Return to index page

Signals - Mechanical Types & Fog Signals



Railway signalling is part science and part art form, it has always relied heavily on dedicated staff and in the early years on the memory of train crews. For details of the ways signals were actually used to control traffic movements see the section on Communications & Traffic Control. This section deals with the appearance and construction of the signals.

It is impractical to cover anything more than the basics of appearance here and the reader is recommended to track down reference books for their particular company. One source is of particular note; the small Peco publication 'Signalling the Layout' gives all the basic information required. The website at signalbox.org provides a lot of information on prototype signalling including numerous illustrations of the equipment used. One small point worth mentioning is that in railway signalling parlance the term 'ahead of' means the track past the signal, the word 'behind' or in 'rear of' means the track before the train reaches the signal. This is often a source of confusion for modellers trying to make sense of official instructions.

The 'bobbies' (signalmen) and gate-keepers at the level crossings first used hand signals, hand-held flags and hand-held lanterns to signal to the train crew. On the railway a white flag or light meant all-clear, red meant danger or stop and green originally meant caution. With the increasing use of gas and electric white lights the significance of green changed to all clear as of about 1893. As with most changes this took several years to apply, the GWR adopted green for all-clear in 1895 but green was still occasionally associated with a cautionary warning until the 1960's. From about the time of the First World War a white light was sometimes used as a danger indication on small track-side signals, discussed below.

Mechanical signals first appeared in about 1834 but there was no common standard. The pointed board signal illustrated below was developed by the Liverpool & Manchester Railway and became the first standard. When turned to face the driver it meant danger or stop, when turned edge-on to the driver it meant all-clear. The Great Western Railway initially used the same design but had a third option of turning the rear face of the board, painted green, toward on coming traffic to indicate 'caution'. In 1841 several companies held a meeting to discuss signalling and agreed to adopt the colours used by the L&MR to indicate 'danger' (red), 'caution' (green) and 'all clear' (white). Three options were required as most lines at the time were operated on the 'interval' system of working discussed in the section on Communications & Control Systems.


Signal designs remained many and varied for quite some time, some systems like the L&M board and the later square iron 'flag' type only giving an indication of danger, no positive indication of all clear. Another popular nineteenth century railway signal employed a circular disc, when turned edge-on and hence not visible it meant all-clear, when turned to face the engine driver it meant stop. If the board, flag or disc fell off its mounting the train crew might not realise it was missing and assume all was well.

The Brunel disc and bar signal appeared on the Great Western Railway in 1840, this had a circular disc about four foot in diameter mounted above and at right angles to a flat bar about eight foot long. When turned so the disc faced the train driver it meant all-clear, when the bar was visible the train had to stop.

Disc and bar signals of various designs were used on several railways, including the Somerset & Dorset, the North Eastern Railway, the Furness Railway and the narrow gauge Ffestiniog Railway, they remained in use in some locations until after the turn of the Century although later designs used smaller discs and bars. Signalling is not however a simple matter of stop or go and variations were introduced on the disc and bar and other systems to indicate such things as which route was being signalled.

There were also signals using balls and kite shapes, hoisted up masts beside the track. These were first used in Britain in 1840 and came from shipping where wicker basket-work shapes are still used today (for example a single black ball, about two foot in diameter suspended from the foremast means the ship has anchored). On the railways the ball hoisted up the mast indicated the 'all clear' so that if it failed (i.e. the ball fell off) the train would stop. This offered a 'fail safe' system which found favour with the French and the Americans. The term 'highball' meant a line which was clear ahead and American express passenger services are still sometimes referred to as 'highball' trains.

The drawings show an early 'fishtailed board' signal, where the entire post was turned, an early 'flag' signal (the flag is in fact an iron plate), and a 'ball' signal. The lower drawings show a Brunel disc and bar (the T shaped ends on the bar indicate it is a level crossing signal) and a Lancashire and Yorkshire Railway level crossing double-disc signal. In each case the signal on the left indicates danger, that on the right shows all-clear.


Fig___ Early signal designs

Sketch showing Early signal designs

Thanks are due to Neil Mackay for pointing out that the signal bottom right was L&Y not NER - He has a book out on NER signals: A History of North Eastern Railway Signalling bv Neil MacKay - published by the NER Association



It is doubtful if many of these early signals survived much beyond 1900, at least on main line railways. On remote branches and industrial lines however non-standard signals survived into comparatively recent times. A Brunel disk and bar type as shown above was in use at a level crossing on a running line in the Forest of Dean into the 1920's and one, again on a running line, survived (unused) in the same area into the 1950's.

The Midland Railway used a diskless bar type signal for shunting or blasting signals, this was adopted by the LMS (who built a few with their standard tubular metal posts) and remained in widespread use into the 1960's. They were used in yards where two sidings crossed one another and there might be a risk of collision and also on lines feeding quarries. In the yard the signal meant stop if the bar faced the driver, on the quarry line the bar was turned sideways to indicate that blasting was in progress. The arm had a multi-lensed oil lamp mounted on top to give an indication at night (signal lights are discussed below). On the Midland example the counterweight for the signal was mounted at the base of the post with a rigid rod mounted diagonally across the post operating the signal.

The signal in the lower left of Fig___ was originally a disk and bar type but the disk had at some point been removed, note the unusual placing of the counterweighted arm near the top of the post. This was still in use in the 1970's on a (steam operated) internal railway at a large steelworks.

One point to note is that modelling a working bar or disk and bar signal is relatively easy, simply use a length of brass tube mounted inside a square section Plastruct tube with a brass wire passing up through the middle attached to the signal arm. The lower end of the wire can extend down through the baseboard where knob or just a simple bend would allow it to be operated.


Fig___ Bar type signals in recent use

Sketch showing Bar type signals in recent use



In the 1790's a French clergyman by the name of Claude Chappe developed signalling system for passing messages between cities. This system was based on naval 'semaphore' and used tall towers on which were mounted pairs of swinging arms which were visible over considerable distances.

This system worked well and in 1841 Charles Hutton Gregory (chief engineer on the London & Croydon Railway) erected the first railway semaphore signal at New Cross (close by 'Telegraph Hill' which had a Chappe type semaphore signalling tower on top dating from the 1790's). Other companies recognised the value of this simple but effective signal and they spread quite rapidly across the system.

Early semaphores were many and varied, most designs used a post with a slot, when the signal arm was down inside the post it meant all-clear, at 45 degrees it meant caution and when lifted to the horizontal it meant stop. The drawing shows a basic slotted post semaphore signal and this design had arms for both directions on a single post, the convention was that the driver looked at the arm on the left of the post.


Fig___ Slotted-post three-position semaphore signal with 'policeman's box'


Slotted-post three-position semaphore signal with 'policeman's box'

The problem with the slotted post signals was that the arm could jam inside the post, and if the arm fell off the signal was left showing the 'all clear'. From the 1950s the rules required that the signal arm was always visible, bringing the common two-position signal in which the arm is always visible.

Early semaphore signal arms were of many shapes and sizes, where the posts were mounted on top of signal boxes the arms could be quite large, perhaps eight foot long by a foot or more high, but the most common designs settled on arms of about three or four foot long. The GWR, which had developed the disk and bar type signal for it broad gauge lines, replaced them quite rapidly after 1870 with semaphore types.


Fig___ Signal arm reverse faces.
Sketch of Signal arm reverse faces
The front face of the semaphore signal arm (facing the driver of the train) was painted red and the reverse side was painted white. At some point (I believe in the late 1850's) a white stripe was added near the end of the front face and a black stripe was added to the reverse. Some companies used white and black dots in place of the stripe and some companies used a horizontal stripe on the reverse face, the Midland Railway is one example.

On lines where trains ran at night a lantern could be mounted on the signal but on most types of signal this had to change colour when the signal position changed. Early systems used lanterns with multiple coloured lenses which were rotated by a linkage to the signal arm to show the appropriate colour to the train crew. In the early years oil lamps burning whale oil and even candles were used in these lanterns (the Great Northern Railway used candle power until the 1870's when a petrol lantern was introduced). Paraffin had been named in 1830 by a German chemist (actually the liquid we call paraffin is technically Kerosene) and in the later 1830's paraffin lamps augmented the whale oil lamps used on signals. The early lamps were thirsty beasts, average consumption of oil was in the order of twenty gallons of oil per lamp per year.

As these early lights did not last very long and sometimes were blown out by the wind they were sometimes mounted lower down the post where they could be easily reached. Quite early on a long-life paraffin lamp was developed which would burn unattended for eight days, a 'lamp man' was employed to changes these lamps once a week. In the early part of this century the design of the burner was improved, reducing each lamp's consumption to about eight gallons per year.

Fig___ Signal & Telegraph Dept. Lamp Man
Sketch of a Signal & Telegraph Dept. Lamp Man
Paraffin lamps, sometimes called 'oil lamps', continued in use into the 21st century as running cables to the older semaphore signals added a lot to the cost. Electric lanterns are sometimes used on more modern semaphore signals. There was a discussion on a news group on this topic, from which I gather than a common option is 12 volt usually 5 watt lamps fitted as these are about the same light as an oil lamp. Other contributors mentions some lanterns running on 110v. However where relevant to the layout signalling a lamp-man walking along the track would add a little interest.

The light on early (pre 1893) three position signals showed red for danger, green for caution and white for all-clear but, because the paraffin lamp produces a rather yellow light, the 'green' lens was blue glass, which gives a green light.

In 1844 the Railway Department of the Board of Trade was established, and this organisation subsequently had to approve all proposed signalling systems. This brought a degree of standardisation but by this time there were a great many signals in place, and with every change in signalling practice it took time to replace or convert the older equipment.

The early signals were all locally operated by the 'bobby' who was usually provided with just a small hut or shelter, resembling a military 'sentry box', about the size of the modern 'porta-loo' used on building sites. By the 1850's signal boxes had appeared, operating remote signals using wires and pulleys, but it was some years before the signal box system was widely implemented. Signals could be mechanically operated out to a maximum of about half a mile or so from the box using only the strength of the signalman's arms.

The three position slotted post signal was not a great success, the 45 degree 'warning' position was difficult to operate reliably and most larger railway companies abandoned the 'caution' position, changing to a two-position system of horizontal for danger and vertical for all clear. The slotted post design remained in service on some minor lines for many years, notably on the Isle of Man railway system. Such a signal would not be out of place on a 1930's light railway, although these would probably be two-position signals, horizontal for danger, inside the post for all-clear.

The rotating multi-lens lamps on signal posts were prone to failures and in 1854 Stephens & Sons signal company introduced a two-position signal with the arm mounted on the side of the post. The arm was set to horizontal for 'danger' and angled down at about forty five degrees for 'all clear', this design meant that if the signal arm fell off the driver would not see it as indicating 'all clear'. The lenses for the lamp were mounted in a kidney shaped frame called a spectacle plate attached to the end of the signal arm, eliminating the need for a rotating lamp.

This idea was soon adopted by most companies but some preferred to mount the lamp and its associated lenses separate from the signal arm and lower down the post where they were closer to the drivers line of sight and more easily serviced. The Furness Railway was another company which stayed with the slotted post signals, although these used the horizontal position for danger and a downward angle of 45 degrees for all clear. Unusually the posts themselves were of metal lattice construction. As a result of the design of the mounting it was impractical to have the lamp and spectacle plate mounted on the end of the arm and these were fitted separately lower down the post. A few of these signals with low-mounted lights and lenses remained in service into the 1930's and rare examples (notable the Furness Railway types) were still in use into the early years of BR.

If the signal could be seen from the signal box a small lens was fitted on the rear face of the lamp and a moving metal plate was added which covered the light when the signal was set to all-clear. This allowed the signalman to check the signal lamp was lit and the arm was correctly positioned at night. The sketch below shows this plate.



Fig___ Early two-position signal
(The drawing is based on a Cambrian Railways original)

Early two-position signal based on a Cambrian Railways original




In some early designs of two-position signal the arm was not weighted and if the signal cable failed the arm could fall to the all clear position. This was evidently a design flaw and a common solution was to make the kidney shaped 'spectacle plate' heavy enough to do the job. The major railway companies adopted this safety feature but is was several years before the Railway Department of the Board of Trade made it mandatory, and that was only after a serious accident.

Most of the early semaphore signals were two-position types with horizontal for danger and angled down for all clear, these are referred to as 'lower quadrant' systems. Some early signals of the side mounted type used three positions as before; horizontal for danger, almost vertical for all-clear and forty five degrees for caution. These new side mounted signal arms thus required a three-lens spectacle plate for the three positions, three position semaphore signals were very rare in Britain but were sometimes seen in America.

In 1876 there was an accident on the Great Northern Railway caused by driven snow jamming a slotted post signal with the signal arm inside the post. Following the accident the company developed a new type of signal, called a 'centre pivoted arm' or 'somersault' signal. These first appeared in late 1876 and were all of the two-position type. Somersault signals gave a very clear indication and the somersault design was used by several companies, notably in South Wales on the Taff Vale, Rhymney Railway, Barry Railway and the Brecon & Merthyr Railway and on the Midland & Great Northern joint railway in Norfolk.

Fig___ Somersault Signals

Sketch showing operation of Somersault Signals
On the early railways the driver of the train pulled up as soon as he could after passing a signal set at danger. In 1846 the adoption of remote controlled signals allowed a change in the rules and the train had to stop at or immediately before reaching the danger signal. This brought in a need for a secondary signal, usually called a 'distant', to pre-warn the drivers that the next signal might be at danger. The 'stop' type signals were thereafter referred to as 'home' signals. Distant signals are usually placed about half a mile down the track before the home signal.

The home signal and indeed any 'stop' signals on a running line would if possible be placed where they would stop a train 'behind' (that is before it passed) the signal box, so the signalman could give hand signals to the crew in the event of a signal failure.

Signal arms were painted red on the side facing the approaching engine and white on the reverse side. The red side had a white strip or dot, the white side had a black stripe or dot. Distant signals were originally not distinguished from home signals by shape or colour, the driver had to remember whether the signal was a stop signal or a warning signal. In the early 1870's the London Brighton & South Coast Railway introduced 'distant' signals with a V shaped end to the signal arm, but they were still painted red with a white bar on them. In 1917 the Board of Trade decided that all distant signals should have the V shaped notch in the end.

The lights on distant signals were still showing red for danger and green for all-clear, the yellow light (replacing the green) was introduced on the Great Central Railway two-position distant signals in 1917 and was soon adopted by other companies. The Great Northern Railway adopted this yellow caution light and they then started painting their distant signal arms yellow with a black stripe. By the mid 1920's most companies distant signals had yellow fronts on the arm and red and yellow lights but it was the mid 1930's before all the older signals were converted.

There were still variations in signal arm colours, for example the Midland Railway, South East and Chatham Railway and the Cambrian Railways all used a circular spot in place of the vertical stripe on their 'home' signals. Some companies, including the Midland Railway, used a horizontal line along the length of the signal arm in place of the more common chevron design on their 'distant' signal arms.

In 1876 the Board of Trade Railway Department decided that signal arms for different tracks should be mounted on separate posts. To save on space the railway signalling engineers developed the 'bracket signal' which had a single supporting mast topped by a platform on which there would be two or more short signal posts.

This was building on the work already done on signal gantries, which are bridge like structures supporting a row of signal posts above the tracks. Gantries were expensive compared to bracket signals and were usually only seen at major railway stations where many tracks ran parallel.

Early gantries were built simply, if sturdily, of wood but by the 1870's metal casting had improved and iron structures were being produced. Early bracket signals used these new cast iron supports for the platforms, usually based on cast circles and arches all bolted together. These castings were heavy and added to the wind resistance, so by the 1920's all companies had switched to lighter structures based on light metal strips called a 'truss' design.

Bracket signal masts were adapted to position a single arm in clear view where normal sighting might be difficult. As an example if there was a large water tank beside the track where the signal had to go the bracket could be jutted out over the track placing the signal arm in clear view.

On multi-arm bracket signals and gantries the convention was that the tallest post referred to the most important route, so for example if approaching a junction where a branch line left the main line the signal post for the main line would be slightly taller than the post for the branch.

This convention became a problem where under-slung signals were used. These were signal arms suspended from the under-side of a bracket or perhaps fitted under the awning of a platform canopy. They were used where some obstruction, such as a bridge, or the platform canopy, would make sighting a normal height signal difficult.


Fig___ Bracket signal and signal gantry

Sketches of Bracket signal and signal gantry




The drawing on the left shows a bracket signal controlling the approach to a station with loop lines for the platforms and a bay platform as well. The numbers next to the signal arms indicate which of the tracks they apply to on the small track-plan above. The support for the brackets is of the older cast iron type.

The sketch on the right is a gantry, this uses the later 'truss' type supporting framework made up of oblong frames with a diagonal 'truss'. The gantry makes sense in this case as now there are two double track lines, the centre pair of lines are for fast trains, the two outer lines handle the slower local trains and goods traffic. This arrangement was particularly useful on outer suburban lines where frequent stopping passenger services ran on the outer slow lines leaving the centre fast tracks for express trains. Note that an alternative arrangement for the fast and slow lines was to have them arranged as two double-tracks side by side. In the drawing shown above the lines on the left would be the 'fast' up and down lines, those on the right the 'slow' up and down lines. This arrangement would require an island platform in the station to handle trains on the centre two lines.

The signals facing you all apply to the 'Up' lines, the circles on the signal arms for the 'Up Slow' line will be explained later. With the ability for trains to cross from the up-fast to the up slow and vise versa each set of signals has posts for both the fast and slow lines, as these are 'through' lines the posts have a second 'distant' arm on them as well.

Ratio offer an excellent kit of a truss type gantry however they do not supply any posts or arms with the kit so you need to add those either from Ratio kits or from home-made parts.

With very few exceptions where you see two arms on a single support they both apply to the same track and the lower one will be a distant signal, giving advanced warning of the next signal on the line. The exceptions are occasional signals with one arm facing each way on a single post (but these were rare) and signals on goods loops. The signal gantry shown above has two arms for the Down lines on the reverse side of the posts (signal arms numbered 5 and 6 on the drawing).

Goods lines were not covered by the bracket signal rules and some rare examples of multi-armed single post signals survived into the 1960's although these double home signals only appeared near points, the upper arm always referred to the track diverging to the left.

To save on the expense of multiple signals a 'route indicator' might be used, these had a separate box which displayed a letter or number to indicate the line the signal arm was currently controlling. For example a single armed signal on the approach to a station might have an indicator to show whether the signal was set for the bay platform or the through line. They first appeared at about the time of the First World War and the principle remains in use today.

Mechanical route indicators would normally be on a single post signal and there were various designs in service. The GWR and some other companies employed a shallow metal box about eighteen inches square and perhaps six inches deep, bolted to the post about three quarters of the way up. This box was normally painted black and contained cut-out metal letters and/or figures. The LMS had a design in which the plates were laid off to one side at an angle and were shifted into the vertical position to be displayed. Personally I would suggest the plain box mounted square on the post, which seems to have been in widespread use, is the easier to model.

These mechanical route indicators might be seen on the 'starter' signals on the platform ends at a terminus, or in goods yards where the driver might be facing a range of complex point-work. They did occasionally appear on running line signals at the approach to a station, but they were associated with lines where speeds were slow so this would normally be a terminus rather than a through station.

The 'works' underneath the indicator box were sometimes open to view, a series of small levers for each letter or number with rods passing up into the box itself. On passenger platforms and on running lines it was normal to box all this in to avoid accidents to inquisitive passengers, the drawing shows such a 'boxed' arrangement. Note the multiple weighted levers at the base of the post.


Fig___ Typical mechanical 'Route Indicator'

Sketch showing various route indicator units



At the signal box there would be several levers associated with this signal, all of them would operate the signal arm but each lever would cause a different letter or number to rise up in the box and be displayed, indicating to the driver which route the signal applied to.

As an example a small country station might have a junction with a branch line, the 'route indicator' on the single signal post might then be used to indicate whether the points were set for the main line or the branch. All of this was protected by mechanical or electrical 'interlocking', so the signal lever could only be pulled after the appropriate points had been set. Mechanical interlocking of points and signals was invented in a crude form by a French engineer but the fully mechanically interlocked signal frame was patented by John Saxby in 1856 and the first installation was at Bricklayers Arms Junction. Interlocking of signals and points meant that signals can only be set when the points are in the right position, in 1889 the Regulation of the Railways Act made interlocking mandatory.

Modern electric colour light signals also use route indicators to indicate the point settings. These are covered later in this section.

By the turn of the century only four companies made their own signals; the Great Western, the Midland, the London & North Western and the Lancashire & Yorkshire. Everyone else used one of the large independent companies; Saxby & Farmer along side the LNWR line at Kilburn, Stevens & Sons in Southwark who supplied most of the Southern lines and in the North the Railway Signal Company based at Fazakerly, close by the L&Y main line.

The LNWR had used Saxby & Farmer until a row between their two managing directors lead to the building of a signal works at Crewe, and the L&Y had originally used the Railway Signal Company products and continued to build very similar equipment when they opened their own signal works.

The supply of signalling equipment on the Scottish lines was split between just two companies; the Highland Railway used McKenzie & Holland, everyone else used the Glasgow factory of Stephens & Sons. Saxby & Farmer and Stephens & Sons were the two really big names in signalling and their products were used by most of the smaller lines.

In 1911 a group of eight railway company signalling experts met to discuss the various problems and in 1913 the Institute of Railway Signal Engineers was founded. In the 1920's the Institute recommended that the warning position on the three-position signals be abandoned due to its unreliability. Most companies had abolished the caution position in about the 1870's but there were still quite a few old signals dotted round the country on minor lines and these were changed to two position operation.

In 1910 the two-position 'upper quadrant' signal appeared, which had the advantage of lighter construction. A horizontal arm still meant danger but the all clear was represented by the arm pointing upwards at 45 degrees. This design was adopted in the 1920's by the post-grouping London Midland & Scottish, the Southern Railway and the London North Eastern. The LNER began replacing the Great Northern somersault signals with new upper quadrant types in about 1923, although some of the latter remained in service for many years.


Fig___ Typical upper and lower quadrant signals

Sketch showing Typical upper and lower quadrant signals

The GWR retained its own in-house lower quadrant design, they tried a three position upper-quadrant signal to an American design at Paddington station but this was the only example of its type.

The railway department of the Department of Transport decided that all the railways should switch to the upper quadrant design sometime in the 1930's and most companies complied as signals fell due for replacement or major repairs. Some companies, such as the Great Western Railway and the Cheshire Lines Committee, continued to use the two-position lower quadrant signals. In the mid 1980's British Railways decided to replace the remaining lower quadrant signals (mainly confined to the old GWR lines) with colour light signals as they fell due for servicing. However there were still lower quadrant signals in use in the first half of the 1990's.

Originally signal masts or posts were of round section wood, and were often quite tall. By the middle of the 19th century square section wooden posts were common, although in Scotland all the companies used metal lattice masts. Company-built signals were often quite distinctive, most posts were of the square section wooden variety but the LNWR favoured heavy section square posts and the GWR used a slender tapered square post. The wooden posts required some form of cap called a 'finial' on the top to protect the end grain from absorbing water, and these were also often of distinctive design. The Midland Railway had a large ornamental spike, several lines including the GWR used variations on the ball and spike design, whilst the LSWR had a distinctive tall spike on theirs.

Considering the equipment from specialist signalling suppliers McKenzie & Holland had a tall spike about three and a half foot tall whereas Stephens & Sons used a more or less standard eighteen inch high ball and spike. In the South of the country signal posts were sometimes built from rail and carried a small square finial on top. These rail-built signals were common in the South by the 1920's.

Also in the 1920's tubular steel posts and square section concrete posts appeared, the latter being common on the Great Northern Railway's 'somersault' type signals. The concrete posts were solid for about the lower third but the upper two thirds were pierced through with oval shaped holes to save weight (see Fig___ below). Concrete signal posts were not the usual bright white, they may have started out white but concrete is absorbent and exposure to the soot and grime of the railway soon toned this down to a grubby grey. Although the Southern Railway made great use of pre-cast concrete, mainly from the Exmouth Junction Concrete Works, they did not favour concrete signal posts preferring to use posts made from lengths of rail.

The Ratio OO range includes GWR, LNER, LMS types and SR (rail built) types, their rather more limited N Gauge range models appear to be of a GWR type square section wooden mast. P. D Marsh offer lattice post signals in N gauge, although these whitemetal mouldings are heavy they do capture the general appearance rather well. The SR rail built signals can be knocked up using Plastruct I section reference BFS-2 as shown in the drawing. Modelling the concrete post type would take a bit of patience. First take a piece of Plastruct I section (BFS-2) and mark up the centre web where the slots will be. The slots are slightly shorter than the width of the signal arm and there are about twenty three of them. Now use a pin or needle held in a pin-vise to scrape through the centre web to make the required holes. Mark the sides of the strip to show where the holes are (pencil will do for that or just make a small scratch with the pin). Wipe a little solvent into the web to attack the surface and form a 'key' and fill the slot on each side with fine grade Milliput. Whilst the putty is still wet poke through the holes with the pin (do not worry of you get lumps building up around the holes) and leave the thing to dry. When the putty has set solid (I would allow a day for this) sand the thing smooth and paint with a 'dirty grey' mix. There were a few cases where the somersault signal arms were replaced with standard LNER upper quadrant signals but the old concrete post was retained. Making just one such post helps identify the location and the remainder of the signals can be of a standard type made up from kits.

Fig___ Various companies designs for signals
Sketches showing signal designs for various companies





Where the positioning of the signal was awkward, such as when close to an over bridge or where background clutter made sighting difficult, some companies employed very tall signal posts, up to sixty feet high, which could be seen against the clear sky and over any local obstruction. These so called 'sky light signals' were more common in the North West than elsewhere and some survived into the 1980's. The LNWR had some particularly fine examples, often with a duplicate or repeater arm mounted low down on the post where it could be more easily seen as the train drew close. In the mid 1980's new Health & Safety legislation was enacted which made these very tall signals illegal, some were replaced with standard British Railways nine foot tubular steel signal posts, but in many cases the opportunity was taken to install the new colour light systems.

The station at Hale in Cheshire had one of these tall signals on a lattice mast with a low-mounted repeater arm close by a road over-bridge. The sketch was made from a rather poor photograph I took in about 1983. This signal was replaced by a standard British Railways upper quadrant semaphore on a nine foot tubular post in the mid 1980's, and then the whole section was converted to colour light signalling in the early 1990's.

When one of these tall signals was used because of background clutter, rather than for sighting over a bridge, the associated lamp was sometimes not mounted at the top but was fitted lower down, where it could be more easily seen at night and more easily tended. The spectacle plate was obviously not attached to the end of the signal arm in this case, instead it was a separate item mounted on a small pivot in front of the lamp.

Where there was a problem with a cluttered background, but a tall signal was either impractical or contrary to company policy, another option was to mount a section of white painted board behind the signal arm so it could be more easily seen or, if the background were owned by the railway (such as a bridge over the line) a white square or quadrant could be painted on it to make the signal more visible.

The third option was to mount a remote 'banner repeater' signal in an easily seen position which mimicked the main signal. The standard design for a banner repeater was a circular disk of translucent material with the signal arm, blunt ended for a home signal, fish-tailed for a distant, painted on in black. Mounted behind the disk was a lantern for night visibility and the disk was electrically or mechanically linked to the main signal so that it operated in concert with the signal arm. The lanterns used were physically larger than those on normal signals, the GWR favoured tall cylindrical lamps, the electric unit on a signal still in use at Stockport has large square lamps with a big circular reflector on the front. The GWR posts were painted a dark colour, the signal at Stockport is all- white. It is perhaps worth noting that the banner repeater operated in the same quadrant as the normal signals on the line, hence GWR banner repeaters were lower quadrant whereas the LMS, LNER and SR favoured upper quadrant.
You may wonder why the signal itself was not positioned where it could be seen, one reason might be that the signal proper, and any train stopped at it, would not be visible to the signal box staff. The banner simply gave an advance warning in case a signal was at danger, the train did not stop at the banner but only when it reached the proper signal. There is a banner repeater signal on the line from Altrincham to Stockport where the line joins the main Manchester to London tracks. The line curves tightly and passes under a road bridge at this point, out of sight of the signal box.


Fig___ Sky signals, painted board signals and banner repeaters

Sky signals, painted board signals and banner repeaters



The sketches in Fig___ are all taken from photographs, the forty foot tall lattice mast signal was in place at Hale near Altrincham in Cheshire into the 1980's. The wooden post sky signal with the separately mounted spectacle plate is based on a GWR broad-gauge signal photographed just before the change to standard gauge.

The upper quadrant signal with the metal sheet was photographed in 1986, on this signal the ladder went right to the top of the post and the rivets holding the plate to the supporting frame had rusted, showing black against the white sheet. The diamond shape on the post indicates that the line is protected by electrical 'track circuits' (these are discussed in detail later).

The junction signal was photographed in the 1940's on the GWR system, the sheets appear to be wood but it is difficult to tell from the photo. The single banner signal is sketched from a photograph in a book on GWR signalling by Adrian Vaughn (for details of which see the bibliography) and the double banner is drawn from a signal still in use on the line from Altrincham to the junction with the main line at Stockport in Cheshire.

Fig___ Photo of a mechanical banner repeater signal still in use in 2001

banner signal repeater outside Stockport
A photograph taken by John Sullivan of the more modern electric light banner signals will be found in the section on Signals - Power Operated and Colour Light Types

As with the earlier disc and bar systems various modifications were introduced to distinguish differing functions for signals, for example a general practice was to fit a large X shape to the arms of signals which were temporarily to be ignored. This practice continues today, colour light signals are covered with a black plastic bag with a white X of tape over the front.

Shunting signals were required to control movement within stations and yards, these were always of the 'home' or 'stop here' type and so that drivers could identify them they were distinctive in some way. Some companies used smaller than normal arms, some had very short and fat arms, others had circle welded on to the arm.

This circle design was generally adopted at about the time of the First World War to indicate signals that applied to a goods siding or goods loop. Again there were minor variations: the LSWR had a white circle added to the standard red and white arm, the LNWR used a black circle on a standard red arm with a white strips where the circle crossed the arm, the SECR had a black circle on their standard red arm with a white dot on the arm itself and the GER and the GWR used a white circle on a plain red arm.

The GWR had used a large letter S on the arm before this, the S arm was subsequently used mounted separately lower down the post on a standard 'home' signal as a 'shunt ahead' signal. If down (clear) it meant 'you can pass the above 'home' signal at danger but only for shunting purposes'. Several companies appear to have adopted variations on this S marked arm for this purpose.

You will occasionally see very small signal arms, sometimes with holes in them, these are usually 'calling on' arms. They were used to allow a train to enter a section already occupied by another train, the driver being responsible for avoiding a collision. These were not common, one application was at long platforms where two passenger trains might be accommodated at one time.


Fig___ Distinctive Signal Arms

Sketch showing a selection of Distinctive Signal Arms



Distinctive signal arms, with the stubby arms, circles, etc. were abolished in 1948, but those in place remained in use for many years (as late as the mid 1980's in some locations).




Detonators

In 1845 the 'detonator' appeared, a small explosive charge that could be clipped to the rail, the weight of the locomotive wheel caused the charge to explode and the loud but harmless bang could be heard by the crew. Even today every engine cab carries 10 detonators. They are used for emergency protection, in the event of an accident the train crew place the detonators on the line to warn approaching trains. They are also used for assistance protection, where a train breaks down they are placed on the line to warn other trains of the obstruction, including the loco coming to assist. Some signals have a mechanical arm which lowers a detonator onto the track when the signal is at danger, lifting it clear when the signal is in the all-clear position. This warns the driver if he or she passes the signal at danger. They are also used in possessions (when the engineers take on a maintenance job on a section of track) to warn train crews that there are men working on the track. Another occasional use for detonators is where a severe temporary speed reduction (TSR) is in effect. See Signals - Ground and Fixed Signals for details of the signs used for TSR's.






Signalling in Foggy Weather

Fog has always been something of a problem for visual signalling systems and has caused several accidents although the clean air acts of the 1950's did reduce the frequency and density of fog.

In foggy weather a plate-layer would be sent out to wait by the signal, placing a detonator on the track when the signal moved to danger, removing it when the signal changed to all-clear. The small fog mans hut with a braisier to keep the man warm close by a signal post was not uncommon. Fleetline offer a small hut for the fog man and a braiser to keep him warm (N74), these were a common sight adjacent to important signals.

Various mechanical aids for placing detonators soon appeared, most used some form of swinging arm with the detonator on the end, they were never terribly successful but some remain in use today.




^
Go to top of page


International Good Guys ~ Doing things properly since 1971 ~ Site maintained by Disabled Access to Computing
All material Copyright © Mike Smith 2003 unless otherwise credited