Linking accidents

One of the many advantages of our data is that it allows us to look at trends and different types of accidents over time. In this guest post, NRM volunteer Philip James looks at problems arising from joining railway wagons, carriages and engines. As ever, our thanks to Philip for this post and all his work on the project!


During the preparation of my last post and subsequently, I have become aware of and greatly appreciated input from others in the project and some outside it. Their observations and suggestions have stimulated research and are a motivating force for further work. There is much I could include but in the interests of brevity, I have decided to focus on accidents related to coupling stock together.

For my earlier articles (found here), I have relied upon observations from the accident reports I personally have processed. While preparing this piece I used the spreadsheet containing entries prepared by other volunteers covering phase one of this project, January 1911 to June 1915. This has expanded my pool of data from about 1,350 rows of data to about 5,000 and may represent a quarter of the reports processed to date. I don’t know which volunteers provided the data, probably all who were working on the project up to December 2017 so my thanks for this.

Accident at Birkenhead

I have seen many accident reports involving coupling. In some cases, men went between waggons to couple by hand, sometimes for defensible reasons but usually without need and in breach of regulations. In other cases, they would use a shunting pole or coupling pole to perform the task from a distance without going between vehicles but this often led to injuries as well.

Typically, men would try to couple two waggons as they were coming together and before the buffers had touched. If things went wrong, then a hand or arm would be caught between the buffers as they closed. Sometimes a body part might be caught between the back of the buffer head and the buffer casing as the buffer was depressed. The question is why did they put themselves at risk when they could wait a few seconds for the buffer heads to come into contact or better still for the waggons to be completely stationary? This point was made repeatedly by the investigating officers.

The accident location

Cavendish sidings, Birkenhead, 1908.
Courtesy National Library of Scotland Maps.

The first 1,100 or so accidents of all types that have fallen to me to transcribe shed no light on this but then this accident at Birkenhead explained it all. The explanation seems so simple that I wonder how I didn’t see it before. If I had been able to do some coupling for real, I suspect it would have become apparent quickly but relying on concise if not terse documentary accounts, things are perhaps less evident.

London and North Western and Great Western Joint Railway.
Date – 17 October 1907.
Place – Birkenhead, Cavendish Sidings.
Name of Person injured – Alexander Chambers.
Age – 23.
Capacity in which employed – Shunter.
Number of booked working hours per day – 11.
How long on duty at time of Accident – 8 hours.
Nature of Injury – Left arm crushed, subsequently amputated.

Description – ‘About 8 p.m., during shunting operations, six waggons were propelled into No. 28 road, in order to attach two others standing there. When the moving waggons wore almost in contact with the standing ones Chambers attempted to couple with his coupling pole. As he was raising the coupling the hook of the pole slipped out of the bottom link, with the result that his left arm was severely crushed between the buffers as they closed up.

Conclusions – In this case there being only two waggons in No. 28 road, if Chambers had waited until the buffers touched before attempting to couple, the weight of the engine and the six waggons attached to it would have driven the two stationary waggons so far forward that he could not have performed the operation, and therefore I consider that the mishap should be classed as incidental to the use of the coupling pole.

Investigating Officer – JJ Hornby.’

Coupling Accidents

As a mathematician, a collision between waggons being shunted reminds me of how you might model collisions between snooker balls. Initially all the balls are at rest and the player puts some momentum into the que ball by striking it. Some or all of that momentum is transferred to one or more other balls when they are struck. Eventually the momentum is dissipated by the frictional forces between the balls and the cushions or table surface and to a lesser extent by the viscous nature of the air surrounding them.

Mathematicians, physicists and engineers use terms like the coefficient of friction or coefficient of viscosity to describe the roughness of a surface or the stickiness of a fluid like air[1] and the snooker ball model will have a coefficient of restitution to describe the proportion of momentum lost on each collision. For this piece we need not go deeper into the theory than this.

Suffices to say that in the real world, shunters and others moving waggons would by experience work out how one group of waggons might behave when struck by another and adjust their coupling technique accordingly. Similar to how snooker players learn to judge and control the movement of snooker balls but likewise subject to errors and mistakes.

In general, if one group of waggons are shunted into another, then the waggons being shunted might rebound slightly, become stationary or continue moving forward depending on circumstances. The group being struck might move forward or remain more or less stationary again depending on the particular circumstances.

Various factors might determine the outcome. How many waggons are there in each group and how heavily are they loaded? Have any of the brakes been applied or have scotches or sprags been used to impede movement? Are there fixed buffer stops preventing movement? Is there a gradient that may retard or accelerate movement after the impact? Were the first group of waggons being loose shunted or were they attached to a locomotive able to apply some braking force throughout the process? How fast were the waggons moving when being shunted? A higher speed would introduce more momentum to be dissipated. Waggons might also be moved using a capstan or a horse.

Without doing any advanced mathematics, the shunters would be aware of these possibilities and through experience know what was likely to happen in any particular case. In particular, they would expect waggons to separate a little after the impact so the chance to couple was a momentary opportunity requiring skill and judgement on the part of those involved. In some cases, it might be possible to wait until the buffers have closed up and all vehicles are stationary. In other cases, it might be sufficient to wait until the buffers have touched and then couple quickly before a rebound can separate them. In a few cases it might be necessary act before the buffers have touched as the rebound will be faster and greater as in the accident above.

Clearly this process was prone to human error. Trying to couple sooner risked being caught between the buffers or at least have a body part pinched between the rear of the buffer head and the buffer casing as the buffer depressed. Coupling later, risked the waggons separating too far to be coupled so a further shunting move would be needed. The accident reports we see reflect the cases where men judged it wrongly and got hurt. We don’t hear about the cases, probably the vast majority, where they got it right or at least nobody was hurt.

Project co-lead Mike makes the observation “this is the sort of assumed knowledge that staff and inspectors had and therefore didn’t usually provide (in reports), but which is invaluable.” He also drew my attention to this short video of shunting in progress. It postdates our era by several decades and includes a type of operation not present at the time of our reports but illustrates waggons moving and efforts to control them. Coupling poles are evident.

NRM Volunteer Coordinator Craig Shaw makes the following observation: “I keep finding various permutations of the “coupling while still moving” or coupling while waggons where being shunted on the line, with references to the rules and Inspectors’ recommendations.  In the 1900s there simply wasn’t the attention paid to the risk, probably due to pressure to complete the work, with often dire results.” The reference to risk and pressure says it all. Men needed to get things done quickly and losing time with additional shunting moves cost money.

Sadly, the young man involved in this accident lost an arm as a result of it being crushed between the buffers. Although being caught between vehicles could be fatal,[2] those attempting to couple with a pole would usually escape with all their limbs. Injury could also be due to body parts being pinched between the back of the buffer head and the buffer casing as two vehicles come together and the buffers are pressed in. Not strictly speaking caught between vehicles but a similar injury and risk for those concerned.

It is also worth reflecting on how this accident report compares with others involving coupling. There were many similar reports, so much so that when transcribing them it was often expedient to copy and paste from one to another and then make minor changes. The accident details, conclusions and inspectors’ recommendations were often so similar that at times they became predictable and it was almost boring to read them.

This accident has a similar description to the others and apart from fatalities, the injury is about as serious as it gets for this category but the conclusions are particularly informative and distinctive, and no recommendation is made. The inspecting officer, JJ Hornby, features in many such reports and is perhaps more assertive than his peers when calling out unsatisfactory working practices and making recommendations for change.

Moving on another two hundred or so accidents and I find a similar report[3] also prepared by JJ Hornby that highlights issues with the couplings themselves. In this case a fireman was forced to couple by hand before vehicles had come together as the coupling on the tender of his engine did not have a draw-bar hook thus making the use of a coupling pole impossible. The relative weights of the engine and waggon meant they would always bounce apart after touching.

Sadly, the unfortunate man lost an arm and also had a leg run over. While attributing the accident to misadventure, the inspector noted that provision of a draw-bar hook might have made it avoidable. He recommended that all engine tenders be fitted with draw-bar hooks as expeditiously as possible; polite but as firm as recommendations get in this era.

Another inspector, C Campbell reported on a coupling accident involving a fatality[4]. In this case a man was using a screw coupling (see images 2 & 3 below) which rose and hit him as the vehicles rebounded and his head was caught between the buffers as they came together for a second time. Unlucky that his head should be between the buffers at that precise moment but illustrative of the mechanics of the coupling process and what can happen if you do it often enough. Inspectors sometimes point out when men were acting with good intentions and this is a case in point as it was not the unfortunate man’s job to couple waggons but he may have thought he was saving time by doing so.

In an earlier piece, I quoted Mike Esbester on the need for automatic couplings. “It was the one that the unions picked up on, largely because there were possible technical solutions, and automatic couplings of freight stock in the USA was mandated in 1893”. Clearly many accidents were incidental to the use of coupling poles and it is not always possible to adopt a way of working where men wait until buffers have closed up or waggons are stationary. The desire for an automatic coupling is understandable.

Brake, Coupling and Shunting Poles

Many accidents feature the use or misuse of coupling poles. A long pole with a hook on the end enabling a shunter or others involved in coupling, to lift a coupling link over a draw-bar hook without having to go between the waggons to couple or uncouple. This might also be referred to as a shunting pole. Another pole used by shunters was a brake pole used to manipulate the brakes on a waggon from a safe distance. This had different dimensions to a coupling pole, being shorter and without a hook on one end. For more on the differences, including some images, see here.

Sometimes a shunting pole would be used as an improvised brake pole with the hook end turned away from the waggon being braked. This could lead to problems as in the confined space between adjacent sidings, the hook end of the pole could get caught on another waggon or the unexpected movement of a waggon in an adjacent siding could strike the pole and unbalance or injure the person wielding it.

Presumably shunters found it easier to carry one pole rather than two, and by and large were, with care, able to get away with misusing a coupling pole for braking purposes. Many reports refer to rules prohibiting such misuse.


Buffers feature in many accidents. Typically, a person or some part of their anatomy has been caught between the buffer heads as they close. In general, such an accident is classified in the database as ’caught between vehicles’.

A variation on this requires a simple description of how a buffer operates. The buffer head is attached to a plunger that fits inside a casing and is sprung. This link illustrates the internal spring; this one shows buffer design.

As vehicles close up, the buffer heads make contact and the springs absorb much of the energy of the impact. This means that the gap between the back of the buffer head and the casing diminishes and a person or part of their anatomy in the gap could be crushed or pinched. Strictly speaking, not ‘caught between vehicles’ but with some similarities injury-wise.

Some buffers were ‘dead’ or ‘dumb’ (meaning that they were solid with no springs) or were defective with the springs not functioning to absorb impact energy. It is not clear what proportion of buffers were of this type but one accident report[5] suggests they were sufficiently numerous for men to make mistaken assumptions that they were. In general buffers on vehicles are of iron or steel construction but wooden buffers[6] have been used. In general, they are on the sides of vehicles but even today, central buffers are used on some vehicles.

On tight curves, buffers might become locked with the heads managing to pass so the backs of the buffer head are in contact. This would be a particular risk if the waggons were of different types, lengths or designs[7]. Such an occurrence would not normally resolve itself and might risk a derailment as the train reaches straight track. Waggons on curved track might be difficult to couple as the buffers force the couplings further apart than on straight track. This link shows buffers sliding off one another.


I have not included any photographs of the accident location but focused on the mechanics of the process. Most pictures and diagrams have come from the Internet and except where stated, must be attributed to unknown providers.

The couplings illustrated are not necessarily a comprehensive list of the types in use but illustrate those frequently referred to in accident reports. The drawbar hook was sometimes used by horse shunters when moving waggons. This was in breach of the rules and separate horse hooks were usually provided on the sides of waggons. This removed the need for the horse or shunter to walk in the four-foot way ahead of the waggon, potentially dangerous given the poor brakes available.

Picture 1 shows a chain coupling being hooked over the drawbar hook of an adjacent vehicle using a coupling pole. The chain on that vehicle is left hanging loose. The buffer heads, plungers and cases are evident.

Pictures 2 (above) and 3 (below) show a screw type coupling. The central section is a bar with a left-handed thread on one end and a right-handed thread on the other. It can be wound around to tighten or loosen the coupling. As before the coupling on one vehicle hangs loose.

Pictures 4 (above) and 5 (below) show different designs of chain link coupling. In practise vehicles with different types of coupling might need to couple. The Railway Clearing House was set up in 1842 and in time took on the role of improving interoperability between different types of waggon including their couplings.

Picture 6 illustrates the effort needed to lift a heavy chain from the level of the ballast. The person doing the coupling would also need to control this process while keeping their hands and arms clear of the buffers.

Picture 7 shows a screw type coupling and buffers.

Pictures 8 (above) and 9 (below) were taken by me. The former in October 2016 at Sheringham shows the buffers between two coaches, the latter at an open day at Ilford Depot in 1985 is of A4 Sir Nigel Gresley, the coupling and buffers are just visible.

Philip James


[1] Liquids also have a coefficient of viscosity. Scientists class both liquids and gases as fluids. In some cases, finely ground solids can also be modelled as fluids and the movement of galaxies in the vastness of the universe likewise.

[2] 1909 Q1 App C, Taff Vale Railway, 27/3/1909, Treorchy, Inspector JJ Hornby.

[3] 1909 Q1 App C, Midland Railway, 11/3/1909, Locomotive Yard, Bedford, Inspector JJ Hornby.

[4] 1909 Q1 App C, North British Railway, 25/1/1909, Kinross Junction, Inspector C Campbell.

[5] 1909 Q1 App C, North Eastern Railway, 3/2/1909, Gascoigne Wood, Inspector A Ford.

[6] 1909 Q1 App C, North British Railway, 5/2/1909, Portobello, Inspector JJ Hornby.

[7] 1907 Q4 App C, Lancashire and Yorkshire Railway, 15/11/1907, Goole, Inspector A Ford.

, , , , , , , , , , ,


  1. Data Analysis and Issues - Railway Work, Life & Death - January 26, 2021

    […] my previous post, I noted the enlarged pool of project data available to me, about 5,000 rows, and thought this […]

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Powered by WordPress. Designed by WooThemes