Twyford GWR station in 1852. Note the two tall disk-and-crossbar signals, the shorter version, and the fantail Caution board at the approach to the platform. The tall signal on the left is indicating 'proceed', while the other - in the centre - is indicating stop. Presumably, it was permissable for the approaching train to pass the stop indication in order to reach the platform. Note also that the platforms have only minimal end ramps.

Great Western Railway broad gauge disk-and-crossbar signals (c.1840 - c.1900)
A range of the characteristic disk-and-crossbar signals developed by I K Brunel for his innovative 7ft broad gauge Great Western Railway.

Time interval and block working
Until the 1870s/1880s the time-interval system for signalling trains was used by the majority of railways in Britain and Ireland. Following a series of accidents, some fatal, and many near-misses, parliamentary legislation was introduced which, eventually, led to the universal adoption of the block system, which is implemented in Trainz Classics 3 and above. In order to recreate the earlier period within Trainz compromise is necessary. Essentially this means using signals designed for time-interval working within the block system.

From the GWR's opening in 1837 until about 1848 company policemen patrolled the line ensuring there was no trespass, that everything in their section was in order, and to give hand signals to trains (lamps at night). From soon after opening, rudimentary signals mounted on posts were introduced and were operated by the policemen if they were within their secrion of track. Policemen usually had a small hut or bothy to occupy in bad weather and where they often had a brazier to keep warm. From about 1848 the policemen's job was split off from the policemen's responsibilities, becoming the sole preserve of signalmen (who were still referred to by the staff as 'bobbies'). Gradually, these hand-operated signals proliferated and eventually various forms of remote-working were introduced so that they could be grouped together and operated from a single place by signalmen pulling levers actuating wires. These early lever frames were quickly enclosed within bothies, the early forerunners of signal boxes. Rapid development followed, partly spurred on by government legislation, and by the late 1870s signal boxes controlled signals and points (switches, turnouts) and were staffed by signalmen. Interlocking frames followed soon after and all these developments, combined with the aboslute block system (lock and block) led to much greater safety on the railways together with the ability to handle a more intensive service.

'Road clear' imperative
However, to return to the early days of the GWR broad gauge, the over-riding expectation in time interval working was that the policemen, as a matter of esteem and pride, could show that their section of line was clear. It was also a matter of prestige. A clear line demonstrated that they were doing their job properly, that the line was secure from trespassers, that it was in full working condition with no damage, and that it was available for use. The ethos was that of the old turnpike roads in that the condition of the road and its availability/suitability for use reflected the work put in by the local commissioners. It was by no means certain, in those early days of the railways, that use of the 'railed-way' or 'railed-road' would be restricted solely to the company's trains. Hence, engineers and railway companies were still finding their way and experimenting to establish the most suitable system for operating the line.

So, under time interval, the over-riding requirement expected of signalmen was to show that their section of track was clear for the passage of trains. Consequently, when a train passed the signalman turned his signal to 'stop' until he believed the train had travelled out of the section for which he was responsible. Consequently, he felt that it was now safe to again show that his section of track was clear, therefore he would turn his signal to show 'proceed'. Usually, five minutes was deemed sufficient for a passed passenger train to clear the route and it was then OK to show 'proceed'. Goods and slower trains were given longer to clear the line, sometimes up to 10 minutes. Hindsight is a wonderful thing and we can immediately see the faults in this system but we should remember that the intensity of traffic was much less than today and that these early railway companies were finding their way, innovating as they went. The Caution board (fantail) was eventually introduced. Its use appears to have been to lengthen the apparent time interval between trains by supplementing the 'stop/go' signals. It was displayed for a longer period, thereby adding a notional time/length between trains when following trains could proceed cautiously with the possibility of encountering the train in front (more info below).

Disk-and-crossbar
For the Great Western Railway, Brunel designed a unique signalling system in which the shape of the board displayed to the oncoming driver indicated whether he could 'proceed' or should 'stop'. Both boards were bolted to a single vertical pivot, which was then rotated through 90° to show either 'proceed' (a disk) or 'stop' (a horizontal bar). Originally, the policeman or signalman responsible simply pulled on a horizontal bar at the base of the signal in order to pivot the vertical rod and thereby show 'proceed' or 'stop'. As signalmen were expected to show that they had everything in order on their section of line and that trains could pass the normal position was 'proceed'. In the block system, of course, signals are kept at danger until such time asa train is permitted to pass, under the strict control of signalmen working as a team along the track.

Unfortunately, time interval working cannot be reproduced in Trainz because it directly contradicts the later and safer block system. Consequently, these models of Brunel's signals act like later block signals even though they are modelled on the time interval signals. Brunel's system was a lot less complex than later signalling so in order to lend variety I have modelled these signals in two different heights (45ft and 30ft) and either with an early Victorian signalman/policeman (made by Bob Saunders) or with a basic crank at the base of the signal as if it was controlled from one of the early primitive signal boxes. Please note that this basic crank is purely conjectural - if anyone can provide details of how these signals were actually controlled from signal boxes then please let me know. I have photos of the signal recreated at Didcot GWS but I felt I wanted to keep it simple.

Up signals (left) and Down signals (right)

Caution
In addition to the basic disk-and-crossbar signal Brunel also provided a Caution board. As far as Trainz is concerned it should be regarded as a Distant signal. However, this was not only how it was actually used. After a train had passed the signalman turned his disk-and-crossbar signal to 'stop' and the Caution board to 'caution'. The disk-and-crossbar was turned to 'proceed' after five minutes but the Caution board remained at 'caution' for another five minutes or so until it too was changed to show 'proceed'. Originally, caution boards were placed close to the disk-and-crossbar signals but after parliamentary intervention in 1852 which led to longer time intervals between trains, the Caution board was moved about 400 yards in front of the disk-and-crossbar signal. Both continued to be operated by hand, with the policeman walking between the signals to set them as required.

Caution board (fantail)

Junctions
Brunel's junction signals are equally problematic for Trainz. Junction and disk-and-crossbar signals held two sets of boards and lamps instead of the usual one set. However, both sets acted like one, their only indication was that there was a junction ahead - they did not indicate which way the junction was set. Each driver, and the policeman were expected to know which way the train was intended to go. Therefore, the junction signals, in Trainz, act just as if a single signal was controlling the junction - if a train approaches it will show 'proceed' no matter to which track the points (switch, turnout) are set. Point indicators showed which way important points were set but these were located right next to the point. They will be incorporated in my animated broad-gauge pointwork.

Up Junction signals (left) and Down Junction signals

Up and Down
Yet another complication and a further difference to later semaphore practice, disk-and-crossbar signals had differently-designed bars according to whether they were controlling a Down line (from Paddington) or an Up line (to Paddington); both designs have been modelled. Finally, Brunel also equipped his new line with special signals which were set either side of level crossings, these being operated by the crossing keeper. Again, the attitude of the time meant that the keeper always sought to show that his crossing and its signals were ready for trains to pass, showing 'proceed' most of the time, except when he passed road traffic across the line and for five minutes after a train had passed. In Trainz these act like ordinary block-working signals and, if positioned correctly near a crossing, will indicate 'proceed' when the gates open, assuming that the line beyond is clear.

Level Crossing signals

Placement
Discussion within the Broad Gauge Society suggests that all of these signals were used sparingly. Disk-and-crossbar signals appear to have been placed beyond the point where we would normally expect Home Starter signals at the ends of platforms. Usually, they seem to have been located some distance beyond stations - although some illustrations show them close in - and also at tunnel entrances and other significant features along the line. Conversely, pictures in Meason's Illustrated Guide to the Great Western Railway of 1852, and also the excellent engravings by Bourne show stations with Caution boards at the ends of platforms and both Up and Down disk-and-crossbar signals positioned just a few yards beyond them, and occasionally in the centres of stations. Evidently, there were no overall rules and signalling was very much a work in progress.

This illustration of Maidenhead station in 1852 also shows Caution boards and disk-and-crossbar signals placed together at the approach the platforms rather than at their departure end - although other illustrations do show Caution boards at the far ends as well. Here there is a short ramp ending the platform on the left.

To create the authentic 1840s/50s broad gauge feel don't use too many. Caution boards can be placed about 400 yards in front of disk-and-crossbar signals (although they weren't present in every case) and also at the ends of platforms where in some examples at least they appear to have acted rather like a Distant signal combined with a Starter. In other words, even if the Caution board was showing 'caution' (pointing towards the track and showing its red face) the train could proceed past it although the previous train had left no less than five minutes before. They seem to have been hand operated until replaced by early semaphore Distant and Home signals as the GWR gradually changed over to block working. Level crossing signals should be placed either side of crossings, and junction signals at junctions of course - there were no splitting distants or, indeed, any indication which way the junction was set!

Thanks to Chrisaw for all his help and patience in getting these working for me, and to Bloodnok for allowing me to use his script.

If you have any queries please send an email and I'll try to answer sensibly.

Steve, June, 2011