UK Heritage Railways - History of Signalling

A Brief History of Railway Signalling

Most people when railway signalling is mentioned would immediately think of the trackside signals. Certainly these are the vital means of informing drivers of approaching trains about the state of the line ahead, but of equal importance is ensuring that the signals shown are correct. This is achieved by interlocking all relevant signals and points to prevent signals being shown for conflicting movements. Good communications between those responsible for operating signals is also vital to operating a safe and efficient railway. Unfortunately developing such a system did not happen without a certain amount of trial and error....
Signals
The first passenger carrying railways began operating in the late 1820's and it was not very long after this that people realised the need for some system to prevent two or more trains attempting to occupy the same space at the same time with disastrous consequences. Early methods relied upon policemen at stations, crossings and junctions using hand held flags to signal passing trains. There was no communication between policemen, they simply relied upon a time interval system to prevent following trains from running into the train ahead. The policeman would show a red flag (for DANGER) to following trains for 5 minutes after a train had passed him, and a green flag (CAUTION) for a further 5 minutes. Only after more than 10 minutes had passed would a white (line CLEAR) flag be shown to approaching drivers. If a train stopped unexpectedly after passing a policeman and out of his sight, then the driver of a following train only had his own vision for warning - not much use on a dark and foggy night! Bear in mind also that, although speeds were slower in those days, the only brakes available were on the engine and guards van. Braking distances were therefore much longer than are required by todays trains. It could take a mile or more to stop even a lightly loaded train from 60 mph. The early flag signals were soon replaced by more permanent signals in the form of variously shaped boards (each railway had its own system) which could be pivoted to face oncoming trains when it would be dangerous to proceed. When the line ahead was clear the boards were simply turned edge on so that drivers could not see them. There was therefore no positive CLEAR signal, merely the absence of a danger signal. These early signals were often placed for the convenience of the policeman operating them. They told the drivers when to stop, but not where! A stop signal at a junction could be placed beyond the actual junction it was protecting. Drivers were supposed to know where to stop as part of their route knowledge. Advance warning of signals could be given by auxilliary signals placed a lttle way before the stop signals they applied to. These were also simple pivoting boards worked by hand with no direct interconnection with the relevant stop signal. Signals at night were given by coloured lights. In early days these would have been hand lanterns. Although later replaced with lamps on signal posts they were often still worked independently to the daytime boards. The colours used were RED for DANGER, GREEN for CAUTION and WHITE for CLEAR.
	Brunel was one of the first to provide a positive line CLEAR signal on the Great Western Railway in 1841. His `disc and crossbar' signals could be rotated to show a red disc to indicate line CLEAR or a red bar to indicate DANGER. To assist the identification of signals at a distance, signals for the down line (ie heading away from London) usually had a downward extension at each end of the bar. Several holes were cut in the disc and the crossbar to reduce wind resistance. A lamp with coloured lenses was used at night and was linked to the same lever as the bar and disc.
Semaphore signals had been in use by the Navy as a means of communication for several decades before they were first used for railway signalling purposes in 1841 at New Cross by the London & Croydon Railway. It wasn't until 1843 that the next semaphore signals were installed at Bricklayers Arms Junction. These were installed into a common lever frame with a crude form of interlocking to prevent opposing signals being cleared together. However there was nothing to prevent points and signals from disagreeing. Further semaphore signals followed and were often set back to back (ie. two arms on a common post controlling movements on adjacent lines). General practice was that the arm was always to the left of the post, if you saw an arm to the right then that signal was for trains travelling in the opposite direction.
	Many of the early signals had three positions, indicating DANGER, CAUTION and CLEAR respectively. Unfortunately the CLEAR aspect was when the signal arm was hiding inside the post, so rather than providing a positive CLEAR indication they were back to the old problem of the line clear signal being merely the absence of a danger signal. Another problem which occured (and caused several accidents) was icing up of the slot in bad weather preventing the signal moving to indicate DANGER.
	Eventually semaphore signals were installed on most UK railways and generally had just two positions, DANGER when the arm was horizontal and CLEAR when the arm was at an angle (which was usually between 45° & 60°). Most were originally lower quadrant, but the majority of railways (The GWR being the most notably exception) later changed to using upper quadrant signals . These had the advantage of being failsafe. If a linkage broke, then the arm simply dropped to the danger position. Seperate arms were used for STOP and DISTANT signals. Additionally, STOP signals had square ends and were painted RED on their faces with a white stripe, the backs being white with a black stripe. DISTANT signals had a notched end and were painted YELLOW with a black chevron on the front, the backs being white with a black chevron.
Another important change which occured around this time was the changing of colours for signals at night. With the increasing use of street lighting it was becoming more difficult for drivers to pick out which white lights in the distance were railway signals and which to ignore. Accordingly, the colours were changed so that GREEN became the CLEAR signal and YELLOW was used for CAUTION. RED continued to be used for DANGER. (You may have noticed that semaphores usually contain a blue filter, this is because the oil lamps which were used, produced a yellowish flame - yellow plus blue is seen as green).
Interlocking
As track layouts became more complicated and the frequency of trains increased, so the potential for accidents also increased. It was not good enough to have clear signals for drivers if the signals themselves were still operated by isolated individuals having to run back and forth between the various signals and points they controlled. There was clearly a need for a central control point for each station and junction. By bringing all controls to a central location it would also be possible to link them in such a way as to prevent conflicting signals being set to CLEAR at the same time. The first signal boxes were installed in the early 1860's by Saxby and Farmer on the approaches to London Bridge for the South Eastern Railway. Further signal boxes followed at London Victoria (which was little more than a hole in a wall overlooking the tracks), and Brighton and Charing Cross in 1864, and Cannon Street in 1866. These signal boxes originally controlled only the main signals and points for an area, with the points for the many local sidings still being controlled from individual levers next to the points. Over time, more and more controls were relocated to the signal boxes until all points on running lines were under the control of a signal box and only those inside yards were still controlled by the man on the ground. Various forms of mechanical interlocking were used to ensure that conflicting movements could not be set up by signalmen, although each installation obviously had to be custom built to fit the requirements of its intended location. Another major benefit of this was that signals could now be located for the benefit of the drivers rather than the convenience of the signalman. Signals could now indicate WHERE to stop as well as when.
Communications
As the speeds and frequency of trains increased, accidents occured more frequently - often with fatal consequences. These accidents were often simply due to a train stopping unexpectedly and the following train running into the back of it. Obviously some method of quickly passing messages between control points was required so that following trains could be given advanced warning of obstructions. In 1837 W. F. Cooke and Prof C. Wheatstone devised a five needle telegraph system which was installed on the GWR in 1839. This system used a diamond grid of twenty letters with five needles arranged across the middle, deflection of any two needles would point to specific letters on the grid. Obviously having 6 letters missing meant a certain flexibilty was require when sending messages, but the main disadvantage was the need to install and maintain 5 wires. As well as operational messages, it was also used for messages regarding the general business of the railway and also for public messages.
Unfortunately the system only lasted a year before deterioration of the insulation rendered it inoperative. It was replaced by a two wire system which received great publicity when it was used to transmit the news of the birth of Queen Victoria's second son in 1844, and again in 1845 when it was used to send a description of a man wanted for murder when he was seen boarding a train at Slough. This system lasted until the end of the 1840's, when it again fell into disrepair and this time was not replaced. Other railways however considered it worth pursuing and by 1845 several railways were installing telegraph systems.
All this was fine for major stations and junctions but was not extended to include the numerous smaller rural stations, even those on main lines. Cost was clearly one factor, but also the accident rate during the 1850's was very low - averaging 1 killed for every 10 million passenger journeys. The thinking prevelant in many railway boardrooms at that time was typified by the LBSCR's response to critisizm following an accident which occured at Clayton Tunnel on 25 August 1861 following a misunderstanding between two signalmen. Capt H.W.Tyler who inquired into the accident was scathing in his findings, not so much of the signalmen's misunderstanding, but of the Brighton stationmaster who had not observed the 5 minute minimum interval between departures, the lack of block telegraph on the line, the inadequate telegraph equipment between north and south ends of the tunnel, and the lack of registers for recording times of telegraph signals and passing trains. He recomended the adoption of the block telegraph system which would provide a means of operating a space interval system between trains and a minimum of three telegraph instruments to be provided for the signalmen at each end of a telegraph section, one for each track to act as a signal which could be pegged at line clear or line blocked and one for more general communications. The LBSCR's reply was: `The Board feel bound to state frankly that they have not seen reason to alter the views which they have so long entertained on this subject, and they still fear that the telegraphic system of working recommended by the Board of Trade will, by transferring much responsibility from the engine drivers augment rather than diminish the risk of accident'. At this time the railway inspectorate could only recommend, not insist. New signalling and safety developments were only adopted reluctantly by cost conscious railway boards. The safe operation of a time interval system was further complicated by the differing speeds of trains. Most routes had only one track in each direction and this was shared by passenger trains running at speeds up to 50mph and goods trains runing at 20mph. This meant that goods trains had to be given sufficient lead time to reach and shunt into the next set of refuge sidings. The dangers of points being left in the wrong position were obvious and facing points were therefore avoided on running lines. Trains entering refuge sidings were therefore required to draw past the sidings and then reverse into them, thus further increasing the time required to clear the line. However, pressure to improve operational safety mounted in 1861 when there were 56 accidents. 32 of these involved passenger trains, and 46 passengers were killed, 38 of these in just two major accidents: On the 2 September 1861, 15 people were killed and 317 injured when a North London Railway passenger train hit a ballast train shunting at Kentish Town. The lack of interlocking contributed to this accident. Just a few days before this, on 25 August 1861, the accident mentioned above at Clayton Tunnel on the Brighton line killed 23 and injured a further 176. This accident highlighed a number of issues and is worth examining in detail. Obviously there was a need to apply the new technologies available throughout the system, not just at major stations and junctions.
For single track lines there was need for even greater caution if head on collisions were to be avoided. For short dead-end type of branch lines it was a simple matter to arrange for only one train to shuttle back and forth along it. Thus establishing the `one engine in steam' principle. An early system used for longer routes was to designate a pilot engine for each section. A train could only move along that section of the line if the pilot engine was attached at the front. A safe system but not very flexible. It was soon realised that a simple token carried on the lead engine of each train would serve the same function as the pilot engine. This was cheaper but no more flexible so the system was modified so that several trains could run over the line in the same direction. The driver of a train would simply be shown the token and could then proceed over the line. The last train of a series would then carry the token to the other end before trains could use the line in the other direction. This was further refined by the introduction of tickets which gave a driver written authorisation to travel over the line when he was not carrying the token.

Last updated 6 September 1999 by Mark Dewell.