In Armagh, Ireland, in 1889, 50-year-old operating procedures, possible tampering, an engine that failed to pull the weight of the train and a staff argument all combined to make the worst ever rail accident in the history of Ireland and a radical change in law dictating railway safety standards.
The Armagh-Warrenpoint Branch
The branch line from Armagh to Warrenpoint is on the east coast of Ireland. It rises up steeply out of Armagh to the summit two miles before Hamiltonsbawn. The line was built in 1864 and was taken over by the GNRI1 in 1879. On 11 June, 1889 the traffic department of the GNRI in Dundalk had arranged for an excursion train to run from Armagh to Warrenpoint at 10.00am the next day. Around 800 passengers were expected so a 13-coach train was planned with two brake vans. The engine booked was a 0-4-2 tender engine No 86. Even before the empty train arrived at Armagh two more coaches were added and when the empty train did arrive there were so many passengers that the stationmaster suggested adding another two coaches. Thomas McGrath, the driver of No 86, protested and threatened to take only the original 13 coaches.
The stationmaster made some brash comments, but McGrath, who had not driven over the stretch of track before, protested and pointed out that he should have been provided a more powerful 0-6-0 engine. Someone suggested that they should transfer some of the coaches to the regular train at 10.35am which was of a much lighter formation. It was also suggested that the engine of the 10.35 could be used to assist the 10.00 train up the gradient. However, McGrath was now sufficiently annoyed at the suggestion that he was unable to get his train up the gradient, so he refused all offers of help and decided to take the 15 coaches.
The Train to Disaster
By the time the special train left Armagh it was 10.15 and there were 940 passengers crammed into the coaches and brake vans; most were children from Sunday school. All the doors on the train were locked2 to prevent 'unauthorised access'. The engine was not short of steam and stormed up the first part of the gradient. After that the train gradually lost speed. Just short of the summit it came to a halt. As with most trains at the time it was equipped with Smiths Patent Non-Automatic Brakes. This meant that the brakes were off until the creation of a vacuum, so that when the brake pipe was disconnected there were no brakes on the entire train, except for screw handbrakes in the brake vans. This was well known by railway staff.
In charge of the train was James Elliot, the line superintendent's chief clerk. Elliot and McGrath had to decide what to do. One option was to split the train, taking the first part to the siding at Hamiltonsbawn two miles beyond the summit and then return for the rest of the train. The front guard told Elliot that the siding at Hamiltonsbawn was already partially occupied by some wagons and that only about five coaches would fit in the remainder.
The Turning Point
At this point Elliot and McGrath had two choices:
Send a man back down the line to warn the following train so that it could assist them up the gradient.
Split the train, taking the first section on to Hamiltonsbawn and leaving 10 packed coaches and a brake van with only a weak handbrake.
Elliot chose the latter to keep the train on time. He went to the back brake van and told the guard to put the handbrake on and to place stones under the wheels; the front guard also put a single stone under a wheel on the sixth coach whilst he undid the screw coupling between the fifth and sixth coaches. As he did this, McGrath unfortunately eased back and by doing so he did what everyone was trying to avoid. The front five coaches only rolled back about a foot but it was enough to push the sixth coach over its stone. The coaches continued to buffer together until they pushed the brake van over its pile of stones. The driver tried to back the train down further to catch up with the other coaches so that one of the staff could reconnect the rear half of the train, but it was all to no avail. On the 1 in 75 gradient the handbrake failed to hold, and the ten crammed coaches quickly picked up speed.
Meanwhile back at Armagh the 10.35 was ready to leave, completely unaware of the dangers that lay ahead. Behind the engine of the 10.35 was a horse box, two vans and three passenger coaches. Under the time-interval rules passenger trains must have at least ten minutes between them and the passenger trains they are following. This time had elapsed. The driver of the 10.35 had been given the token for the following section of track and, four minutes late, the 10.35 was given the all clear.
The 10.35 was soon climbing the 1 in 75 gradient steadily, and was one and a half miles from the summit when the fireman saw the coaches from the 10.00 train hurtling towards them. The driver applied the brakes, but the train had not stopped when the runaway coaches smashed in to the 10.35 at over 40mph. Tragically all of the coaches were wooden, small, light and packed with passengers. The brake van and first two coaches were destroyed, leaving the rest to pile up behind the wreckage. The passengers in the brake van had a chance to jump clear but the rest were locked in their coaches.
There was now the danger of there being another runaway. As the engine of the 10.35 was knocked onto its side by the impact, the coupling bar to the tender had snapped, and the tender and horsebox began to roll away. Through sheer luck the driver of the 10.35 was thrown in to the tender by the impact of the collision and was able to apply the handbrake. The guard managed to apply the brake on the rear part of the train.
The Human Loss
There were bound to be casualties. Out of the 600 people in the rear half of the 10.00, 80 were killed or died later, including 22 children. A further 262 were injured.
The inspecting officer assigned to the inquiry for this disaster was Major General Hutchinson. After extensive trials Hutchinson found that the engine assigned to the 10.00 train was more than capable of hauling the 15-coach train up the incline and that the brake in the brake van was strong enough to hold the 10 coaches. He could find no reason why the train couldn't reach the summit. He looked at the possibility of the brake being tampered with by one of the passengers being carried in the brake van. Hutchinson concluded that McGrath should not have been assigned to the train because his knowledge of the line was poor. He placed most of the responsibility on Elliot for authorising the train to be split in two because that released the brakes on the rear portion of the train.
Hutchinson's findings provoked a stark reaction. Eight passengers were killed in other rail accidents across the British Isles in 1889, and there had been a number of fatalities over the preceding years, including an estimated 75 killed in the 1879 Tay bridge disaster, and Parliament decided that it had had enough. For years the Board of Trade, who were responsible for the railways until 1919, had been urging railway companies to adopt better safety standards. Many of these companies had not wanted to pay for the expensive equipment and had pleaded poverty, but now the law meant that they had no choice - they had to pay up. Had the Absolute Block system been in operation, the 10.35 would not have left Armagh. If Automatic Vacuum Brakes had been in use, the brake blocks would have been hard on the wheels of all the coaches when the 10.00 train was split. The horrific nature of this accident meant that the issues involved were addressed with uncharacteristic speed and decisiveness.
Absolute Block system
To operate the Absolute Block system the line is divided into blocks with a signal box at the boundaries of each section. Each box has single stroke bells and single needle block telegraph instruments. Each line has two block instruments, one for the section to the rear and one for the section ahead. If the box controls a junction there will be another bell and two block instruments for the branch and for any additional running lines. With a complicated system of bell codes3 each signalman was able to make sure that the block ahead of his box was clear before allowing a train to proceed onwards. This system was introduced in 1889 after this accident and is still in use today where semaphore signals are still used.