by Ben Russell
‘The Blacksmith’, from The Book of Trades (1824).
A model of John Wilkinson’s mill for boring steam engine cylinders, 1775. The cylinder is held down on its side with chains, and sectioned so that the cutter working inside can be seen.
Having made the unfinished components, and the cylinder, it was now time to fit them together. The first step was to gather all the engine parts at the site where they would be assembled. As far as possible, heavy parts like cylinders were moved by water, from Birmingham by canal, from Wilkinson’s Bersham works via the River Severn, and then around the coast until as close the final destination as possible. But after that the final leg was completed by road, with huge teams of draught horses, supplemented by teams of men if, as in Cornwall, the mines were perched on cliff-tops or on rugged moorland.
The arrival of all the parts on site marked the first time the complete engine would have been united at a single location. The engine erector and his team had the first chance to inspect everything, and parts would often be missing, or the wrong size. Erecting the engine was not just a matter of final assembly, but of completing a major part of the actual manufacturing process. Assuming the engine house walls were in place, the first step would be to install the engine beam, lifting it about 18 feet to the top of the engine house using pulleys and ropes, first raising one end, then the other, and then placing it approximately in position.63 The next major step was to fit the cylinder and align it perfectly on its foundations. This needed careful use of a weighted plumb-line to achieve perpendicularity, wedges for adjustment and long bolts to hold the whole securely. With these two main parts positioned, the rest of the engine could be assembled. It might take four or five weeks to complete, and came with attendant risks to the men involved – Boulton wrote home from Cornwall in 1779, ‘Tom Bowden has had two of his fingers burnt one quite broke off James Darlestone his hand much hurt, so that out of seven we have but four that can work.’64
The opportunities for injury reflected the nature of the engine-erecting process. Fitting the components together would emerge in time as a distinct trade. Much of the fitter’s work would be concerned with the accurate cutting and shaping of metal components using a range of basic hand tools: hammers, chisels, files and drills. Hammers and chisels were used to chip away metal. The chisel was up to 8 inches long, with a convex cutting edge. It was used to make a series of cuts across the surface to be removed, about one-thirtieth of an inch deep. The first layer of cast iron chiselled away could be the most difficult, as it would be impregnated with sand from the moulds it was cast in, which destroyed the chisel’s cutting edge.65 If a lot of metal had to be removed, huge ‘flogging’ chisels, over a foot long, were used, with one man gripping the chisel in both hands and another wielding a sledgehammer.66 Tallow, rendered from mutton or beef fat, would be used to lubricate the chisel’s motion and protect the cutting edge, but the hammer and the end of the chisel were kept scrupulously clean to avoid blows glancing off and causing injury.67 Having chipped off as much as they could, engine erectors then reached for their files to complete the surface to a fine finish. Even in the 1820s, the cost of producing a perfectly flat cast-iron surface was twelve shillings per square foot – one of Boulton & Watt’s engine erectors in Cornwall might earn nine shillings per week.68
The View of Botallack Mine by Philip Mitchell, 1840. Botallack was famous for its precarious position, and the workings extended some way under the sea.
Drilling holes in metal was equally time-consuming and labour-intensive. For the size of work needed on a steam engine, the brace would be the most effective tool, made of metal as opposed to the wooden braces used by carpenters. The brace would turn a flat drill rather than the spiral fluted twist drill used today; the flat drill had a pair of cutting edges meeting in a point, and worked not by cutting the metal but by scraping it. To assist in this onerous task a lever or heavy weight could press down on the top of the brace to apply pressure onto the drill, but often two men would be needed to turn the brace, sometimes spooning linseed oil into the hole formed to cool the drill.69 A larger hole could be made by drilling through a number of times and rounding out the hole with a file; or, more commonly, it was easier to cast a hole bigger than needed and then use wedges to secure whatever passed through it into position.
Watt would have been acquainted with the techniques of foundry, forge and fitting even before he began his partnership with Boulton. From 1765, as well as making scientific instruments in Glasgow, and his experimental engine at Kinneil, he had built a small number of atmospheric engines for customers in Scotland. And as the engine grew from a benchtop model to an industrial machine, Watt grew as well, noting in 1769, ‘I am not the same person I was four years ago when I invented the fire engine . . . the necessary experience in great was wanting; for acquiring it I have met with many disappointments.’70 By ‘in great’, Watt meant the ability to build a full-scale machine, and this ability was now to be tested to its fullest extent: between 1776 and 1778 orders for 22 engines came from Cornwall, along with almost the same number again elsewhere across Britain, and the practical burden of constructing them fell on Watt’s shoulders. Such were the calls on his time and expertise that he wrote to Boulton in February 1778: ‘I fancy I must be cut in pieces and a portion sent to every tribe in Israel.’71
The physical and mental exertions of erecting engines placed Watt under enormous pressure; here was a man who had previously admitted he would ‘rather face a loaded cannon than settle an account or make a bargain’.72 But the pressure to design and build machines, struggling against the limitations of tools, materials and workmen, and working under adverse conditions a long way from home, also told in Watt’s family life. In July 1776 he married his second wife, Annie McGrigor, daughter of a Glasgow-based bleacher – illustrating Watt’s continued associations with the chemical industry. The opportunity was taken to move into a bigger house in Birmingham, and bring from Scotland Watt’s children from his first marriage, Margaret and James. Their mother was dead, they had been left with relatives when aged only six and four respectively, and now they faced losing their father again as he drove forward the engine business in Cornwall. Two new children arrived: Gregory in October 1777 and Jessy in May 1779. But with their father often absent from home, and Annie vociferously advancing her own children, Watt’s family life thereafter would not be happy. By 1785 Watt’s relationship with his elder children had broken down: Margaret married without his consent in 1791 and moved to Scotland, having four children of her own but dying in childbirth in June 1796. She had not heard from her father for two years. James junior remained on difficult terms with his father, finding Matthew Boulton a more supportive figure, culminating in a period of rebellion, the context and consequences of which will be explored later.73 Any consolation from the presence of younger Gregory and Jessy would be short-lived: Jessy succumbed to tuberculosis in 1794 aged only fifteen, and a decade later that disease also claimed Gregory, a promising and talented young man intended to follow Watt into the engine business. It is likely that Watt could at least have salvaged his relationship with Margaret and James, if only he had been a more permanent presence in their lives; such could be the cost of making machines.
Matthew Boulton was also heavily engaged in Cornwall. He rolled up his sleeves and dived into the fray, writing to Watt that ‘of all the toys and trinkets which are manufactured at Soho, none shall take the place of fire engines in respect of my attention’.74 He could be found greasing the piston on an engine, complaining of a pair of engine erectors being ‘drunken, idle, stupid, careless, conceited rascals’, even delving into the technical detail of how to build the engines more quickly: ‘We are desirous if possible to lessen the expense of fitting up our nozzles, which at present is very considerable . . . we are at present obliged to chip and file a great deal.’75 Presently Boulton and Watt appointed William Murdoch as their chief representative in Cornwall, his technical skill and ability to take on fractious Cornish mi
ners with his fists if so required – an option hardly open to the reticent Watt – giving him considerable authority, and freeing Boulton and Watt for other activities.76
For Watt this meant working at his home in Birmingham with an assistant on calculations, correspondence and drawings, and only occasionally venturing onto the shop floor. This way, he could play to his strengths, leaving the outward-facing part of the business to Boulton, and turning his attention to some of the other innovations that would support the engine making process. The main challenge facing Watt was that, unlike making buttons, building engines remained dependent on a relatively small, multi-skilled group of men. Good mechanics were scarce: Watt told his engineer friend John Smeaton that he wished he could ‘find operative engineers who can put engines together according to plan, as clockmakers do clocks – we have yet found exceedingly few of them’.77 Watt’s use of clockmakers as an exemplar is interesting, because in clock-making one man still made a complete clock from start to finish – unlike buttons, there was little, if any work specialization.78 So Boulton & Watt’s men tended to be jacks of all trades, and they were expected to excel in all of them, producing ‘the well-known “Soho workmen”, whose services were sought directly or indirectly wherever their fame had spread’.79 Ensuring that these men could work as effectively as possible required a range of supporting tools and techniques, and this became Watt’s preserve.
This comes with a caveat, however. Although Watt has received the credit for many of these improvements, we should not forget that others were the responsibility of his relatively unsung colleagues on the shop floor at Soho and elsewhere. For example, William Murdoch had quickly become Boulton & Watt’s leading pattern maker before being sent to Cornwall, and there his abilities led him to make small, largely unauthorized adjustments to the engines as they were built – often to Watt’s irritation. Later the success of Boulton & Watt’s engine for driving factories as well as pumping water depended on Murdoch, as did much of the machinery required to build engine components at Soho. With men of this calibre engaged by the company (and Watt described Murdoch as ‘the most active man and best engine erector I ever saw’80), there was a constant flow of changes that did not necessarily begin life in Watt’s hands.
First came new paper systems. In 1779, Watt wrote his Directions for Erecting and Working the Newly-invented Steam Engines. This was the first book anywhere devoted to the steam engine, and it codified all the practical experience of engine making acquired over the previous three years. It was intended to counter a steady flow of enquiries from engine men working across the country, and came with printed checklists of engine components to use when working on site – Watt later used spare copies to wrap up delicate items in his workshop. Printed forms also detailed which iron founders would supply which components for each engine, with room left for dimensions and other notes and comments.81
Of particular importance was Watt’s method of letter copying. In order to record the voluminous correspondence that building each engine entailed, everything had to be copied longhand. Now, to save time and effort, Watt developed the first practical office machine, which he patented in 1780, and which Joseph Black praised for its ‘neatness and propriety’.82 Its working premise was simple: a letter was written using ink specially formulated by Watt. Then thin copying paper was laid over the top, and both were squeezed between rollers, after which the duplicate could be read through the back of the thin copy-paper. The ink was formulated to stay wet for around 24 hours so, in theory, several copies could be taken from a single original. In practice this relied on the ink being very carefully prepared, and often the original might be smudged when first copied. The copying press was hugely successful, simplifying the administration of the engine-building business. A copy made by Watt, and preserved in his workshop, records that ‘Time, labour & money are saved, dispatch & accuracy are attained, and secrecy is preserved by this newly invented art of copying letters and other writings.’83 Boulton and Watt formed a subsidiary company marketing the press to others; 630 were sold in its first year. Well might Watt claim that ‘Every gentleman engineer must now wear a pocket rolling press.’84
Roller press for copying letters, by J. Watt & Co., 1818.
Paper processes were matched by technical improvements to the engine itself. There was a move towards using standardized parts, which meant quicker initial construction and easier provision of spares in the event of a breakdown. Boulton realized that otherwise, ‘if any misfortune should happen’ and, for example, the pumping engine for a mine was halted, floodwater might drown the workings, which ‘might be ruined before a piston rod could be made’.85 Even a standard design of engine house was adopted. Quality was ensured by making at Soho all the wooden patterns which might be used for casting iron components, even if the casting itself was carried out elsewhere.86 And improvements were supported by careful experimentation. For example, there survives from the firm of Boulton & Watt a set of four models of engine beams of different designs, which appear to have been tested to determine which was the strongest. Two have reinforcing wooden ‘queen posts’ and iron straps, together with over-scale hooks or rings to hang weights on, and the other two are wooden latticeworks, one of which actually broke during the course of experiments.87
Finally, making measurements associated with the engine led to the development of new precision instruments. The first of these was the micrometer, reputedly made and used by Watt from c. 1776, which used the rotation of a fine screw-thread to measure to within 1/1,900th of an inch.88 The instrument had originally been developed by William Gascoigne for astronomical use, both measuring the diameters of stars and the angular distances between them. Its adoption in engineering suggests the degree of accuracy to which Boulton & Watt aspired. The second instrument points to the business model they adopted to make and sell the engine: it is a revolution counter. Remember Boulton & Watt’s distinction between making the engine and exploiting it. Despite their best efforts they made very little money from the former: just a few late-arriving components, or a faulty casting, could turn a profitable job into a loss-making one. Exploitation was where money could be made: Boulton & Watt’s engine was more efficient than the equivalentsized atmospheric engine, so the partners charged a royalty for its use, initially a flat rate based on the saving in coal fuel consumed. However, this caused disputes between Boulton & Watt and their customers, because engine use fluctuated: for instance, a mine pumping engine might stand idle in the summer when groundwater levels fell. Boulton complained: ‘It is rather hard to work without profit and then not get paid.’89 The answer was to record the number of working strokes made by each engine and then send a bill based on that, for which purpose Boulton & Watt had built a number of engine counters. They sat on the engine’s beam and, as it rocked up and down, a pendulum inside clicked back and forth, its movements captured by a set of dial wheels. The box was tamper-proof, and only Watt’s agents had a key. Here was an early application of power metering, the idea of which remains in use today.
Watt’s engine counter, c. 1781. An early batch of the counters was made by John Wyke of Liverpool, who had sold Watt instrument making tools earlier in his career.
It is tempting to say that Boulton’s declaration that he was making engines like buttons was canny marketing. But it was not empty rhetoric: he had to take a bold leap in the engine business for it to succeed because no one had undertaken such an ambitious project before. That leap had to be attended with measures to minimize risk. Returning to that first engine, erected at the Bloomfield Colliery, the newspaper report of it beginning work breathlessly noted, ‘All the iron foundry parts were executed by Mr Wilkinson; the condenser with the valves, pistons and all the small work at Soho by Mr Harrison and others, and the whole was erected by Mr Perrins conformable to the plans and under the directions of Mr Watt.’90 Wilkinson was at Bersham, Harrison at Soho, Perrins at Bloomfield and Watt at his house – separated by miles, but all working on the same project. B
oulton created what we will call a ‘dispersed factory’, and it was what a business analyst today might call ‘scaleable’: once the basic infrastructure was in place, it could be expanded as circumstances dictated, without a detrimental effect on Boulton & Watt’s core interests. This was no mean feat – in fact, it cannily exploited the fact that locally made components were cheaper than those made at Soho and could be shipped out to where they were needed.91 By 1800 Boulton & Watt had built 440 engines, each of which, with its attendant boilers for raising steam, was the size of a modern detached house. And by going out and hustling for the engine, Boulton freed up Watt to do what he could do best – working on the practical innovations that the business depended upon. We can close this chapter in Watt’s career by returning to those practical matters which took up so much of his time, and that of his colleagues.
First, just as Watt was intrigued by heat, and measured it with thermometers, mechanics used heat to work metal, and measured it by the metal’s colour. They hardened and tempered their tools to give them a good cutting edge by heating them to a pale straw yellow for lathe tools, or a yellow tinged with purple for chipping chisels and saws.92 Foundrymen needed the right temperature to pour molten metal – too hot and the mould would be damaged; too cool and the metal would not pour smoothly: they watched for the sparks dancing atop crucibles of molten iron, cooling after pouring into the mould until the surface was effaced with glowing lines, as if ‘covered with thousands of wire-worms in great activity’.93 Even today, the phrase ‘to strike while the iron is hot’ derives from the smith’s ability to gauge the welding heat of iron and effect a weld before the temperature fell. Heat was as much a mechanic’s practical tool as a subject of philosophical research.