The Most Powerful Idea in the World

by William Rosen

  But it was the second engine, a 38-incher intended for the blast furnaces at the New Willey ironworks that really deserved all the attention. The Bloomfield colliery engine was still, after all, a water pump, and Boulton knew that pumping water out of mines was only a stepping-stone—a profitable one, to be sure—to something much larger. Boulton wrote, “I rejoice at the well doing of Willey Engine13 and now hope and flatter my self that we are at the Eve of a fortune,” and he was right to rejoice. The engine not only gave the new design an extremely profitable seal of approval and showed off the versatility of the new engine, but it introduced Boulton & Watt to the remarkable figure of John “Iron-Mad” Wilkinson.

  WILKINSON WAS A MEMBER of one of the dozen or so families with Nonconformist beliefs (in Wilkinson’s case, Presbyterianism) that dominated the iron trade in eighteenth-century England. His father, Isaac, an apprentice foundryman who became a master ironworker, was yet another of the fraternity of onetime artisans who pulled his family into the upper classes by dint of hard work, government contracts, and patent protection. In 1753 he acquired the Bersham Furnace, a foundry built by the family of Abraham Darby just across the Severn from Coalbrookdale, and became a favored supplier of cannon, shot, and shell casings—“engines of mortality of all descriptions”14—to the Office of Ordnance, a customer previously neglected by the Quaker Darbys. Four years later, in 1757, he applied for, and received, protection for his design for a cylinder-blowing engine, which used the weight of several columns of water to increase the pressure in a smelter’s bellows, and in 1758, a patent for a new method of casting and molding iron.

  From the time Isaac’s son, John, was sixteen years old, he was a partner in the iron business and in 1757, father and son founded the New Willey Company, a collection of furnaces and smelters on the Severn, about three miles south of Coalbrookdale. Four years later, John was running the entire complex, and doing his best to earn his nickname. He was, indeed, mad for iron; Wilkinson may have appeared grim in person—the profile he had stamped on the copper coins he used as payment in many of his factories practically glowers—but apparently he had a large capacity for love. Not so much for his wife, or his many mistresses; but the man loved his iron. He loved it so much, in fact, that he built an iron pulpit for his church, produced iron writing tablets and pens for his children’s school, and, in legend at least, had an iron coffin constructed in which he intended to be interred and that he kept on view in his New Willey office. He even promised to visit his beloved blast furnaces seven years after his death—and his promise was so credible that, on July 14, 1815, thousands of foundrymen showed up expecting to see his ghost. But the real business of New Willey, as with Bersham, was the production of cannon.

  There are few shapes simpler than a muzzle-loading cannon: an iron tube, closed at one end, with a small hole perpendicular to the long axis. Through that hole, a bit of flame ignited enough gunpowder to send a solid ball of—you guessed it—iron, weighing anywhere from ten ounces to forty-eight pounds, up to a mile and a half downrange. That kind of controlled explosion put a lot of stress on even the simplest of tubes, and early cannon had a disconcerting tendency to blow up, often enough because when the tube was cast, imperfections, known as “honeycombs,” were introduced that remained invisible until the gunpowder was ignited.

  Wilkinson’s answer was to cast a solid cylinder of iron and bore out a perfectly circular hole, which created a tighter seal around the cannonball and eliminated the casting imperfections. But drilling a hole in a twelve-foot-long cannon called for a very long drill bit indeed—so long that gravity caused it to deflect downward while drilling.

  Wilkinson’s great insight was to drill a pilot hole and suspend the boring head at both ends, thus guaranteeing a perfectly round tube, and to rotate the cannon instead of the drill bit. On January 27, 1774, Wilkinson turned his insight into a piece of legally protected property, patent number 1063: “a cylinder attached to a spindle15 driven by water with the drill stationary except in forward and back motion, as it is mounted on a carriage attached to rack-and-pinion.”

  The result was phenomenal accuracy, a maximum error estimated at only one-thousandth of an inch. The accuracy was purchased at what seems, from the perspective of the twenty-first century, to be a huge amount of time: In 1800, boring a 64-inch cylinder16 required twenty-seven working days, and even cannon-sized cylinders could take a week. But New Willey’s customers, the Royal Office of Ordnance, were willing to wait, since the alternatives were even worse; without access to Wilkinson’s technique, the boring of a forty-inch cylinder for the Philadelphia Water Works took more than four months.

  Wilkinson’s patent was intended to serve the arts of war more than those of peace, but, like radar, penicillin, and the interstate highway system, it took an unexpected turn. While the finer parts of Watt’s engines—the valves, governors, and especially the condenser—could be made by Boulton’s craftsmen at Soho, an economical engine needed a large cylinder, made of iron, and that demanded the expertise of others. Darby’s Coalbrookdale works, and the Carron Ironworks owned by Roebuck, had been the source for the earlier engines and other experimental versions, but Watt was regularly disappointed with the quality of their work, which he called, with his characteristic perfectionism, “unsound, and totally useless.”17 Even with the earlier innovations in smelting and casting, no one in Britain seemed able to cut the bore of a cylinder in the precise shape of the piston that needed to move within it—ideally with no friction and no leakage of air. Wilkinson’s nearly perfect boring system solved this problem.

  Boulton and Watt may have encountered Wilkinson as early as 1768, when all three were, briefly and simultaneously, in Birmingham. Certainly they were corresponding not long after the new boring machine was patented; one of the earliest letters to Watt from Wilkinson—who had been hired, in April 1775, to produce the cylinders for both of the demonstration engines—reveals that the ironmaster had as much visionary zeal as Boulton himself: “I wish to do all in the best manner18 and to start fair. Let us only succeed well in these first engines, particularly in mine, and I will venture to promise you more orders than will be executed in our time…. Our time in this world (at best) is but short and we must be busy if you intend that all the engines in this Kingdom shall be put right in our day.” For twenty years, beginning in April 1775, when Wilkinson was enlisted to make his first steam engine cylinder for Boulton & Watt, he would remain virtually their sole supplier.

  In many ways, Wilkinson had as much invested in the success of the first Boulton & Watt engines as the eponymous firm’s owners. He had seen the revolutionary impact of the New Willey engine not only on his own business—it would eventually run not only his bellows but also the forge’s stamping hammers and presses—but on all of manufacturing. If the new design caught on,19 it would create a gigantic new market for symmetrically bored cylinders—a product on which he had, as a result of his 1774 patent, a de facto monopoly.

  Boulton was delighted with the ironmonger’s enthusiastic embrace of the engine, and shared his belief in the steam-powered factory. However, because he had successfully extended the Watt patent until the year 1800, he had a slightly different timetable, and his plan “to make [engines] for all the world” ran through Cornwall.

  CORNWALL, THE SOUTHWESTERN TIP of the British Isles, has been dotted by tin and copper mines since at least Neolithic times. The two components of the alloy that gave the Bronze Age its name were being exported from Cornwall not later than 1000 BCE. When Julius Caesar invaded Britain in 55 BCE, his objectives included Cornish tin and copper.

  The men who worked in Cornwall’s mines might have thought coal mining in Yorkshire something of a vacation: The sedimentary sheets of coal could be mined with horizontal tunnels off a single mineshaft, while the copper and tin of Cornwall, originally formed from fissures in granite, could only be mined out of individual shafts. By the middle of the eighteenth century, those shafts were the deepest in Britain, as much as e
ight hundred feet down. As a consequence, the typical Cornish miner traveled to and from his place of work either by whim—placing his foot in a loop of rope lowered by a mule—or by ladder, which was not only dangerous, but exhausting. A contemporary traveler described deep mining in Cornwall thus:

  With hardly room to move their bodies,20 in sulphureous air, wet to the skin, and buried in the solid rock, these poor devils live and work for a pittance barely sufficient to keep them alive, pecking out the hard ore by the glimmering of a small candle, whose scattered rays will hardly penetrate the thick darkness of the place.

  The work was brutal, the business eccentric. In Cornwall, landowners rarely worked the land themselves, instead leasing a “sett” for a period of twenty-one years. The lessors were generally consortia known as “adventurers” (“venture capitalist” is cognate) who put up all costs, and paid the property owner a royalty of between 1⁄15 and 1⁄32 of the value of the ore they were able to extract. Adventurers, in turn, appointed “captains”21 who were responsible for mining operations, including the hiring and management of frequently unruly miners.

  Because of this system, Cornish mines were legendary for accounting practices so arcane that, in the words of a correspondent of the Cornish Telegraph, “shareholders might grumble22 over the price of a pennyworth of nails, and pass over without comment the price charged for a steam-engine.” The wide range of expertise and interest among those shareholders meant that in what was one of the world’s most class-stratified societies, miners, ironmongers, and nobles sat down for dinner four times a year at quarterly “count house dinners.”

  Inevitably, drainage topped the agenda of these meetings. Because of their depth, draining Cornish mines was absolutely essential, and pumps were almost always needed, in some cases supplemented by drainage tunnels, or adits, as elaborate as the mines themselves. One of them, the Great County Adit,23 covered more than twenty square miles of Cornish mine country, and in the second half of the eighteenth century, more than half of the most advanced steam engines in the entire world were situated over the Great Adit. As Matthew Boulton realized, the demand for steam pumps, and the distance from coal mines, made Cornwall the perfect laboratory for any invention seeking to reduce the cost of mining.

  Boulton’s insight would produce a genuinely innovative (not to say genuinely weird) business model. Boulton & Watt took shares in Cornish mines in return for providing engines to drain them, negotiating royalties of one-third of the difference in cost between a Newcomen-type atmospheric engine and a Watt machine doing the same work. This in turn demanded a more precise way of calculating the “same work.” The most common calculation of performance was in terms of the so-called “duty,” a measurement of the pounds of water raised one foot by a bushel of coal (confusingly, 84 lb. of coal in Newcastle, 88 lb. in London24). A high-performing Newcomen-style engine typically generated a duty of between five and nine thousand pounds; that is, a bushel of coal could lift that many pounds of water. A 1778 Watt engine, with separate condenser, achieved a duty of 18,900 pounds.

  This was fine for comparing two kinds of steam engines, but less persuasive for customers who were still using wind or animal power to pump water. Watt himself came up with an alternative, measuring the pressure produced by a traditional atmospheric engine, which averaged about seven pounds per square inch, and comparing it with a Boulton & Watt engine using the separate condenser, which generated ten and a half. He then, in his careful way, recorded the effective load lifted by the two engines, using beams of the same length and weight, for the same time period. Finally, he converted it into something he called “horsepower.”

  The term by then already had a long and fairly inexact history. For millennia, animals (and, all too frequently, other humans) had done most of humanity’s work and were the only power sources with variable costs. The outlay for waterpower, which was, by the eighteenth century, even more widely used and easier to measure, was entirely fixed: the construction of a waterwheel. As a result, horses were a favorite way of comparing one sort of pay-as-you-go energy with another. In the first decade of the eighteenth century, Savery himself had promised that his “impellent force” engine could “raise as much water as two Horses25 working together,” and others frequently used horse equivalents as an engineering shorthand. Getting precise about it, however, was a bit more difficult, even when everyone agreed about the need for a standard number. J. T. Desaguliers, the experimental philosopher and lecturer, thought one horse could lift 27,500 pounds a distance of one foot during one minute (or 2,750 pounds five feet in thirty seconds … you get the idea); Watt came up with 33,000 or 550 foot-pounds a second (approximately the same as 1,000 pounds at 3 mph for twelve hours a day), which is essentially the same number in use today. Rough though the measure was, horsepower was a relatively effective way to calculate the work done by pumping engines, since the weight and height of the water lifted are the only measurements required.

  Watt’s horsepower offered a reasonably fair way to balance one engine against another. Boulton’s royalty system, however, tilted the balance heavily toward the engine seller, since it obliged buyers to pay one-third the difference for engines based on an ideal annual performance, while the actual engines could be out of commission for months at a time. Even when operating, they frequently weren’t replacing water or horse wheels, but supplementing them, yet intermittent use still demanded constant payment. It was a powerful disincentive against buying a bigger engine appropriate for the future needs of a business, since buyers had to pay a royalty based on an increase in use right away.

  Boulton & Watt could afford to ignore any problems the system engendered so long as they were the sole suppliers of the world’s most efficient steam engines. In 1795, though, the Birmingham manufacturers discovered that John Wilkinson had been not only providing Boulton & Watt with steam engine parts, but using the same design to sell them on his own, without seeking permission or—far more important—paying a royalty for the privilege. The violent correspondence and litigation that followed is an object lesson in the principle that while ideas may want to be free, their creators very much prefer to keep them leashed.

  As a result of the hostility26 between Wilkinson and his onetime Birmingham customers, they started casting the components of Boulton & Watt engines themselves, building their own ironworks at the Soho Foundry, about a mile from the original Manufactory. But whether the pieces were cast at New Willey or Soho, buyers of Boulton & Watt engines were not getting, except for the cylinder and condenser, a finished engine. In essence, the company sold a kit—no charge for the directions—with “all the cast iron,27 hammered iron, brass, & copper work … including the screw bolts of the cistern, fire door, grate barrs [and] the flywheel with its shaft.” They didn’t include the boiler, the framing for the engine house, or any of the masonry to hold the contraption, which buyers had to supply themselves, though Boulton & Watt would supply plans and estimates. Even the wooden beam itself that operated all the pumping engines was the responsibility of the buyer, though they did provide engineers to see to the final installation.

  It gets better. The royalty charged by Boulton & Watt was based on the difference between the cost of the new engine and the cost of a Newcomen-style pump: the greater the savings, the higher the patent royalties. But even though running the engines was at least as demanding as building them, Boulton & Watt did not supply operators, who might have simultaneously seen to the (usually) smooth operation of the engines and their usefulness to the buyer.

  The commercial relationship was further complicated by the usual problems of any industry in its infancy. To cite one critical problem, all the heavy goods needed to be transported from foundry to site at a time when the canal system was barely started—at least one Boulton & Watt engine was too large28 to travel by canal through the Harecastle tunnel—and Rocket was still fifty years in the future, to say nothing of a national rail network. In even shorter supply than transportation were skilled workmen—
and they were needed not only to build the “kits” provided by Boulton & Watt (each one custom made, and no two parts assembled anywhere before they arrived on site) but to do so with extreme precision, with the shaft set at an exact right angle to the main beam and perfectly level beneath it and the pistons needing to touch either the top or bottom of the cylinder. It was in response to these demands29 that Boulton & Watt became what we would today call professional publishers, producing Directions for Erecting and Working the Newly-Invented Steam Engines in 1779.

  Boulton’s plan depended on achieving at least a dozen difficult objectives, each one independent of the other, and each necessary: building a factory that could both cast and machine iron; persuading a group of adventurers to partner with him on a hypothetical cost advantage in fuel (whose future cost couldn’t even be predicted with any certainty); training a new workforce to build and operate his engines. Against all odds, it worked. Over the course of five years, Boulton & Watt’s steam engines became the pump of choice in the majority of Cornish mines.

  Beginning with the installation of an engine at the Ting Tang Mine in November 1776, and the Chacewater Wheal Busy mine a month later, Watt and Boulton succeeded in impressing Cornwall’s mine operators, and even the engineers who had been building and servicing Newcomen-style machines for decades. The Chacewater engine, particularly, had been heavily promoted by Boulton as yet another demonstration of the greater efficiency of the new design—he invited hundreds of engineers and mine owners from all over Cornwall to witness the trials, and Watt himself supervised the engine’s erection. “All the world are agape,”30 wrote Watt, “to see what it can do,” and it gave “universal satisfaction to all beholders, believers or not…. Our success here has equaled our most sanguine expectations [and while] our affairs in other parts of England go on very well, no part can or will pay us so well as Cornwall.”