by Ben Russell
Wedgwood obtained the work for himself by cunningly suggesting they should ‘call in some able Potter to their assistance . . . Would you think it? they took me at my word & I have got a fine jobb upon my hands in consequence of a little harmless boasting’.40
Although the most attention was devoted to these very expensive objects, made in relatively small quantities, antiquity also found its way into the output of Boulton’s toymakers. Chatelaines for holding keys, scissors and other useful items incorporated intricate outlines of acanthus leaves with classical imagery, flowers and putti – tiny winged people – stamped into them, and buttons used polished steel frames around tiny Wedgwood jasperware medallions representing figures from Greek mythology or the signs of the zodiac.41 Boulton employed at Soho one P. J. Wendler, who wrote from Naples that he had obtained a copy of Hamilton’s Antiquities, comprising ‘about 456 prints in folio divided into 4 volumes [comprising] handsome Designs & patterns for the Birmingham Manufactorys . . . I am certain, that . . . all our chief manufacturers . . . would be very glad to have those Prints, for they may be very usefull for Birmingham.’42 This would suggest a wider utility for such source material beyond Boulton’s works.
So, antiquity pervaded the environment in which machine-makers and manufacturers like Boulton & Watt operated. If they were successful enough (as the elder Boulton and Watt were) to commission portrait busts of themselves, they were often portrayed as antique figures: the eighteenth-century industrialist recast as a Roman clad in a toga.43 Even their hair might be cut in the ‘Brutus’ style, cropped short and brushed forward, replacing a formal wig. Boulton and Watt were too old for this new fashion but their sons were not.44 The men who worked for them, while not necessarily possessing the same detailed command of the literature explaining ancient worlds, might nonetheless have buttons, pieces of Wedgwood or other possessions referencing Greece and Rome. As far as engine making at Birmingham went, with components being designed and in part constructed in a complex of buildings designed in the antique style, in proximity to objects designed after the antique, it was almost inevitable that at some point those components would begin to look antique too! The next step for engine builders was to take up Piranesi’s challenge and use their inventing, creative genius to apply the antique to new forms – the steam engine.
William Chambers’s Treatise on the Decorative Part of Civil Architecture provides a springboard for exploring how they did so. First published in 1759, Chambers’s work was republished in 1825 by the English architect Joseph Gwilt. Gwilt’s analysis from his introductory essay on architecture’s ‘elements of beauty’ can be transposed to the new world of machines. He outlines his belief that artists should ‘study the effects that flow’ from those works of art ‘which by the common consent of ages are esteemed beautiful’. By doing so, they would understand those works’ qualities ‘which act on the understanding and excite our affections by means of the beautiful result they exhibit’.45 Gwilt believed that we construct meaning about a piece of architecture by looking at it, by striving to understand it and by devising an emotional response to it, and we can frame an analysis of machine making and engine making in these terms. As we saw in the previous chapter, the new technologies of the 1780s had made their mark in terms of visual impact. The steam engine in particular had amply conveyed what Gwilt would call its ‘magnitude and strength’, not least through the approach pioneered by Boulton & Watt. The two areas where progress was more constrained – or, perhaps it is more accurate to say, where greater codification of practical machine-builders’ tacit knowledge would be most advantageous – were in achieving a more widely held understanding of the engine’s internal operations, and in establishing how it could further appeal to the emotions of those who saw it. The former could be addressed by greater experiment and theoretical work on the proportioning of parts, and how they formed together into an integrated machine – what Gwilt might have called the engine’s ‘order and harmony’. The latter could be addressed by placing greater emphasis on the engine’s aesthetic design, the qualities Gwilt described as ‘richness and simplicity’, which were strongly influenced by prevailing fashions and taste, and its overall ‘design or disposition’, or how it was decorated.
Hand-coloured etching by Peter Fabris showing the Temple of Isis at Pompeii being unearthed, published in William Hamilton’s book Campi Phlegraei (1776). The unearthing of the temple is suggestive of engineers’ mining of antiquity for inspiration.
Regarding how the engine worked, Boulton & Watt had taken great care with their standardized range of machines to define the physical characteristics of each – so much so that John Farey devoted an entire chapter of his magisterial Treatise on the Steam Engine (1827) to discussing them. The engines were constructed in 23 different sizes, ranging from 4 to 100 horsepower. For each size, the diameter of the steam cylinder, the length of stroke and number of strokes per minute of the piston, even the volume of steam expected to be consumed per minute, were all carefully defined.46 Farey expended a huge amount of effort studying individual engines in the flesh and added even greater detail, all of which was rendered into simple rules of thumb, ‘stated in the most concise terms which could be chosen’ for ‘practical engineers’. For example, ‘the length of the great lever should be rather more than three times the length of the stroke of the piston’, and ‘the diameter of the aperture opened by each valve’ had to be ‘one-fifth of the diameter of the cylinder’.47
The same year that Farey’s book was published, Thomas Tredgold produced his own book, The Steam Engine. Reading both tomes together, Tredgold’s is much more theoretical – as Tredgold described it himself, it is about ‘the application of science to art’, and descriptions, drawings and rules of thumb are matched by complex equations and algebra.48 As historian Tony Woolrich has described him, Tredgold was ‘a compiler of formulae and rules and not a practical mechanical engineer’.49 But together, the books represent an immense quantity of knowledge about steam, codified and made available to readers. And in this respect they exemplify an explosion in the number of books about the engine written in the 1820s.50
However, the sudden expansion of publishing about steam raises an important question: Boulton & Watt’s partnership had ended in 1800 and since then, two decades had elapsed with no clearly formulated guides on how to make those engines beyond those materials already written by the firm of Boulton & Watt. Tredgold claimed in 1827 that ‘the effects that may be obtained by engines of different species, have now been reduced for the first time, to definite measures, and their proportions referred to scientific principles’. John Farey had produced an authoritative article in Abraham Rees’s Cyclopedia of 1816, but that was too short to provide the detailed technical guidelines that many engine builders would have wanted.51 So the question is: what was the received wisdom underpinning engine building in the critical two decades after 1800?
As a first step, we return to the influence of antiquity. After all, just as Farey considered the proportions of engine components, and the millwright Andrew Gray was declaring that ‘proportion is the foundation of all good mechanism’ as early as 1806, so proportions were long prescribed in the classical orders of architecture, from the ‘plain and robust’ Tuscan, to the Doric, Ionic and the ‘virginal slenderness’ of the Corinthian.52 The component parts of each were described in immense detail: a column divided into a pedestal, the shaft itself and the entablature resting on top. In turn the pedestal comprised the base, trunk and corniche; the column consisted of the base, shaft and capital; and the entablature was formed of the architrave, frieze and a final corniche surmounting all. Each of these components in turn resolved into tiny details: ovolos, astragals, fillets, larmiers, ogees, reglets and annulets, each contributing its mite to the appearance of the whole. And the relative size of all these was expressed in ‘modules’, one of which usually equalled the diameter of the column; an ideal Doric pedestal was two and a half modules in height, the column eight modules and the en
tablature two.
It is tempting to suggest that here was a system ready-made to be appropriated when designing engines. Certainly the classical column quickly found its way into mechanical designs: Boulton & Watt were designing a 52-horsepower engine for the cotton spinner James Kennedy of Manchester in 1806 for which the valve gear controlling the flow of steam to the cylinder was in the shape of a pair of fluted columns, topped by a finely detailed entablature.53 Tellingly the pair were in exactly the proportions required by the Tuscan order. And some of the columns employed in another engine, built for Samuel Bridge of Manchester, match the Doric order exactly.54 However, these are just two examples out of hundreds where there was no precise association with the classical orders. Even the slender columns supporting the engine beams (which became a signature feature for Boulton & Watt and were widely copied by other engine makers) are generally designed in a ratio of height to diameter of about 18:1 – which has no equivalent in the world of antiquity.
We must not forget that as understanding of its internal workings grew, the engine was increasingly perceived not as a static structure, but as one subject to dynamic stresses and strains, which would change continuously as it worked. John Bourne carefully differentiated between ‘the dimensions proper to be given to the various apertures, pumps, and vessels in connection with the engine, to ensure the maximum amount of efficiency’ and ‘the necessary sizes to ensure strength which should be given to the fixed and working parts’.55 The former, necessary to safely and economically ease the flow of steam around the engine, was not an issue that Boulton & Watt’s forebears in antiquity had to consider, but it played a decisive role in engine design alongside the latter.
This is not to entirely discount the influence of antiquity on engineering design. Flashes of it occasionally illuminate engineers’ letters. John Southern, head of Boulton & Watt’s drawing office until 1815, once wrote, with a dry sense of humour,
We have rummaged all the ancients and moderns from Palladio to that great aristocrat and Knight Sir William Chambers in quest of tables and propriety and having altered and altered and altered again and again various ingenious and highly meritorious designs have at last concentrated the essence of our ‘labours’ in one which for taste, beauty, magnificence, and utility will vie with the most renowned products of any genius of any age or nation before the conquest of Egypt.56
The maker of this small beam engine, c. 1840, is unknown, but its excellent design includes many classical design motifs, including finely detailed columns.
Later James Nasmyth wrote that ‘viewing abstractedly the forms of the various details of which every machine is composed, we shall find that they consist of certain combinations of six geometrical figures, namely, the line, the plane, the circle, the cylinder, the cone, and the sphere.’57 Nasmyth’s comment mirrors Plato’s view of the beauty of ‘abstract geometrical’ shapes which, he explained, did not refer to ‘living things or pictures’, but to ‘straight lines and circles, and shapes, plane or solid, made from them by lathe, ruler, and square’.58 But if the precise nature of antiquity’s influence is hard to quantify, what other factors influenced engineering’s evolution after 1800?
First and foremost, Boulton and Watt continued to wield considerable moral influence, even though they were personally no longer involved in building engines. John Farey wrote that ‘all those essential forms and proportions which affect the performance of the machine were so ascertained by the first inventor, that no improvement has since been made in them’, and that ‘the engines made at the Soho Manufactory, for some years after Mr Watt retired from the business, continued to be proportioned by his scale.’59 Even Thomas Tredgold, one of Watt’s critics, wrote:
An almost innumerable quantity of schemes for improvements on the steam engine have been crowded on the public eye within the last ten years, but except a few for improvements in construction, of small importance, there has been nothing done that is worthy of detaining the reader to notice.60
As late as 1868, John Bourne was still publishing Watt’s rules of thumb for engine making – for determining the thickness of the connecting rod, the diameters of shafts and the size of beams, for instance – alongside more recent theories.61
How far engineers actually stuck to Boulton and Watt’s plans is questionable. Tredgold noted that there was
no indication of a settled rule for the proportions of the cylinder, when the length of the stroke is unlimited by convenience. The proportions followed at different times by the firm of Boulton & Watt, in cases where the stroke was not limited, vary from 1 3/4 to nearly three to one [that is, length of stroke to diameter], the most common about 2.7 to one, the changes having no regularity . . . Equally irregular are Maudslay’s proportions but approaching to 2 to 1; Fenton, Murray, and Wood’s about as 2 1/2 is to 1.62
The occasional effects of these disparate approaches were seen by William Creighton in Glasgow, where he wrote in April 1803 of having ‘visited several of the engines here . . . they are as usual dirty and from the insufficiency of thin framing & construction of the frames are all distorted’.63 Even John Farey, with his very wide practical knowledge of who was building what, lamented that ‘owing to defective workmanship, and want of knowledge of the true proportions, it was generally found that the engines first executed by these new makers, fell very short of the performance of the pattern engines.’64 This last quote gives us a way out of this impasse: Boulton & Watt set the benchmark for how the engine performed. The challenge to other engineers was how to fully realize what John Farey called ‘the permanence of the machinery’, its practical embodiment.65 This was the subject of widespread and creative experimentation in machine making.
Not all engineers were single-minded, deadly rational beings like Watt. As may be surmised from the preceding debate, there was a tremendous range of opinion on the nature of engine making. As John Farey declared that ‘every departure from [Watt’s] forms and proportions has impaired the performance, to a greater or lesser extent’, so Thomas Donaldson argued that ‘old machines, when they were originally invented, had not any beauty of form: they were of large proportions’, and they had to be relieved ‘from cumbrous proportions’.66 Even a hardened theoretician like Tredgold could note with approval that ‘appropriate forms, good proportions, and excellent workmanship should be attended to in all machinery; and in many instances it is desirable that they should be beautiful.’67 The historian Lewis Mumford, in Technics and Civilisation, calls Britain’s machine makers at the turn of the nineteenth century ‘a new race of artists’.68 It is as artists that we should review their products. For the first time, after all, they were as much concerned with the outward appearance of the engine as with its internal workings. Here is Gwilt’s richness and simplicity, design and disposition, applied to making machines.
As we left the engine, being built on a new standardized scheme at the Soho Foundry, it was very much a functional machine, accurately built but with little conscious thought given to its outward appearance. Matthew Boulton stated his preference for ‘a simple, clean, orderly, convenient, modest building with good machinery’ compared to ‘that which is magnificently bad’.69 Form followed function, and the engine’s function was to facilitate the manufacture of other products, from spun cotton to beer. But the engine was becoming thought of more as a desirable product in itself. Engines built by Henry Harvey at Hayle, for example, were ‘more like ornaments for a show-room than machines for draining a mine’.70 Rather than the traditional division between process and product, what we have here is the process as product.
From the 1790s cast iron was being employed in greater quantities, and the means of shaping it offered new aesthetic possibilities. Beyond the cylinder, one of the earliest, and largest, of the component parts of the engine to employ cast iron was its working beam. This was no mean feat, given the huge stresses that it was subject to, and the early cast beams by Boulton & Watt were monolithic, not dissimilar to the baulks of seasoned timber t
hat they replaced. But gradually, with the realization that the greatest strains were at the centre, the beam was transformed into an ellipse, deepest at its centre, where more strength was needed, and tapering towards each end. This was a defining moment in the engine’s transformation, giving it ‘a grace of motion not hitherto perceived’.71 The properties of cast iron afforded gradual change to other components: the connecting rod linking the beam to the crank, and from there the flywheel, gradually took on a cruciform section, principally because it was the shape that let the cast iron cool at the most uniform rate, maximizing its strength.72 Returning to the beam, experiments showed that the greatest extending and compressing strains were withstood by the upper and lower edges. These could be increased in breadth by half at the expense of the inside of the beam, which could be made ‘open’ using a carefully designed latticework in place of solid material, for example.73
Boulton & Watt’s ‘Lap’ beam engine, 1788. This is the oldest essentially unaltered rotative engine in the world, and incorporates all Watt’s design improvements into a highly functional design.
All this work depended on improved pattern making and foundry work. The pattern maker required ‘all the methods, care, and skill, of good joinery or cabinet-making’ and a detailed appreciation of the characteristics of cast iron down to making an allowance ‘of about one-eighth of an inch per foot . . . for the contraction of the metal in cooling’ and a bevel of ‘about one-sixteenth of an inch in six inches’ to allow the pattern to be easily removed from the sand mould prior to the molten metal being poured in.74 The foundryman had to avoid air bubbles in castings and sought to ensure toughness by carefully cooling the finished product over a long period.75 This degree of care and precision beyond what had been done before made possible Samuel Clegg’s contention that with cast-iron construction, ‘It does not cost one farthing more to construct an elegant figure, than it does to create a deformity.’76