PART FOUR: COMMUNICATION AND CALCULATION 712 A station agent telegraphed the page number with his first signal, and the number of the word on the page with his second. Phrase and place-name vocabularies, each again of 92 by 92 items, were developed; to distinguish them, an initial signal indicated the code; the second, the page; and the third, the item on the page. No public excitement over the semaphore system occurred until 1 September 1794, when Condé was recaptured. The Convention then ordered the extension of the line from Lille to Ostend, and a second line to be built to Strasbourg, which was completed in 1798 with 46 towers at a cost of 176,000 francs. Despite the practical success of these telegraph lines, they did not bring in revenue, but rather entailed great expense; maintenance and service in the eighth year of use cost Figure 15.2: Three stages in the evolution of the telegraph: (a) Chappe’s 1794 semaphore; (b) von Soemmering’s 1809 electrolytic telegraph. INFORMATION 713 434,000 francs. Napoleon, losing enthusiasm for the venture, reduced the appropriation to 150,000 francs a year, and cancelled a planned Paris-Lyon line. Claude Chappe thought that his system might benefit business, as well as give an advantage in war. During the Restoration, interest in the telegraph increased and Chappe’s dream was at last realized—but he committed suicide in 1805, despondent over the slow progress and suffering from bladder trouble. However, his brothers continued to perfect the system; Abraham worked at overcoming the fog problem, anticipating British use of hydrogen fires during the Second World War. Semaphore systems remained in use well into the nineteenth century, even after the electric telegraph was developed (see p. 714). Figuier, writing in the 1860s, claimed that aerial telegraphers could send a dispatch in two minutes from Lille to Paris (a distance of 240km (150 miles), requiring 22 stations); in three minutes from Calais (270km (168 miles) and 33 stations); in eight minutes from Brest (600km (373 miles) and 54 stations); and in twenty minutes from Toulon (more than 1000km (620 miles) and 100 stations). The equivalent rates of speed would be, respectively, 7200, 5400, 4500 and 3000kph; even the worst case is three hundred times better than the then fastest system—a relay of horses and riders, such as the famed Pony Express in the USA. The electric telegraph The idea of the electric telegraph preceded its practical possibility by many decades. Chappe’s semaphore had demonstrated the need for rapid Figure 15.2: (c) Cooke and Wheatstone’s 1837 5-needle telegraph. PART FOUR: COMMUNICATION AND CALCULATION 714 communication of messages, but until electricity was better understood, little progress was possible. There were attempts to do the job with static electricity, but the high voltage and very low current characteristic of this form of electricity did not allow reliable transmission beyond a few metres. The inventor of the earliest electric telegraph did not wait upon the epochmaking discoveries of Oersted, Faraday and their contemporaries (see Chapter 6). In Germany, S.T. von Soemmerring observed that electric current passed through an acid solution caused bubbles to appear (electrolytic decomposition of water into its elements, hydrogen and oxygen). He invented a telegraph system using this principle in 1809, which used 26 parallel wires to transmit letters of the alphabet a distance of up to two miles (Figure 15.2 (b)). He even designed an ingenious alarm to alert the receiving operator (this was really the first relay, although not electromechanical). However, the expense of so many conductors made the system economically impracticable. André-Marie Ampère, in 1820, invented the galvanometer which enabled electricity to be measured, and suggested using a galvanometer needle for telegraphy. W.F.Cooke and Charles Wheatstone invented a five-needle telegraph, which was patented in 1837 (Figure 15.3 (c)). Although still a parallel data-communication system like Soemmerring’s, this reduced the number of wires to only six, by representing alphabetic characters with a 2- out-of-5 code. By 1839, they had set up a 13-mile telegraph for the British railways using this system. It was not long before the cost of multiple conductors stimulated reduction of the number of needles to two, and finally to one. So the needle telegraph became the first serial data-communications system, with codes defined by sequences of needle deflections to make up each character. Samuel F.B.Morse, already a noted American painter, became interested in the possibilities of electrical communication in the 1830s. He built a telegraph system in 1835: the sender used notched metal strips to encode the alphabet, and the receiver was an electromagnetically-driven pendulum with a pencil attached which wrote the coded signals on a moving roll of paper. Morse’s knowledge of electricity was rudimentary, but by 1837 he had patented the telegraph, replaced the type-bar sender with what we now call a telegraph key, and simplified the receiver to a pen which put marks (dots and dashes) on paper tape. Early telegraphs used two wires, but it was soon found that one would do, the earth acting as the return path. With some government support, Morse started a telegraph service between Washington and Baltimore in 1844. When the line was extended to New Jersey, it attracted customers in the financial community who appreciated the commercial value of instantaneous communications. The visual receiver was replaced by a sounder, because operators could transcribe aurally transmitted codes to paper faster than the marks on paper tape. This early human-factor discovery shows the importance of inventor-user interaction in bringing an INFORMATION 715 idea to practical fulfilment. The so-called Morse code was actually the work of Morse’s assistant, Alfred Vail. The telegraph was one invention which did not have be marketed; the public was eager and ready to pay for good communications. By 1852, more than 18,000 miles of wire covered the eastern third of the USA, with service provided by many small companies. In 1856, Western Union was created, and extended the telegraph to the west coast in 1861 with government support; a message cost senders $1 per word. There had been several attempts to develop submarine cables for telegraphy, but the first, under the English Channel, ruptured in 1850. Improvements in cable-making, and the development of specially equipped cable-laying ships, enabled construction of a reliable link from Dover to Calais in 1851. A much more audacious undertaking, a transatlantic cable, was proposed by an American, Cyrus W.Field, in 1856, and was successfully laid by HMS Agamemnon between 1857 and 1858. However, a few months afterwards, operator error put 2000 volts across the cable which rendered it unusable. It was not until after the American Civil War, in 1865, that Field attempted another link; this time, the cable was spun out from a single ship, Brunel’s Great Eastern. This sail-and-steam leviathan carried almost 5000km (3,100 miles) of cable weighing over 5000 tonnes. However, after laying 2000km (1250 miles) between Ireland and Newfoundland, a defective length snapped; after ten days’ unsuccessful grappling, the spot was marked and the attempt abandoned. Another cable was manufactured by the Telegraph Construction and Maintenance Company of the UK, and laid by the Great Eastern in 1866: the commercial communications link so long sought between Europe and the United States was at last established. The first telecommunications network—organizational rather than physicalwas that of the Associated Press, begun in the 1840s by a group of New York newspapers to share telegraph expenses. Many technical improvements were made in both the United States and Europe, including the paper-tape perforator of Wheatstone (1855), and the multiplex telegraph of Emile Baudot (1874). However, the most radical departure, which led the way to the modern era of telegraphy, was the invention of the teletypewriter (or teleprinter) by E.E. Kleinschmidt in the US in 1928. This allowed operators to compose messages using a typewriter-like keyboard to punch paper tape, which was then torn off and fed into a tape reader for transmission. At the receiving end, the message was printed out on paper strips (later, directly on a roll of paper). Thus, the days of the telegraph operator, sending with a lightning touch on the key, and listening to the clicks of the sounder were ended. However, Morse is still used in radiotelegraphy, where it can get a message through static and difficult transmission conditions when electromechanical and electronic alternatives are unworkable. PART FOUR: COMMUNICATION AND CALCULATION 716 Telex The sending and delivery of a telegram has always been a hybrid type of service, part electronic and therefore very rapid, part manual and therefore slow and labour-intensive. The sender had to go to a telegraph office and print out his message in capital letters on a special form; the clerk had to count the words and compute the amount due from a tariff schedule; then the telegraph operator had to key in the message. At the other end, the reverse took place; the message would be printed out on a paper strip, cut into segments and pasted on a form, and then had to be delivered by hand to the recipient. By the time most customers had their own telephones (the 1930s in the US and Canada; later elsewhere), the US telegraph carrier Western Union would accept outgoing telegrams over the phone, and read incoming telegrams to recipients who were phone subscribers. However, the unique feature of the telegram among public telecommunications services has always been the delivery of a written message, providing legal proof to sender and receiver. Most customers would insist on physical delivery even though the message had been read to them over the telephone; in such cases, telegraph officers would resort to the postal system. For business users, the whole system seemed archaic. In the US they turned increasingly to the telephone whenever written proof was not essential. In Europe, with its many languages and complex cross-border tariffs, the telephone did not have the same ease of use or economy. Therefore, when the first telex network was put into service in Germany in 1933, a great unmet demand was released. From its start, with only nineteen subscribers in Berlin and Hamburg, telex service had an explosive growth. In Germany subscriptions grew at more than 100 per cent annually up to the Second World War; similar growth, but at lower rates, was experienced in other European countries. By the early 1980s, the number of telex subscribers world-wide exceeded 1.5 million, more than half of them in Europe. Today, telex is a world-wide, switched public teleprinter service. Therefore it is like the telephone, in that subscribers are loaned terminal equipment, can dial up other subscribers themselves, can receive messages, and are charged on the basis of time and distance. However, unlike the telephone, messages must be keyed in and received on teleprinters. Up to the 1970s, the usual practice was to keypunch a series of messages on paper tape, dial up the recipient’s teleprinter, and put the strip of tape in the sender’s teleprinter; this store-and- forward process has been greatly enhanced by the use of computers in telex officers, and by microprocessors and electronic memory capability in terminal equipment. Also, in store-and-forward operation, the same message may be sent to many recipients (a type of message broadcasting), and re-dialling done automatically when receiving terminals are busy. INFORMATION 717 One of the greatest advantages of telex is that teleprinter equipment and service available to subscribers is uniform throughout most of the world. Unfortunately, a faster service, Teletype, was introduced in the US side-by-side with Western Union’s telex, and until the FCC forced AT&T to relinquish their public switched message service, conversion was difficult. However, since Western Union took it over, computers do this conversion without customers having to worry about it. The international telex system uses the five-level Baudot code, which requires a shift from letters to numerals and back again, and there is no provision for indicating letter-case. The speed of transmission is very slow: 50 bits per second, equivalent to 66 words per minute (a word is defined to be six characters, including space). In international practice, the telex system still has these limitations, but within certain countries faster and more flexible equipment is allowed. In the early 1980s, a completely new public switched message service was introduced called teletex. It operates at 2400bps, 40 times faster than telex, and allows upper-and-lower-case letters, formatting codes and a host of other features: teletex can best be described as communicating word processing. However, new equipment is required in both central offices and subscribers’ premises; total cost is higher for low-volume users; and in the mid- 1980s teletex service was only available in a few countries. Therefore, it is unlikely to displace telex in international use until the 1990s. THE TELEPHONE In addition to an understanding of electromagnetism, a knowledge of acoustics and human hearing was essential to the development of the telephone. However, with the exception of Bell, most of the early experimenters were physical scientists, engineers or dedicated tinkerers, rather than biologists, physicians or psychologists. Acoustic transmission The science of acoustics developed out of the theory and practice of music, and the refinement of musical instruments. An understanding of the laws of harmony started with Pythagoras in ancient Greece. Shorthand symbols had been used to record Greek and oriental speech (ekphonetics), and between the fifth and seventh centuries AD a system was developed to show melodic movement using symbols called neumes. The staff was created in the ninth century—a single coloured line. Guido d’Arezzo suggested using three and four lines, the latter becoming the standard for recording Gregorian chants. Today’s five-line staff appeared in the eleventh century, but did not come into general use until the seventeenth century; some composers would like to have six or even eight lines. PART FOUR: COMMUNICATION AND CALCULATION 718 In 1796 a German student, G.Huth, musing on ways to overcome the disadvantages of Chappe’s semaphore (see p. 711), such as excessive operating cost and interruption of service by fog and rain, hit upon the ideas of using large speaking-tubes (megaphones). Specially selected and trained operators would listen to the message from one tower, and turning their megaphones in the direction of the next, repeat the message immediately. Huth wrote that ‘this… difference might deserve a different name. What could be more appropriate…than the word derived also from the Greek: Telephone or Fernsprecher?’ Actually, Huth’s idea was less practical than the semaphore, because visual signals are usable over much further distances than voices or even whistles or drum beats. The speaking voice, even using a megaphone or electronic amplification, is audible only up to a few hundred metres; a human whistle, screams and yodelling carry further. Mechanically aided sound-generation systems, such as noisemakers and musical instruments are also useful; African drums can carry messages for many kilometres. Today, aerosol-powered shriekalarms can carry at least a kilometre, and police and ambulance sirens can be heard over a radius of several kilometres even in noisy cities. Sounds also can be sent through hollow tubes, and through denser matter (liquids or solids) they can be transmitted much further and at higher speeds than in air. In 1838, Romershausen proposed a telephone which would transmit speech through narrow tubes. Modern hydrophones listen in on the cries of dolphins and whales, and detect other underwater sounds over a range of frequencies far beyond human hearing. Electrical speech transmission However, although such means could carry intelligible messages, there is less privacy, greater interference from noise, and greater annoyance to the public at large than in visual signalling. What was needed to make sound transmission a practicable means of communication, just as with the telegraph before it, was a better understanding of the nature and application of electricity. In 1837, about the time that Cooke and Wheatstone were building the first commercial telegraph system, Professor Page in Salem, Massachusetts, discovered that an iron rod which was suddenly magnetized or demagnetized would emit sounds; this came to be called Page’s effect, but no practical applications emerged. In 1854, Charles Bourseul in France predicted the coming of speech transmission, and outlined a method for its realization that was correct except in one critical feature: he expected that it would be achieved using make/break circuity (like the relay and telegraphy, which is digital) rather than with circuits capable of handling continuous current (which is analogue). Philip Reis in Germany constructed a telephone along the lines suggest by Bourseul in 1861, INFORMATION 719 but was able to transmit only tones rather than speech. In 1868, Royal E.House, who had invented a printing telegraph in 1846, patented what he called an electrophonetic telegraph, but did not realize its capabilities to transmit speech. Ironically, we are today on the threshold of ISDN (Integrated Services Digital Network) which will transmit all signals—speech, video, text and data—digitally. The breakthrough came in 1876 when Alexander Graham Bell and Elisha Gray almost simultaneously—and independently—filed patents for successful speaking telephones. Although, after long patent litigation, Bell was awarded priority, Gray’s system was technically superior. Bell was an early example of the brain drain from the Old World to the New. His parents had emigrated from Scotland to Canada after having lost two sons to tuberculosis, and Bell later settled in Boston where he became a teacher of the deaf. From this background, rather than as a physical scientist or engineer, Bell approached the problem: as a teacher trying to find better ways to teach deaf students to speak. For this purpose he had invented a special notation for recording speech. Bell’s earliest working models used no external source of electricity, and transmitter and receiver were almost identical. This system—the voice- operated telephone—is still used over short distances. Switched telephone systems require external power. On 7 March 1876 the famous command, ‘Mr Watson, come here. I want you!’ was uttered by Bell to his assistant, who instead of only hearing Bell’s voice from the other room, heard it over the primitive induction device with which they were experimenting. This, the first transmission of a human voice over wire, happened only three days after the first telephone patent had been issued to Bell. He had filed for this patent on 14 February, only three hours before Gray had filed a caveat that he was working on a similar device. Bell’s patent was probably the most valuable ever granted, giving birth to what has become the world’s largest private service organization, the Bell Telephone Company. Bell Telephone was formed only a year afterwards, the first two telephones being leased to a Boston banker. Also in 1877, Bell, who had married the deaf daughter of his first backer, Thomas Sanders, went to England on his honeymoon; mixing business with pleasure, he gave many demonstrations, and presented Queen Victoria with a pair of telephones. However, although the British Post Office’s chief engineer recommended obtaining rights to manufacture and use the new technology, the Post Office claimed that this was not necessary to protect its monopoly—because the telephone was just another form of telegraph. In 1883, Edison discovered what eventually became the basis for signal amplification in both wire and wireless communication—he observed that when he added another element (electrode) to a light bulb, that current could flow across the evacuated space between filament and electrode. This came PART FOUR: COMMUNICATION AND CALCULATION 720 to be called the Edison effect, but was only a laboratory curiosity until 1904, when Fleming invented the two-element rectifying diode: shortly afterwards, in 1906, Lee de Forest added a third element which enabled small voltage changes to control the flow of large currents, giving birth to the triode and the age of radio. THE GRAMOPHONE All sounds, whether voice, music or simply noise, depend upon physical vibration of some medium—if our ears had the ability to detect extremely low frequencies and faint sounds, we might even hear the ripples in a calm lake and not just the roar of ocean waves. Unlike the extremely limited ability of our eyes to sense the immense range of the electromagnetic spectrum directly (less than one per cent), we can hear about ten per cent of the sound spectrum which is inherently limited because physical molecules must vibrate to emit and transmit sound. It has been reported that a visitor entering an ancient Chinese temple is greeted with the Chinese equivalent of ‘close the door’ upon entering and, when he does, ‘thank you’. The mechanism for this automatic doorman is said to be a pointed object which moves over a serrated strip when the door is opened, and then moves backwards when it is closed, the Chinese for ‘thank you’ being the reverse of ‘close the door’. If true, this invention is similar to noisemakers, where one stick is rubbed over another which has a regular pattern of notches cut into it; however, the ability to cut notches so that a semblance of the human voice is reproduced strains credulity. Practical systems for the automatic recording and reproduction of sound have become available only recently, and voice synthesis and recognition are still in their infancy. Mechanical sound recording and reproduction Techniques of recording vibrations occurred before means were found to reproduce them. In 1857, Leon Scott de Martinville, designed a device for tracing soundwaves on a cylinder covered with lampblack, which he called a ‘phonautograph’; however, he could not find a way to replay the sound. In 1877, Charles Cros in France outlined a scheme to convert Scott’s waveform trace into a physical groove by photoetching, and play the sound back by a point attached to a diaphragm which followed the groove. He dubbed this device the ‘paleophone’, but lacked funds to put the idea into practice. The first machine actually built which could both record and reproduce sound was Edison’s 1878 cylinder phonograph, for which he filed the first detailed patent application in England. This had been preceded by an 1877 INFORMATION 721 patent for a means to record Morse code by making indentations on paper stretched over a disc or cylinder, and another which outlined a method for transmitting the human voice over telegraph wires. The seminal 1877 Edison patent contained in its specifications not only details of his original tinfoil-on-cylinder model, but also claimed disc recording, wax materials, electromagnetic (rather than acoustic) recording and reproduction, mass production of recordings and amplification. Unfortunately, Edison failed in his efforts to make the equivalent claims in the US. As a consequence, a tangle of litigation marred the early decades of the phonograph, just as it did the contemporary invention of the telephone. In fact Bell, with his brother and Charles Tainter, patented a rival system, the graphophone, in the mid-1880s which used removable waxed-paper cylinders. The US patent, dated 19 February 1878, was titled ‘Phonograph or Speaking Machine’. The drawing showed a spiral-grooved cylinder wrapped with tinfoil; on one side was a short horn (the mouthpiece) containing a diaphragm to which a rounded needle (the stylus) was fastened at right angles; opposite, a larger horn with an upward-facing opening was similarly outfitted (the speaker). Thus, when Edison spoke his famous ‘Mary had a little lamb’ into the mouth-piece, he was startled to hear his voice reproduced on the other side a fraction of a second later. Bell’s first telephone (see above) was similar in its simplicity, being able to act either as sender or receiver. In 1888, Edison also patented a wax-cylinder machine, but one year earlier, Emile Berliner had patented his gramophone; although depicted in the drawing as a cylinder machine, by the time he built his first model a year later, it was a disc machine. Another innovation introduced by Berliner was lateral recording of sounds in spiral grooves. Scott’s tracings had been lateral, but Edison and his followers had employed vertical (hill-and-dale) recording. By 1895, Berliner had combined his flat disc, Scott’s lateral recording method and Bell and Tainter’s wax coating, and formed the basis of the modern mass-production recording industry. Edison never exploited the alternatives of his 1878 British patent, sticking with cylinders almost until the First World War; reluctantly, in 1912 he brought out a disc phonograph. He was right about some of the advantages of the cylinder: it provides identical grooves over the entire surface, rather than spiralling to a smaller and smaller radius; and hill-and-dale recording can handle a much greater dynamic range without shortening recording time. The original application of the graphophone and phonograph was not in entertainment but in business. In the late 1880s, the North American Phonograph Company was formed to rent out both types of cylinder machines to record dictation (for a similar machine see Figure 15.3); however, the graphophones proved unreliable, and the confusion of offering two incompatible machines to customers led to failure of the company. Rather, it was as an entertainment medium that the phonograph and gramophone scored their early successes. In 1888, Gianni Bettini in New York used a phonograph . developed out of the theory and practice of music, and the refinement of musical instruments. An understanding of the laws of harmony started with Pythagoras in ancient Greece. Shorthand symbols. until the 1990s. THE TELEPHONE In addition to an understanding of electromagnetism, a knowledge of acoustics and human hearing was essential to the development of the telephone. However, with the. segments and pasted on a form, and then had to be delivered by hand to the recipient. By the time most customers had their own telephones (the 1930s in the US and Canada; later elsewhere), the US