The Bear’s Lair: How steam engines entered the economy

Conventional wisdom has it that James Watt invented the steam engine, which became economically important in the 1780s, setting off the Industrial Revolution. All three of those beliefs are wrong. This column will lay out the true process, which was a 3-stage one over a century in duration, intensifying the Industrial Revolution but not directly setting it off. Understanding that process properly has important implications for how we adopt new technology today.

Apart from Ancient Greek and Chinese precursors, there were several attempts during the 17th century to use steam power to assist in pumping water out of mines, already identified as a serious problem not easily solved by other means. Spain’s Jeronimo de Ayanz y Beaumont developed a steam-powered pump for the silver mines in Seville, for which he was granted a Royal patent in 1606. Then the 2nd Marquess of Worcester in 1655 wrote “The Century of Inventions” in which he described a steam engine that could be used for pumping water – both Ayanz and Worcester appear to have built their contraptions and got them to work. Thus, the similar device built by Thomas Savery in 1698, which drew on Worcester’s book, should not have been patentable, let alone have had its over-general patent extended to 1733.

All three of these devices worked, albeit not very well and had a tendency to explode if asked to pump water up more than 34 feet, the height equivalent to atmospheric pressure. Most important, all three were not really “engines” – more like coffee pots; if you had inserted ground coffee at the appropriate point in their mechanism you would have got a decent cup of java. However Savery, as well as having the connections to get a patent extended for 35 years instead of the normal 14, was an energetic marketer, christening his device “The Miners’ Friend” and getting himself recognized as the country’s leading expert on steam power, even if he did not sell many pumps.

Savery’s luck turned when he got a Naval job that took him in 1706 to Dartmouth, where he met Thomas Newcomen. Newcomen, a prosperous ironmonger who also did fabrication work, had already been working for about four years on a steam pump to solve the flooding problem at Cornish tin mines, but unlike previous designs Newcomen’s was a genuine engine, with a massive beam rocking to and fro. His design work was bedevilled by scale problems; since it relied on atmospheric pressure to power the mechanism, and machining was very primitive, the engine needed to be very large to work properly – table-top models lost too much power. (James Watt would discover this problem with the University of Glasgow’s Newcomen engine in 1763, which was one third the size of a full-scale engine).

Once Savery had met him, Newcomen was forced to work within Savery’s patent, because attempting to acquire his own would involve litigation far beyond his means – anyway, Savery’s patent extended to 1733, well beyond the timescale of an ordinary patent taken out in 1710 or so. This also gave him the advantage of Savery’s marketing work and connections – probably useful to secure his first contract in 1710 at the Wheal Vor Cornish tin mine, owned by Sidney, Lord Godolphin, effectively prime minister at the time. The Wheal Vor engine was not entirely satisfactory, but Newcomen’s second engine in 1712, at the coal mine attached to Dudley Castle (which he marketed through a Baptist connection at Bromsgrove church) was highly successful and much admired.

Newcomen’s next problem was building a nationwide distribution network. His engines had a jerky motion, so could not be used to power machinery directly. Further, since the price of coal varied wildly between locations because it was expensive to transport across 1712’s miserable infrastructure, his very fuel-inefficient engines were most useful either in coal mines or in other mines located beside the coast where coal could easily be shipped in. The distribution problem was solved by forming a partnership in 1715: “The Proprietors of the Invention for Raising Water by Fire” with capital of about £20,000 and 80 shares outstanding, at least one of which was held by Newcomen himself. That partnership marketed the engines extensively, through agents around the country and on Wednesday afternoons at the Sword Blade Coffee House in the City.

The coffee house reference suggests that the £20,000 was raised by the Hollow Sword Blade Company, a proto-merchant bank that was later responsible for the South Sea Bubble – certainly it was far beyond the direct means of either Savery or Newcomen. Notably, Newcomen’s own shareholding was small – but before we get too indignant, we should remember that in 1715, there was no expectation that the inventor of a world-changing technology would become a billionaire. For Newcomen, it provided publicity and business for his ironmongery operation, and doubtless earned him fees on the jobs (for example in Cornwall) where logistically he could build and instal lthe engines himself. An early customer Sir James Lowther described him as

“a very honest good man, who I believe would not wrong anybody to gain ever so much.”

We should admire him as such and give him proper historical credit even without the billions – he was probably just as happy in comfortable blue-collar prosperity as any billionaire today. If we date the Industrial Revolution purely by steam engine technology, it began with Newcomen in 1712.

“The Proprietors” was moderately financially successful; it appears to have paid out in dividends before the 1733 patent expiry more than its initial investment. By that time, 110 Newcomen engines had been installed, a number that becomes more impressive when you discover that only 48 steam engines were operating in all of France in 1816, nearly a century later. One such engine, installed by Newcomen’s pupil Martin Triewald in Dannemora, Sweden in 1726 was the first steam engine to power machinery; it did so by pumping water back up a mill race, thus supplying a water wheel – an inefficient solution. By 1800, 1,400 Newcomen-type engines had been installed in Britain, three times as many as those of the Watt type, in full production from 1783.

After Newcomen, there was effectively no technological progress on steam engines for 60 years – whether or not due to the repressive social policies and dozy economic policies of the 1714-60 “Whig Supremacy” can be debated. One problem was the primitive machining technology of the period. Cylinders for Newcomen engines were generally hammered into shape and welded together, leaving structural weaknesses and gaps of up to 1 inch where steam escaped – it was said that while you could hear a Newcomen engine in action, you could not see it because of the dense fog of steam surrounding it. James Brindley, the canal builder, attempted without success to build a wooden Newcomen engine (which would have made shaping the cylinder much easier) and John Smeaton in the 1770s produced modest improvements. However, there was no technological breakthrough until James Watt’s condenser in 1769 and (just as important) John “Iron-mad” Wilkinson’s cylinder boring machine of 1774.

Meanwhile, by economic output measurements, the Industrial Revolution had moved into full swing, with patents doubling in the 1760s, Brindley’s “Grand Cross” canal system halving the price of coal in several cities, country banks making financing easier and Wedgwood, Wilkinson Boulton and Arkwright all getting into full-scale business.

Watt’s condenser did not alone produce the next breakthrough – cylinders still leaked too much, so that the unfortunate John Roebuck, who was financing Watt’s experiments, was forced into bankruptcy in 1772. However, Wilkinson’s boring machine, powered by a gigantic water wheel, solved the problem and produced cylinders accurate to 0.1 inches. Watt engines, produced from 1776, were then three times as efficient as Newcomen engines, having a “duty” of 18.9 million pounds, compared to a Newcomen engine’s 5.1 million pounds and a Smeaton “Newcomen plus” engine’s 9.1 million. (The “duty of an engine was the amount of water it could pump up 1 foot on a bushel of coal; it didn’t help mensuration that Newcastle and London bushels were different!) Watt engines were quickly used in many Cornish tin mines, where the less efficient Newcomen engines had been too coal-hungry.

Although Arkwright and other mill-owners began to use Watt engines and cheaper coal to pump water through the water wheels of their factories, the final Watt breakthrough, achieved in 1783, was the rotary engine that could power machinery directly. From that date, textile mills and eventually other factories began to use Watt engines as power sources. After first subcontracting their manufacture to Wilkinson, the Boulton and Watt partnership, fortified by Watt’s patent on the condenser that Boulton had got artificially extended to 1800, became Britain’s main engine supplier.

The advantage from using Watt engines, except for pumping water, was still only moderate, however – many textile mills remained water-powered through much of the 19th century. Since Watt engines were still “atmospheric” – relying only on atmospheric pressure to produce power – they were inevitably large and clumsy. Even though the Frenchman Nicholas Cugnot produced a steam carriage as early as 1769, it was slow, heavy and impracticable and railroads in particular were impossible without a better engine.

A steam engine using higher pressure was first patented in 1781 by Jonathan Hornblower, one of a three-generation family of Newcomen engine installers (his uncle Josiah had imported Newcomen know-how to the American colonies in 1753). Hornblower produced a “compound” engine in which steam was expanded in two cylinders, one at high pressure and a second at lower pressure. Regrettably, Hornblower’s innovation was quashed by Boulton using the Watt patent, and further innovation was delayed until its expiry in 1800. Shortly after that Richard Trevithick produced a high-pressure engine and Matthew Woolf another compound engine. Trevithick was a poor businessman and lacked the financial backing of a Boulton or a Hollow Sword Blade Company, but Woolf went into production of “Cornish engines” around 1815, which had a “duty” of 50 million pounds, about three times that of a Watt engine. With these engines, steam powered railways were fully practicable, and appeared within the next decade.

To produce high-pressure engines a further upgrade in machining precision was necessary, made possible by the work of Joseph Bramah and his pupil Henry Maudslay in the decades after 1780. Bramah began by producing high-quality locks, then invented the hydraulic press, forming an engineering business in which first Maudslay and later Woolf were trained. Maudslay invented the screw-cutting lathe, the bench micrometer and in 1807 the first table-top steam engine, broadening the use of steam to applications where room-sized Watt engines were impracticable. From the work of Bramah and Maudslay descends the machine tool industry, and the plethora of steam engine applications of the 19th century (Richard Roberts, a pupil of Maudslay, produced the first perfected power loom in 1822, for example.)

The century-long process by which steam engine technology entered the economy has lessons for us today. Some of these are as follows:

  • New technologies often have only narrow applications and only marginal advantages at first; they probably will not make their protagonists ultra-rich. Artificial Intelligence and Genetic Engineering may be hugely important in the long run, but they will not become so immediately.
  • Pauses of decades or more can occur between different generations of a technology — it took 64 years for the Newcomen engine to be succeeded by the Watt engine. Thus, it is not necessarily damning that a similar period has seen no huge improvement of the Apollo space program technology, and no mass adoption of it. (Apollo may in any case have appeared early because of the massive government subsidies involved.)
  • The true inventor of a technology is often unclear, even afterwards. In steam engines, was it Watt, Newcomen, Woolf or even Wilkinson? In computers/IT, was it Turing, von Neumann, Jobs, Berners-Lee or somebody else?
  • The big money is probably made in the second and third generations of the technology, when its uses spread beyond a narrow field to substantial sectors or the economy as a whole.
  • Inventions beyond the technology itself may be necessary for the technology to become ubiquitous – Wilkinson’s boring machine and Bramah/Maudslay’s machine tools, for example. Before those inventions appear, the technology may not be “fit for purpose” for broad applications.
  • True innovators generally do not emerge from large organizations or conventional sources, but a system must exist for their innovations to be funded and a marketing/distribution network set up – they need not become rich, but they should remain satisfied, fulfilled and happy.

This has been a lengthy historical digression from this column’s usual subject matter, but worth making because the conventional picture of how steam engines were adopted is inaccurate. By knowing the true picture, we can make better technological decisions today.

Martin’s upcoming book “Forging Modernity – Why and How Britain Got the Industrial Revolution” to be published by Lutterworth Press in March 2023.

(The Bear’s Lair is a weekly column that is intended to appear each Monday, an appropriately gloomy day of the week. Its rationale is that the proportion of “sell” recommendations put out by Wall Street houses remains far below that of “buy” recommendations. Accordingly, investors have an excess of positive information and very little negative information. The column thus takes the ursine view of life and the market, in the hope that it may be usefully different from what investors see elsewhere.)