Abermule Disaster 1921


David Burkhill-Howarth is writing a book about the rail disaster which occurred at Abermule in 1921 when seventeen people were killed. In April 2007 he wrote an article outlining the background to the crash:

"The Cambrian Railways was the largest, though not the busiest, of the Welsh independent railway companies which opened up the Principality in Victorian days. It came into being on the 25th July 1864 when four small companies amalgamated.

The lines stretched from the English border at Whitchurch to the coast of Cardigan Bay between Pwllheli and Aberystwyth, and through mid-Wales and the Wye valley to Brecon. The connections to the North-west of England were via the London and North Western Railway, whilst the Great Western Railway provided links for stations between London and North Wales.


For a high proportion of the Cambrian's mileage trains ran in both directions over the same set of rails; this was potentially dangerous, but a very safe operating system had been developed over the years.

Each single line section, or block, of a few miles in length had a token which the engine Driver had to have in his possession before he drove along it. This was in addition to him obeying the signals.


Each block had an instrument at each end, and these were electrically connected in such a way that only one token could be obtained from either instrument at a time, no matter which way the train was running. To appreciate the circumstances which led to the terrible accident on January 26th, 1921, the station working at Abermule in the upper Severn valley needs to be considered.

 One of the single track sections in question started at Welshpool, then continued in the 'down' direction towards Montgomery station, which was situated two miles from the town of Montgomery. After 3 miles, the line reached the crossing loop between the platforms at Abermule. The route then aimed westwards, four miles up hill towards Newtown, the next block station.


The staff at Abermule station on that fateful day was as follows. Relief-Stationmaster Frank Lewis from Montgomery was deputising for the regular Stationmaster, John Parry, who was on leave. Lewis was well acquainted with the routine at Abermule, having previously acted there on two or three other occasions as relief Stationmaster and a similar number of times as Signalman. Although well known to the other members of staff, he was without doubt an outsider and considerably junior in service to the Signalman, who was a possible choice to take charge in the regular Stationmaster's absence.

Signalman William Thomas Jones (Bill) was 60 years old, and had spent all his 32 years service at Abermule. Starting as a Porter in 1888, he became a Signalman-Porter in 1896; there was insufficient traffic to warrant a Signalman who had no other duties.

Bill had been on duty since 3 a.m. with a break taken from 8.20-9.20; this crossover was the last train movement of his shift.


It is interesting to surmise that Bill Jones felt some hostility towards Lewis. Bill as the 'old hand' on his home ground, probably felt that he was quite capable of looking after the station without the assistance of upstart ex-goods guards. No doubt the extra pay would have come in handy as well.

Also working at the station that day were: Booking Clerk, Francis William Thompson of Llandyssil; he was a 15-year-old with two years service. He was described at the Inquest as 'being a very intelligent lad and very sharp at school'.


Porter, Ernie Percy Rogers of Garthmyl, who was a 17-year-old with four years' service.

Neither of these youngsters was officially trained, nor permitted, to take part in the operation of the signalling system. We must assume that Thompson and Rogers were both ordinary country lads who fancied a life with a modicum of excitement, a bit of a uniform, and a small but steady wage instead of being a farm hand. They might have made it to Signalman or Stationmaster in the course of time. We may take it as certain that the errors committed at Abermule station on the 26th January 1921 were by no means unique on the Cambrian, or any other railway. At many another single line crossing station, unauthorized staff had worked the tablet instruments designed to protect the trains from disaster, or had passed the safety tokens from hand to hand, bypassing those instruments.

Other Stationmasters had, for one reason or another, failed to be present at the arrival of a train and then authorized the 'right away' without proper assurance that all was as it should be.


Many a Driver had received a 'tablet' authorising him to proceed, and had not examined it.

However, at Abermule on that disastrous day, those errors joined one another until the outcome was certain. Ironically, it did not help that the system was well-tried and apparently fool-proof. Rail travel had become a lot safer around the turn of the century. The last passenger train collision on a single track in the UK to result in fatalities had been at Radstock on 7th August 1876, over 44 years previously. In view of the Company's explicit instructions to Stationmasters to give personal attention to the crossing of the two trains, the first mistake Relief Stationmaster Lewis made on this particular day was in taking his dinner later than usual, although this was caused by attending to Company business.

As a result, he did not return to the station building until ten minutes after the usual time, and uncomfortably close to the expected arrival of the local that was to halt at Abermule to let the express pass.


His second mistake was immediately to leave the office on another Company errand without first finding out where the express was, and without arranging to be called as soon as it was 'belled'. The accident would never have occurred if he had been told on his return from lunch that the express was already well past Moat Lane Junction. Lewis acknowledged full responsibility for his failure to examine the tablet. He also agreed that he had made no alteration in the station's working practices when he took over, although he could see that they did not conform to Company practice.


For instance, it was common for the young Booking Clerk, Thompson, to be unlawfully involved in the 'tablet' changing, but he, Lewis, had merely cautioned him to be careful.

It can be supposed that this lack of 'pulling the staff up by their boot-laces' might, in part, have been because of a strained relationship with Signalman Jones, the 'old hand' on the station. Knowing that he was only there for a short period, it was presumably easier for Lewis to work with the Abermule system of operating a railway, than the Cambrian's. Jones, from what can be deduced from this incident, made no attempt to assist or coordinate his knowledge of a situation with the Stationmaster.


Jones appeared to consider himself entirely free from blame in the event. Yet, it was his duty, in the absence of the Stationmaster, to receive the tablet from the 'down' train, and, if he had performed that duty, the accident would not have happened.

He knew well in advance that the express had passed Moat Lane Junction 'on time', but made no attempt to pass on the information to Lewis. Jones did not subsequently take any action to find out what had happened to the express when the crossing plans were changed, though he knew that unless a breakdown had occurred, it should pass the local train at Abermule as planned. The crew of the local train from Whitchurch, through lack of attention at the station and on the track, for whatever reason, held at least half the responsibility for the crash. Between them all, they were responsible for a significant number of deaths and injuries of Cambrian passengers and employees through their slap-happy working and, perhaps, a little bit of resentment or an uneasy working relationship.


The only operating staff immediately involved who came well out of the event were the Driver and Fireman of the Aberystwyth express who did everything they could to mitigate the disaster; they only had a few brief moments after sighting the local steaming hard towards them on the same line. The inevitable collision occurred at a point one mile south-west of Abermule Station, close to a place called Red House crossing. The railway from Abermule leading to this point was on a low rising embankment, and there was nothing to obstruct the local crew's view of the line. Newtown is four miles from Abermule, and the express had accelerated in good style. It was running at about 45-50 m.p.h. when it reached the usual point for shutting off steam prior to slowing down for the tablet exchange at Abermule.

The express passed under the bridge on a short length of right-hand curve; this lay in a cutting whose slopes fell in height from about 16 feet at the over-bridge to three feet at the site of the accident.


It was possibly the only place on the whole main line where the view was so restricted. The combined impact speed of the two trains would have been about 60 m.p.h., approximately 30 yards per second, so, depending on the efficiency of the braking there would have been only 10-12 seconds before they crashed. The crew of the 'local' train did not appear to see the express because the column of smoke continued to pillar skywards as their engine laboured hard against the rising grade. The Driver of the express claimed that he never whistled, his priority was to reduce his train's speed; however witnesses, including his Fireman, heard a whistle just before the collision. Did the Driver of the local catch sight of the express bearing down on him at the last minute, and take some belated action?


There might have been a small chance of averting the disaster if both trains had braked immediately; certainly the impact would have been greatly lessened if either member of the local's crew had been keeping a look-out. Driver Pritchard Jones and Fireman Owen had stepped outside of their cab and were riding on the footsteps knowing that they had done all they could. The brakes of the express train had bit hard against the wheels, but not hard enough to save the situation; the 254 tons express was running on a falling gradient of 1 in 123, still making about 25-30 miles per hour when it rammed head-on into the 217 tons of stopping train, labouring hard at about the same speed.The crash occurred at approximately 12.06."


Article by David Burkhill-Howarth.