The Railways

  Mechanical signs with moving parts no longer play much part in everyday life, so that there is something marvellous about the survival of this Victorian technology into its third century. The twitching of the control wire stretched taut alongside the rails, the shudder of the signal as it hits the clear position, and the gentle diminishing bounce with which it reverts to the level alignment after the train has passed, all brought about by a human being who may be glimpsed at work behind the windows of a neat little building a few feet from the track: these movements are like a polite rebuke by an elder to the callow assumption that nothing much ever gets done without the help of electrons or hydrocarbons. And there is a winning cheerfulness in the smart signal shades of red and yellow, now always of imperishable enamelled metal (painted arms made of timber, like the ‘wonderful wooden razors set aloft on great piers, shaving the air’ noted by Dickens en route to Carlisle in 1857, were usual until the 1920s).

  That the shapes, colours, markings and light codes of signals once differed bewilderingly across the various companies will come as no surprise. Standard colours and markings for signal arms became mandatory as late as 1928.* Even the codes used between adjacent boxes might differ. Uniformity for most of these was imposed in the wake of a suburban misadventure on 10 December 1881. A signalman on the Great Northern at Finsbury Park, prompted by a code received from a North London Railway signalman up the line which meant something completely different on each company, blithely sent four trains into Canonbury Tunnel, one after another. Far from having a clear road, the first train was held at a signal at the other end, and those following bumped into one another in the darkness. No serious casualties resulted, only a great deal of distressed milling about in the murk, until a guard made his way out of the tunnel mouth to warn an approaching fifth train not to join the crush.

  One big regional difference remains, a relic of the special tenacity of Great Western traditions: the signals on western lines still drop by forty-five degrees to show a clear road ahead, whereas those in other areas are raised upwards by the same angle. The drooping kind was once preferred across the country, but in their original form they had a serious drawback: should a fault occur, the signal arm might fall to the clear position even when the road ahead was unsafe. The Great Western got round this risk by adding a counterweight, so that gravity pulls the arm back up to the danger position. On other lines, the signal drops back to the horizontal after the train has passed. One sublime cinematic instance may be remembered: the climax of The Ladykillers (1955). The creepy fake professor played by Alec Guinness, last survivor of a self-destructive gang of bank robbers whose safe house stands above the tracks leading to King’s Cross, is felled from behind by the bright red arm of a home signal, falling blindly back to ‘danger’.

  Semaphore signals and signal boxes belong together in other ways. It was among the signalman’s duties to ensure that the lamps for each signal remained lit, so they were normally placed within sight of the box. Signals that faced away from the box had apertures at the back to indicate that the light was still burning. There was also a limit on how far points could be located from the signal box without exceeding the range of control levers and rodding. Until 1892 this limit was habitually set at 150 yards. The distance later increased considerably, but it was not until power-operated points became common that mechanical constraints could be escaped completely. A comparable breakthrough for signals was the invention of the pyrometer, which sent an electrical signal to a ‘repeater’ display within the box, confirming that the lamp remained lit. Repeaters allowed signals to be spread out much further and still remain within mechanical control by wire. The Southern Railway in the 1930s organised a competition to identify the longest ‘pull’ on its lines; its upper range at that time included Worgret Junction box in Dorset, where the remotest signal stood almost a mile off (specifically, 1,680 yards). Where semaphore signals survive, however – as along the Cumberland coast, or on the main line through Cornwall – they tend to stand much closer to their parent boxes. The traveller who keeps an eye out for them soon acquires a sense of being handed on from one operating point to the next, traversing long expanses of unsignalled track between.

  These clusters of signals make visible another principle of safe operation that was finally enforced after the Armagh disaster of 1889, known as absolute block working. Under this system, the network is divided into ‘blocks’ or sections of track by means of signals that interlock with the points, each section being permitted to take only one train at a time according to telegraphic indications received. Put another way, no train is allowed to enter a section until the train ahead of it has passed through, having first been accepted by means of an exchange of bell-codes by the signaller (the gender-neutral, rather military-sounding term now officially adopted) of the section beyond. Instruments within the box and linked to the telegraph allow a visual record of the state of traffic to be kept constantly in view. Mechanical and electro-magnetic interconnections make it physically impossible to change a signal from danger to clear unless the relevant instrument already shows an unobstructed path, which it can do only after the next block signal down the line has already been cleared, which depends in turn on the indicator covering the block after that – and so on. The Railway Club Journal in 1910 summarised the end in view: ‘[I]f, at any given junction, a man entirely unacquainted with railway working were to enter the signal box, he might safely be allowed to manipulate the levers as he chose, in any manner, rational or irrational, without any collision between trains occurring, so long as the signals were obeyed.’

  The beginnings of the block system, as illustrated in a children’s book of 1962. The gates represent signal boxes, staffed by boys who allow only one movement through the section at any time

  Absolute block put an end to the routine use of the time-interval method, which permitted trains to follow one another at fixed minimum intervals. The Metropolitan Railway, intensively worked in smoky tunnels, had the good sense to install the absolute block system from the line’s beginning in 1863. Anyone could see that the system was safer than the alternatives, and the Board of Trade had been urging its adoption on the sometimes reluctant railway companies for years before the Armagh disaster brought matters to a head. Safety did not come cheap: adopting absolute block on the Lancashire & Yorkshire Railway, its chairman complained in 1883, had cost £581,000 and had increased annual operating costs on the company’s busy and intricate network by £80,000. Long sections that previously used the time interval had to be chopped into shorter blocks, each with its box, signals and continuing extra outlay on labour. The chairman calculated that all this was enough to knock half a percentage point off his company’s dividend.

  One alternative was so-called permissive block working, which allowed multiple trains to tailgate one another within a single section, each driver keeping an eye out for the rear lamp of the train in front. The method suited goods lines with steady flows of traffic, such as those serving docks or mining districts, and was sometimes also used where passenger trains moved slowly around busy stations. So the old ways lingered here and there.

  The working principle that only one train may occupy a section is still apparent on many single lines. Here the supreme danger is not one train running into the back of another, but a head-on collision between departures that have mistakenly been admitted from opposite ends. Railways of the 1830s and early 1840s were nearly all double-tracked, so the risk did not arise. Telegraphy permitted safe two-way operation over single lines, which were also much cheaper to build. Extra safeguards were provided nonetheless. One method was to require a pilotman to travel with each train. Since he could not be in two places – on two trains – at once, a collision could not occur. To replace the pilotman with a symbolic item such as a wood or metal staff was a straightforward economy. Protection by means of a single staff was awkward, however, because it presumed that trains would pass through consistently in alternate directions, ferrying the staff up and dow
n in unbroken sequence. This was no use when one or more trains had to follow each other, and a method was found to relax the rule in these cases: the driver of the first train (or trains, if there was more than one) was simply shown the staff, as an assurance that the line was under the protection of the box in question, and a ticket was issued for the journey instead. In the 1880s a more sophisticated system was developed by which a supply of tablets or tokens was kept at both ends of a section, secured within instruments that were connected electrically in such a way that only one could be released at a time. A fresh token could be issued at either end only after the first one had been surrendered by the departing train and reinserted in the apparatus.

  To save trains having to stop at every box even when the line was clear, the transfer of these tokens or tablets – the fireman handing over one for the section just completed, while accepting another for the section following – was commonly managed by a skilful manoeuvre. The tablets were kept in leather wallets attached to big metal loops that could be slipped over the arm as the train moved past. On the Exmouth branch, as remembered by the poet and railwayman’s daughter Patricia Beer (1919–99), the exchange was made ‘with dramatic intensity and a fine swinging action while the train was moving quite fast’. In some places the process was eased by the use of fixed receivers and holders alongside the track, which could cope with speeds of up to 60 mph (thrillingly, mailbags used to be picked up and dropped off by comparable methods, but this ceased in 1971). Now that cabs no longer have a crew of two, passengers on single lines are more likely to find that their train stops briefly a little way short of the station platform, or just beyond it, so that the driver can perform the necessary exchanges.

  Where no signals or tablets are apparent, there may indeed be none in force. Lightly used branches may operate on the principle that only one train may enter at a time, after which it can run safely up and down as the timetable directs. Or there may be signals of an invisible kind. A method of train control by radio was developed for use on the lines to the far north and north-west of Scotland in the 1970s–80s. Tokens are still exchanged under this system, but of a virtual kind: a driver requests permission by radio, and the signal that is sent in response can be issued to only one train at a time.

  Such developments make it still more extraordinary that the wires, arms and rods invented by Victorian signal engineers are still sending trains safely along substantial parts of the national network. What remains is rudimentary stuff, however, by comparison with the bewildering displays that were once common in the busiest locations. Signals here might stand in groups like ship’s masts in harbour, a few arms projecting from each. Or they were combined into enormous gantries spanning multiple tracks, with twenty or more arms and a multitude of the lesser arms and sundry special lights and markings required to cover shunting and reversing in and out of all the various running lines, passing loops and sidings.

  All such structures had to incorporate walkways and ladders for tending and replacing the lamps. These had to be placed with care, so that the lamp-man or signalman would not obstruct vital sightlines when about his business. Wicks on quieter lines might be lit and extinguished daily, but it was easier to deploy lamps that contained enough oil to burn for eight days, sufficient to allow a rota of weekly attention. Among the personnel who regularly trudged the lineside were the lamp-men, patiently carrying fresh lamps and fuel to the signals. Even after the most careful trimming, it was not unknown for lamps to blow out in between their visits. Climbing up and down a spindly, shuddering post or gantry on a windy night to rekindle the flame in a ‘black light’ was not the most popular task on the railway; if a signalman could not be spared from duty, a platelayer might be called out to do the job. Oil was also the usual fuel for the headlamps and tail-lamps carried on the trains themselves, so that the railways continued to depend on the lamp huts and lamp rooms dotted all over the network, even after gaslight or electricity conquered stations and carriages. Dickens’s fictional lamp-man in Mugby Junction (1866) had a ‘greasy little cabin’ for the purpose, its walls and bench marked by oily smears and smudges left by shoulders and velveteen trousers. The man himself had even begun to resemble the element of his labours, with a ‘peculiarly shining transparent complexion’, and hair that stood up so straight that ‘the top of his head was not very unlike a lamp-wick’.

  Work at a large station or intersection also exacted a busier and more demanding routine from the signalman. This in turn meant better pay: as remembered by one of the men employed at St Helens Junction No. 1 Box in the 1920s, an extra five shillings a week could be earned in the busiest roles on the London & North Western, on top of the basic signalman’s wage of £2 10s. His own box had sixty-four levers and the traffic required him to memorise thirty or forty bell-codes – a typical tally on an intensive route – in order to communicate with the boxes on either side.

  The largest boxes tended to be very long, and sometimes also very high, without being proportionately deep, all the levers inside being in a single row. Sometimes the gantry principle was adopted for the signal box itself, as at Clapham Junction in 1907, when a weird structure like an elevated bowstring bridge was set up across the tracks at the London end. The operating part was housed in an assembly of little timber cabins on the bridge. Like other boxes on vital routes, it was given a protective steel hat during the Second World War, the framing for which remained in place until the whole contraption was taken down in 1990. Many men might work side by side in boxes of this size, as along a bar in one of the giant urban pubs of the era. The largest box of all, at York Locomotive Yard, required a staff of eight to operate its 297 levers.

  Large or small – occasionally very small indeed, like the tiny cabin on the platform at Damems on the preserved Keighley & Worth Valley Railway in Yorkshire, which looks almost like something meant for growing tomatoes in – each signal box carried a nameboard to identify its location. The naming of junctions in particular did not flinch from admitting unglamorous truths: Gas Works Junction, Coke Works Junction, Sheet Stores Junction, Factory Junction. More common were titles taken from nearby settlements, helping the curious traveller to guess at the identities of village churches or country houses visible by the line. Larger stations usually needed more than one box, so their names were distinguished by local features, suffixes of numbers or compass points, or by occasional flights of fancy such as the North Eastern Railway’s ‘Erimus’ at Middles-brough (‘We shall be’, adopting the young town’s motto) and ‘Severus’ at York (after the Roman emperor who died there in AD 211). As it happens, the world’s largest surviving mechanical box goes by the name of Severn Bridge Junction, and stands a few hundred yards from the platform ends at Shrewsbury: a 35ft-high monster with an original complement of 180 levers, built by the London & North Western in 1903.

  The electrified lines that converge at Clapham Junction were typical of the intensively worked suburban lines where colour-light signals made early headway. Colour lights, rather than semaphore arms, were the obvious type to use in the permanently dark tunnels of London’s Tube lines, which ran well below the sub-surface routes taken by the Metropolitan Railway and its kindred. One such was that odd anomaly the Waterloo & City Railway, which did not pass into the control of the main underground system until 1994. The line opened in 1898, which means that all-electric signals have Victorian roots. It proved harder to develop lights powerful enough to be legible at a distance under the open sky. By nice symmetry with the Waterloo & City, the first successful installation was raised above the earth-bound world: it belonged to the Liverpool Overhead Railway, an astonishing electric facility that began running in 1893. The only British parallel to the elevated urban railways of New York and Chicago, the Liverpool line ran on a wrought-iron viaduct five miles long, along the frontier between the city and its enormous docks. Its colour-light signals were introduced in 1920 and lasted until the entire concern went for scrap in 1956 because there was no money to keep it in repair.


  Colour-light signals could be placed anywhere within clear sight of passing trains. Their green, yellow and red lamps abolished the semaphores’ cumbersome splitting of the ‘distant’ and ‘home’ features between separate signal arms, nor did they depend on the lamp-man for replenishment. They also made extra settings possible. Once again, London suburban routes proved fertile ground for innovation. Holborn Viaduct station received four-aspect lights in 1926, in which the change from red to green was made by two stages, yellow and double yellow. Multiple aspects of this kind allowed more information to be conveyed by fewer signals, which meant in turn that fewer boxes were needed to operate them. The route into central London from New Cross, near where the first semaphores had waved their arms back in 1841, went over to lights of this type shortly after the Holborn Viaduct conversion. The previous arrangements were hideously congested: ‘In rush-hour conditions often near to chaos the trains groped their way from one manually-operated box to another: New Cross A and B, North Kent East, Southwark Park, Spa Road, London Bridge A and B, Borough Market, Metropolitan Junction, Union Street, Waterloo A and B, and finally into Charing Cross.’ All this in a journey of less than five miles: thirteen signal boxes each endlessly resounding with call-and-response codes for every train passing from one section to the next. The replacement boxes made use of pneumatic or electric power in other ways, as the weighty levers needed to pull signal wires and point rods securely into place gave way to banks of much smaller hand-sized versions.