Seeing the Chemical Steam through the Historical Fog: Watt's Steam Engine as Chemistry

Summary

James Watt (1736–1819) is best known as an engineer who dramatically improved the efficiency of the steam engine. What we take to be his chemical interests are conventionally seen as peripheral to his main line of work. He is usually treated as a chemist in three main contexts: his ‘practical’ chemical work relating to chlorine bleaching, varnishes, pottery, and so on; his work with Thomas Beddoes on the medicinal uses of various ‘airs’; his, much disputed, claim as a chemical discoverer in the case of the composition of water. In this paper, I argue that Watt himself, and his contemporaries, saw the centrepiece of his steam engine work—the separate condenser—as a chemical invention. I also suggest more broadly that Watt understood the steam engine as a chemical device. For Watt and his Scottish friends, the study of steam and heat was a chemical enquiry. The subsequent changes in the place of heat in chemical enquiry in the early nineteenth century led to a reclassification of Watt's chemical investigations as ‘physics’. This, in turn, produced the sharp separation of his chemical and engineering activities characteristic of modern historiography. Watt's steam engine, which is usually placed in the lineage of machines understood as heat engines, and explained by the laws of thermodynamics, is better seen in context as a chemical device. Watt's ‘indicator diagram’ is reassessed in the light of this.

Acknowledgements

This paper benefited from the critical attention of audiences in Sydney, Leeds, and Vancouver who heard earlier versions of it. I am grateful to them and, especially, to Ben Marsden for his constructive scepticism. Trevor Levere and two anonymous referees also helped to improve it. I acknowledge support from the Faculty of Arts and Social Sciences of the University of New South Wales. This research was partially supported under the Australian Research Council's Discovery Projects funding scheme.

Notes

¹‘Boulton & Watt versus Hornblower & Maberley. Copy from Mr Gurney's shorthand Notes of the proceedings of the Trial of this case in the Court of Common Pleas, Guild Hall by a Special Jury before the Right Honourable Lord Chief Justice Eyre. December 16th 1796’, JWP 3/9, 23–25.

²Key documents include: James Watt, ‘A Plain Story’, in James Patrick Muirhead, The Life of James Watt with Selections from his Correspondence (London, 1858), 83–91; Eric Robinson and Douglas McKie (eds), Partners in Science: Letters of James Watt and Joseph Black (Cambridge, MA, 1970); A.E. Musson ‘Editor's Introduction’ in Musson (ed.), Science, Technology and Economic Growth in the Eighteenth Century (London, 1972), 1–68; Milton Kerker, ‘Science and the Steam Engine’, Technology and Culture, 2 (1961), 381–90; Donald Fleming, ‘Latent Heat and the Invention of the Watt Engine’, Isis, 43 (1952), 3–5.

³See D.S.L. Cardwell, From Watt to Clausius. The Rise of Thermodynamics in the Early Industrial Age (Ithaca, NY, 1971), 42.

⁴Cardwell, From Watt to Clausius, 54. This assimilation of Watt to thermodynamics can be done well or not. For a particularly egregious effort, see the account in ‘James Watt and the Science of Thermodynamics’ in Keith J. Laidler, To Light Such a Candle. Chapters in the History of Science and Technology (Oxford, 1998), 13–67.

⁵See Richard L. Hills, ‘How James Watt Invented the Separate Condenser’, Bulletin of the Scientific Instrument Society, 57 (1998), 26–29 and 58 (1998), 6–10; Richard L. Hills, ‘The Origins of James Watt's Perfect Engine’, Transactions of the Newcomen Society, 68 (1996–97), 85–107. See also Richard L. Hills, James Watt. Volume 1: His Time in Scotland, 1736–1774 (Ashbourne, UK, 2002), 363–80. Watt himself uses the term ‘perfect engine’ in his retrospective account of his early improvements. However, this is not, as I argue below, decisive evidence against my argument in this paper.

⁶I owe this notion of rear-vision mirror historiography to my colleague John Schuster, who has long applied it productively to his studies of the Scientific Revolution or, more properly, of the evolution of natural philosophy.

⁷Watt's natural philosophical interests were broad, including meteorology and geology. An important future step will be to explore the guiding ideas of what we might call Watt's ‘cosmology’ in which chemistry of airs, steam engines, the weather, and geological processes were closely interconnected. My suggestion is that his chemical theory of heat is likely to be a central feature of this cosmology.

⁸Patsy A. Gerstner, ‘James Hutton's Theory of the Earth and his Theory of Matter’, Isis, 59 (1968), 26–31.

⁹Arthur Donovan, ‘James Hutton, Joseph Black and the Chemical Theory of Heat’, Ambix, 25 (1978), 176–90 at 176–77. The title of the current paper obviously o wes much to Donovan's historiographical imagery.

¹⁰In eighteenth-century natural philosophy, the term ‘principle’ was deployed to designate one of the simplest forms of matter from which other substances are formed through combinations with other principles. The term was also used, as in Hutton's case here, to refer to a source of agency. See D.R. Oldroyd, ‘The Doctrine of Property Conferring Principles in Chemistry’, Organon, 12–13 (1976–77), 139–55 and Rhoda Rappoport, ‘Rouelle and Stahl—The Phlogistic Revolution in France’, Chymia, 7 (1961), 73–102.

¹¹See especially John Playfair, Illustrations of the Huttonian Theory of the Earth (Edinburgh, 1802) and David Oldroyd, Thinking about the Earth: A History of Ideas in Geology (London: Athlone, 1996), 86–107. Oldroyd notes that Hutton was much impressed by the new steam engines which, with their complex, large-scale movements driven by a single source of heat, may well have been one of the inspirations for his theory of a heat-driven earth machine (p. 93).

¹²A useful overview of Watt's chemical activities of various sorts, but one which typifies the lack of interest in delving into his chemical philosophy or its relations with his engine work, is Jennifer S. Pugh and John Hudson, ‘The Chemical Work of James Watt, F.R.S.’, Notes and Records of the Royal Society of London, 40 (1985), 41–52.

¹³See Discussing Chemistry and Steam: The Minutes of a Coffee House Philosophical Society 1780–1787, edited by Trevor Levere and Gerard L'E. Turner (Oxford, 2002).

¹⁴Whewell to Harcourt, 11 February 1840, in E.W. Harcourt, The Harcourt Papers (Oxford, 1880–1895), vol xiv, 105–106 (my italics). See also David Philip Miller, Discovering Water: James Watt, Henry Cavendish and the Nineteenth-Century ‘Water Controversy’ (Aldershot, UK, 2004), 146–47.

¹⁵This has had a lasting effect in the sense that even the most sympathetic students of Watt's life and work find it hard to treat his philosophical chemistry seriously, while tending to separate it from his practical chemical work with bleaching, pottery glazes and the like. This is true, for example, of Richard Hills’ treatment of Watt's chemistry in his recent three-volume biography. I am aiming in this study, and in others forthcoming, to explore the ways in which Watt's philosophical and practical chemistry closely informed each other.

¹⁶See Miller (note 14), 18.

¹⁷Much attention has been lavished, of course, on the relationship between Black's discovery of latent heat and Watt's improvements of the steam engine. Robison's writings on this question, though unreliable, have been enormously influential historiographically. See Cardwell (note 3), 41–42. Watt himself denied this connection in the sense that he repudiated the suggestion that he had been a pupil of Black. Watt did not deny, however, that the issue of latent heat and his steam engine work were substantively related. In denying the specific indebtedness to Black, Watt, and his son after him, sought to portray the great engineer's work as completely sui generis. See Miller (note 14), 92–104. My suggestion is that, in understanding Watt's intellectual world, his views on heat, his work on steam, and his ideas about the composition of water are best seen as all of a piece.

¹⁸Lecture, 17 November 1766 in Blagden MS ‘Notes of Dr Black's Lectures’, 1766–1767, Wellcome Institute for History of Medicine, as quoted in Henry Guerlac, ‘Joseph Black's Work on Heat’, in Joseph Black 1728–1799. A Commemorative Symposium, edited by A.D.C. Simpson (Edinburgh, 1982), 13.

¹⁹This section relies on Guerlac (note 18).

²⁰David Ralph Dyck, The Nature of Heat and its Relationship to Chemistry in the Eighteenth Century, Unpublished Ph.D. thesis, University of Wisconsin, 1967.

²¹The idea is expressed in Johann Theodor Eller, ‘Seconde Dissertation sur les Elemens’, Mémoires de l'Academie Royal des Sciences et Belles Lettres [Berlin], 2 (1746), 25–48 at 43–44.

²²See Hélène Metzger, Newton, Stahl, Boerhaave et la Doctrine Chimique (Paris, 1930).

²³Hills (note 5, James Watt), 143. As late as 1769, Watt needed help in reading German (Hills, note 5, James Watt, 148). Much of Eller's work was published in French in which Watt was reasonably proficient.

²⁵The manuscript is published, with extensive, valuable background and explication in Russell McCormmach, Speculative Truth: Henry Cavendish, Natural Philosophy, and the Rise of Modern Theoretical Science (Oxford, 2004), 87. The first version of the material theory was held by Black and by Watt and also by William Cleghorn in his inaugural dissertation at Edinburgh in 1779. See Douglas McKie and N.H. de V. Heathcote (editors and translators) ‘William Cleghorn's De Igne (1779)’, Annals of Science, 14 (1958), 1–82.

²⁴The manuscript is published, with extensive, valuable background and explication in Russell McCormmach, Speculative Truth: Henry Cavendish, Natural Philosophy, and the Rise of Modern Theoretical Science (Oxford, 2004).

²⁶Quoted in J.P. Muirhead, The Origin and Progress of the Mechanical Inventions of James Watt, 3 vols (London, 1854), II, 167.

²⁸James Watt to Joseph Black, 13 December, 1782, transcribed in Robinson and McKie (note 2), at pp. 117–18.

²⁷James Watt to Joseph Black, 13 December, 1782, transcribed in Robinson and McKie (note 2), 117–19.

²⁹James Watt, ‘Thoughts on the Constituent Parts of Water and of Dephlogisticated Air; With an Account of Some Experiments on that Subject. In a Letter from Mr. James Watt, Engineer, to Mr. De Luc, F.R.S.’, Philosophical Transactions, 74 (1784), 329–53.

³⁰Watt discussed many of these proposed changes with the celebrated French chemist Claude-Louis Berthollet to whom the essay was sent. They had an extensive correspondence. For part of their nomenclatural discussions, see Watt to Berthollet, 18 [December] 1788 (JWP LB/1, 268), Watt to Berthollet, 24 January 1789 (JWP LB/1, 279) and Berthollet to Watt, 28 December 1788 (JWP W/11).

³¹David F. Larder, ‘An Unpublished Chemical Essay of James Watt’, Notes and Records of the Royal Society of London, 25 (1970), 193–210, at 194.

³²David F. Larder, ‘An Unpublished Chemical Essay of James Watt’, Notes and Records of the Royal Society of London, 25 (1970), 202.

³³David F. Larder, ‘An Unpublished Chemical Essay of James Watt’, Notes and Records of the Royal Society of London, 25 (1970), 203.

³⁴See Thomas Beddoes (and James Watt), Considerations on the Medicinal Use and the Production of Factitious Airs, 3rd edition, enlarged(Bristol, UK, 1796 ), and discussion in Miller (note 14), 53–54. I am researching with Trevor Levere the theories of Beddoes and Watt about airs in relation to their collaboration in pneumatic medicine.

³⁵Other, better known, extenders of Black's work in this sense were, of course, William Irvine, Adair Crawford, and William Cleghorn, on whom see Arthur Donovan, Philosophical Chemistry in the Scottish Enlightenment (Edinburgh, 1975), 271ff. See also Robert Fox, The Caloric Theory of Gases from Lavoisier to Regnault (Oxford, 1971), Chapter 1.

³⁷The Notebook is published in Robinson and McKie (note 2), 431–90. The original is in the Doldowlod Papers, Birmingham Central Library, JWP W/14 (new reference MS 3219/4/170). This section draws upon, but significantly extends, my account of this question in ‘True Myths: James Watt's Kettle, his Condenser, and his Chemistry’, History of Science, 42 (2004), 439.

³⁶The Notebook is published in Robinson and McKie (note 2), 431–90. The original is in the Doldowlod Papers, Birmingham Central Library, JWP W/14 (new reference MS 3219/4/170). This section draws upon, but significantly extends, my account of this question in ‘True Myths: James Watt's Kettle, his Condenser, and his Chemistry’, History of Science, 42 (2004), 333–60.

³⁸On the workings of the Newcomen Engine, see Richard L. Hills, Power from Steam: A History of the Stationary Steam Engine (Cambridge, 1989), 20–30.

³⁹Watt, ‘Plain Story’ (note 2), 87, as quoted in Donovan (note 35), 261–2.

⁴⁰Hills (note 5, 1996–1997, 1998, 2002). Another very clear exposition of Hills’ approach is in Ben Marsden, Watt's Perfect Engine: Steam and the Age of Invention (Cambridge, 2002), 43–64.

⁴¹Richard L. Hills, James Watt. Volume 2: The Years of Toil, 1775–1785 (Ashbourne, UK, 2005), 27. While ostensibly escaping presentism here, Hills’ account actually embraces it in the judgement that Watt's understanding of the ‘perfect engine’ concept was ‘flawed’ rather than merely different. Watt's concept was only flawed on anachronistic criteria. It was precisely the chemical interpretation that distinguished Watt's ‘perfect engine’ concept from that of nineteenth-century thermodynamics.

⁴²Marsden (note 40), 51.

⁴³Robinson and McKie (note 2), 459.

⁴⁴This is the Table devised by Watt in 1814 as rendered in Robinson and McKie (note 2), 473. Watt had made some corrections to the results as originally recorded in the Notebook, but these do not materially affect our concerns here.

⁴⁵Donald Fleming, ‘Latent Heat and the Invention of the Watt engine’, Isis, 43 (1952), 3–5.

⁴⁶The above treatment of these issues relies heavily on, while differing in emphasis from, Donovan (note 35), 250–65.

⁴⁷The confidence that historians often have in the source of this statement is misplaced. The statement, or variants upon it, is attributed most often to Lawrence Joseph Henderson, the Harvard physiologist and sociologist. James B. Conant, Science and Common Sense (London, 1951), 39 tells us that Henderson was ‘fond of remarking that before 1850 the steam engine did more for science than science did for the steam engine’. Conant was possibly the source for the quote in Charles C. Gillispie, The Edge of Objectivity (Princeton, NJ, 1960), 357, where it is treated skeptically, and the admission of ‘some truth’ in it by R.J. Forbes, ‘Power to 1850’, in A History of Technology, edited by Charles Singer et al. (Oxford, 1958), 165. Web-based sources frequently date the quotation to 1917, but it does not appear in Henderson's The Order of Nature, published that year. None of the attributions to Henderson that I have seen refer to an alternative printed source. The source was almost certainly Henderson's lectures on history of science given at Harvard from 1911–1912 to 1917–1918. (See Arnold Thackray and Robert K. Merton, ‘On Discipline Building: The Paradoxes of George Sarton’, Isis, 83 (1972), 472–95 at p. 484). J.D. Bernal, Science and Industry in the Nineteenth Century (London, 1953), 27, and The Social Function of Science (London, 1939), 129, expresses similar sentiments but does not deploy a catchy phrase. Lord Kelvin is sometimes attributed with a variant substituting ‘thermodynamics’ for ‘science’, but the source of this I also find remarkably fugitive, and I have seen it credited to J. Willard Gibbs.

⁴⁸Crosbie Smith, The Science of Energy: A Cultural History of Energy Physics in Victorian Britain (London, 1998), 34.

⁴⁹See R.L. Hills and A.J. Pacey, ‘The Measurement of Power in Early Steam-driven Textile Mills’, Technology and Culture, 13 (1972), 25–43.

⁵⁰ee R.L. Hills and A.J. Pacey, ‘The Measurement of Power in Early Steam-driven Textile Mills’, Technology and Culture, 13 (1972), 40–41.

⁵¹Émile Clapeyron, ‘Mémoire sur la Puissance Motrice de la Chaleur’, Journal de l'Ecole Polytechnique, cahier 23, 14 (1834), 153–90. An edited translation of this paper is available in Reflections on the Motive Power of Fire by Sadi Carnot and other Papers on the Second Law of Thermodynamics by E. Clapeyron and R. Clausius, edited by Eric Mendoza (New York, 1960), 71–105.

⁵²Cardwell (note 3), 220–21, is convinced that Clapeyron had encountered the indicator diagram but is not sure how. John Farey had seen an indicator diagram being taken in Russia. It is possible that the same thing happened to Clapeyron, on whose presence and activities in Russia at this time see Margaret Bradley, ‘Franco-Russian Engineering Links: The Careers of Lamé and Clapeyron, 1820–1830’, Annals of Science, 38 (1981), 291–312.

⁵³N.L.S. [Sadi] Carnot, Réflexions sur la Puissance Motrice du Feu et sur les Machines Propres a Developer cette Puissance (Paris, 1824).

⁵⁴See Crosbie Smith and M. Norton Wise, Energy and Empire. A Biographical Study of Lord Kelvin (Cambridge, 1989), 53, 290–91, 357–58.

⁵⁵Peter Guthrie Tait, Heat (London, 1884), 298–301.

⁵⁷Baird (note 56, ‘Instruments’), 121. Baird is drawing, and building, on recent emphasis by some scholars upon the importance of ‘practices’ in science by way of critique of the tyranny of theory-centrism in accounts of science dynamics. Given that instruments cannot speak for themselves, it seems to me that the idea of ‘hardware-discourse couples’ is a better concept than ‘Thing Knowledge’ in pursuing such questions. See, for example, John A. Schuster and Graeme Watchirs, ‘Natural Philosophy, Experiment and Discourse: Beyond the Kuhn/Bachelard Problematic’, Experimental Inquiries, edited by Homer LeGrand (Dordrecht, 1990), 1–48.

⁵⁶Davis Baird, ‘Instruments on the Cusp of Science and Technology: The Indicator Diagram’, Knowledge and Society, 8 (1989), 107–22; Davis Baird, Thing Knowledge: A Philosophy of Scientific Instruments (Berkeley, CA, 2004), 41–43, 170–88.

⁵⁸John Robison, ‘Steam’, Encyclopaedia Britannica, 3rd edition, 1797, vol. 17, 733–43 at p. 735.

⁵⁹John Robison, A System of Mechanical Philosophy, 4 vols (Edinburgh, 1822), vol. 2, 14.

⁶⁰It is also important to realize, however, that there were early and late models of the ‘Indicator’. The early models simply had a pointer that would register the pressure. It was later models that incorporated the sliding board and produced the trace throughout the cycle. It seems likely that over the years, Watt had two uses for the indicator, one as a measure of the work that an engine was performing, the other as a gauge of the elasticity of the steam.

⁶¹Henry Cavendish and Charles Blagden, ‘Computations and Observations in Journey 1785’, Devonshire Collections, Chatsworth, Cavendish Mss X(a) 4, 36–37.

⁶²John Robison, ‘Steam-Engine’, Encyclopaedia Britannica, 3rd edition (1797), vol. 17, 758–59.

⁶³Robert Brain and M. Norton Wise, ‘Muscles and Engines: Indicator Diagrams and Helmholtz's Graphical Methods’, in Universalgenie Helmholtz: Rückblick nach 100 Jahren, edited by Lorenz Krüger (Berlin, 1994), 124–45; M. Norton Wise, ‘Work and Waste: Political Economy and Natural Philosophy in Nineteenth-Century Britain’ (unpublished essay).

⁶⁴M. Norton Wise, ‘Precision: Agent of Unity and Product of Agreement. Part II—The Age of Steam and Telegraphy’, in The Values of Precision, edited by M. Norton Wise (Princeton, NJ, 1995), 222–36. Wise (p. 233) describes the hard work involved in the case of Helmholtz thus: ‘Conceptually speaking, Helmholtz's adaptation of engine indicators to physiological motive forces is extremely simple. They are of particular interest because they move across immense difference in pressure and timescales. It is easy enough to say that the work of engines replaces that of muscles, and therefore both should be subject to the same kind of analysis. It is quite another to compare a fifty-horsepower engine running at fifteen revolutions per minute with a frog muscle developing a few ergs of energy almost instantaneously. Part of Helmholtz's genius, as Olesko and Holmes describe it, lay in his ability to design the extraordinarily sensitive mechanisms that could realize the conceptual analogy, rendering it quite precise in terms of the time development of the muscle's energy’. Referred to is Frederic L. Holmes and Kathryn M. Olesko, ‘The Image of Precision: Helmholtz and the Graphical Method in Physiology’, in idem, 198–221.

⁶⁵It is very likely that Watt himself played some part in this. Especially in his retirement after 1800, he was actively involved (with James Watt Jr, John Southern, John Playfair, David Brewster and others) in retrospectively recasting the history and character of his inventions. I have touched on this in ‘“Puffing Jamie”: The Commercial and Ideological Importance of being a “Philosopher” in the case of the reputation of James Watt (1736–1819)’, History of Science, 38 (2000), 1–24. There is much more to learn about that phase of Watt's ‘self-fashioning’.

⁶⁶See Iwan Rhys Morus, When Physics Became King (Chicago, 2005), 128–39.