The LMSR "New Lines" signalling

Last modified: 2020-10-09

History

The LMSR "New Lines" signalling system was introduced in 1932-3. Despite the name, it was actually found on two separate suburban lines in the London area. Both were owned by the LMSR, both were used by London Underground trains, and both saw an intensive service as a result.

The first such line runs from Camden to Watford Junction. In 1912-3 the LNWR (one of the LMSR's precursor) opened a third pair of tracks (hence the "New Lines") over this section to keep suburban services separate from the busy long-distance lines. For most of the length they paralleled the main line, but between Bushey and Watford Junction they swung to the west to serve Watford High Street. Between 1915 and 1917 the lines were electrified on the four rail system and the London Underground Bakerloo Line was extended on to it.

The other line runs from Bromley-by-Bow to Upminster. In 1858 the LTSR (another LMSR precursor) opened a line from Bow to Barking to meet its existing route to Tilbury and Southend, and in 1885 it built a cut-off between Barking and Pitsea via Upminster. In 1902 the Metropolitan District Railway, to become the District Line, was connected to it just west of Bromley-by-Bow station, with trains running as far as Upminster. In 1905 the line was electrified on the four-rail system to East Ham and the District service cut back to there, returning to Barking in 1908. In 1932 the line was quadrupled, electrification was extended to Upminster, and the new signalling installed on the new tracks.

In 1960 London Underground took over the electrified tracks from Bromley to Upminster and replaced the LMSR signalling with their own standard system. The signals on the New Lines were replaced by the British Rail standard 3-aspect system in 1988 as the old system had worn out.

Principles of operation

The basic characteristics of the signalling system were:

• the line was track-circuited throughout, so signals could be made automatic on plain line;
• all stop signals had train stops;
• trains were mostly electrical multiple units with good brakes, so braking distances could be short (other trains were severely restricted in speed);
• the intensive service meant that it was necessary to handle failures (particularly of track circuits) efficiently.

On plain track there were two kinds of signals: stop signals and repeaters, with a repeater at braking distance before each stop signal. A stop signal showed green if the line was clear to the end of the overlap of the next signal (a section of track, typically 183 metres long, following the signal), and red otherwise. In the latter case, a second red light was shown below the first to indicate that it was a stop signal. When the stop signal was green the corresponding repeater was green, while when it was red the repeater was yellow. However, if there was a train between the repeater and the end of the overlap of its corresponding stop signal, the repeater showed red. To make it clear that it was a repeater (although trains should not normally be coming across these signals, because the previous stop signal would still be red) this also had a second red light below the main signal, but this time offset to the left. The second red light, in both cases, was called the "marker light".

As mentioned above, there was a need to recover from failures, which the design assumed were most likely to be track circuit failures. Therefore, when a train came to a stop signal showing red, after one minute - and if the overlap was clear - the train stop would lower and the marker light would change to a small yellow. The train could then proceed cautiously, with the driver looking out for obstructions or, of course, another train ahead of him. In this way the train could reach a repeater showing red. If so, the driver had to wait another minute, then sound the whistle and continue towards the stop signal (there were no train stops at repeaters).

The standard signal was a "searchlight" (a single lamp at the top changed colour by moving different lenses in front of it), so the signal aspects in normal operation were:

Stop signals			Repeaters
Clear	Stop	Calling-on	Clear, next stop signal is clear	Caution, next stop signal is red	Stop for one minute

Variations

As might be expected, the simple theoretical design didn't always work in practice. Here are some of the variations that might be found at various times and places.

• Some signals, particularly those installed at Camden and Watford Junction when the old signal boxes were replaced by power boxes in the 1960s, were multiple aspect rather than searchlights. That is, the signal had separate red and green lamps above the marker lamp. Unlike normal signals, the red was usually above the green.
• If signals were closely spaced, a repeater for one stop signal would need to be at or near the previous stop signal. In this case, the latter signal did the duty of the repeater as well; when the line was clear to the next stop signal (and the end of its overlap), it showed yellow or green as appropriate. In extreme cases a double yellow aspect would be used to provide sufficient warning. Again, at some locations these signals were multiple-lamp rather than searchlights.
• At Watford Junction, and in the case of the bay at Willesden Junction, the small yellow light on the last stop signal was used to indicate that the train was being signalled into an occupied platform. While the aspect had the same meaning, trains did not have to wait for a minute for the signal to change. If the platform was clear to the buffers, the last stop signal showed yellow (it could never show green).
• There was no marker light on the last stop signal before a change to conventional signalling. Of course, since this signal would not display a marker when the train was allowed to pass it, this was less than ideal in indicating this fact to the driver. The main point was that the marker could not change to a small yellow light and the signal must not be passed at danger.

Junctions

The "New Line" signalling predates the introduction of the junction indicator (a line of white lights showing that a train is taking a diverging route), and so signals at facing junctions consisted of two or more heads, side-by-side. If the routes were of the same importance, the signals would be level, while if one was more important than the other (usually meaning that it had a higher speed limit) the corresponding head would be raised.

Stop.	Stop.	Stop.	Clear on right-hand route.	Calling-on on left-hand route.
Both routes equally important.	Left-hand route more important.	Right-hand route more important.	Both routes equally important.	Right-hand route more important.

In some cases one route at a junction would lead to a line using conventional semaphore signalling or to a line where allowing trains to pass the signal at danger was not appropriate. In these cases the corresponding marker light would be absent, or would not be accompanied by a small yellow lamp.

Stop.	Clear on left hand route.
Left-hand route more important.	Right-hand route more important.
Left-hand route has no calling-on facility.	Left-hand route is not this signalling system.

In semaphore signalling, if a driver needed to know the route set at a junction before reaching the junction signal, then there would be two distant signals, one for each route. Other colour light signalling schemes of the time, such as those on the LNER, replicated this by giving the previous signal two heads, one showing yellow and one green if the junction was clear. This approach was known as the splitting distant, and it has recently come back into favour.

Stop.	Caution, prepare to stop at junction signal.	Clear, junction clear for left-hand route.	Clear, junction clear for right-hand route.

However, both the LMSR schemes eschewed this approach in favour of directing lights. Two yellow lamps were added to the signal before the junction (which was always a stop signal), one on each side and lower than the main aspect. If this signal or the junction signal was at danger, the directing lights remained off. However, if the left head of the junction signal was green, the right directing light was lit, while if the right head was clear the left directing light was lit, in each case giving the impression (particularly at night) of a splitting distant:

Stop.	Calling-on.	Caution, prepare to stop at junction signal.	Clear, junction clear for left-hand route.	Clear, junction clear for right-hand route.

If the signal spacing beyond the junction was particularly tight, the signal before the junction might need to show double yellow. In this case the result would be a triangle of yellow lights:

Preliminary caution, prepare to pass junction signal at restricted speed for left-hand route.	Preliminary caution, prepare to pass junction signal at restricted speed for right-hand route.

There were only three places with directing lights: on the Up line at Watford High Street, on the Down line at Colne Junction (just north of Bushey), and on the Up line in Primrose Hill tunnel. The first two (signals HS4 and BE3) could show three yellows but the last one (signal CD2.16, later EN148) couldn't. (HS4 was later renumbered one of WJ128, WJ129, WJ131, or WJ137, but I have not been able to find out which.)

Other features

Signals that worked automatically carried a white board with a black A on it. This told the driver that he could pass the signal after three minutes even if it didn't change to the calling-on aspect. Of course, the train stop would still "trip" the train and need to be reset. If the signal was controlled from a signal box but could be left to work automatically when the box was closed, the A was an illuminated indicator.

Automatic signal	Semi-automatic signal under control of the signal box	Semi-automatic signal working automatically

Signals with a telephone nearby carried a similar board with a T on it.

Signals immediately before a gap in the power supply (a "section gap") carried a sign that illuminated with the words TRACK DEAD when the power is off. If the signal is a repeater, it was fitted with two marker lamps (this was not part of the original system, but was added later). In normal circumstances the one on the left was lit when the signal was red. But if the power was off beyond the gap then the main red and the marker directly below it would be lit, indicating that the driver must not pass this signal.

Power on		Power off

Signal numbering and signal boxes

Signals were numbered with a prefix and then a number. With two exceptions the prefixes were two letters. A prefix was used for the signals leading up to a station (and occasionally just beyond); signals on the Down line used odd numbers counting down to 1 and those on the Up line used even numbers counting down to 2. So approaching South Hampstead on the Down would be SH7, SH5, SH3, and then SH1 as the station starting signal, while approaching Harlesden on the Up would be HD6, HD4, and then HD2 as the starting signal, followed by WL6. There were exceptions: for example, QP3 was the Queen's Park Down starting signal with QP1 beyond the junction with the Bakerloo, protecting a crossover. At both Watford and Camden the signals were just numbered by the lever operating them (though Watford Junction also had automatic signals WF7 and WF5 on the Down line) and later the signal number for the controlling power signal box. Repeater signals had an R in front of the number (e.g. RWF7).

Notes:

• Camden No.2 only had two New Line signals: CD2.16 and CD2.18/25 on the Up line. The signal box was replaced by Euston PSB in 1965 and the signals were renumbered as EN148 and EN147.
• On 1977-01-30 Willesden New Station took control of the Stonebridge Park and Queen's Park areas, the junctions being operated by remote control. Queen's Park was not renumbered but I have not been able to determine whether the signals at Stonebridge Park always carried SP prefixes or whether they previously had PH prefixes and were renumbered. SP6 is positioned to have previously been a PH signal; SP1 is north of the junction to the new depot and would probably have also been a PH signal.
• QP7 was on the Bakerloo northbound track; QP10 was a down direction signal at the north end of the Up platform.
• KG4 was within Kensal Green tunnel.
• WL12 was the bay platform starting signal; WL41 and WL42 were the signals on the Down line from Kensal Green Junction.
• SP10, SP12, SP14, and SP16 were the signals on the depot exit.
• Signals controlled by Colne Junction were not numbered by the lever that worked them. Instead:

  Line	Up			Down		Up branch		Down branch
  Signal	CJ4	CJ2	BE4	BE1	CJ1	CJB4	CJB2	CJ15
  Lever	1	2	7	17	14 (to Watford) 16 (to Croxley)	4	5	15

• When Watford No.4 and Watford High Street were replaced by Watford Junction PSB, the signals were renumbered with WJ prefixes.
• For some reason Watford No.4 had levers numbered 26 to 80 rather than starting at 1. 27 to 32 were the platform starting signals (27 and 28 were alternative routes from platform 5); the only other Up signal was WF33. In the Down direction were WF7 and WF5 (both automatic), WF80, then a junction signal with two heads, then two separate inner homes. The numbering was complicated: the left head to the left inner home was 61, the left head to the right inner home was 67, and the right head to the right inner home was 79. The left inner home was 59 to platform 1 or 60 to platform 2. The right inner home was 62 to platform 1 through to 66 to platform 5.

Rules

The original rulebook for the "New Lines" is available on a separate page. These rules were aimed at operating staff and provide a different view of the system. When reading the rules, it is worthwhile to bear the following points in mind.

• Instruction 80(a): the "certain point beyond the next signal in advance" is the end of the overlap.
• Instruction 80(a) refers to a special aspect "yellow over red", though the last two words are struck through. I have been unable to find any further information on this. The situation sounds like Regulation 5 working, where the driver is warned before the signal is cleared so that he knows that there is no protective overlap beyond the next signal. The implication is that initially the marker light remained lit when the signal changed to yellow, but at some later date this was replaced by a simple delayed yellow (which is modern practice for this situation).
• Instruction 80(b) describes the directing signals found before a junction, but doesn't describe junction working itself.
• No examples are known of the double red version of the signal in Instruction 80(c).
• Instruction 83 means that light engines will go round the triangle in a clockwise direction.
• Instruction 88: at the time, tail lamps only needed to be lit at night or in fog or falling snow.
• The 3 minute timing in Rule 136 would seem to match the 3 minutes in Instruction 81(c), suggesting that any such move will foul the overlap of the previous signal and prevent the calling-on aspect from appearing.

Back to the LMSR index. Back to the signalling index. Back to the Rail index. Back to Clive's home page.