An Atmospheric Conspiracy?

22in diameter atmospheric pipes on display at Didcot Railway Centre. These pipes were unearthed at Goodrington Sands in 1993. It was Brunel's intention to use these larger pipes on Dainton Bank. (Photo from Wikipedia website)

Mention the name Brunel and we immediately think of the genius who pushed civil engineering, mechanical engineering and marine engineering to new limits. His innovation was never exhausted.

Brunel was the engineer for the Great Western Railway, he designed steam locomotives, he designed the largest ships and was the inventor of the screw propeller. He was, and still is a man regarded in the highest of esteem for his achievements.

Clearly Brunel was a man of exceptional ability but I am led to ask why did he promote the Atmospheric system of propulsion for the South Devon Railway? It was a system which the simplest of arithmetic clearly shows was physically impossible.

The South Devon Railway was the fourth railway built with the atmospheric system of propulsion patented by Clegg and Samuda in 1838. The first line was a demonstration track at Wormwood Scrubbs on what is now the West London Line. The others were the London and Croydon Railway and the line between Dublin and Dalkey in Ireland.

The more I read about the building of the South Devon Railway with the Atmospheric system the more I doubt the accepted historical account. Essentially the popular story is as follows:

Having completed the construction of the Bristol and Exeter Railway, Isambard Kingdom Brunel was appointed as chief engineer by the South Devon Railway for a line between Exeter and Plymouth. The route along the southern foothills of Dartmoor was difficult country in which to build a railway necessitating steep gradients. Brunel had been impressed by the Atmospheric system and proposed its use for the South Devon. He proceeded to survey and design an alignment for the railway with total disregard for gradients, which would not be a problem for the system. Subsequently there was difficulty keeping the atmospheric pipe airtight and the rats ate the leather flaps attracted by the use of tallow as a sealing agent. The system was abandoned and the rats took the blame. The railway reverted to steam haulage, struggling over severe gradients never intended for that form of traction. It is also said that had the Atmospheric system been abandoned at an earlier date Brunel could have changed the alignment at the Plymouth end to ease the gradients.

Well that all might be perfectly true but I have a suspicion there is much more to the story than has been recorded! I suspect that the railway, which was built, together with steam locomotive propulsion, was exactly what was intended from the beginning and that Brunel and the directors of the South Devon Railway conspired to deceive objectors, landowners and Parliament into thinking that the Atmospheric system would avoid noise, smoke and burning cinders from being thrown out across farmland. One has only to look at a modern Ordnance Survey Landranger map to see that the route is lined with country residences belonging to the gentry. Today there are no witnesses to be called in evidence. The only evidence is circumstantial and is to be found in the historical record. Unfortunately historians do not agree entirely on dates of the various events. I have examined books, videos and the Internet and find that historians sometimes differ by a year in their accounts. I have compiled the following chronology, which is a consensus of several sources. However, we can never be sure that one source is not just copying another.


21 Nov 1843    Brunel’s proposed route agreed by shareholders
1844    South Devon Railway Act passed by Parliament
1845    Atmospheric trains began to run on the London Bridge and Croydon Railway
30 May 1846    South Devon Railway opened between Exeter and Teignmouth
Jul 1846    Atmospheric trains abandoned on the London – Croydon Railway
30 Dec 1846    Teignmouth to Newton Abbot opened
20 Jul 1847    Newton Abbot to Totnes opened
13 Sep 1847    Public service with Atmospheric system started to Teignmouth
10 Jan 1848    Public service with Atmospheric system extended to Newton Abbot
5 May 1848    Totnes to Laira Green opened
9 Sep 1848    Atmospheric trains withdrawn
18 Dec 1848    Branch to Torquay (Torre) opened
2 Apr 1849    South Devon Railway opened throughout to Plymouth Millbay


  • The London Bridge to Croydon Railway had not commenced operation when the Act for the South Devon Railway was passed by Parliament so it is unfair on the part of some historians to comment that Brunel should have learnt from the lessons of that railway. The SDR was already under construction between Exeter and Teignmouth when the problems of the Atmospheric system became serious for the L&CR. However, Brunel was a very competent engineer. He had taken civil engineering to new limits, with for example his arches across the River Thames at Maidenhead and Box Tunnel. He was the inventor of the Broad Gauge, a designer of steam locomotives and an architect of ships. He had engineered the railway all the way from London to Exeter, which was no mean feat considering the hundreds of people armed with computers that were employed in Tottenham Court Road to design the Channel Tunnel Rail Link. This was a man of genius; there is no question about that! But, it is inconceivable that he did not make the elementary calculations to establish the suitability of the Atmospheric system. Atmospheric pressure is about 14.7 lbs weight per sq. inch. On a piston of 15 inches diameter as used on the L&CR this provides a force of 2598 lbs weight. Brunel opted for a larger pipe on the SDR of 20 inches in diameter giving a force of 4618 lbs weight, which suggests that he did actually do the calculations. However, these figures are only the maximum theoretical values for stationary trains. They assume absolutely no leaks and that a primitive steam powered reciprocating pump is capable of achieving a perfect vacuum in a pipe up to five miles away. Once the train starts moving they assume no friction between the piston and the cast iron pipes also that the pressure in the pipe is restored instantly behind the piston through an inlet valve located at the previous pumping station to the rear of the train up to five miles away. The probability is that the system was only able to achieve a force of about 1,000 lbs weight at the connection to the piston. To put this into perspective a Brighton 0-6-0 Terrier Tank of 1872 has a tractive effort of 10,695 lbs at the drawbar. So in reality there was no way Brunel could have seriously expected a system with only a tenth of the tractive effort of a Terrier Tank was ever going to take a meaningful train over Dainton, Rattery or Hemerdon banks.
  • Taking the calculation a little further the maximum theoretical gross weight, which the system could hold stationary on the 1 in 37 gradients of the SDR was 76 tons at the very best but the figure was more likely to be less than 20 tons in practice. So quite possibly the system was incapable of moving even a single loaded wagon up the 1 in 37 gradients. Brunel had designed locomotives for the GWR. He knew what loads were to be carried. He had argued for the Broard Gauge on the basis that the 4’-8.5” gauge was inadequate for the loads to be carried. So why was he now advocating a system, which needed lightweight rolling stock and had no capacity to shift a meaningful load?
  • Brunel had extensive experience of track layout design and operational requirements, so why did he advocate a system, which could not have pointwork? The Croydon line overcame the problem by building the world’s first flyover at Norwood Junction to carry the atmospheric tracks over other lines. The South Devon was built as a single line, so how did he envisage that trains would pass at stations? It could only be done with a steam locomotive in attendance. The atmospheric trains would reach the ends of the pipe sections and be disconnected from the pistons. A steam locomotive would haul one and then the other into a passing loop. It would then shunt across to the rear of the first train and propel it forward to connect to the piston in the next section of pipe. It might even have given the train a push to see it on its way! It would then return to the second train and repeat the operation. On the arrival of trains the pipes would be at atmospheric pressure. Once the trains were connected to the pistons in the next sections and sitting waiting with the brakes on there would be a lengthy delay while pumps up to five miles away pumped out the air from the pipes. Not to wait with the brakes on would have given an unpredictable start and risk of injury to passengers.
  • Historians record that the atmospheric pumping stations were either three or five miles apart. However, when one considers the locations chosen further questions are raised. Why was a pumping station constructed at Totnes? It is at the bottom of two inclines. Towards Exeter we have Dainton Bank West and towards Plymouth is Rattery incline. Atmospheric trains would have coasted down the inclines to Totnes from both directions. A smaller pipe would have been laid alongside the railway at Totnes station with a piston, which could be coupled to rolling stock with a rope for shunting. So what other purpose would the Totnes pumping station have fulfilled?
  • Why did Brunel propose the vulnerable route around the coastline via Dawlish? The ravages of the sea were no less before the railway was built than they are today! The alternative was to have taken an inland route much the same as was eventually constructed via Heathfield to Newton Abbot. If gradients were not a problem for the Atmospheric system why take a vulnerable route?

Above: A London bound HST descends Dainton Bank East on 12th August 1982

The Route

The route proposed by Brunel was only one of several proposals considered by the promoters of the South Devon Railway. Several options were proposed by James Meadows Rendel a former apprentice to Thomas Telford. Rendel’s preferred option was a route right across the centre of Dartmoor a distance of just under 43 miles. The cost was estimated at £770,781 whereas Brunel’s proposal, which he first surveyed as early as 1836, was 50 miles in length and estimated at £1.8m. So Brunel’s proposal was not the cheapest option by any means, exceeding Rendel’s proposal by over a million pounds at the value of money in 1840! Clearly there was something impracticable with Rendel’s proposal. To go over the Moor would have meant climbing to an altitude exceeding 1000ft. Simple calculation reveals that to rise 1000ft in half the distance (21.5 miles) required an average gradient of 1 in 57. Rendel claimed that once up on the Moor the valleys would be easier to cross. If we remove the section across the Moor and the approaches to Exeter and Plymouth from the above calculation the gradient would be significantly steeper. The proposal most likely included some impossible gradients, which were rejected by the promoters.

It is also claimed that the directors favoured a coastal route running via Torquay, Dartmouth, Torcross, Kingsbridge and Modbury. Interest in this route was probably generated by the recommendation to the Government by an Admiralty Committee in 1840 for Dartmouth to become the port for mail steam packets. Subsequently the Government chose Southampton. On closer inspection this route would have needed several gradients of a similar order to Dainton and Rattery, i.e. 1 in 36. Torcross is a long way south of a direct line between Exeter and Plymouth. It would have been another “Great Way Round”! A major bridge across the River Dart would have been required together with tunnelling south of Dartmouth. There would probably have been considerable opposition to routes through the towns of Dartmouth and Kingsbridge.

The allegation that Brunel disregarded the severity of the gradients, as they were not seen to be a problem to the Atmospheric system is almost certainly untrue. The route follows a watercourse from Newton Abbot to Stoneycombe and Dainton Bank West follows another watercourse all the way down to Totnes. Really there was no scope for any other alignment through this area. Possibly Dainton Tunnel could have been deeper and longer. West of Totnes Rattery incline follows another watercourse and then the Harbourne River to Marley Tunnel. Again, there was little or no scope for an alignment with an improved grade through this area. Finally Hemerdon Bank follows another watercourse down to Plympton. It is very unlikely that Brunel’s alignment could be improved on even with the use of computers and modern alignment design software. Possibly curves could be eased a little to increase the line speed at some locations but there would be a penalty to pay in more extensive earthworks and bridges.

The Conspiracy!

At a meeting held on 21st November 1843 the shareholders decided to abandon Rendel’s proposal. However, these shareholders were not your average punters! The project was largely funded by the Great Western Railway, the Bristol and Exeter Railway and the Bristol and Gloucester Railway. It is most likely that the directors of the South Devon Railway also sat on the boards of these companies.

The problem facing the directors was to silence the local opposition. The successful passage of the London and Croydon Railway Act probably inspired someone on the board to suggest the inclusion of a proposal in the South Devon Railway Bill to use the Atmospheric system but really without any serious intention of using it. When the South Devon Railway Act was passed by Parliament on 4th July 1844 the directors were faced with another problem. If they declared their intention to abandon the Atmospheric system they risked a legal challenge under the Act. There was also the possibility of a private members bill to change or nullify the Act. The directors needed to delay the abandonment of the Atmospheric system until the railway was substantially complete and open for traffic while at the same time not spending a penny more than necessary on installing equipment. Once the railway was open it was unlikely that a legal challenge would succeed.

The historical record of the next few years provides the circumstantial evidence to support the suggestion of a conspiracy. The first section of the South Devon to be constructed was between Exeter and Teignmouth. The choice of the near level but vulnerable coastal route to Newton Abbot avoided challenging the Atmospheric system on gradients, which it could not cope with. The atmospheric pipes were laid with the track to minimise installation costs. The line opened on 30th May 1846 using steam locomotives and the Atmospheric system was not brought into use for public trains until 13th September 1847. Did it really take another 16 months to bring the pumping stations into use or was this a deliberate delay to buy time? The abandonment of the Atmospheric system on the London to Croydon line in July 1846 would have raised concerns in Devon regarding the intentions of the South Devon Railway, which in turn would have put pressure on the company to begin operations with the system as soon as possible. During this period of delay construction of the line forged ahead. Public service opened to Newton Abbot on 30th December 1846 and to Totnes on 20th July 1847. So a public service between Exeter and Totnes was in operation with steam locomotives two months before the first Atmospheric train even ran in service. The Atmospheric system was extended to Newton Abbot on 10th January 1848. Possibly this was as far as it was ever intended to use the Atmospheric system. Works to provide the Atmospheric system beyond Newton Abbot were deliberately stalled keeping outlay to a minimum. Every penny spent now was being wasted. Any attempt at using the system to draw trains up Dainton Bank would be doomed to failure and a legal challenge was still possible. Meanwhile construction of the railway continued.

On 5th May 1848 the line opened for public service to Laira Green but the company continued to stall the extension of the Atmospheric system beyond Newton Abbot. The pipes had not been installed at the same time as the track beyond Newton Abbot and the installation of pumping equipment was being held up. By 9th September 1848 the railway was nearing completion into Plymouth Millbay and the pumping station at Dainton was about to come into operation. The whole route was now secure and it was time to abandon the Atmospheric system. Brunel would have liked to conduct trials on Dainton Bank but a public failure would have brought his integrity into question and roused suspicion of a conspiracy. No, it was better to abandon the system at this time and let the Devon rats take the blame.

The line finally opened all the way from Exeter to Plymouth Millbay on 2nd April 1849.

So were they right or wrong to do what they did? They gained nothing by their actions. The Atmospheric system cost the South Devon Railway between £300,000 and £400,000 from which it never really recovered. In the end it has been the ordinary people of Devon and Cornwall who have gained with a railway providing them with services to Bristol and beyond. The railway survived the Beeching cuts although many stations and branches were closed. It has retained double track even though other main routes have been singled and it has benefited from recent residential development with a new station opened near Ivybridge. The line will probably continue to serve the community until long into the future. Maybe they were right!

Paul Plowman - 14 January 2020

Reader Comments