RAILWAY AND HISTORICAL SOCIETY

EARLY RAILWAY GROUP Occasional Paper 256

[ editor’s note: this paper is in reply to a query from Ian Martin in Circular 37:

“Can anyone identify this rail from the Kidderminster Railway Museum? It has either a W or M cast into it, but there are no further details.” From Chris Lee:

“The rail at Kidderminster Museum is a Jessop edge rail. There is a photo of a similar rail on page 9 with description on page 10 (top) in The Railway, British Since 1804, by Andrew Dow (Pen and Sword, 2014), NRM archive record photo 1892-1071. According to Dow, the “W” indicates a Wide tread. I have no special knowledge of this rail, but yesterday I was trying to reference some track for a different subject (Penydarren) and recalled seeing the photo during my search.”

And see the related Railway & Canal Historical Society, Early Railway Group Occasional Paper [ERG OP] 255, Michael Lewis, ‘The Leicester Navigationʼs Forest Line: a myth debunked’.]

______

Butterley Company Edge Rails: their use at Belvoir Castle and elsewhere

Rowan Patel

The Company of Ripley, , originally named & Co., is well known as a supplier of rails for use in the construction of early iron railways. Butterley began casting rails in the late 18th century and continued to do so until cast-iron rails fell from favour, having been superseded by wrought iron as the material of choice. The company was especially active as a rail supplier from 1794 through until the early 1820s, a large fraction of the period when cast iron was the favoured material for .2 Philip Riden concluded that Butterley ‘occupied an important, possibly uniquely important, position in the railway scene between the introduction of the plate rail and about 1815’.3

1 [According to the NRM catalogue, 1892-107 is, “Two lengths of fish-belly cast iron rail, Loughborough” – and see Lewis, ERG OP 255 - Ed]. 2 P. J. Riden, ‘The and Railway Construction, 1790-1830’, Transport History, vol. 6, no. 1 (Mar. 1973), 30-52. 3 Riden 1973, 49. 1

Butterley was primarily a supplier of plate rails rather than edge rails. Indeed, the company’s founder Benjamin Outram became the major exponent of this type of rail, having adopted plateways with enthusiasm less than five years after they were first used on the surface (they had previously been used below ground).4 Despite its association with plate rails, however, Butterley also cast edge rails, which the company’s management came to see as superior to plate rails in later years. Riden has stated that ‘large quantities of edge rails’ cast by Butterley ‘went to the North East to John Buddle and John Goodchild’.5 Butterley first supplied rails to Buddle in early 1809, for use at Benwell Colliery just west of Newcastle. These rails were fish-bellied, 3 feet in length and with holes at either end. They were also butt-jointed and sat in chairs containing two holes in each cheek, such that the rails could be secured using cross pins.6 Rails of this type were entirely typical of North Eastern practice at this time,7 but were unknown in the vicinity of the Butterley works, for that area was then dominated by the plate rail, as was most of .

In less than 5 years however, Butterley was to cast edge-rails for local use, and these were of the company’s own design. Rails of this unusual pattern were produced by the Butterley Company at various times from 1813 onwards, becoming something of a speciality for the ironworks in the later years of cast-iron rails. These rails were associated with a number of railways engineered by (1781-1826), the son of senior (1745-1814), civil engineer and an early partner in the Butterley Company. The rails in question saw use on a number of railways in the midland counties of England. They are fish-bellied in form and did not use chairs. In all cases, one end of the rail incorporates a rectangular plate – best termed the ‘foot’. This foot contains a single hole, through which the rails were spiked to stone blocks. The other end of these rails terminates with a complex joint, which interlocked with the foot of the next rail, and in so doing, hid from view the spikes which held the rails down.

It is a rail of this type which Ian Martin asked about in Circular 37. Although these rails are all broadly similar, there are notable differences between the rails as used at different locations – forming an interrelated family of rails, which show subtle differences. These differences are distinctive enough, whilst these Butterley edge rails are sufficiently idiosyncratic, that the rail at Kidderminster Railway Museum can be identified with confidence as originating from the Belvoir Castle Railway (Figure 1). Some dimensions of a rail from this line will be provided in this Occasional Paper, and by comparison with these, it should be possible to prove beyond doubt that Belvoir Castle is the source of the Kidderminster rail.

Figure 1: A Butterley Company edge rail cast for the Belvoir Castle line (courtesy of Ian Castledine)

4 Rowan Patel, ‘The early development of the Outram-pattern 1793-1796’, Journal of the Railway & Canal Historical Society, vol. 39, pt. 6 (Nov. 2018), 326-30. 5 Riden 1973, 32, 40-1. 6 Information courtesy of Michael Lewis, citing North of England Institute of Mining and Mechanical Engineers (NEIMME), Buddle vol. 23, fos. 55-835 passim (letters to Butterley) and Buddle vol. 4 (letters from Butterley). 7 M. J. T. Lewis, ‘Bar to fish-belly: the evolution of the cast-iron edge rail’, Early Railways 2 (Newcomen Society, 2003), 105. 2

The variations displayed by these rails, as used at different locations, have not been discussed previously. That being the case, it is worth placing some details on record of all the known examples of this rail type, considering the similarities and differences shown by the known variants, along with the date of each use. Rails of this type are known to have been used at the following five locations:-8

 Belvoir Castle Railway (1813-1815)  Mansfield & Railway (1817-1819)  Portland Railway, an extension of the M&P (1823)  Park, Butterley Co. lines (c. 1825)  & High Peak Railway (1825-1827, completed 1831)

This pattern of fish-bellied rail was first used on the Belvoir Castle Railway, engineered by Josias Jessop and opened in 1815. This line connected the ancestral home of the Dukes of to a wharf on the Canal. The rails were probably designed by William Jessop junior (c. 1783- 1852), who took over management of the Butterley Company following Benjamin Outram’s untimely death in 1805.9 Josias Jessop, as both the line’s engineer and brother of Butterley’s manager, was presumably consulted, and may also have been involved in developing this type of rail, although the relative roles which the two brothers had in designing these rails cannot now be determined.

The rails for the Belvoir Castle line were supplied by the Butterley Company, and information on the rails cast for the line are detailed in Furnace Ledger B, the company’s second-earliest extant ledger. The earliest relevant entry is on 8th December 1813, when the first ‘1257 Gang Rails’ and ‘Four Thousand Rail Blocks’ are listed. Rails continued to be cast throughout 1814, and up until 1st March 1815, when the ledger lists ‘4 Short Gang Rails’ and ‘Gang Nails’.10 In Butterley Company records a ‘Gang Rail’ is the standard terminology for a rail, generally used in reference to plate rails, but here evidently referring to edge rails. The rails listed in the ledger weighed 40 lb each, and this weight is consistent with surviving rails from the Belvoir Castle line.

Butterley did not generally provide stone blocks and wrought iron track spikes, for they simply supplied the rails which they cast.11 However, the company did supply these items for the Belvoir Castle line, since they constructed this railway under contract, rather than simply acting as a rail supplier. Butterley Furnace Ledger B once again provides some details, in the account for 'His Grace the '.12

The details given are:- forming making & completing the Railway from Muston Wharf to Belvoir Inn per contract - £2577 do do do from the Inn to the Castle - £695 do do do from the main road at the Wharf to the Brick Kiln 183 yds long - £198 10s

These particulars, amounting to £3470 10s, appear to cover the full length of the line, which ran from Muston Gorse Wharf on the to Belvoir Castle itself.

8 Lewis 2003, 104-5; Martyn Taylor-Cockayne & Stuart Saint, The Portland Path (The Portland Path Project, 2012), 30, 49. 9 Riden 1973, 31. 10 Derbyshire Record Office (DRO), D503/41/1, fo. 733. Further examination of the Butterley ledgers would doubtless shed additional light on the rails discussed in this Occasional Paper. 11 Riden 1973, 39. 12 DRO, D503/41/1, fo. 109. The list comes to the end of the page in Ledger B, which finishes 'Carried to fo. 92 Ledger C', but that ledger (DRO, D503/29/3) has not been consulted.

3

The rail shown in Circular 37 has all the distinguishing features of a Belvoir Castle rail. It also has an ‘M’ cast into the side, which is known to be the case for some rails from this line, there being surviving examples which bear this letter. Other rails from Belvoir Castle are identical in appearance but are marked with the letter ‘W’ in the same place. A rail on display at the National Railway Museum and attributed to the line bears one of these letters, although unfortunately I cannot remember which. Alongside the rails which have letters cast into the side, a third variant is unmarked and does not bear any letter.13

To return to the previously mentioned rails which Butterley supplied to John Buddle in 1809, these were of two types which varied slightly in their dimensions. Rails for a level were marked ‘L’, whilst those for an incline were marked ‘H’. The purpose of these letters was such that the two types of rail could be easily identified when laying the track.14 It is possible that the letters designating Belvoir Castle rails ‘M’ and ‘W’ served a similar purpose, allowing for rails which differed in some respect to be easily distinguished. If so however, it is strange that letters which are superficially similar, differing only in orientation, were chosen for this purpose.

Alternatively, it has been proposed that the ‘M’ refers to ‘Manners’, this being the family name of the Duke of Rutland, for whom the line was built.15 This is possible, yet family names of dukes not being widely used, it is curious that the rail would allude to ‘Manners’, when even Butterley’s own records referred to the duke as ‘His Grace the Duke of Rutland’. This explanation cannot account for the rails marked ‘W’. If the rails were supposed to be marked ‘M’ in order to distinguish ownership (which generally entailed specifying a name more definitively), then the only explanation for the rails marked ‘W’ could be a mistake made with some of the patterns used to cast the rails, which seem improbable.

Some dimensions for a Belvoir Castle rail are:- 16

 length – 3 feet  top flange width – 1.3125 inches  bottom flange width (in centre) – 2 inches  rail depth at mid-span – 5 inches  rectangular foot at rail-end – 3½ inches x 6¼ inches

Andrew Dow has noted that there are two rails of this type at the National Railway Museum. One of these is said to have a top flange which is 0.75 inches wide. The second example, ‘identical in all other respects’, has a top flange width of 1.625 inches and is marked ‘W’ – a detail which certainly implies a Belvoir Castle origin. Dow speculates that ‘W’ could stand for wide.17 It is notable that the upper flange width of both of these rails, differs significantly from the figure given above. These details were measured from a rail which certainly originates from the Belvoir Castle line and which is marked ‘W’. Any conclusion as to the meaning of these letters, if a conclusion can even be reached, would be assisted by a thorough examination of all surviving rails.18 It is also possible that the Duke of Rutland’s manuscripts at Belvoir Castle would be illuminating in this respect.

13 An example is in the Lewis Collection, Ironbridge Gorge Museums, Accession No. 1983.3133. 14 Information courtesy of Michael Lewis, citing Buddle correspondence at the NEIMME (see note 6). 15 I owe this valuable suggestion to Martyn Taylor-Cockayne. 16 In the forthcoming text measurements of Belvoir Castle and Codnor Park rails are taken from specimens which I have examined, but I have not had the opportunity to inspect M&P or C&HP rails myself. 17 Andrew Dow, The Railway: British Track Since 1804 (Pen & Sword Books Ltd., 2014), 9-10. 18 Charles Hadfield and A. W. Skempton, William Jessop, Engineer (David & Charles, 1979), 289, notes 16 and 17 mention rails with Science Museum Inventory Nos. 1892-108, 1900-218 (ascribed to Belvoir Castle) and 1892- 107 (erroneously ascribed to Loughborough). All of this material was almost certainly donated by Clement 4

Regarding the variant forms of this type of rail, as used at different locations, the exact nature of the rail joint underwent several design changes, with significant differences between the interconnecting components. The Belvoir Castle rails have a substantial bottom flange, and this flange flares outwards at the end of the rail, increasing to a maximum width of 3 inches (Figure 2). The feet of these rails have a corresponding depression, designed to receive the flared end of the rail (Figure 3). This specific joint design is unique to the Belvoir Castle line.

Figures 2 and 3: The ends of a Belvoir Castle rail showing the flared end (left), and the foot (right) within which there is a depression designed to receive this flared end (courtesy of Ian Castledine)

The rails used on the Belvoir Castle Railway demonstrate an understanding of the properties of cast iron, a material which is strong in compression, but weak in tension. When a load is applied to one of these rails, the narrow upper flange is compressed, whilst the bottom flange is placed under tension. That the rails have the widest flange on the underside, shows these rails to have been designed in the knowledge that this provides a strong section, the iron being distributed so as to strengthen the rail where it is weakest.19 The cross section of the Belvoir Castle rails is shown by Figure 4.

Figure 4: The cross section of the Belvoir Castle rails (Lewis 2003, 111)

Stretton, who donated material to the Science Museum in 1892-1894 and 1900-1904, see: Stephen Duffell, ‘Clement E Stretton: Railway engineer, historian and collector’, JRCHS, vol. 35, pt. 3 (Nov. 2005), 167. 19 Lewis 2003, 107-112 discusses the influence of beam theory on cast iron rail design. 5

After being used on the Belvoir Castle Railway, rails of this type were next used on the Mansfield & Pinxton Railway (M&P), which also employed Josias Jessop as its engineer. This line was originally authorised by an Act of 1817 and opened in 1819.20 The rails used were closely related to the Belvoir Castle permanent way.

Riden noted in relation to the permanent way of the M&P that ‘the contract for the rails was split between Butterley, Oakes and several small ironworks, and for no clear reason Butterley refused to execute their share of the job, although they had previously supplied rail patterns to the other ironworks concerned. Only in 1823 did Butterley first sell rails to the company’.21 The circumstances behind this peculiar scenario remain unclear. However, given Butterley supplied patterns to other ironworks, the original design for the rails would seem to lie with Butterley, irrespective of whether they originally cast rails from their patterns.

Like the Belvoir Castle rails, the M&P rails were 3 feet long. The upper surface of the rails was 1½ inches wide, very similar to the width measured from a Belvoir specimen of 1.3125 inches. Whereas the Belvoir rails have a very substantial bottom flange, this is not the case with the M&P rails (cross section in Figure 5). This means that, considered as a beam, the Belvoir rails are technically stronger. In terms of cost, it may have been determined that the M&P rails were strong enough without the large bottom flange, which contained a significant amount of iron. Another notable difference is the design of the rail joint. Whereas the bottom flange of the Belvoir Castle rails flared outwards, the width of the M&P rails remains constant. These rails have a projection at the end, which slots into a corresponding inset behind the rail’s foot. The nature of the joint used on these rails is well illustrated by Figure 5.

Figure 5: The rails used on the M&P (Birks and Coxon 1949, 226)

20 John A. Birks & Peter Coxon, ‘The Mansfield & Pinxton Railway’, The Railway Magazine, vol. 95 (1949), 224-6; see also: John Vanags, The Mansfield and Pinxton Railway (Old Mansfield Society, 2000). 21 Riden 1973, 49. 6

Figure 6: A rail from the M&P in the Midland Railway Centre (courtesy of Martyn Taylor-Cockayne)

It is unclear exactly why the interconnecting components of the rail were redesigned for the M&P rails, but it is most likely that some aspect of the rail joint used on the Belvoir Castle line was found to be unsatisfactory. Possibly the rails were susceptible to vertical displacement as the wagons passed over them, a defect which would be remedied by the M&P design. Figures 7 and 8 show the adjoining components of a M&P rail.

Figures 7 and 8: The ends of a rail from the M&P line (courtesy of Martyn Taylor-Cockayne)

7

The M&P was associated with ‘various other branches...to nearby collieries’ which must also have used edge rails.22 One of these branches was built by the Butterley Company, with Josias Jessop as its engineer. This was the so called ‘Portland Railway’, which branched from the M&P at Butterley’s Portland Collieries and ran to Codnor Park, opening in 1823.23 The rails used must have been like those used on the M&P, or some further variant. Fieldwork on the line is not known to have yielded any examples of the rails, but sleepers from the route prove that the rails were of this type (Figure 9). If the rails were not of the kind described from the M&P, then they are likely to have been of a variant type known from Codnor Park.

Figure 9: A stone sleeper from the Portland Railway. This bears a rust- stained rectangular depression left by the foot of the rail, a single spike hole, and a groove consistent with the rail’s fish-bellied span (courtesy of Martyn Taylor-Cockayne)

The Butterley Company mined and ironstone at Codnor Park from the late 18th century, and a network of railways developed to serve these mines. The Butterley Company obtained new mineral rights from the Ormonde Estate in 1824, subsequently expanding their mining operations at Codnor Park and building new railways using edge rails.24 Edge rails found at Codnor Park are different to those from the M&P or Belvoir Castle lines. I found a complete example of one of these rails near Codnor Castle in 2012.25 This is identical to other edge rails discovered at Codnor Park and weighs 36 lb.

Some dimensions for a Codnor Park rail are:-

 length – 3 feet  top flange width – 2.125 inches  bottom flange width – 1.0625 inch  rail depth at mid-span – 5 inches  rectangular foot at rail-end – 3½ inches x 7 inches

The rail discovered was found in two pieces, the foot being found first, followed by the rest of the rail. Remarkably, these two pieces are a perfect match. Although there has been corrosion since the rail was broken, they appear to be pieces of the same rail, as shown in Figures 10 and 11.

22 Birks & Coxon 1949, 226. 23 Taylor-Cockayne & Saint 2012, 30. 24 Taylor-Cockayne & Saint 2012, 49. 25 Rowan Patel, ‘Early Railway Artefacts in the North Midlands’, Backtrack, vol. 29, no. 1 (Jan. 2015), 12-3. 8

Figure 10: An example of a Codnor Park rail (scale: 8 inches)

Figure 11: A second view of a rail from Codnor Park (scale: 8 inches)

The dimensions given show that for the Codnor Park rails, the widths of the top and bottom flanges differ significantly from earlier rails. As described, the width of the bottom flange for the Belvoir Castle rails is much wider than the top flange, due to the weakness of cast iron when placed under tension. In contrast, the rails from Codnor Park have a very broad top flange, and a narrow bottom flange. This is shown by the cross sectioned Codnor Park rail pictured in Figure 12.

Figure 12: The cross section of a rail from Codnor Park (courtesy of Martyn Taylor-Cockayne)

The Codnor Park rails, considered as a beam, are of a weaker section than the Belvoir Castle rails. These rails appear to have evolved to a technologically inferior form on the grounds of practicality. The Belvoir Castle rails were extremely narrow, with the rail head measuring just 1.3125 inches across (or something similarly narrow). Although the width was increased slightly for the M&P rails, the width of the rails must still have caused problems. The most likely difficulty encountered with these rails, is that which Josias Jessop explained in a letter to Robert Stevenson in 1823. The rails are

9 said to have led to ‘the destruction of the wheels which running on a small surface were worn out into grooves’.26

In Josias’ letter, the rails which he says caused this problem were those used on the lines of the Leicester Navigation company. However, as Michael Lewis has demonstrated in OP255, it is debatable whether Josias himself ever saw these railways, whilst all of the evidence suggests that they must have used a quite different form of edge rail. It seems most likely that Josias filled in the gaps in his knowledge of the Leicester Navigation rails, by assuming that they had presented similar problems to the edge rails which he had used on his own lines.

That the rails used at both Belvoir Castle and M&P were unsatisfactory, and that the narrow rail head must have destroyed the wheels, just as Josias described, is shown by the redesigned rails used at Codnor Park. Here, this potential problem – as had been encountered on earlier lines – was avoided, by increasing the running surface of the rail to over 2 inches. With a top flange of this width, it would require a large quantity of iron to provide a wider bottom flange. Consequently, to keep the cost of the rails down, the large bottom flange was discarded, the rails instead having an adequate running surface, but presumably to save iron – only a small bottom flange.

In redesigning these rails for use at Codnor Park c. 1825, the rail joint was altered yet again (Figures 13 and 14). The rail-end adjoining the foot now incorporated a vertical slot ¾ inch wide. At the other end of the rail, the final ¾ inch of the rail is flangeless such that it fits into this slot. Additionally, either side of the slot, square holes are cast in the iron, 0.375 inches across, whilst the end of the rail has a groove in it of the same width. This arrangement allowed a cross pin to be inserted after consecutive rails had been adjoined, which would have stopped any vertical movement.

Figure 13 and 14: The two ends of a Codnor Park rail showing the features described (scale: 4 inches)

There was a final use of these rails on the Cromford & High Peak Railway (C&HP), for which Butterley supplied the permanent way.27 These rails were very similar to those used at Codnor Park, but of more robust dimensions. The C&HP rails were 4 feet long rather than 3 feet, and weighed 84 lb. The top flange of these rails was 2½ inches wide, whilst the bottom flange was of a similar width. The depth of the rails at the mid-point of the fish-bellied span was 6½ inches. The form of the rail joint was exactly as was used at Codnor Park, as shown by Figures 15 and 17.

26 Martyn Taylor-Cockayne, ‘Josias Jessop’s Railways Observed’, ERG OP 223, 2012, 2. 27 Riden 1973, 39. 10

Figure 15: A sketch of a C&HP rail showing all of the major dimensions (Courtesy of the Railway & Canal Historical Society, Baxter Collection)

The engineer of the C&HP was Josias Jessop, although dying in 1826, he did not live to see this line completed. Many of the rails, with the company’s initials cast in the side, remain to be seen at Middleton Top (Figure 16).

Figure 16: C&HP rails at Middleton Top (courtesy of Martyn Taylor-Cockayne)

The C&HP was the last use of these unusual edge rails. Indeed, this was the last use of cast-iron rails on any major railway project, for such rails were outdated technology even when the C&HP was newly finished. By the time the C&HP opened in 1831 cast-iron rails were, to all intents and purposes, consigned to the history books, having been superseded by the all-conquering wrought- iron rail. From this point onwards, Butterley played no part in the development of permanent way, and the C&HP rails represent the final fruits of the company’s involvement in cast-iron rail design, an involvement which began over three decades previously.

11

Figure 17: Preserved rails from the C&HP showing details of the joining components (courtesy of Wikimedia Commons)

I would like to express my gratitude to both Michael Lewis and Martyn-Taylor Cockayne for their generous help. Martyn-Taylor Cockayne also kindly provided many of the photographs.

April 2020

______

12