by A. H. JOHN, B.Sc. (Econ.), Ph.D.


Lecturer in Economic History

The London School of Economics and Political Science

University of London.






Excerpts on tramroad development from pp 144 - 146



In 1750, much of the coal was still brought out of the levels in wheelbarrows, and transported from the colliery to the ports in carts or, where the roads were exceptionally bad, in horse panniers. The first waggon, way was laid down about this time at Neath by Sir Humphrey Mackworth, but his example was not followed by other coal masters for several decades. In 1756, he constructed what was probably the first tramroad in Wales. The whole line ran from the coalface to the quay side, a distance of nearly a mile and a half, 1200 yards of which was underground. At each wall of coal he employed three cutters and four waggoners, who needed "great skill to keep the waggons on the rails through the turnings and windings underground which are so intricate and do so greatly vary in every street of coal that a good waggoner in one street will not serve in another; and the work is so laborious that few men will undertake it."4 Each waggon contained almost a ton of coal. When finally he was imitated by other large coalmasters in the neighbourhood, the waggons appear to be of a smaller size, for in the accounts of the Briton Ferry Collieries in 1775 there are to be found entries which refer to " extra allowances " to the drammar " boys. The employment of juvenile labour would imply a small waggon more suitable to their strength. In 1776, some five years after Richard Reynolds first used iron rails, Lockwood and Morris, coalmasters in the neighbourhood of Swansea, ordered one hundred tons of cast-iron plates, each 4 ft. long and 5 ins. wide, weighing 56 lbs., from the Coalbrookdale Ironworks, and it was stated that " by 1788 there were about 240 tons of cast-iron tramroad plates underground at Landore Colliery."5 It is not known when iron rails appeared in the northeastern part of the coalfield, but the problem of carriage from the levels to the works was partially solved at Cyfarthfa by improvising a canal between the level mouth and the coal bank near the furnace.6 With the increased output of the puddling forges, tramroads above and below ground rapidly became general at most collieries of any size. It was estimated in 1811 that 150 miles of tramroads had been built in Glamorgan, Monmouthshire, and Carmarthenshire, to serve the iron, copper and coal industries. In addition, many miles of tramroad were laid down solely for the purpose of conveying coal by means of a system of inclined planes to the adjacent canals.


It is hardly necessary to emphasise the importance of coal and metallurgical tramroads in the development of railways. As early as 1804, Trevethick had experimented at Merthyr; Gurney's steam carriage was tried at Hirwaun in 1830, and in the same year a rack-and-pinion locomotive built by the Neath Abbey Foundry was tried at Penydarren.7 The same firm supplied Thomas Powell in 1830 with the first locomotive used in hauling coal. This engine, on its first trial, drew 50-1/2 tons of coal from the Blaencyffin Isha Colliery to Newport and the empty waggons back to the colliery, a distance of 30 miles, in twelve hours after being delayed, it was stated, for three hours by horse waggons. It accomplished in one day work which took six horses two days to perform, thus very considerably reducing transport costs.8


4. G. G. Francis: Neatb and its Abbey (Swansea, 1845)

5. W. H. Jones : The Port of Swansea (Carmarthen, 1922.).

6 N.L.W. : Crawshay Deeds, Box 4, No. 2.

7. The Cambrian, 25 February, 1804 and zo March, 1830.

8. The Cambrian, 31 July, 1830.


Excerpts on Iron Making from pp. 153 - 157



In the iron industry the influence of cheap coal and ore was as important as in copper smelting, if less apparent. Unlike the latter industry, however, it was most significant in the postsmelting stage, where it permitted the extensive development of a process in which there occurred considerable wastage. Situated at the head of a valley, where there was sufficient water power to drive the machinery, the blast furnaces were usually built against the face of the hill so that by the use of terraces they could be easily filled.3 This arrangement also enabled the kilns for the calcining or " roasting " of ores to be placed immediately behind, and level with, the tops of the blast furnace. These kilns were made in the form of cylinders, six feet high, set upon inverted cones about four feet high; the calcined ore was drawn through a door set in the side of the cylindrical part of the kiln. On the same terrace, too, were the heaps of limestone and coke. At a lower level, and in front of the furnaces, were the remainder of the works, the refineries, forges and mills.


Much had been accomplished in increasing the size of the furnaces since the beginning of the century: an expanding market had enabled the economies of large production to be exploited, as they were further to be exploited in the decades following 1827. In this respect the iron industry differed from the copper industry where technical considerations prevented much development in the size of the furnace. The expansion was facilitated by a steady improvement in technical knowledge and in blast-furnace practice. The use of large cinders, discovered by Anthony Hill of Plymouth Works in 1811, the introduction of haematitc ores, and the use of the hot blast, contributed in varying degrees to increasing the output of the smelting operation. In addition, Hill improved the iron by remedying its tendency towards being " red shot."


In 1800, most of the furnaces in Glamorgan were turning out 1100 - 1250 tons of pig iron a year, although some exceptionally large ones erected at Cyfarthfa Works in 1794-5 had a capacity of 1800 - 2200 tons. This latter figure became a common make for the coke furnaces in the north-cast of the coalfield in the first decades of the next century, though the output of the furnaces in the western part of Glamorgan was probably smaller. In 1813, the furnaces at Cyfarthfa were yielding 60-70 tons of pig-iron weekly,4 a make which by 1820 was fairly common in most ironworks in the neighbourhood ; and the 8o tons obtained at the Dowlais and Plymouth Works in 1827 became the average blast furnace yield in the two following decades.5 At Dowlais in 1844, the fifteen furnaces averaged 801 tons a week for the 52 weeks and, in the boom of the following year, the weekly output was pushed up to 94-2/5 tons. At the Rhymney Ironworks, the weekly average yield for the 52 weeks in 1846 was 93-2/5 tons and in the following year 106-2/5 tons."5 Thus at the middle of the century furnaces on the bituminous coals in the eastern parts of the coalfield were producing ninety to a hundred tons of pig-iron weekly, almost twice the output of the blast-furnaces situated further west which used anthracite or semi-anthracite coal. The double impetus of competition from the Scottish iron industry and the size of the market for railway iron had induced, and continued to induce, the iron masters to increase the size of their furnaces, and in 1857 so great had been the advance that the President of the South Wales Institute of Civil Engineers was able to state " not so many years since 120 tons a week was considered a good work for a blast furnace, now 200 tons is a common make, and we are looking forward to 300 tons as the next step in this direction.6


Apart from the economic advantages of increasing the size of the blast-furnace, there were also several technical reasons why such furnaces were larger in South Wales than in the Midlands. There was first the difference of products: for as most of the iron in the Midlands was destined for the foundry, a higher quality pig-iron was necessary, and this could be obtained only from smaller and more manageable furnaces. In South Wales on the other hand, as most of the metal was used in making bars and rails, the pig-iron underwent a refining process in which there was considerable loss. It was, therefore, desirable to obtain as large a smelting product as possible. Secondly, the development of large furnaces in the Midlands was hampered because Staffordshire coal produced a soft coke which crushed under a heavy burden, and thus obstructed the passage of air through the furnace. To some extent this was obviated by obtaining hard coke from other districts, but nevertheless the extra cost involved acted as a deterrent in the building of such furnaces.7


Side by side with the increase in the size of the blast-furnace went a great reduction in the amount of coal used to produce a ton of iron. At the Dowlais Works in 1791, for example, some eight tons were required for this purpose, but forty years later the equivalent amount was three and a quarter tons, including coal for the engines, the lime and ore kilns.8 Efforts were also made in other directions to reduce costs by fuel economy. At the end of the thirties a number of ironmasters in the neighbourhood of Merthyr, especially those using semi-bituminous coal, found that by enlarging the "throat" of their furnaces they could use raw coal, or half coal and half coke.9 Experiments were also carried out at the Ebbw Vale, Abersychan and Aberdare Works in the use of blast-furnace gases for heating purposes, especially in connection with the boilers of the blowing engines, but these were eventually abandoned when it was found that the saving of fuel was not substantial.10


Cheap coal, and coke which even with the wasteful methods of coking represented 70% of the original product, tended to prevent the rapid development of hot blast. There existed a strong opinion, too, that for heavy purposes hot-blast iron was too soft and tender to bear great strain. Only twenty of the furnaces in blast in South Wales in 1839, approximately one-sixth of those situated on the bituminous coals, had adopted this invention. Under the pressure of Scottish competition, the numbers increased considerably by the middle of the century though the belief in cold-blast iron was still largely unshaken.


Hot blast achieved its most interesting success, however, when Crane at the Ystalyfera Works, the last in a long line of experimenters, succeeded in smelting iron with an anthracite coal. Great expectations were immediately entertained - "never .... since the first introduction of pit coal for smelting iron ..... did iron manufacture present so interesting an aspect as the present time, never was there a fairer field for exertion nor a more powerful stimulus to render it effective.11 They were not fulfilled. The iron was found to be extremely strong but not malleable, and the prospects of this new branch of the iron trade immediately declined. A number of works were established, but in 1849 their total output was 46,704 tons and five years later 54,500 tons, not equal in either year to the output of the Dowlais Works alone.12


Most of the pig-iron produced on the coalfield was made into rails and bars of various kinds, by first refining and then puddling and rolling. The earlier method was the slow, laborious refining of iron under hammers and its subsequent drawing into bar by the hammermen. If the continuity of technical development needs emphasising, it can be shown that in at least one case, at the Clydach Works, puddling, as late as 1805, was sandwiched between the finery and the old type of hammer forge. Here, pig-iron was refined, puddled, shingled into half blooms and made into bars by the chafery and hammermen.13 But at Merthyr Tydfil it is clear that the full process of puddling and rolling was in existence in 1790 and had probably completely superseded the old forge operations from the beginning.


The preliminary refining of iron without which, it was alleged, Cort's process was of little value, was added by Horinfray of Penydarren Works. This addition was effected soon after the establishment of the first forge in South Wales in 1787, for it was in use at Cyfarthfa in 1790.14 In the puddling process there was only one other invention of importance, the substitution of in iron floor in the puddling furnace in place of a sand one as originally designed by Cort. This was invented in 1818 by the manager of Nantyglo Ironworks, and as a result, the output of each furnace was more than doubled and the loss in the process considerably reduced.15 Their inability to improve or replace the puddling process, which even after the introduction of iron bottoms accounted for 25% of the cost of bar iron, remained a constant source of annoyance to the ironmasters.


3 Encyclopaedia Britannica, 7th Edition, Vol. X11, P-443

4. N.L.W. : Crawshay Papers, Letter book 18 13-17, 3 0 January, 1817.

5. Bute Estate Archives : Dowlais Papers.

6. Menelaus : Transactions of South Wales Institute of Engineers, 1857, p.6. The average yield of the furnaces at Dowlais in the year ending March 18 5 7, was upwards of 212-1/2tons per week (each ton 21 cwt. of 120 lbs.).

7. Victoria County History of Worcester, Vol. II, p.270.

8.W.Truran: The Iron Manufactures of Great Britain, p.169 (London, 1855.)

9. OP. Cit: p.113.

10. OP. Cit: p.89-95

11. Mining Journal, 19 August, 1837.

12. Symons : Industrial Capacities of South Wales, in the Cambrian Journal, Vol. I and 11.

13. N.L.W. John Lloyd MSS. 5 3.

14 N.L.W. Cyfarthfa Works Ledgers, 1791-8.

15. J. Percy Iron and Steel, p.252. (London, 1869).


Go to the Merthyr Tydfil page


About The Hopkin Thomas Project