A Welsh Ironworks at the Close of the Seventeenth Century


L J Williams, National Library of Wales journal. 1960, Summer Volume XI/3

Extracted onto the pages of GENUKI with the kind permission of the National Library of Wales

This is a complete extract of this article plus appendix (Gareth Hicks Feb 2003)

The places featured are Tredegar, Machen and Caerphilly

See also footnote dated June 2011

The industrial history of Wales before the mid-eighteenth century still remains largely unchartered ground. Our knowledge of the iron industry, for example, is frequently confined to the names and locations of a number of the forges and furnaces; some more or less precise dating of their period of activity; and occasionally one or two names of the people who were connected with them. A series of accounts amongst the Tredegar papers, 1  however, makes it possible to reconstruct in a little more detail the operations of one enterprise in the Welsh iron industry at the close of the seventeenth century. Unfortunately no letters and virtually no other documentary material have survived to elucidate and expand upon the accounts : nonetheless the essential features of the business emerge fairly clearly.

The business consisted of a furnace at Caerphilly and two forges, one at Machen and the other at Rudry. The accounts cover the years from 1690 to 1701 but the furnace and forges had then already been in existence for some time. The forge at Rhyd-y-gwern in the Parish of Machen had, indeed, been established over a century before 1690, being first referred to in 1567. 2  Amongst the Tredegar papers there is a document, 3 which is undated but which can---from internal evidence---be placed at shortly before 1676, that gives further evidence of the earlier existence of these undertakings. The furnace was then in poor condition: the walls were 'indifferent good', the shaft was 'decaying but may hold for a blast more', whilst 'the stream that belongs unto it is not a third part of that in former tyme'. The furnace had been held by one Hart, 4  at first for 15 years at £5 p.a. and then for a further 4 years at £30 p.a. --- the increase arising because at the end of the first term he had still had considerable unused stocks at the furnace so that he had either to sacrifice these or take a new lease. This furnace was almost certainly the old Taff furnace--- it is described as being 'upon the brink of the River Tave'--- which had been working in 1564. About 1680, however, this site was abandoned and the furnace moved to Caerphilly: 5  it is this Caerphilly furnace which figures in the Tredegar accounts. The two forges--- one in Machen and the other in Rudry--- are described as being in good condition. The former commanded (c. 1676) a rent of £50 p.a., the latter £10 p.a., whilst a storehouse at Newport was also rented for £8 p.a.

The partnership which was established in 1690 was, then, formed to conduct an undertaking which had already had a considerable history. The partners were John Morgan of Tredegar, Roger Williams and Roger Powell. Of the total capital ................

  • 1. Tredegar MSS. 76/1-25.
  • 2. H. R. Schubert, History of the British Iron and Steel Industry (London, 1957), p. 298.
  • 3. Tredegar MSS. 76/1.
  • 4. Presumably George Hart, merchant, of Bristol who in 1677 was granted a 21 year lease of coal mines and slate quarries in Senghenydd, Rudry and Whitchurch. C. Wilkins, History of the Iron, Steel and Tinplate Trades, (Merthyr, 1903), p. 24.
  • 5. Hist. MSS. Comm. Lord De L'Isle and Dudley, vol. I, p. 318; Schubert, op. cit., pp. 176, 389.

............................ of £3,000 John Morgan was to hold two-thirds--- conforming to a common pattern of the time whereby a large part of industrial capital was frequently provided by local landowners--- and the others between them were responsible for the remaining third. John Morgan paid his share by providing the money to purchase the stock which already existed at the furnace and forges. This consisted of pig iron, hammers and anvils valued at £6 5. 0d. a ton; broken iron and old plates valued at £5 a ton; wood at 2s. 6d. a cord, plus 1s. 7d. a cord for wood already cut; charcoal at 23s. a load; braises (i.e. small or broken charcoal) at 20s. a cord; and such equipment as helms, arms, rabbets hammer beams and bellows' woods. The stock at the Tredegar (or Rudry) forge was bought for £972 8s. 9d.; that at Machen forge for £1,075 12s. 1 1/2d. ; and at the furnace for £488 7s. 6d. Since their total value was £536 8s. 0d. over John Morgan's share of the capital he was credited with this sum in the first account. Although the other partners provided no actual cash initially, their £1,000 was doubtless intended as the source of working capital for the business. 1

The picture which emerges from the evidence of the way in which the enterprise was organised is somewhat blurred. Roger Powell, who with Roger Williams took up a one-third interest, seems to have played no part at all. 2  John Morgan of Tredegar, the major partner, was more directly involved: it was through him that some of the customers made their payments and he also settled some of the debts. Most of this aspect of the business was, however, conducted through Roger Williams who also regularly attended the twice-yearly Bristol fairs and drew up the accounts. His more active role was recognised by the payment of a fee of £20 a year. 3  The actual management of the works was in the hands of one Nathaniel Tainton until 1694 and thereafter seems to have been shared between Edward Harry and Samuel Pratt.

The general profit record of the business is summarised in the table below (the figures are in £s to the nearest £).


Tredegar forge

Caerphilly furnace and Machen forge


Taken out by partners

Value of Capital plus Undistributed profits

A/c. to Sept. 1690  






do                 1691






do                 1692






do                 1693






do                 1694






do                 1695






do                 1696






do                 1697






do                 1698






do                 1699






do         Sept 1700






do                 1701






  • 1. Tredegar MSS., 76/2.
  • 2. It was, however, his wife, Mary, and their son, Roger---describing themselves as Executors of the late Roger Williams and Roger Powell, deceased--- who finally settled the accounts with John Morgan in 1701. Tredegar MSS., 76/24.
  • 3. There is, however, no record of his being paid this fee in the accounts ending Sept. 1962 and Sept. 1693. Tredegar MSS., 76/8 and 76/11.

This table calls for a few comments. In the first place the form of the accounts prevents the separation of the profit made by the production of pig iron at the furnace from that made by bar iron manufacture at the forges. This is because, in the words of the first account, 'the disbursements at Machen forges and the furnace are too intermixt that they cannot be perfectly distinguisht to make a particular Accompt of each by itself. 1 There is, however, a separate account for the Tredegar forge. Secondly, it is clear that the partners did not regularly each year divide out between them some part of the profits which had been earned. It is, however, unsafe to conclude from this that they were anxious to plough back profits into the business. It seems more likely--- especially as neither the output nor the capital value of the equipment showed any tendency to increase--- that their forbearance arose more simply from the fact that there was no cash in the business upon which they could draw. In 1697, 1698 and 1699 the separate individual accounts for Roger Williams and John Morgan show that they had each paid out more money on behalf of the business than they had received : the book profits of the business derived from the substantial debts owed by their customers. In effect, the partners were lending money to the business during these years--- and it is perhaps a reflection of their comparatively unbusinesslike approach that nowhere do they credit themselves with interest for these loans.

Finally, the last column, showing the cumulative total capital of the enterprise by adding the amount of undistributed profit to the initial capital, is to some extent notional in character. That is to say the book profits which are recorded for the furnace and forges are partly dependent upon, for example, the valuation of the stock at the end of each year and on the extent to which the sums which were owed to the partners turned out to be bad debts. Thus in one year, 1698, when Roger Williams estimated the 'clear' value of the stock at the furnace and forges he put it at £5,338 compared with the £6,850 entered in the accounts under this head. 2   Something of this nature may lie behind the substantial fall in the value of the capital between 1700 and 1701. It is difficult to be sure because the account for 1701--- apparently made up to dissolve the partnership after the death of Roger Williams during that year--- is on a different basis from the earlier accounts. The 1701 account simply adds up the total sums owing to the firm together with a ......................

  • 1. Tredegar MSS., 76/2.
  • 2. Tredegar MSS., 76/16.

......................... valuation of the equipment at the forges and furnaces, deducts from this the payments made during the course of the year, and calls the residue the 'ballance to be divided'. 1 There is thus no rough profit and loss account comparable to those for previous years yet it seems highly unlikely that the difference between the capital value in 1700 and 1701 is to be explained in terms of so large a loss made during that year's current working.

A more likely explanation is perhaps involved in the somewhat obscure relationship which existed between the Tredegar enterprise and that at Neath. In the payments made by Roger Williams from 1694 onwards there regularly occurs the item 'Disburst at Neath'. In so far as the partnership was buying pig iron from Neath furnace this calls for no comment : but the payments made greatly exceed the debt thus incurred. Between 1694 and 1701 the amounts paid out at Neath furnace (£8,823) exceeded the value of pig iron bought from this source (£5,647) by £3,176. There is no indication of any other debts incurred at Neath: the partners thus seem to be investing directly in the Neath enterprise. There is, however, no explicit indication that this was the case---although when Roger Williams sat down in 1698 to work out the 'clear value of the stock' he included £600 for the value of stock at Neath  'after charge of building is taken out which was above £600'. 2 But the account of 1701, compiled after the death of the two minor partners---Roger Williams and Roger Powell---to assess the value of the enterprise, makes no mention of any assets at Neath. It may be that it is the absence of this item which accounts for the fall in total capital value between 1700 and 1701.

What is certain, however, is that the profits actually drawn out by the partners up to 1700 totalled £4,928, amounting to an average yield of 16.4 per cent. on their initial capital. Moreover the sum to be divided in 1701---£4,642---represented a further increase of £1,642 (or nearly 55 per cent.) over the original investment. Even though this latter figure was dependent upon how much of the outstanding debts of £1,817 was collected, and even though this way of expressing the matter somewhat overstates the return, it is clear that the Welsh iron industry at this time could yield a healthy return.

This is roughly confirmed by the information on prices and costs which the documents provide. In the year ending September 1691 the price received for bar iron 3  was £15 10s. a ton; during the following year the price fell first to £14 15s. and then to £14 8s. before recovering to £14 10s. This recovery continued through 1693, 1694 and 1695 when the price had reached £17 10s. It was then comparatively stable at £17 a ton in 1696 and 1697, falling to £16 by September 1698. Thereafter it was again mainly steady, varying only between £15 10s and .................

  • 1. Tredegar MSS., 76/24.
  • 2. Tredegar MSS., 76/16.
  • 3. The accounts do not state the precise type of bar iron that was produced, except that osmund iron is always designated as such. It seems probable, however, that it was mainly best mill bar which was made from a blend of tough pig with coldshort pig iron. These, at least, were the types of pig iron which were used at the Tredegar and Machen forges.

........................... £16 for the three years 1699-1701. One other grade of iron was produced---osmund iron   1 ---which was made only at the Machen forge. The price of this followed the price movements of merchant bars fairly closely although---as one might expect of a higher quality material produced only to meet a specific order---the fluctuation in the price of osmund iron was less marked, the total range of variation during this decade being from £18 to £19 10s. as compared with £14 10s. to £17 10s. for merchant bars. The documents also include three small scraps of paper---dating about 1699---giving some information on the cost of production per ton of bar iron. These suggest a total cost of production for bar iron of 13  10s. a ton, 2  including carriage, which---on the prices of this decade---would yield an average profit of £2 to £2 10 s. a ton.

There is no direct information about output, but it is possible to extract from the accounts figures for the number of tons of bar iron sold each year and this is sufficient to give a good indication of the scale of operations. 3  The forges were each producing an average of between 150 and 200 tons a year which, approximating to the contemporary forge output in Staffordshire and South Yorkshire, seems to have been about the normal forge output of the period. 4  The output of the forges found its market almost entirely outside the district, local sales being insignificant. The focal point of the sales process was Bristol, the twice-yearly fairs of that city providing the occasion for making new sales and collecting old debts. The costs---of getting the iron weighed, providing gratuities for the weigher and boatsman, and Roger Williams's expenses---incurred by Roger Williams in attending these fairs occur regularly in the accounts.

The ultimate markets, however, frequently lay beyond Bristol. All the osmund iron, for example, was sold to Thomas Foley, and since this quality of iron was particularly suitable for wire-making it was doubtless used in Foley's wire works at Tintern. 5  In addition much of the bar iron went to feed the growing finished iron trade of the Midlands. Prominent amongst the customers for bar iron were, for example, Samuel Banner of Birmingham, and Thomas Foley, who at this time sat at the centre of an intricate network of ironworks in the Forest of Dean, ......................

  • 1. Osmund iron, originally obtained from Sweden, was of a particularly high quality and combined tenacity with a high degree of ductility.
  • 2. These estimates were based on the use of 2 1/2 loads of charcoal (at about 24s. a load) and 26 cwt. of pig iron (at a little over £6 a ton) to produce each ton of bar iron. The estimate of cost was: 'Cole for a Ton £3; pig iron £8; Makeing £1; Petty charges £1; Carriage 10s; Total £13 10s.
  • 3. See Appendix I
  • 4. B L C Johnson, 'The Foley Partnerships; The Iron Industry at the end of the Charcoal Era', Economic History Review, n.s. vol 4, 1951-2, p339; A Raistrick and E Allen, 'The South Yorkshire Ironmasters (1690-1750)'Economic Hist Rev, vol 9, 1938-9, p184
  • 5. Johnson, op. cit., p333

........................... Staffordshire (especially the Stour valley) and north-east Worcestershire. 1 The Tredegar enterprise thus fitted naturally into the existing pattern of an important section of the contemporary iron trade which rested upon the blending of tough pig iron---which the Forest of Dean furnaces characteristically produced in excess of the requirements of their own forges---with an inferior coldshort iron. This blending was normally performed in the forges of the Stour valley but---especially as the Stour valley furnaces did not produce sufficient coldshort iron---there was an obvious opening for similar activity in South Wales. 2

That this was the basis of trade from the Tredegar and Machen forges receives some confirmation from an examination of the sources of the pig iron used at these works. In the ten years up to 1700, for example, Tredegar forge used 2,560 tons of pig iron, of which a little less than one-half (1,100 tons) was imported from Tintern and Flaxley furnaces in the Forest of Dean. The remainder was Welsh pig iron most of which (916 tons) came from their own furnace at Caerphilly, but a further 534 tons were also drawn from Neath furnace. The presumption that the forge was mainly occupied in blending these different types of pig iron is, to some extent, fortified by a consideration of the prices of the pig iron. The price of Forest of Dean pig iron at between £6 2s. 6d. and £6  10s. a ton was consistently cheaper than lower quality pig obtained from Neath at £6 10s. a ton; the regular purchases which were nonetheless made from Neath perhaps derived from the need for this type of pig iron for blending purposes. Too much weight should not, however, be placed on this last point since, as we have seen, there is some obscurity about the relationship between the Tredegar concern and that at Neath.

Some of the themes which these documents illustrate are familiar enough:  the part played by landowners, like the Morgans of Tredegar, in providing capital during the early years of industrial development; the instability of enterprises founded on a partnership basis; and the central role of Bristol as a market stimulating activity in Wales. More important, however, is the extent to which they suggest that the industrial prospects of the region were becoming more attractive. There is a tendency to think of the growth of the iron industry in South Wales towards the end of the charcoal era as developing only because of the exhaustion of other areas, particularly Sussex: being, that is, more or less reluctantly pushed towards South Wales. This short paper is based on the experience of only one firm over a comparatively short period and is thus inadequate to bear the weight of any broad generalisations. So far as it goes, however, it suggests that the iron industry in South Wales could be reasonably profitable and was suitably located to provide the type of iron required by the growing Midland market : it suggests that the iron industry was being strongly pulled, as well as pushed, towards South Wales.



University College of Wales, Aberystwyth.


  • 1. Tredegar MSS., 76/23: Johnson, op. cit., p. 322.
  • 2. Johnson, op. cit., pp. 331-4.



   (figures to nearest ton)

                 Tredegar        Machen           Total

1691             117                173              290

1692             227                162              389

1693             227                180              407

1694             208                212              420

1695             157                174              331

1696            183                 184              367

1697            180                 106              286

1698             54                    87              141

1699           167                  201              368

1700            227                 133              360

1701              -                      -                283


This is a useful article, I have carried on a half hearted study of the Caerphilly Furnace for many years, but H P Richards and William Rees have mined most of the sources.
I think John Williams may have confused the forges. There were three, two at Machen and one at Bassaleg in Tredegar Park. I think the the latter is the Tredegar Forge, not Rudry as he states. The two Machen forges cause confusion, neither were in Machen, on the east side of the river Rhymney, both were on the west bank in Glamorgan, one was in Rudry Parish and one in the Hamlet of Rhyd y Gwern, in the parish of Machen, but in the county of Glamorgan.The matter is confused by the name Ffwrwm Isha, which is now a public house. Machen proper was Lower Machen around the 13th century church. Originally Upper Machen was called the Ffwrwm,(Gwynedd Pierce suggests this name means Forge and may refer to the housing associated with the forge)  and on the 1829 OS Map there is a Ffwrwm Ucha and a Ffwrwm Isha, which may refer to two separate forges, a distance apart,or may merely refer to separate areas of the Ffwrwm district, as in the 1841 Census returns.  

John Owen (June 2011)