South Devon Railway
Newton Abbot

In which Joe sees the end of the atmospheric South Devon Railway, but not the end of construction and not the end of these pages.

Explorations: 25 June and 11 October 2018, and 28 June and 21 October 2019.
Journey: Paddington to Newton Abbot.

The eighth engine house on the South Devon Railway was at Newton, later called Newton Abbot. It was adjacent to the station and was the company's center of operations. This was the farthest point to which atmospheric trains ever worked. Further atmospheric-related construction was undertaken but not put into use.

Newton was the home base of the South Devon Railway. Its location very roughly midway between Exeter and Plymouth was important, but even more so was that it was the only place on the line with available land around it that was both flat and not flooded. It was also at the boundary between the almost level line to Exeter and the very hilly line to Plymouth that would need either larger atmospheric pipes or, otherwise, more locomotive power. Here at Newton the company established workshops for locomotives and other rolling stock on the ground east of the main lines. The last of the maintenance work here ended in 1981. The property is now covered by other commercial buildings.

Newton Abbot OS map
— Newton Abbot, Ordnance Survey 25-inch, 1890. CC BY-NC-SA 4.0 courtesy National Library of Scotland.
Annotations by Joseph Brennan.

Peter Kay's research has the Newton atmospheric buildings in place until "1893?", so the OS map of 1890, and surveyed within a few years earlier, includes them. I think I have the boiler house and chimney in the right places, but judging by this map, the engine house was either gone or included in a larger group of buildings. The reservoir was of course gone. The present-day station is much larger than the one on the map, so I have shown the extent of the platform area and also of the station house on Station Road that was completed in 1927.

William Dawson provided a plan of the arrangement during the brief atmospheric period. North is to the left. The "Tank" reservoir was of course long gone by the date of the OS map. The Torquay Road, seen toward the right, was the original proposed station site.

Dawson print
— William Dawson, plan at Newton, via Brunel's Atmospheric Railway.

Dawson's view from the north, below, is more informative, if somewhat foreshortened. On the left we a horse and cart on the Quay Road coming from the dock on the Teign, about to cross under the railway. The engine house is raised to track level, to which it connects by a two-arched stone bridge that may have carried the siding for coal delivery. The yellow building beyond appears to be the locomotive shed that is seen in the plan, with a turntable near its entrance.

Dawson print
— William Dawson, view of Newton station, via Brunel's Atmospheric Railway.

The arrangement of atmospheric pipe and track in Dawson's drawing is presumably correct, but how it was worked is puzzling, since there must have been points and crossings to fan out the single track to what we see here. The only track with atmospheric pipe is at the center, and it runs into a wooden passenger station similar to the others on the SDR, with one track in the trainshed and one platform on the town side. This track is flanked by ordinary tracks, and outside of those are the smaller pipes used for starting ropes, but why they are where they are and parallel to the large pipe is not obvious. On the far right on the ground are two unused atmospheric pipes. In the shed on the right seem to be a piston carriage and a freight wagon. Closer,we see a signal, a telegraph post (wires not shown), a policeman's booth, and a tall signal post like those seen previously.

Below is a recent picture of the station looking in the opposite direction. In general, Plymouth trains are on platforms 2 and 3, and Paignton branch trains on 1, but switches on each side of the station create a very flexible arrangement in which any train can use any platform. For example trains from the Paignton branch to distant points beyond Exeter often come in on 3.

Newton Abbot stationt
— Newton Abbot station, looking north, July 2009. Mattbuck, CC BY-SA 3.0

The pre-atmospheric South Devon Railway

On the Teignmouth page I recorded the opening of the railway to that point on 30 May 1846 with locomotives, and from there to Newton on 31 December and Totnes on 20 July 1847. Construction to Newton was delayed by a change of route requested by Newton people in 1845 and not approved by an Act until August 1846. The object was to bring the station closer to the town center by having the railway cross the Teign farther west and locating the station farther north than originally planned to a point from which a new street, now Queen Street, could be opened. Construction west of Newton, intended for atmospheric operation, had begun in 1845 and was completed to a temporary terminus at Laira Green, on the east side of Plymouth, on 5 May 1848.

The atmospheric South Devon Railway

Atmospheric operation to Newton began on 17 December 1847. Trains continuing to Totnes, open 20 July that year, had to change power at Newton, but since Newton was also the passing point for up and down trains, there was necessarily a longer stop at Newton anyway. From 23 February 1848 almost all trains ran atmospherically as far as Newton, the exceptions being the earliest and latest each way.

The South Devon was the only railway to run freight trains atmospherically. In September 1847 such a train was recorded with a weight of 120 tons passing in 15 minutes between Exeter and Starcross, roughly 32 miles per hour. Earlier in the year a "lightweight" train was reported as making 70 miles per hour. The usual passenger train speed was about 40.

Sekon quotes a traveller: "We tested the steadiness of the piston carriage when the train was at its greatest speed, by placing a half-penny on a narrow ledge of the sliding sash, and for a distance of 3 miles it never shook or moved from its position." Many other accounts mention the smoothness of atmospheric propulsion.

The end of the atmospheric South Devon Railway

After about six months of good service, problems with the atmospheric system began to appear. For a time it was not so much how often it failed, but that when it did, locomotives had to be sent to rescue the trains, resulting in hours of delay. Meanwhile locomotive trains to Totnes had trouble on the 2½ per cent grades that had been constructed with atmospheric operation in mind. Brunel's dream was in trouble.

The leather of the linear valve began to break alarmingly by May 1848. At first it cracked, causing leakage, and then it tore. This had not happened on the Dalkey or Saint-Germain lines, opened earlier and still in operation. The explanation from Samuda, the supplier, was that the valves had been kept too long in the packing crates, from one to two years. But another explanation was the salt air that wet the valve despite Brunel's prediction that sea water would not top his sea walls. Hadfield thought that it was the iron, "in contact with saturated leather on one side and air on the other", rusting around the bolts and breaking off. Samuda's answer to this was that the leather had not been regularly greased, a practice known to have been done faithfully on the Saint-Germain line. At any rate by June two miles of valve were in need of replacement.

Brunel was asked in June to report to the SDR Board, but he did not do so until August, when he proposed replacing all of the valve between Exeter and Newton, if Samuda would do so. He no longer proposed atmospheric operation beyond Newton, the very portion where he had designed the steep grades that challenged locomotives. Samuda would not agree to replace all of the valve and operate the atmospheric section at his own expense. The situation deteriorated, and the Board voted against continuing. With a few weeks' notice, the last atmospheric train arrived at Exeter the night of 9/10 September 1848. It was all over.

There were a couple of other factors at play when the decision was made at the Board's August meeting. Although a telegraph was installed along the railway, Brunel had instruments placed only in stations, and yet in August Brunel blamed excessive fuel consumption on not having instruments in engine houses to alert them to train movements— as if he had no part in creating the problem. A second factor was a misleading report to the Board of an operating loss of £2,489 for the first half of 1848, which did not include a large pending payment from the Post Office that made it correctly a £2,009 profit. The second half of 1848 was even better, an unprecedented £18,217 profit boosted by running to Plymouth.

To put the financials into perspective, Sekon quotes a shareholder at the meeting who stated, "On the system which is now to be abandoned, the company have already expended from £300,000 to £400,000. Mr Brunel had received warning upon warning from engineers as eminent as himself, but had recklessly entered upon this extravagant expediture, not at his own expense, but that of others." Sekon himself puts the investment at "over £300,000" and the income from sales of the disused equipment and buildings as "about £50,000", a horrifying loss for what was a relatively small railway company. Peter Kay puts the expense at £430,000 and sales around £85,000, which is even worse. Construction already in hand was brought to completion, namely the branch to Torquay (Torre station today) on 18 December 1848 and the final few miles to Millbay terminus in Plymouth on 2 April 1849. Nothing else was added for ten years, a period in which the South Devon Railway company survived only by reducing train services and postponing capital expenses.


The dates I give at the top are when I changed trains at Newton Abbot or changed between bus and train. I never took a photograph there, in the belief that nothing remained from atmospheric days.

In relatively modern times the station had more than three platforms. Platform 4 made it a four-track station with two island platforms. The fourth track was removed in 1987 with the happy result that northbound platform 3 can be reached on the same level as the station entrance. Until 1959 passenger trains from the Moretonhampstead branch line, joining north of Newton Abbot, came in on their own track along the west side of the main line to a dead-end platform 6 on the north side of the station house. The missing platform 5 has inspired the ironic names of Platform 5 Publishing (Peter Kay's book) and the Platform 5 Brewing Company who operate The Railway Brewhouse on Queen Street just north of Newton Abbot station.


From what we learned of Teignton we may believe, tun meaning enclosure or farm, that Newton means a "new farm" or perhaps a "newt farm". The latter is unlikely, since except for the limited market in eyes for witches, the amphibian is of little commercial value. In this case neither is considered correct. Rather the name refers to the New Town of Torre Abbey, a place given the rights to a weekly market in the thirteenth century. The abbey's town was on the east side of the River Lemon, and so it included the area of the station and of the town center a short distance west. On the other side of the Lemon was Teignwick, called Newton Bushel since about the same time, after a local land owner. The abbey was dissolved under Henry VIII, and the two markets were combined in 1633. The South Devon Railway used the name Newton until 1877, and Newton Abbot thereafter, but the two towns were not combined under that name until 1901. The name Bushel persists as a ward of Newton Abbot. I like to imagine Mr Bushel's wife pronouncing the name as Bouchelle.