Friday, July 22, 2011

Innovation and Trolls: The Auld Alliance

My internet buddy Brett Holman was musing about counterfactual history today. Specifically, that one about steam engines in the Roman Empire, which, to be fair, sounds  reasonable and gets a great deal of play in science fiction. (I was going to find a few examples, but someone's gone to a great deal of the necessary effort already.)

But, well, this is the problem with counterfactuals. You like the ones you like, and you pitch a fit about the ones you don't. And, truly, the haggis is in the fire in this one. That's a reference to one of Star Trek's Lieutenant-Commander Montgomery Scott's more excessive stage-Scottish moments. Although there's probably complicated sociological reasons that we associate Scots with ship's engineers (I'll let you know if I ever extract something useful from my prosopographic sample  of Institution of Mechanical Engineers obituaries), a lot of it does come back to James Watt and the steam engine.

As the title suggests, I've got some issues here. By now any regular reader will suspect that I see Scotland and Scottishness as a Trojan Horse in many dimensions. So here's my take on what actually happened with the steam engine, in a season when we're talking about how we can never, never, raise tax rates, lest it do awful things, like stop innovation.

But first, some Nineteenth Century American music, as interpreted by a Star Fleet engineer played by an actual Irishman for someone with a Scottish name who fell for a Trojan horse.

For starters, it is important to remember that Britain is a long-time mining country. Veins of lead and silver have been exploited since Roman times, and the tin mines of Cornwall need no further mention here. Mining, if you go back far enough (and that is the Paleolithic) starts on the surface, and follows the veins down. It doesn't have to be invented. As it happens, British mining got a great deal more aggressive in the late 1700s, because of the decision to copper the bottom of British warships. The copper veins of Wales are deep and rich, and this presented technical challenges, not least in draining them. That's part of the story, and I could jump to the denouement if I didn't want to inflict lots of screen-reading on you.

Because Britain is also a coal-rich country. Many of its great provincial towns lie in, or next to, coal fields. This has allowed them to burn coal for fuel for a very long time. Far from a new invention or discovery, coal-burning was limited by logistics. One could only make a profit by landsale so far from the pithead. It happens, though, that Britain is also an island, and some of the collieries were at, or near, the sea. This made "seasale" possible. Coal could descend the Tyne and ascend the Thames on the tide. That made Newcastle coal reasonably cheap in London. Some 3 million tons of coal in 1700, 5.2 million in 1750, 15 million in 1800, 23 million in 1815 and 31 million in 1830. (I would argue that the numbers create an exaggerated impression of a   “takeoff” at the turn of the century by taking this as the only war year  sampled. Source.)

This didn't happen in a particularly "bourgeois" way. If you owned land around Newcastle, chances were that you had a colliery on your land. To be a great landowner was to be a great coalowner, and the same logic that pushed large landowners to improve their fields encouraged them to wonder about a huge drift that disappeared under a thick layer of stone south of Newcastle.  At his Worsley colliery, Francis Egerton, 3rd Duke of Bridgewater, in the course of the 1750s, built a canal into the mine drain, and then extended it via an aqueduct overland to take coal to Manchester, and for 4 kilometers underground into the colliery. It was extraordinary engineering from start to finish, that Jenny Uglow has neatly described and tied into the "Lunar Men," the association of scientific minds from around Birmingham that included Watt's capitalising partner, Matthew Boulton, as well as Erasmus Darwin, among others.

Three points: first,  if you're wondering why now, I can only point to the government's policy of subsidising coal exports (or at least relaxing excises) to win exchange, so as to subsidise foreign wars. Higher prices for coal mean better returns for investors. Second: Bridgewater needed capital and a plan, but he also needed acts of parliament to clear the way for his building. And he got them, and this should not surprise us. This is not the story of pure free enterprise triumphant. It is a story of power and patronage, linked to a war that Pitt's ministry could not afford to lose. Third, James Brindley, who did much of the work of designing the canal, started out as a millwright. Of course he did. On the one hand, who knew more about pumping machines. On the other, who more than a millwright knew about surveying land, finding elevations, and directing water in order to run mills? This is the reason, IMHO, why the functions of "surveyor," and "engineer" converged during the fortress-and-canal-and-beaver-pond building age that began around 1500 on the continents that surround the Atlantic basin.

And, yes, I do want to give anonymous Eastern Woodland Indian sachems agency in the birth of the Industrial Revolution here, but that's rather too much to tackle in a blog posting that's supposed to be about James Watt.

So back to relevance, here: specifically, the new Admiralty dockyards in the West Country, which by the Seven Years War employed 6500 employees on shore, including more than 3000 rated as skilled tradesmen. That's before the yards launched their coppering project or Brunel Senior's early experiments in mass production, but well after some famous breakthroughs in research and development, such as producing the first domestic sailcloth up to navy standards. (Source; 889--2.) I could take the story of State factories back a long time if I chose (and I'm sort of committed to doing it), but, again, James Watt. This is his context: big manufacturing concerns that pump water out of places (in this case, drydocks) and do a great deal of metalworking.

By this time, that included steam engines. I like to romantically imagine Thomas Savery, Admiralty Commissioner of Sick and Wounded as of 1699, impressing members of the Whig Junto with the way that his ward vapourisers eased the suffering of sailors stricken with pneumonia. Be that as it may, he was close to the Junto, and got a patent (extended to 1733) by act of parliament for a machine that burned slack coal to run a piston. "Fire engines" were no great novelty, but great mounds of slack coal were, due, again, to the subsidised export of coal. Thanks to his patent, Savery's machines got rid of the whole unsightly mess, and by 1733 still had their place down at the bottom of collieries. Savery's steam engines were famously not very fuel efficient, but when we recall that they were bought at least in part to get rid of great piles of useless, crumbled, impossible-to-transport, fire-hazard coal, we can see that  as a feature rather than as a bug. (This argument is from Flinn, again.)

It also included some pretty spectacular metal-working. The Admiralty, like fortress-builders, didn't care how heavy cannons were, and was willing to take a fling on cast iron over bronze. By the wars of the 1700s, it had no choice, and the problem was now one of producing iron guns in volume. Casting with a core in place is pretty difficult, and, whether the makers fully understood this or not, the very act of boring out a solid-cast gun caused stress-hardening in the layer next to the bore. And, as we did not fully understand until the advent of quantum chemistry, stress-hardening is thermal "hardening." Isaac Wilkinson, an ironmaker of Sheffield, is the famous example of an English cannon-maker who is willing to go to the extra trouble of boring out a solid-cast gun, and he will soon enough be making the cylinders of Watt steam engines. This is why some people say that the Romans didn't have steam engines for "metallurgical" reasons; the tolerances between cylinders and pistons have to be quite fine, and it is possible that the Romans couldn't do that. "Possible" isn't much of an argument, though, see below.

And then there's James Watt (1736--1819); inventor, not so much of the steam engine, as of the condensor. I'd like to be all contrarian, here, but, as usual, Wikipedia has by now read the same sources as I have, and almost all traces of the Smilesian "Lives of the Engineers" rubbish is gone from that wonder of our age. (He still gets "educated at home," implicitly kept clear of  the evil influence of the Established Church in the Wikipedia version, though.)  Grandson of a bailiff who in turn was father to a vastly successful Clydesyde entrepeneur, Watt gets a measure of "self made man" credit by virtue of a temporary setback in the family fortunes in the crucial years after 1753. Otherwise, he was indisputably a gentleman, a natural philosopher first, and craftsman second, his enthusiasm for theoretical mathematics and scientific inquiry manifest in his letters. All accounts of his breakthrough to the condensor reflect this by making it clear that he was closely associated with University of Glasgow Chemistry Professor Joseph Black, who carried traditional Presbyterian interest in "pneumatics" forward into the idea that the heat required to convert water into steam remained present in it as “latent heat” that could be recovered by reversing the phase change. (Never mind how we get from salvation through grace to latent heat. I think that the connection can be made, but I'm not that interested in doing it.)

The thing is, that while prosecuting a crucial patent infringement case in the 1790s, Watt edited his friend John Robison's deposition to make Watt the "inventor" of the condensor. He does this as pure, inspired tinkering, before Black appears to explain the theory behind it. This is because, if the condensor is an application of Black's theory, the patent that Watt took out, and which Boulton had extended by Act of Parliament (see? Power and patronage!) is an illegal patent under the Act, which forbids patenting a principle of nature.
 Indeed, we can hardly push the influence of Black's new theory too far. The whole country had gone “engine mad” in the heat of the Seven Years War. The well-established mania for electricity had abundantly justified theories of an imponderable electrical fluid. Hydraulic theory was making all the difference in designing bridges and canals, and perhaps even guns. It seems to me that a theory that denied the fluid nature of heat might have a home in theology by analogy, but that would only make it even more controversial. After all, the major audience for "natural philosophy" at the time was the clergy, and they cared about this stuff.  It is perhaps no coincidence that, in the heat of controversy, Black and his close associate Watt turned to a model Newcomen engine owned by the university.

(Sources: Eric Robinson, “James Watt, Engineer by Act of Parliament,” 13—26 in Bicentenary of the James Watt Patent for a Separate Condensor for the Steam Engine: Proceedings of a Two-Day Symposium, September, 1st and 2nd, 1969, arranged at the University of Glasgow by the co-operation of the Scottish Universities, the City of Glasgow, Industry, and Other Bodies, ed. Robert Donaldson ([n.p.]: University of Glasgow, 1969); R. S. Silver, “Some Aspects of the Development of the Condensor from the time of James Watt,” 27—38 in Ibid., manages to make it amazing that Watt succeeded at all; further biographic details and documentary extracts, see Eric Robinson and A. E. Musson, James Watt and the Steam Revolution: A Documentary History by Eric Robinson and A. E. Musson (London: Adams and Dart, 1969): 1—21.)

When Watt finally came to apply for his 33 year patent in 1769, with the assistance of Doctor Roebuck and Roebuck's occasional partner, Matthew Boulton, it was an impressively large and vague patent. The kind of patent that makes the fortunes of modern patent trolls. Why actually build steam engines when you can get a patent on the ones that the Hornblower, Trevithick and Dudley clans  were already building in Cornwall? Why build steam engines when you can make your fortune charging people for licenses to, well, innovate? That is certainly what the alleged "Cornish clique" was supposedly paying Edmund Burke to allege in Parliament. Watt and Boulton indignantly replied that the Cornish clique was after the condensor. Certainly some were, as it turns up independently in other people's designs, or at least in Joseph Bramah's fire pump. Bramah was called to court when Burke tried to have the Boulton-Watt-Roebuck patent set aside. It was hindering innovation, the well-connected engineer argued. But, for whatever reason, the patent was left to expire in 1792. Perhaps it was that Boulton had friends in the Ministry and the court. Perhaps it was that the partners were seen to deserve some kind of reward for making the condensor public, known, practical, and universal.

That period of patent-won security, and the income it generated, may well be the key point. The Boulton and Watt steam engine was like many other inventions, in being one percent intuition, theory, or inspiration, the rest very expensive perspiration. The cylinder of the Savery/Newcomen engine was sealed against atmosphere by a cistern of water lying over the piston head, the water that percolated down into the cylinder being no great problem, whereas Watt's engine lost all rationale unless it had a dry end. That meant sealing steam in and atmosphere out with some kind of seal or packing that did not absorb the piston's work in friction. This is, in effect, a problem that has never been solved, and you can hear all you want and more about it from engineers, if not historians. Leather, rubber, grease, heavy oils, contaminants, corrosion.... 

So much of the 7 years in which Watt should have benefitted from his patent were spent finding the right materials for packing, lubrication, and the condensor. Wilkinson's cannon-boring machinery might have helped, but their rivals did not depend on Wilkinson's borers, and it appears that his main contribution was in this area. Wilkinson certainly designed a steam jacket, may have helped with the condensor material problem, and solved the neglected problem of broaching the cylinder for the condensor valve. The final result was a fine steam engine with many parents, for which Boulton found the customers. It was just that it was scarcely the only one out there.
Indeed, Boulton's great Soho works built only a few steam engines. Rather, it designed them, produced a few crucial parts (notably the packing house), and oversaw their installation. In practice that meant sending out Watt (and, soon enough, more junior assistants) to train local “engine smiths," although he still had time to invent a governor, which he never patented, because it worked in an entirely mysterious way (it was a feeback-controlled machine and points to a new age of innovation.) 

After waving at the mysterious realm of partial differential equations that implies the computer age, I think that I will end things in 1800. Watt has invested his engineering money in real estate and retired to be a landlord. Some 2000 steam engines representing an aggregate 8000hp  have been installed, mostly in the last third of the century, including 496 Boulton and Watt plants. 

The really interesting question was where the Watt engines were installed. The easy answer is, "where fuel economy was important." If you're thinking "factories," you're wrong. In general, there wasn't that much demand. They might be mistaken for  factory installations, but what they were doing at places like the Royal Dockyards or the Carron Ironworks was supplementing existing water mills. The engines pumped waste water back up to the head of the falls.

So it is clear that the steam engine had transformed the possibilities of industry, but by making up for the fact that heads of water were running out. And why is that? Because water was also needed for water-fed agriculture, mainly water meadows in this climate, and for canals to bring the forage produced there to market. Although not directly, the steam engine industry is rising on the back of agriculture, and is vitally linked to horseraising.

So that's my organic story of the rise of the steam engine. It was possible because of coal; because the ministries played favourites and gave out bad patents; because at the same time, British agriculture was using up key resources; and because of theoretically-led scientific development of new industries. The root of that, I've suggested, if perhaps not in the clearest of ways, is to be found in the state's drive to power, and the great wars of the Eighteenth Century. Could it have happened in the Roman Empire? Barely conceivably --in a very different, and much more developed Roman Britain. Certainly the Romans tried to develop Roman Britain. The real question is why they failed.


  1. A slightly weird thing is that in the 17th century the coal industry did lead to the development of railways, but they were horse-drawn.

  2. And mill-drawn, and some really impressive use of inclines....But, yeah, mostly horse-drawn. Which comes back to the fact that early industry was an aspect of agricultural intensification.

  3. Is it too easy to simply link Scottish ships' engineers to the Clyde shipbuilders? I know Don Leggett at Kent is big on 19th-c. shipbuilding; I'll ask him when I next see him.

  4. The problem is that it turns out that a term of service as ship's engineer qualifies as an apprenticeship. So while there are a great many Scottish ship's engineers, they're young men with no prior employment. Even if they go on to work in the shpbuilding (or, more likely given the sample base) the marine engineering sectors, it is after they come off the ships.

    I've got this great database --but it has so many variables in it that I still haven't been able to detect any trends that don't hit you over the head anyway when you explore things like apprenticeship requirements.

  5. It's too easy to simply link to the Scottish engineers vessels in shipyards Clyde? I know that Don Leggett in Kent big 19 c.; I ask when I see him more.Marketing agency Sheffield

  6. You made the following comment "Why actually build steam engines when you can get a patent on the ones that the Hornblower, Trevithick and Dudley clans were already building in Cornwall?" What do you know of these families efforts in engineering in Cornwall? I can find heaps on the Hornblowers and Trevithick but nothing on the Dudley family, although I do know that Thomas Dudley was one of Watt's erectors (controversial) and have long suspected his father, John, to have been an engineer also...

  7. I really did a number on myself in terms of burying the reference to Dudley engineers, but it seems to come from Ian McNeil, Joseph Bramah: a century of invention, 1749-1851 (Newton Abbot : David & Charles, 1968), 89--90. The main evidence seems to have been depositions from one or another patent infringement trial.