into the cable.
The cable across the American and European continental shelves has already been laid and buried by AT&T on this end and by French and British companies on the other end. It is buried because the greatest enemy of submarine cables is mankind. (The first cable across the English Channel was ruined after one day of service by a fisherman who pulled up the coated wire, cut it, and took a section home under the impression that it was seaweed with a gold core.)
This summer AT&T buried a hundred-mile length of TAT-8 in the mud of the continental shelf, using a sea plow dragged by the cable ship, with the robot sub Scarab available when the plow couldn’t handle the terrain. A sea plow is a heavy framework of steel tubing that slides along the bottom, cutting a two-foot-deep trench and feeding in cable coming down from the ship. Television cameras inform the engineers of any obstacles or problems. The chief purpose of the burial is to prevent fishing trawlers from snagging the line with their bottom-dragging nets. Steel armoring wires, wrapped around the polyethylene, also help defend the cable against fishing gear and small dragging anchors.
There’s another hazard on the continental shelf: sharks. AT&T has discovered from earlier trials off the Canary Islands that sharks will occasionally grab a section of fiber-optic cable and try to bite through it; something about the line’s electromagnetic field appears to trigger a feeding reflex. (The earlier copper-cored cables never attracted sharks). The solution: a double layer of tooth-resistant steel tape atop the armoring wires. The deepsea portion of the cable, which AT&T and partners will unroll in 1987, won’t require armoring or burial. Virtually nothing can damage it down there.
The high cost and high technology of TAT-8 are a reminder that getting information across the ocean is still a huge undertaking. Transatlantic cables have always demanded the most reliable communications technology. When their owners cut the shore ends and abandon them to the ocean, it’s not because they’ve completely worn out, like old shoes, but because new cables with enormous capacities have taken over. The owners of TAT-8 plan for only twenty-five years of service. They know it will still be in good condition after that period, but they also know something bigger and better will come along.