Darwin Among the Machines (27 page)

The development of the electric telegraph during the first half of the nineteenth century represented the culmination of principles incubated over many years. As early as July 1729, Stephen Gray of London transmitted an electric charge a distance of 765 feet. In July 1747, following the invention of the Leyden jar, or charge-storing capacitor, in 1745, William Watson succeeded in communicating an electric charge by iron wire across the Thames at Westminster Bridge; in 1748, Benjamin Franklin did the same across the Schuylkill River; on 5 August 1748, Watson and Henry Cavendish extended the distance to 12,276 feet. The Abbé Jean Antoine Nollet (1700–1770) succeeded in electrifying 180 of the king's palace guards, who “felt the shock at the same time” at Paris in 1746. Later, using lengths of iron wire, he formed a 900-foot chain of Carthusian monks and reported that “the whole company at the same instant of time gave a sudden spring.”
²⁸

In February 1753, a Scottish inventor identified only as C.M. (believed to have been either one Charles Marshall, of Paisley, or Charles Morrison, of Renfrew) described “an expeditious method for conveying intelligence” by means of twenty-four parallel wire conductors supported by glass insulators about “every twenty yards.” Observing that “electric power may be propagated along a small wire, from one place to another, without being sensibly abated by the length of its progress,” C.M. detailed the construction and operation of an electric telegraph, though there is no evidence that the machine was ever built. “Let the wires be fixed in a solid piece of glass, at six inches from the end; and let that part of them which reaches from the glass to the machine have sufficient spring and stiffness to recover its situation after having been brought in contact with the barrel.” By depressing individual conductors in sequence, any alphabetic message could be conveyed, or, by reversing the process, on an agreed signal, a reply could be received. “Close by the supporting glass, let a ball be suspended from every wire; and about a sixth or an eighth of a inch below the balls, place the letter of the alphabet, marked on bits of paper, or any other substance that may be light enough to rise to the electrified ball; and at the same time let it be so contrived, that each of them may reassume its proper place when dropt.” Less practical was the suggestion to employ electrically actuated bells, by means of which the two correspondents “may come to understand the language of the chimes in whole words.”
²⁹

Telegraphs surfaced like weeds. In his
History of Electric Telegraphy to the Year 1837
, John J. Fahie reviewed the work of some forty-seven inventors whose telegraphs succumbed to technical obstacles, lack of support, or a combination of both. In the 1790s, Don Francisco Salvá of Barcelona (1751–1828) enjoyed the support of the king and was rumored to have constructed a single-wire telegraph line over the twenty-six miles between Aranjuez and Madrid. Salvá experimented both with electrostatic signals and with the transmission of faint pulses of direct current, indicated by the convulsion of frog legs as much as 310 meters apart. Working before news of Volta's battery was received, “Salvá employed, as his motive power,” reported Fahie, “the electricity produced by a great number of frogs.”
³⁰
He delivered a paper entitled “Galvanism and its application to Telegraphy” on 14 May 1800 to the Academy of Sciences in Barcelona, followed by a second treatise in 1804 showing how the frogs could be replaced as both transmitters and receivers of electric signals by electrochemical cells. In England, Francis Ronalds of Hammersmith, London, working at his own expense, transmitted electrostatic signals over eight miles
of wire in 1816. Envisioning “electrical conversazione offices, communicating with each other all over the kingdom,”
³¹
but foreseeing the vulnerability of overhead lines, he proposed a network of buried cables, testing his apparatus over 525 feet of insulated conductor buried 4 feet underground. He then wrote to the first lord of the Admiralty, “soliciting his lordship's attention to a mode of conveying telegraphic intelligence with great rapidity, accuracy, and certainty, in all states of the atmosphere, either at night or in the day, and at small expense.”
³²
The official response was that “telegraphs of any kind are now wholly unnecessary, no other than the one now in use will be adopted.
³³
A section of Ronalds's cable was excavated intact in 1862. Success depends on timing—not necessarily on being first.

In March 1800, Alessandro Volta announced his invention of the voltaic pile to the Royal Society, and electric batteries were duplicated around the world. In 1819, Hans Christian Oersted outlined the essential principles of electromagnetism, brought into coherent mathematical form by André-Marie Ampère, who delivered his first paper on electrodynamics, as he termed it, on 18 September 1820, after only seven days of work. In October 1820, some fifteen years before he first put the word
cybernétique
into print, Ampère followed a suggestion of his colleague Pierre Laplace in proposing an electric telegraph. “One could, by means of as many conducting wires and magnetic needles as there are letters, and by placing each letter on a different needle, establish by the aid of a voltaic pile . . . a genuine telegraph, writing all the details one might wish to transmit, across whatever obstacles,” noted Ampère. “By connecting to the voltaic pile a keyboard whose keys bear the same letters and establish contact by their depression, this method of correspondence could be performed with great facility, in no more time than necessary to touch each letter at one end, and read it at the other.”
³⁴

Commercially successful telegraphy slowly evolved. Electrochemical indicators that displayed signals by releasing bubbles of gas were invented, as was a system that inscribed electrostatic signals as marks on a moving strip of litmus paper, demonstrated by Harrison Gray Dyar over 8 miles of wire encircling a Long Island racetrack in 1827. Construction of a line between New York and Philadelphia was terminated when charges of bank conspiracy brought the project's financing to a halt. A five-channel needle telegraph was developed by the Russian baron Paul L. Schilling in 1823 and exhibited as a working model to Czar Alexander I in 1825. Carl Friedrich Gauss and Wilhelm E. Weber constructed a 1.5-mile galvanometer telegraph in Göttingen in 1833, used daily to communicate between the physics department
and the observatory until 1838. In the United States, there were the proof-of-principle experiments of Joseph Henry, appointed first secretary of the Smithsonian Institution in 1846, who rang a bell at a distance of 1,000 feet in 1830, followed by a 1-mile relay-actuated communication link between his house in Princeton, New Jersey, and Princeton College, where he worked.

Returning in 1832 from Europe, where electrical developments were in the air, Samuel Morse entertained visions of electromagnetic telegraphy, making his first notes toward the binary dot-dash representation of numbers and letters later known as the Morse code. A professor of art history, not an electrical engineer, Morse relied on the assistance of Joseph Henry and Alfred Vail to develop a system that worked. Vail became rich, while Henry, who asked for nothing, was personally attacked when the circumstances surrounding his contributions threatened to undermine the patents held by Morse. The Morse telegraph incorporated relays for pulse regeneration, closely following Henry's lead. Telegraph relays, manufactured in large quantities and adapted for switching purposes, would spark the proliferation of binary logic gates that led to the age of digital machines.

Morse established the first long-distance line, between Baltimore and Washington, which opened on 24 May 1844 with a message selected by Miss Annie Ellsworth, daughter of the commissioner of patents—the last phrase of the twenty-third verse of the twenty-third chapter of the Book of Numbers: “What hath God wrought!” Success attracted both converts and competitors, and by 1851, the year the progenitor of Western Union was established, there were over fifty telegraph companies in the United States. That same year the first telegraph cable linked England and France; by 1852 there were some twenty-three thousand miles of telegraph lines in existence, enough to encircle the world. In 1861 the first line spanned the North American continent, and in 1866, after many failures, a durable connection linked England to the United States. India was reached in 1870, Australia in 1871, and in 1874, Brazil.

The obstacles shifted from establishing the physical connections constituting each leg of a telegraph circuit to switching, regenerating, and encoding and decoding the messages at either end. Telegraph signals were digital signals, whether conveyed by the on-off state of a fire beacon, the twenty-four-symbol alphabet of Robert Hooke, the ninety-eight-state signal of the Chappes, a series of positive-negative voltages, or the dot-dash sequences of Morse code. To process these signals requires discrete-state machines, whether the machine is a
human operator looking through a telescope and referring to page and line numbers in a book of code or one of the punched-tape teleprinters that soon came to dominate telegraphy throughout the world.

Telegraph engineers were the first to give substance to what had been shown by Leibniz two centuries earlier and would be formalized by Alan Turing in the century ahead: all symbols, all information, all meaning, and all intelligence that can be described in words or numbers can be encoded (and thus transmitted) as binary sequences of finite length. It makes no difference what form the symbols take; it is the number of choices between alternatives that counts. It takes five binary alternatives (2
⁵
= 32) to encode the alphabet, which is why early needle telegraphs used five separate two-state indicators and why teletypewriter tape is five holes wide. (Polybius had specified two separate five-state indicators, a less efficient coding that makes sense if you are keeping a naked eye out for torches in the dark.)

Every message had to be encoded, decoded, stored, reencoded, and retransmitted many times as it made its way from one node to the next. In 1858, Charles Wheatstone introduced perforated paper tape as a means of automatic signal transmission; receiving perforators, reperforators, and perforated-tape-driven printers soon followed. In the 1870s, Jean Maurice Émile Baudot introduced both time-division multiplexing (interweaving several code sequences over a single circuit) and the 5-bit alphanumeric code that bears his name (Wilkins was long forgotten by this time). Although beginning and ending its journey as alphabetic text, the message was represented as either sequences of pulses over a wire or sequences of punches in a strip of paper through the many stages along the way.

The telegraph system soon evolved store-and-forward procedures at its nodes—the ancestor of the packet-switching protocols used in computer networks today. An incoming telegram arrived at the switching node as a sequence of electrical signals, converted to a series of holes in a strip of paper tape, identified by its origin, its priority, and its destination address. The station operator considered this information along with the state of the outgoing lines in making a decision as to when and via what routing to relay the message, retranslated back into electrical pulses by a machine that sensed the pattern on the tape. When the message was acknowledged by the next station, the tape could be discarded, and the transitory state of mind that it represented was erased.

High-speed automatic telegraph instruments were the ancestors of modern computers and gave the electromagnetic data-processing industry its start. “Computing machines,” explained John von Neumann
in 1949, “can be thought of as machines which are fed, and emit, some medium like punched tape.”
³⁵
This definition works both ways. Like the molecules that convey hereditary information between living cells, telegraph equipment performed the function of recording, storing, and transferring sequences of code. Most early digital computers—from the Colossus to the IAS machine—used paper-tape teletype equipment for input and output between the computer and the outside world, augmented by the ubiquitous punched-card equipment from Hollerith and IBM. It was only natural that the first computers incorporated high-speed telegraphic equipment, and it is no accident that the genesis of the Colossus within the British Telecommunications Research Establishment was mirrored in the United States by early steps taken toward computers by Claude Shannon and others within Bell Laboratories and RCA. Only later did the communications industry and the computer industry become temporarily estranged.

The solitary computers of the early 1950s exchanged code sequences by means of mutually intelligible storage media and, before the end of the decade, by connecting directly or, in language that has now been extended to human beings, on-line. But no matter what the medium of exchange, the code itself and the protocols that regulate its flow remain directly descended from the first strings of telegraphic bits. Evolution of computer-to-computer communication, like previous advances in telecommunications, was closely related to defense. In the early 1950s, a computer project known as Whirlwind evolved to support an integrated air-defense system developed for the U.S. Air Force by the Lincoln Laboratory at MIT. Whirlwind led directly to the SAGE (Semi-Automatic Ground Environment) air-defense system, a real-time interactive data-processing system, which led in turn to the development of time-sharing computer systems and, eventually, to computer networks as we know them today. John von Neumann's legacy to computer networking was to be found not only in the architecture of individual computers, but in the proliferation of weapons against which networks of computers offered the best hope of defense.

In 1955, less than three years after the explosion of the first hydrogen bomb, von Neumann was able to announce that “we can pack in one airplane more firepower than the combined fleets of all the combatants during World War II.”
³⁶
Von Neumann was also thinking of long-range missiles, which, as chairman of the Strategic Missiles Evaluation Committee established in 1953, he referred to as “nuclear weapons in their expected most vicious form.”
³⁷
Light-weight,
high-yield thermonuclear weapons would soon be available that could be lofted by rocket into space and guided to earth many thousands of miles away. After the Soviets exploded their first hydrogen bomb in 1953, the push to develop an intercontinental ballistic missile, or ICBM, became a race.