The 1838 Gwaun-cae-Gurwen railway : an abandoned feeder to the Swansea canal

By Paul R Reynolds




This article was originally published in the journal of The Railway & Canal Historical Society in March 1998. (ISSN 00338834)
It is reproduced here with the kind permission of the publisher and author who retain copyright - it must not be downloaded and re-published in any format without their express permission.

The village of Gwaun-cae-Gurwen lies at the head of the Amman valley in the extreme northwestern corner of the county of West Glamorgan. It is 14 miles from Swansea and 19 miles from Llanelli, and on the northern crop of the coalfield which at this point is anthracite in quality. It was the existence of coal which led to the development of the village in the first place in an area of rather inhospitable moorland and hill pasture, and it is coal which still provides most of the employment in this area. There is a large opencast site on the northern side of the village, while to the south is the modern deep-level colliery of Abernant. A single-track branch off the Central Wales line connects these two traffic sources with the rest of the British Rail network.

There are spasmodic references to coal-working at Gwaun-cae-Gurwen in the eighteenth and early nineteenth centuries, but it never developed to any great extent because of the isolated nature of the area and the distance to the nearest port. As was invariably the case with inland parts of the coalfield, such as this, in the period before the construction of canals or railways, the costs of transport were so great that it was impossible to sell the output at a price which would repay an entrepreneur for his investment. Even after the opening of the Swansea Canal in 1798 had caused a major expansion of mining activity at the head of the Swansea valley and in its tributary, the Twrch valley, Gwaun-cae-Gurwen still lay too far from the canal to be able to take advantage of the new opportunities it afforded. It was only the extension of the Llanelly Railway, authorised in 1835, that opened up this remote area and made the exploitation of its mineral resources an attractive proposition.

The first industrialist to realise the potential of Gwaun-cae-Gurwen now that its isolation was about to be ended was Roger Hopkins. Hopkins was a local man, born in Glamorgan ( Camb. 2.5.1840), and by the time he turned his attention to Gwaun-cae-Gurwen in 1837 an experienced and successful engineer. His earliest recorded appointment was as engineer to the Monmouth Railway in 1810   1. The following year he was seconded by them to the Severn & Wye Railway 2, and subsequently he established himself in the west of England. There he designed and constructed the Teignmouth bridge which, when it was opened in 1827, was claimed to be the longest in Britain 1. In 1831 he surveyed the Bodmin & Wadebridge Railway which was built under his direction in 1832-34. The rails for this line came from Ebbw Vale, which is a pointer to the next stage in Hopkins' career, as manager of the Victoria Iron & Coal Company which Hopkins established in 1836 and which was based at Bath.

Shortly after he set up the Monmouthshire Company, Hopkins also created the Swansea & Gwaun-cae-Gurwen Anthracite Company to lease and work the coal at Gwaun-cae-Gurwen. Like the iron company, it was based at Bath and the same man, J.J. Skinner, was secretary to both companies. The first half-yearly meeting was held at Bath on 24 April 1838 ( Camb. 5.5.1838) which puts its formation in 1837. It was in 1837 that the company obtained a mineral lease at Gwaun-cae-Gurwen. Altogether 717 acres of coal and iron ore were leased from Capel Hanbury Leigh (1776-1853) of Pontypool. He had acquired the property through his marriage in 1797 to Molly Mackworth, the widow of the last survivor of this family of landowners and industrialists of Neath.

There were two parcels of land, Gwaun-cae-Gurwen itself (680 acres) and the adjacent but much smaller Tyr Nant Hyr (37 acres). The lease was for 99 years from 25 May 1837 at a rent of 7d per ton, subject to a minimum of 583 6s 8d per annum. This was the equivalent of an annual output of 20,000 tons, a fairly modest amount   3.

GCG tramway map

In 1838 work started on the construction of a railway from the colliery down to the Swansea Canal. It was constructed under the powers conveyed by the 8-mile clause in that company's Act of Parliament. An advertisement seeking tenders for the first -and as it was to turn out, only - section of the line appeared in The Cambrian (17.3.1838):-

To Excavators and Others

All Persons desirous of TENDERING for the EXCAVATION and FORMATION of the CUTTINGS and EMBANKMENTS and other WORKS required in the construction of the first mile and a quarter of the RAILWAY intended to be made by the Swansea and Gwaun-caeGurwen Anthracite Company, from Gwaun-cae-Gurwen to the Swansea Canal, are requested to attend on the spot at twelve o'clock on Monday and Tuesday, the 19th and 20th instant, when the Works to be performed will be pointed out, and the Specification thereof may be seen.

The Company will provide such Wheelbarrows, Planks, Rails, and Waggons, as may be required for the execution of the Works.

The Tenders are to be delivered to Messrs. Roger Hopkins and Sons, at the house of Mr. Samuel Morgan, adjoining Gwaun-cae-Gurwen, before eleven o'clock on Thursday morning, the 22d inst., when the parties are expected to be in attendance.

J. J. SKINNER, Secretary

Office, 3, Harington-place, Bath, March 10, 1838.

At first it might appear rather strange that Hopkins should have chosen to construct a railway feeder to the Swansea Canal when steps were already being taken to carry out the powers of the 1835 Llanelly Railway Act. This Act had authorised the extension into the heart of the coalfield of what had previously been a purely local system serving Llanelli docks. Tenders were invited in August 1837 for the extension of the railway to within a mere two miles of Gwaun-cae-Gurwen at Garnant ( Camb. 26.8.1837). Since Hopkins did not put his Swansea Canal link out to tender until the following March he cannot have been acting in ignorance of the plans of the Llanelly Railway. He must have known what its implications for his business were likely to be, yet even so, he still set to work on the line down to the canal. What were the reasons that led him to embark on what, with a locomotive railway in the building from a good modern dock or to a mile or two from his colliery, was an anachronistic and troublesome mode of transport?

There were two inherent flaws in this system, both of which would have been absent had Hopkins decided to trade exclusively on the Llanelly Railway. In the first place a railway to the Swansea Canal would involve trans-shipment of the coal between the two modes at Pontardawe which would add to the cost of transport and cause damage to the coal and so reduce its value. Secondly a major drawback of the Swansea Canal - or any other canal - was its tendency to freeze over in a hard winter, thus bringing traffic to a standstill. This was a complaint made about the Swansea Canal so long as it remained the main form of transport in the valley. In addition, there was the further drawback in the canal link project in that connecting the new colliery to the Llanelly Railway involved less than two miles of new track whereas the line to the canal would be six miles long and so cost about three times as much.

The plans of the Llanelly Railway were public knowledge, and the flaws in the railway/canal proposal so obvious that an experienced engineer like Hopkins must have known exactly what he was doing in even considering a railway feeding traffic onto the canal. The one great advantage of this idea was that it gave him access to Swansea rather than to the much smaller port of Llanelli. At this time Swansea was still just about the largest coal-shipping port in south Wales (but with Newport on the point of overtaking it). In 1837 a total of 526,961 tons of coal, culm, etc., had passed through the port, about twice as much as through Llanelli and this must have led to a more buoyant market and better prices at Swansea. On the other hand the shipping facilities at Llanelli were better than Swansea. At Swansea coal was still shipped from wharves and staithes along the river whereas at Llanelli there was a public floating dock, the first in Wales, opened in 1834. Against this, however, the approaches to Llanelli were difficult, since the Burry Estuary was subject to shifting sand banks. A further point in favour of Swansea was the distance from Gwaun-cae-Gurwen, about five miles less than the distance from Llanelli to Gwaun-cae-Gurwen.

Swansea was probably just too important in the coal trade for Hopkins to be able to ignore it, despite the less than ideal means he would have to resort to to get his coal there. He did not want to be bound to the smaller port of Llanelli but wanted the freedom to send his coal out through either Llanelli or Swansea. This flexibility would enable him to switch his traffic from one port to the other so as to obtain the best price on offer and avoid any delays that might occur if he were dependent on a single port. Both ports had factors in their favour, and this was brought out by the catalogue of the colliery when it was offered for sale in 1845: the existing Llanelly Railway and the canal feeder, when completed, would between them "connect the colliery with the best shipping ports in Glamorganshire and Camarthenshire" 3. Since Hopkins was not having to pay for the construction of the Llanelly Railway he may have felt better able to spend money on the line to Swansea which he hoped would improve his marketing capacity.

Work on the railway must have started fairly soon after the contract was let, although the identity of the successful contractor is not known. The line started at Hopkins' shaft at Gwaun-cae-Gurwen (later known as the Old Pit) and followed a generally SSW direction towards Pontardawe. The earthworks of the first 1.25 mile contract were finished but no further tenders were sought for the remainder of the line, nor is there any reason to suppose that permanent way was laid on the first stretch. Altogether about 4000 were spent on the works that were completed 3. Part of what was achieved in 1838/40 still survives, including an impressive embankment and cuttings. The northernmost part of the works, however, has been removed by land reclamation and at the southern end it is no longer possible to trace the line, since the works involved in its construction were much less spectacular and have merged back into their surroundings.

By 1840 Hopkins must have realised that his interests would be better served by a short line linking his colliery to the Llanelly Railway rather than by continuing with the Swansea Canal link. The canal line was not positively abandoned but it was deferred in favour of the shorter and cheaper Llanelly Railway connection. Even as late as 1844 in the sale catalogue of that year its completion was still envisaged:  "... a Branch Railway has been commenced from the Colliery at a cost of 4000, which, from various facilities, may be completed at a moderate outlay; and a junction made with the Swansea Canal, at the distance of 6 miles only " 3. However, on 10 March 1840 the Llanelly Railway's Garnant branch had opened, terminating less than two miles from Hopkins' colliery. Llanelli might be second best to Swansea as a shipping port but the arrival of a railway straight from the dockside to a point so close to the colliery was a powerful argument. In June 1840 Hopkins was awarded the contract for the short two-mile extension from Garnant for 4400   4 and by October he was obviously well at work, for in that month the Coal Company offered ten draft horses for sale, since "a railway now building makes them unnecessary" ( Camb. 3.10.1840). The line was formally opened on 6 May 1841. The only feature of any significance, in terms of engineering, was an incline up from the valley floor at Garnant.

Meanwhile, as the reference to draft horses shows, work had been going ahead on sinking and by now the colliery was producing. Coal was obviously being taken out by horse and cart to the railhead at Garnant. Sinking is said to have started in 1837 5 and in January 1840 the Great Vein, 5 ft 3 ins thick, was reached at 87 fathoms ( Camb. 25.1.1840), which made it the deepest pit in the anthracite belt. The shaft was oval in section, 16 ft by 8 ft, and bratticed to divide it into upcast and downcast. There were two engines and ventilation was by a surface furnace 3,5. In addition to the shaft at Gwaun-cae-Gurwen, a second shaft was started a quarter of a mile to the east on Tyr Nant Hir. By 1844 it had been sunk to 30 yards   3 and the Llanelly Railway branch was extended to it, but it never seems to have developed further.

Proving the Great Vein improved the prospects of the business and in March 1840 reserved shares were being offered at a 5 premium ( Camb. 7.3.1840) - perhaps as a management puff! There was also talk of Hopkins starting an ironworks: his retirement as manager of the Victoria works at Ebbw Vale was announced in May 1840. He was said to be returning to Glamorganshire, "the place of his nativity", there to erect an extensive ironworks ( Camb. 2.5.1840). Although Gwaun-cae-Gurwen is not named, it must almost certainly be what was intended. Veins of iron ore were included in the lease and the 1844 sale prospectus emphasises the suitability of Gwaun-cae-Gurwen as the site for an ironworks. It must be remembered that a few years previously, in 1837, a method of using anthracite for iron-smelting had been perfected at the nearby Yniscedwyn ironworks in the Swansea valley, and this led to a proliferation of speculative ironworks on the anthracite coalfield.

By 1842 coal production was up to about 30-40 tons a days but this rate fell short of the 20,000 tons p.a. needed to work out the dead rent. Perhaps it was this disappointing return that led Hopkins to sell out a couple of years later. The colliery was offered for sale by private treaty in 1844 ( Camb. 25.5.1844, etc.) but cannot have found a buyer, for on 8 January 1845 it was sold by auction at Swansea ( 3; Camb. 4.1.1845). The new owners, the Blaencaegurwen Colliery Company, undertook a number of improvements, including the sinking of a second shaft to the north of Hopkins' pit and the re-ordering of the ventilation systems, but they did no further work on the 1838 railway. With the Llanelly Railway running from the pithead to the dockside the day had passed when there was any need to construct a horse-worked railway feeding into the Swansea Canal.


The foregoing differs in several respects from other accounts of Hopkins' venture at Gwaun-cae-Gurwen. A version that has found some circulation is that contained in J.H. Davies, History of Pontardawe and districts which was followed by J.D.H. Thomas in his unpublished MA thesis ' Social and economic developments in the upper Swansea valley' (MA, University of Wales, 1974). Davies gives no source for his version but it appears to derive, in part at least, from a brochure published by the then owners of the colliery in 1927, the Gwauncagurwen Colliery Co., to mark the completion of a new sinking, the Steer Pit. This brochure was probably also the source of three items in the Iron and Coal Trades Review (23.9, 21.10, 28.10.1927). These articles in turn cite no authority, but the assumption must be that they derive from documents then in the possession of the company.

The principal points on which I differ from Davies' account are these:-

1.The start of Hopkins' undertaking. According to Davies (o.c., p. 104), Hopkins' first attempt to win the coal at Gwaun-cae-Gurwen was made in 1832. He sank a pit and started work on his railway to the Swansea Canal, but had to abandon work when he struck water in quantities too great for the pumps to overcome. He then made a second, successful attempt in 1837, again involving the construction of a railway. I am dubious about the historicity of this 1832 venture, which appears to be simply a doublet of 1837. I have not come across any contemporary evidence for work by Hopkins in 1832, whereas, as I have shown above, there are firm dates for his lease in 1837 and for the start of construction work on the railway in 1838. Further, the ICTR articles cited above contain no references to any project in 1832. The supposed 1832 attempt can perhaps be put down to a misreading of his source material on the part of Davies or to confusion on his part between Hopkins' 1837 project and some other short-lived venture involving another entrepreneur in 1832.

2.The winning of the Great Vein. Davies (o.c., p. 44) states that this was in March 1839. This is almost certainly an error. The Cambrian (25.1.1840) has an unambiguous report that "the Great Waynecaegurwen Anthracite or Stone Coal was reached last week". Davies' terminology seems to indicate that his date derives from an advertisement in The Cambrian (7.3.1840) which refers to the "recent" winning of the Great Vein (i.e. in January). This he has mistakenly attributed to 1839, and at the same time he has failed to realise the full force of the word "recent".

3.The seams being worked. Davies (o.c., p. 44) states that sinking was to the "Big or Milford vein". In fact it seems fairly certain that Hopkins only reached the Big (or Great) vein and not the Milford vein which is a totally distinct seam about 30 or 40 fathoms below the Great Vein. This is born out by the paragraph from The Cambrian (25.1.1840) cited above and by information in the advertisement in which the colliery was offered for sale ( Camb. 25.4.1844, etc.). Davies' statement again seems to derive from the advertisement in The Cambrian (7.3.1840) which describes the company as having "just reached the Big or Milford Vein". This wording may have been chosen by the company because they were not quite sure then as to just which vein they had reached: in neighbouring collieries the Great Vein lay much nearer the surface (e.g. 36 yards at Hendreforgan), and they may have wondered whether, at 87 fathoms, they had missed the Great Vein altogether and struck the lower Milford vein.