• 1  See for example the discussion in David S. Landes, The Unbound Prometheus, Cambridge, Cambridge Un (...)
• 2  See Jacques Payen, “La technologie des machines a vapeur in France de 1800 a 1850” in L’Acquisitio (...)

1Comparative assessments of the use of steam power technology in Britain and in France during the first of the half of the nineteenth century have a curious, almost paradoxical history. On the one hand, economic historians, in contributions devoted to the study of the origins and diffusion of industrialization from England to continental Europe, have typically described France as a technological laggard in many areas, including the use of steam power. This assessment is mainly based on comparative estimates of the number of steam HP installed in the two countries such as those reported here in figure 1.1 On the other hand, historians of technology have pointed out the prolonged resilience of the Boulton and Watt low-pressure engine design in England, despite the fact that throughout the 1810s and 1820s, a number of engineers such as Trevithick and Woolf had successfully developed more fuel efficient high pressure expansive designs that were more fuel efficient. In Britain, these high pressure expansive designs did not become popular for industrial applications until at least the late 1840s. Several contemporary engineers considered this situation as an expression of irrational technological conservatism. Interestingly enough, the same designs were instead quickly adopted in France, so that in 1820s the Woolf compound design had become the standard design for steam engines used in French factories and workshops.2

• 3  Nick Von Tunzelmann, Steam power and British industrialization to 1860, Oxford, Clarendon Press, 1 (...)

3The first technically successful use of high-pressure steam can be ascribed to the “puffer” engines designed by Richard Trevithick around 1800. In these engines, high-pressure steam, after the stroke, was discharged into the atmosphere, rather than being conveyed into the separate condenser. The chief advantage of these engines was their compactness and cheaper cost of installation (due to the elimination of the condenser, the air pump and the beam). However, this high-pressure non-condensing engine did not make use of early cut-off and was less efficient than Watt low-pressure engines, as it tended to consume about 25% more coal.3 In Britain, the compactness of this engine design would make it the most natural option for railway use, but it did not find widespread use in industrial applications where the Watt low-pressure engine remained dominant.

• 4  Later on, this type of boiler would be termed “Cornish” in the contemporary engineering literature

4The use of high-pressure steam for attaining fuel economies was the outcome of the parallel efforts of the Cornish engineers Richard Trevithick and Arthur Woolf. In the early 1810s they developed engines in which high-pressure steam was employed expansively. These engines also made use of the separate condenser which permitted the exploitation of a larger range of operating temperatures (although the rationale for this design feature was not fully understood until the 1850s with the formulation of classical thermodynamics). The layout of the high-pressure condensing engine developed by Trevithick, making use of beam, separate condenser and air pump, was substantially the same as a Watt low-pressure engine, with the key exception of a new form of tubular boiler for the generation of high-pressure steam.4 Woolf instead adopted a so-called compound design in which steam was expanded subsequently in two cylinders.

• 5  The first three reports were published in West Briton, a local newspaper. From 1812, Lean’s Engine (...)

5It is not surprising that these pioneering developments of the high-pressure expansive engine took place in the Cornish mining district. By comparison with other locations, one of the distinctive features of the Cornish mining economy was the high price of coal. As a result, Cornish mining entrepreneurs were keenly interested in improvements in the fuel efficiency of the steam engines that could curtail their costly fuel bills. From 1811, they sponsored a monthly publication containing detailed reports on the performance (measured in millions of lbs. of water lifted one foot high per consumption of a bushel of coal, or, as it was termed by contemporary engineers, the “duty” of the engine), technical details and operating procedures of the steam engines at work in the county. The explicit intention was twofold. First, the publication would permit the rapid identification and diffusion of best-practice techniques. Second, it would create a climate of competition in the Cornish engineering community with favorable effects on the rate of technical progress. Joel Lean, a highly respected mine “captain” was entrusted with the compilation of the reports and the publication was generally known as Lean’s engine reporter.5

• 6  Robert C. Allen, “Collective invention”, Journal of economic behavior and organization, 4, 1983, p (...)

6In a previous article, I have argued that the Cornish mining district in the first half of the nineteenth century can be seen as an example of what Robert Allen has termed “collective invention settings”. Within “collective invention settings”, rival firms or independent individual inventors freely release to one another pertinent information concerning the solution of technical problems, rather than appropriating it by means of patents or secrecy. Each firm, in turn, makes use of the received information to incrementally improve on a basic common technological design. In Cornwall, the chief channel through which information concerning the technological characteristics and the performance of the engines was released was clearly Lean’s engine reporter.6

• 8  John Farey, A treatise on the steam engine, Historical, practical and descriptive.Volume II, Newto (...)

9Similarly, John Farey vigorously denounced a widespread and culpable “state of apathy as to consumption of fuel” in the “great manufacturing districts of the North”.8

• 10  In the same article, G. H. Palmer, on the basis of the caloric theory of heat, fixed the maximum d (...)

11These passages suggest that, despite numerous solicitations, many engineers and practitioners had re-mained extremely skeptical, at least till the late 1830s, about the fuel advantages of using high-pressure steam expansively. Since the superior fuel efficiency of the high-pressure expansive engine had remained theoretically mysterious, the dramatic early rise of the duty of the (best-practice) Cornish expansive engines (in the 1810s up to more than 40 millions and by the late 1820s to more than 80 millions) was not easily accepted outside Cornwall. In fact, several doubts were voiced on the actual levels of efficiency achieved by Cornish engines, actually denying the existence of a Cornish technological lead. In 1838, G. H. Palmer published an article in the authoritative Transactions of the Institution of Civil Engineers, in which he contended that the levels of fuel efficiency claimed for the Cornish engine were undoubtedly exaggerated (because in open contrast with the caloric theory of heat):10

• 12  Robert Armstrong, An essay on the boilers of steam engines, London, 1839, p. 76.

12The most strenuous defender of Lancashire technical practice was perhaps Robert Armstrong. In his Essay on the boilers of steam engines published in 1839, he declared that the Cornish duty figures were undoubtedly “gross exaggerations”, the real duty probably being equal to about 30 millions. He concluded that “there is nothing in their [Cornish] system of management that can be profitably imitated here [Lancashire]”.12

• 13  See, in particular, Donald Cardwell, The Fontana history of technology, London, Harper, 1994.
• 14  Robert C. Allen, The British industrial revolution in a global perspective, Cambridge, Cambridge u (...)
• 15  Robert Fox, “The challenge of a new technology. Theorists and the high pressure steam engine befor (...)
• 16  R. Fox, op. cit., p. 151. In the 1810s the Cornish engine reports were reprinted regularly in Anna (...)
• 17  R. Jenkins, “A Cornish engineer : Arthur Woolf, 1766-1837”, Transactions of the Newcomen Society, (...)
• 18  N. Von Tunzelmann, op. cit. (3), p. 281 cites this particularly enlightening passage from the answ (...)

13If we turn to the French case, it is possible to characterize the history of steam engineering from 1700-1850 as divided in three stages: i) a first phase of autonomous experimentation, ii) a phase of technological dependence and imitation (going from the years of Watt’s invention of the separate condenser to the Restoration), iii) a phase of creative adaptation of British inventions and successful catching up (marked by the widespread adoption of Woolf compound design). Historians of technology have properly given credit to Thomas Newcomen for pioneering the development of steam power technology, by designing the first fully operational steam engine.13 To be sure, Newcomen’s contribution to steam technology was a major one and of probably unparalleled historical significance. However, it is important to recognize, that Britain was not the only country involved in early experimentation with steam engine designs. For example, Cugnot in the 1760s proposed an interesting design of a high-pressure steam tractor to be used for pulling cannons. According to Allen, the key-difference between these early experimentations in England and France was not so much one of different inventive or engineering abilities, but one of different economic environments. Steam engine design and construction did represent a substantial investment and, given the low levels of fuel efficiency of early steam engine designs, this investment was likely to turn out to be profitable only in locations with low coal prices. Thus, the abundance of coal in Britain and its scarcity in France are the key-determinants of the different commitment of the two countries to the development of steam power technology during the eighteenth century.14 In this perspective, it is not surprising that the key breakthroughs in steam technology of the eighteenth century (Newcomen’s atmospheric engine and Watt’s separate condenser) occurred in Britain. As Allen aptly points out, both inventions did require substantial economic outlays for being put into practice. Interestingly enough, the minimal role of steam power technology in the French economy, did not prevent French scientists and engineers from becoming keenly interested both at theoretical and practical level in steam engineering. As pointed by Robert Fox, one of the main factors accounting for the interest of French scientists in the British achievements in steam engineers were their implications for the experimental and theoretical study of the thermal properties of matter in a gaseous state, a prominent area of research for many French scientists.15 The Revolutionary and Napoleonic wars resulted in a forced interruption of most of the information channels used by French scientists and engineers to keep abreast with the evolution of steam engineering in Britain. Thus, especially after 1815, the news of the successful design of high pressure expansive engines by Trevithick and Woolf in Cornwall and of their staggering improvements in fuel-efficiency was received with enthusiasm, which was not tempered, as in Britain, by skepticism and attachment to James Watt’s designs and reputation.16 In the early 1810s Humphrey Edwards broke his partnership with Woolf and moved to France where, under the protection of a brevet d’importation, he introduced the rotative compound design. As mentioned, this type of design became immediately very popular. According to a contemporary estimate, by 1824, already some 300 of this type engines had been installed.17 After the expiration of Edwards’ patent several other firms took over the construction of the Woolf engine and the rotative compound design became the standard option in French manufacturing application of steam power. As a result of these developments, from the 1820s, the practice of steam engineering in Britain and France diverged. We do not see anymore French engineers importing and imitating British low pressure designs, as during the eighteenth and early nineteenth century. Instead, we see the successful adaptation of a design that was not popular in Britain to French local conditions.18

• 19  These profitability assessments are discussed in detail in Alessandro Nuvolari and Bart Verspagen, (...)

15It is possible to examine in some detail two cases of adoption of the high-pressure engines in the late 1830s and early 1840s in England and carry out some simple profitability assessments.19 In the case of the purchase of a new engine, an entrepreneur will be indifferent between a high-pressure and low-pressure engine when

• 20  Thomas Wicksteed, An experimental inquiry concerning Cornish and Boulton & Watt engines, London, 1 (...)
• 21  Ibid.
• 22  N. Von Tunzelmann, op. cit. (3), p. 73. In several cases steam engines in water-works were worked (...)

20We will examine first the profitability of adopting a high-pressure expansive engine for pumping applications (the user context most similar to the Cornish case). One of the first high-pressure expansive pumping engines installed outside Cornwall was erected at the East London Waterworks as late as 1838. The installation was preceded by a travel of Thomas Wicksteed to Cornwall where he conducted a detailed research on the merits of the Cornish engine.20 Although Wicksteed heartily encouraged the adoption of Cornish engines, the management of the waterworks remained reluctant. Finally, in 1838 a second-hand Cornish engine was purchased for £ 7600 under the condition that it would perform a duty of 90 millions over twelve consecutive months, otherwise a penalty had to be paid. Once the engine was installed, Wicksteed made a systematic comparison of the performance of the new Cornish engine with that of a Boulton and Watt engine.21 In table 1, we assess the profitability of adopting high-pressure expansive engines for pumping applications in the late 1830s using Wicksteed’s data. As it is hard to estimate a typical number of working hours for engines employed in water-works, we have computed the threshold coal price for a reasonable range of possible values.22 The upshot of the results of table 1 is rather striking (although probably not fully surprising for a reader acquainted with the contemporary engineering discussions). At the high price of coal (around 20 s.) prevailing in London, it would have been worthwhile to scrap all the installed low-pressure engines replacing them with new high-pressure engines.

• 23  Von Tunzelmann, op. cit. (3), p. 91.

23Von Tunzelmann has calculated the “threshold” coal price at which, for rotative applications around 1835, it would have been economically worthwhile to install a new high-pressure expansive engine, instead of a low-pressure one, as 12 s. per ton.23 This result, according to Von Tunzelmann, goes some way in the direction of rehabilitating Lancashire entrepreneurs from the allegations of entrepreneurial failure to which contemporaries, such as Farey, had condemned them.

• 24  Our calculation suggests that threshold price computed by von Tunzelmann for 1835 is overrated. Th (...)

25In Table 2 we report Hick’s prices and estimates of coal consumption and our calculation of the threshold coal prices. When installing a new engine, for the most common sizes in this period (i.e., 40 and 50 HP), it would have been economically justified to adopt high-pressure expansive engines even in locations with a low coal price. The calculated threshold coal price for engines of 30, 40 and 50 HP is equal to slightly more than 1 s. a ton, which is even lower than the cost of “slack” coal at the colliery pithead. When a low-pressure engine was already installed, results are less clear-cut and possibly consistent with maintaining the low-pressure engine as favourite technical choice if coal prices were lower than 7 s. Overall, our results indicate a greater cost effectiveness of the high-pressure expansive engine than the one originally estimated by Von Tunzelmann.24

• 25  See, Von Tunzelmann, op cit. (3), p. 277-278.

26What are the implications of these calculations if we take into account the French context? In the 1830s, the average price of coal in French manufacturing districts was probably around 36 s. per ton and nowhere in France, the price of coal was less than 28 s. per ton.25 Given these different circumstances of coal prices, and against the background of the profitability assessments carried out above, the quick shift to the high pressure expansive engine in French manufacturing does not come as a surprise.

• 26  Paul A. David, Technical choice, Innovation and economic growth, Cambridge, Cambridge University P (...)

28In our judgment, the model proposed by Paul David in his reassessment of the Rothbarth-Habakkuk debate provides an interpretive framework that can be fruitfully applied to the case of the different practices in England and in France in the first half of the nineteenth century.26 The essential feature of David’s model, in comparison with the traditional neoclassical model of technical choice, is the idea that technical progress is, to a large extent, “localized” around specific techniques (that is to say, improvements in one techniques do not “spill over” to other points of the unit isoquant). In case of localized technical change, different factor endowments can lead to persistent differential rates of technical progress between different environments (regions or countries).

• 27  Allen, op. cit. (14), the case of the steam engine is analyzed in chapter 7.

32To conclude, we can summarize our re-assessment, which is consistent with the arguments put recently forward by Allen for explaining the onset of industrialization in England and its successive diffusion in Europe, as follows.27 During the eighteenth century early steam power technology was an innovation that was actually suitable of substantial application only in the British context of low coal prices. However, in retrospect, the curiosity and the interest of the French scientific and engineering establishment in this technology, despite its limited cost effectiveness in the French circumstances, was not entirely frivolous. The early episodes of importing or designing steam engines in France and the speculative efforts of the French scientific establishment provided a fundamental basis that was probably an indispensable pre-requisite for the later successful adaptation of the high-pressure expansive engine in France. Thus, when in the early nineteenth century, the use of high-pressure opened the doors for a fuel-saving trajectory of improvement that was well suited with French circumstances, France was ready to seize the new opportunity.