© Railway Wonders of the World 2012-23  Contents  |  Site Map  |  Contact

Railway Signalling

The Development of Signalling on Britain’s Railways

LONDON BRIDGE SIGNAL-BOX controls one of the most important railway stations in Great Britain. The electrically-operated cabin contains over three hundred levers, and all train movements in the vicinity are indicated by coloured lights on diagrams in front of the signalmen. A total of two thousand four hundred trains is handled every day at this station, which has six through tracks as well as fifteen terminal platforms.

THE railways of Britain are guarded by signal cabins spaced at least every five miles and on lines with heavy traffic, particularly in the vicinity of large stations and junctions, the cabins are sometimes within a few hundred yards of each other. Pulling and pushing the heavy signal and point levers for 12 hours or more at a stretch is arduous.

Signals in the very early days of the railway generally were given by flags, lanterns and hand and arm signals. “Policemen” stationed at intervals alongside the track were responsible for the safety of trains by signalling the time which had elapsed since the preceding train had passed his position. For many years in Britain signalmen were called “Bobbies” because their original duty was to police the line. However, the first passenger steam railway, the Liverpool and Manchester, installed rotating “disc” signals within four years of its opening in 1830 and on the London & South Western Railway disc signals were introduced in 1840.

As train speeds increased it was found that a man with a flag or lantern standing at track level could not be seen in time by the driver of an approaching train and means were sought to improve signalling by erecting rotatable masts with shaped boards. Although the boards were not all disc-shaped (the actual shape varied according to the railway company) “disc” is a convenient group description of this type of signal to distinguish them from the semaphores which came later.

One alternative tried in 1840, but not perpetuated in Britain, was a ball hoisted to the top of a mast to indicate when the line was clear. In North America this type of signal was used for many years and from its use came the expression “highball” to indicate “proceed”.

Brunel’s Great Western disc-and-cross-bar signals, first used in 1838, were of exceptional size, sometimes sixty feet above the ground with the crossbar eight feet or more long; some remained in use until the 1890s. In the “stop” position the crossbar was presented. If the line was “clear” the disc was displayed by turning the post through ninety degrees. For many years these signals, and similar types on other lines, were operated by the policemen moving a lever at the base of the pole. Control by a lever at a distance through wires did not come into use for some years. It is said that a policeman at Watford on the London & North Western Railway in 1846 devised the first system of operating a signal at a distance. He used a weight on the signal and a wire running to his hut.

The “discs” on rotatable posts were gradually superseded by the semaphore signals which were to become a distinctive feature of British railways. The first semaphore signal, supposedly derived from the signal arms used by ships of the Royal Navy (but more likely derived from the Chappe land relay chain) was installed by C. H. Gregory on the London and Croydon Railway in 1841. At first the semaphore arms could be set to three different positions: the arm horizontal to indicate “stop”; at 45 degrees to show “caution” and vertical, concealed in the slotted post, “all clear”. Although in general these indications, or aspects, replicated the displaced policemen’s arm signals, there was little standardisation between individual railways, nor indeed any need for it in early years, of signal aspects and their interpretation.

STATION POST SIGNAL of c.1844 set at caution and danger.

An essential part of a signalling system is communication between signal cabins. Despatching trains on the time-interval system had to be abandoned gradually as traffic and speeds increased above certain low levels, and the electric telegraph was introduced into railway operation to provide the answer. However, the single-needle telegraph, in which letters were represented by the deflection of the needle, could not be adopted extensively for railway use in the 1840s because the greater number of station masters and signalmen could neither read nor write. This situation was emphasised when I. K. Brunel gave evidence on railway safety before a Parliamentary Select Committee in 1841. The North Eastern Railway adopted the “speaking” telegraph in 1846, although even until as late as 1851 illiteracy among staff restricted its usefulness as a means of communication between stations and signalmen. However, in that year C. V. Walker, of the South Eastern Railway, introduced the single-stroke bell telegraph system using a simple code of bell strokes which could be learnt easily by the signalmen. For example: one stroke - “attention”; two - “train entering section” and so on.

Around 1850 there was significant progress in block telegraph systems, notably by Edward Tyer who simplified the equipment and reduced the number of wires and batteries required to operate it. He also anticipated the modern track-circuit by developing a treadle rail with electrical contacts for announcing the position of trains.

In 1861 he invented the step-by-step “alphabetical” telegraph for communicating between signal boxes. This was the precursor of the train describers with which signalmen could supplement the single/block bell code by “describing” each train forward to the next box. For example, in addition to “offering” and getting “accepted” from the next signal box and advising “train entering section”, the describer could be used to pass details of a train, eg, “Up fast parcels train for branch line”.

The single-needle telegraph came into use alongside the block bell telegraph as a method for passing traffic information and public messages. Signalmen could “listen with one ear” to the clicking telegraph needle at the same time concentrating on the track diagrams, the signal indicator lamps and the signal levers.

During the 1870s more and more signal and point levers were being concentrated in frames at one point so that they could be interlocked with each other to prevent the signalmen displaying signals which conflicted with the position of the points. Instead of signalmen who operated the signal levers and pointsmen who stood at the point levers beside the track, all were brought under the control of one man in a signal cabin. This became the standard British arrangement and one which was encouraged by legislation administered by the Board of Trade. It was not, however, adopted in other parts of the world even when there were traffic conditions sometimes as complicated as those of this country.

In the 1890s the three-position semaphore went out of favour because, for one reason, it was not easy to arrange the mechanism so that the arm was accurately set. Sometimes the arm took up a position between stop and caution. It was replaced by the two-position signal with the arm falling vertically into the slot in the post, as before, to indicate “clear”. However, the semaphore arm moving into a slotted post had the serious drawback of occasionally becoming frozen in the “off” position by ice or snow. Such a “failure to danger” could hardly be considered satisfactory although in the signalling systems accepted in some countries fail-safe practices have yet to be enforced.

In Britain the lesson was learned early enough. At Abbots Ripton in 1876 the signalman acted correctly and set the signals for the trains passing through his section. In driving snow and freezing air temperature, the southbound “Flying Scotsman” was rapidly overhauling a slow freight train on the same line. Blinded by snow, the driver of the freight overran the stop signals at Holme, where the signalman intended to shunt the freight into a siding out of the way of the Scotsman, which was only 16 minutes away. It was not until the freight train had run past two more signal cabins, which were not on the telegraph line and therefore were unaware of the developing situation, that the signalman at Abbots Ripton, advised by the telegraph, stopped the train and instructed its driver to back into the siding.

The signalman to the north of Abbots Ripton had correctly set his signals to danger behind the departing freight train and would keep them “on” until he received the bell code, “train out of section” from Abbots Ripton. Within a few minutes the block bell told him that the Scotsman was approaching his signals. To his surprise it did not slacken speed but swept past and disappeared to the south, lost in the swirling snow. At Abbots Ripton it crashed into the side of the freight train which had not completely cleared the main line. The driver of the express had seemingly ignored the distant signal for Abbots Ripton; the lever for which and that for the home signal were both in the “on” position. The disaster was worsened because of poor communications and the appalling weather conditions, which prevented a northbound express from being stopped before it crashed into the wreckage.

The driver of the “Flying Scotsman” had been lured into danger by signals giving a false indication that the line was clear. The signalmen concerned had acted correctly but the signal arms were frozen in the clear position by snow packed in the slots in which the arms moved. The balance weights designed to pull the arm to the stop position in the event of a broken wire could not overcome the resistance of the frozen arm and the pull of wires weighed down by frozen snow.

Abbots Ripton highlighted inadequacies in the signals and their equipment of the 1870s and the difficulties of drivers and signalmen working in adverse conditions. However, as has happened before and since, the report of the Board of Trade Inspector on the accident contained recommendations intended not only to avoid a recurrence but to improve the safety of the railways generally. Among them was the abandonment of the type of semaphore signal which fell to “clear” into a slot in the signal post.

THE RAILWAY ACCIDENT AT ABBOT’S RIPTON. The scene on the line after the double collision.

Even more important however was the abandonment of the principle of keeping signals normally in the “off” position and putting them to “danger” after a train had passed. Instead, in British railway practice, the signals were to remain normally in the “on” position and were to be cleared only after the necessary bell signals had been exchanged and the line proved clear.

Had that been the practice at Abbots Ripton then the signal arms might have frozen in the danger position which, while inconveniencing operations, would have been a “fail-safe” method. Another outcome of that historic accident was the adoption by the Great Northern and by some other lines, of the balanced or somersault signal designed to be unaffected by such things as the weight of snow.

Safety of train movements had been the subject of a number of Governmental regulations which were introduced successively from 1839 onward. The Board of Trade and its inspectors were empowered to investigate operating practices and the provision of satisfactory methods and equipment for the safety of railway passengers. However, it was in the Railway Regulations Act of 1889 that the Board of Trade finally consolidated and codified regulations and orders applying to all British railways. The regulation covered the adoption of the absolute block system of working for all passenger lines, ie one train only in a section at a time; the interlocking of all points and signals on passenger lines; and the adoption of continuous brakes, operated by air or vacuum, on all passenger rolling stock.

Signals proliferated as railway companies complied with the Board of Trade requirements that all passenger train movements had to be under the control and protection of signals interlocked with the points and with other signals which might conflict. Even every exit from little-used sidings had to have the appropriate shunting signals to govern movements. There were of course variations in the way each company arranged and operated its signals. On some lines there was a separate signal for all possible moves; on others only the minimum to satisfy Board of Trade regulations.

The development of the electric telegraph in early Victorian times and its adoption by the railways laid the foundation of modern railway signalling, yet it took 60 years of haphazard railway operation and many accidents before it was made compulsory. Some railways adopted a primitive form of telegraph for sending messages about trains running over dangerous sections of line, tunnels, for example, as early as the late 1840s. But the equipment was cumbersome and messages were spelt out letter by letter. Soon the railways adapted telegraph instruments into a simpler form to show the state of the line between two stations merely by the position of the telegraph needle supplemented by bell code signals so that signalmen could advise each other of the movement of trains.

Some of the most progressive railways quickly realised the advantages of the telegraph system for it meant that signalmen had a positive indication of whether the section of line they controlled was occupied by a train or not, a far better state of affairs than the old time-interval system in which signalmen had no means of communication between one station and the next. The sections of line, which generally ran from one signalbox to the next, were known as block sections, and the method of controlling trains by the electric telegraph became known as the block system. The principal rule was that there should not be more than one train in a block section on one line at one time and signalmen were able to advise each other as trains passed into and out of each block section by the electric telegraph and bells.

A system was evolved whereby a signalman, before he could clear his signals for a train, “offered” it by bell code to the next signalbox ahead and if the block section was clear the train was “accepted” by repetition of the bell code back to the offering signalbox. As the train passed the first box the signalman sent the “train entering section” bell signal to the box ahead, and when it arrived or passed the signalman there sent back the “train out of section” bell signal. The block indicator needle, worked by the signalman at the exit end of the block section and electrically repeated at the entry end box, was used to show the state of the line and whether a train had been accepted or was actually in the section. This system remains in use today on lines still controlled by old-style mechanical signalboxes.

Some railways were slow to adopt the block system but it was finally ordained by Parliament in the Regulation of Railways Act 1889 that it should be installed compulsorily. This Act also required that all signals and points controlled from one signalbox should be interlocked with each other at the lever frame so that conflicting indications could not be given. This meant that a signal could not be cleared unless the points were correctly set, and at junctions two signals from different lines leading to a conflicting route could not be cleared at the same time.

Facing points, that is points giving a direct running move from one track to another, were avoided wherever possible except, of course, at route junctions or large stations; at small country stations facing crossovers from a line in one direction on to the line in the opposite direction were almost unknown. Indeed, this sort of crossover was nearly always provided as a trailing connection so that if a train had to cross to the opposite line it had to run beyond the points, stop, then reverse back through the crossover. By this means the risk of head-on collision by trains being accidentally diverted on to the opposite line was virtually eliminated.

Signals themselves by that time were almost entirely of the two-position semaphore pattern which denoted stop with the arm horizontal and clear with the arm lowered at 45 degrees. In the danger position at night they showed a red light and in the clear position a white light. Arms were generally painted red with a white vertical stripe near the left hand end; “distant” signals, that is the advance warning signals which repeated the indication of the next stop signals ahead, had a vee notch cut out of the left hand end. Signal arms on most lines were by then pivoting outside the post instead of within a slotted post, but some railways adopted the somersault arm which was formed of a centrally balanced arm coupled to a separate glass spectacle casting designed to defeat the unbalancing action of a snow build-up. A number of accidents had been caused because snow had affected the working of normal semaphore arms and caused them to show a false clear indication.

SINGLE LINE WORKING METHODS. This illustration shows a single line token fitted to a lineside post to facilitate picking up and setting down. The inset shows the electrically controlled token holding instrument. With this instrument a second train token cannot be removed in error, either from that end of the section or from a similar instrument at the other end, until the first token is replaced in either instrument.

Other safety devices included facing point locks in which a bolt positively locked the point blades fully home in their normal and reverse positions so that there was no danger of a point blade standing away from a rail and causing a derailment. The point lock bolt was connected to a steel bar longer than the space between any two pairs of wheels on locomotives and coaches laid against the inside edge of the rail approaching and through the points. Before the points could be moved they had to be unbolted by a point lock lever in the signalbox, during which operation the bar, called the fouling bar, rose up to rail level. If a train was standing or passing over the fouling bar the wheel flanges prevented the bar from being lifted and in turn the points from being unlocked and moved.

On single lines it was thought necessary to have something more than the block system to prevent head-on collisions and the staff system was used. At first this consisted of a large wooden staff, one for each block section, and the driver of every train passing through the block section had to carry the staff as his authority to be on the single line. As there was only one staff for each section it followed that if everyone obeyed the rules there could be no collision. The basic system proved inflexible because it meant that ideally trains ran alternately in each direction and this did not always happen in practice.

From it was developed the staff and ticket system in which if several trains had to follow each other in the same direction over the single line the drivers of all but the last train were shown the staff but given a written ticket to proceed, while the last train of the group carried the staff. The tickets were kept in locked boxes opened by a key on the staff. Even this system was cumbersome; its place was taken on many single lines by the electric staff instrument or its similar but smaller related systems - the tablet and key token. The signalboxes at each end of a single line section were equipped with electrically interlocked staff instruments with several staffs in the system. Electric locks were fitted to the two instruments so that only one staff could be taken out of either instrument at one time, thus again ensuring the security of the single line.

With the virtual standardisation of the two-position semaphore signal and the abandonment of the old three-position type which, at night, showed red for “danger”, green for “caution” and white for “clear”, the green light was gradually adopted for the clear indication instead of white, which could be confused with ordinary lineside lighting, and particularly gas lights, then being adopted more widely in towns and on stations.

By 1900 the wonders and use of electricity were becoming known. The early years of the present century saw the introduction of track circuits, in which a weak electric current was passed through an insulated section of the running rails and when short-circuited by the wheels of a train occupying the line could be made to operate electro-magnetic relays, which in turn were employed to illuminate lights of signalbox track diagrams or to operate locks on signal levers.

Electricity was also used to power electric motors connected to points or signals which could be controlled a greater distance away from a signalbox than by mechanical means. Indeed there was (and is) an absolute limit of 350 yards on the mechanical operation of points by rodding from a signalbox and the practical limit for a wire-worked signal was about three-quarters of a mile.

The earliest application of automatic signalling in Great Britain was on the Liverpool Overhead Railway in 1893 where electrically worked semaphore signals were placed to danger by an arm on the back of a passing train striking a lever on a contact box fixed on the lineside. As the train progressed the operation of the striker arms altered the electrical connections to place the signals immediately protecting the train to danger and cleared the signals at the previous station to allow another train to proceed.

During the next 20 years, which saw the construction of most of the deep-level underground lines in London, improvements in signals and signalling systems were evolved gradually. At first some of the tube lines used the normal block system with signal cabins at each station controlling miniature semaphore signals illuminated by oil or gas lighting. Later, electricity was used for signal lights, still retaining a moving arm or vane containing coloured glass, but eventually the colour-light signal, which employed separate bulbs for each colour indication and had no moving parts, was adopted. Track circuits were gradually introduced to allow automatic operation of signals, which permitted the closure of many signalboxes at intermediate stations where no points existed.

A MODERN ALL-ELECTRIC SIGNAL BOX AT WATERLOO. In this signal box there is a total of 309 levers arranged on three frames of 75, 159 and 75 levers. Each of the four signalmen faces an illuminated diagram of the tracks concerned, on which the sections occupied by trains are shown by the glowing of small red lights.

Main-line railways also took advantage of power operation in some new signalling installations; in 1898 the London & North Western introduced electric working of signals and points in the Crewe area. The lever frames no longer needed to be fitted with the long levers necessary for mechanical operation to obtain adequate movement of wire or rodding, and instead short levers, still with mechanical interlocking, were used to operate electrical contacts which transmitted the current to the signal and point-operating equipment. In 1902 the London & South Western Railway introduced automatic signalling controlled by track circuits, with semaphore arms powered by low-pressure compressed air through electro-pneumatic valves, on its main line between Woking and Basingstoke.

The track circuit was also gradually adopted to provide better protection in mechanically signalled areas controlled by normal block working, for it could be used to prevent a signalman clearing a signal leading to a section of line already occupied by a train.

Other allied safety devices similar in effect, if not in principle, were also used, including the lock and block system. This pre-dated the track circuit and was used extensively on some lines in Southern England and in North East London. It employed treadles situated along the rail edge which were depressed by the wheel flanges of a passing train and used to actuate or release locks in signalbox equipment. Generally the system provided complete locking between the block instruments and the signals in such a way that a train had to pass the signals at one signalbox and the signals had to be restored to danger behind it before the block instrument for the section the train had just left could be released and cleared for a second train. In turn, the signals at the previous station could not be cleared for a second train until the block instrument for that particular section was put in the clear position. In this period also the telephone had been perfected and was installed widely for giving information on train running to supplement the block system and the needle telegraph still employed in its original form for sending general train messages.

In the years before and after the First World War more changes became apparent in signals, particularly distant signals. One or two railways had begun to install three-position signals again but this time the arms worked in the upper quadrant, that is to say, from the horizontal danger position they were raised to 45 degrees for caution and upwards vertically for clear. In the “caution” position at night they showed a yellow light which conflicted with the normal caution indication of two-position distant signals which on most railways still showed red; gradually, however, some companies adopted yellow for the night-time indication of a distant at “caution” and at the same time painted the arm yellow with a black vee stripe.

Indeed, because of the possible confusion between the two types of signal, the growing introduction of colour-light signals and automatic signalling, and the possible need for additional signal indications, a signal engineers’ committee was set up in 1922 to examine signalling needs for the future. It advised against the adoption of the three-position semaphore and found that the normal two-position semaphore would be adequate for ordinary working. However, in closely signalled areas with a frequent service the committee recommended the adoption of the four-aspect colour-light system using red for “danger”, one yellow for “caution”, double-yellow for “preliminary caution”, and green for “clear”.

The Great Central Railway was the first to adopt automatic colour-light signalling out of doors (as distinct from the underground lines in tunnels) in 1923, and two years later the first re-signalling schemes on the Southern Railway’s South East London approaches embodying the four-aspect colour-light system were introduced. The final abandonment of the three-position semaphore left the way open for the adoption of the upper-quadrant two-position semaphore which was raised to 45 degrees for the clear position. The upper-quadrant type of signal was of lighter construction than the lower quadrant, since it returned by gravity to danger instead of being weighted to ensure that the arm returned to the horizontal position, both when the lever was put back to normal in the lever frame or in the event of a broken wire. As a point of later railway history the upper-quadrant signal did not completely oust the lower-quadrant type, for the Great Western Railway continues to use the lower-quadrant semaphore, and even on other lines isolated lower-quadrant signals survive in one or two places. The recommendations of the committee were so far reaching that there has been no radical change in signalling indications in Britain from then until the present time, nor is there likely to be while lineside signals exist.

The 1920s and 1930s saw the gradual extension of colour-light signalling but still largely worked by signalboxes controlling limited areas. While the old mechanical limits no longer applied, the power signalbox of that period still controlled signals and points by individual miniature levers - one lever per function - and control areas were generally not much more than a mile or so from the signalbox. Sometimes lengths of automatic signalling intervened between adjacent signalboxes. In 1927 the Great Western took signalling control a stage further by the adoption of route levers at Newport (Monmouthshire). In this system the levers controlling the signals (which were electrically operated semaphores) also set the points to which the signal applied in the one operation. The movement of the miniature lever by the signalman was in stages, first to a check-lock position as the equipment proved that the line was free to be used and that no other conflicting move had been set (this was achieved by track circuits and normal interlocking), then to a position which operated the points, and, finally, when they had been proved in the correct position, the lever movement was completed which cleared the signal.

During the 1930s the LNER introduced the first all-electric signalbox control panels. Instead of being controlled by levers, signal and point movements were initiated by thumb switches. The turning of a switch on the signalman’s control desk, like the Great Western’s system at Newport, proved the section of line free for use, operated the points and cleared the signal, from one signal to the next; in this case, however, the signals were colour-lights. By now the locking between the signalman’s control switches was achieved electrically by relays instead of mechanical locks.

You can read more on

“Automatic Safety”,

“The Magic of Modern Signals”

and

“Ruling the Traffic Routes”

on this website.