Up until the 1780s, steam engines were used almost exclusively for pumping water. To the extent that they drove industrial machinery, it was almost always indirectly, by lifting water uphill from whence it could run back down and turn a waterwheel.
Industry thus remained dispersed in villages and rural areas, where easy access could be had to running water. So too did the steam engines, which mainly worked to drain mines in the countryside. To the extent that coal besmirched the skies and walls and British cities, it was a product of home heating, not “dark, Satanic Mills.” Three forces drove the transformation of the steam engine into an urban and industrial machine – the new levels of fuel efficiency achieved by Watt, an explosion in demand for power, and a new style of steam engine that produced a rotary motion.
The demand came from a new industry that spread across the British countryside, especially in the county of Lancashire around the industrial town of Manchester: the making of cotton cloth. For centuries, Britain had been a country of woolmakers. Sheep thrived on the rolling green fields of England and Scotland, and wool cloth had been a major domestic and export industry since the Middle Ages. In the middle of the eighteenth century, roughly one in every ten Britons was still involved in the woolens industry – raising and shearing sheep; spinning yarn; and knitting, weaving, fulling, or dyeing cloth.
Cotton, on the other hand, remained an exotic import. Unlike sheep, cotton did not thrive in the British climate, nor did it usually make economic sense, given the cost of British labor, to import raw cotton and spin and weave it locally. Instead, the bulk of cotton cloth in Britain came as a manufactured product from India, and often went by the name calico,after the Indian city of Calicut (now known as Kolkata). Like linen (the only other readily available fabric except for luxurious silk), cotton was more comfortable against the skin than wool, and it could hold detailed patterns and luminous colors better than wool or linen. It therefore became popular and fashionable as soon as imports began in the seventeenth century, especially among those who could not afford silks. This increasing popularity infuriated the domestic wool interests and silk weavers, leading to riotous attacks on calico-wearers and, eventually, the 1700 and 1721 Calico Acts, which banned first the import and then the sale of imported cotton cloth.
They did not, however, forbid the import of raw cotton. And so, a domestic cotton cloth industry slowly developed, despite the combined cost of importing fibers by ship and then paying expensive British workers to turn it into cloth. Meanwhile, slowly, over the course of the eighteenth century, several entrepreneurs figured out new ways to improve the efficiency of the production steps required to turn raw fibers into cloth. For example, John Kay’s fly shuttle of 1733, which allowed a single artisan seated at a loom to sweep the bobbin holding the weft yarn from one side of the warp to the other with a quick flick of the wrist. By eliminating the need for an assistant on either side of the loom to pass the shuttle back and forth, this invention reduced the cost of making British broadcloth.
Richard Arkwright was the first to exploit the potential of cotton making on an industrial scale. As a young Lancashire man, Arkwright plied the cephalic trades of barber and wig-maker, and made a tidy profit from his invention of a new water-proof wig dye. In 1767, he decided to invest his earnings in the cotton business. With the help of another John Kay (no relation) he devised a new machine for spinning cotton fibers into thread. In fact, the ratio of helper to helpee almost certainly ran quite in Kay’s favor, given his expertise with fine machinery as an experienced clockmaker.
Creating a spinning machine presented a serious challenge because it had to combine two motions – first a clump of short cotton fibers (called the roving) had to be drawn out into a longer piece, with just the right speed and force – too hard, and the fibers would simply pull apart, too gentle and they would remain clumped. Then they had to be twisted around one another to form a thread. Kay and Arkwright’s machine, borrowing from but improving upon other machines from earlier in the century, used a combination of weighted rollers to draw the cotton and rotating flyers to spin it. Their invention made it possible to drive spinning machinery with water power on a large scale.
Having perfected his machines, Arkwright enlisted some partners, and together they erected the first large cotton-spinning factory at Cromford near the River Derwent between Sheffield and Nottingham. It began operating in 1772, employing two-hundred people (mostly women and children) twenty-four hours a day (in two shifts) to help operate and tend to the machinery. The new spinning machines, being the first of their kind to be driven by a waterwheel, became known as “water frames.”
Within the decade, however, Arkwright’s new cutting-edge factory was already obsolete, a pace of change all but unknown to the previous five millennia of human history, but soon to become commonplace. Samuel Crompton was another son of Lancashire, but with rather humbler beginnings than those of Arkwright. His father died young, and with no trade to his name, the young Crompton helped support his family by operating a “spinning Jenny,” a machine that allowed a single spinner to operate many spindles at once. After growing quite familiar with the machine, Crompton could clearly see its defects, and set out to eliminate them. The machine he built acquired the name “the mule,” because it blended the traits of Jenny and water frame. Like the water frame, it required only machine tenders, not operators, but it produced far superior thread in its fineness and regularity – superior in fact to what most hand spinners could accomplish.
Unlike Arkwright, who quickly secured his patent rights, the naïve Crompton all but gave away his invention and never found fortune for his achievement. And such an achievement it was. By 1811, over four million spindles were gathering thread from Crompton-style machines in the Lancashire hinterlands around Manchester. Britain could now make far more cotton cloth domestically than it used to import from India. In the 1780s alone, imports of raw cotton into Britain increased from five million pounds to thirty million, then quintupled again in the following decade. Prices for cotton thread plummeted at the same time – a pound of 100-count thread would set one back 38 shillings in 1786 to but only 9-and-a-half by 1800. The result was to transform cotton from a luxury for the gentry and mercantile class into an everyday fabric for the entire citizenry.
Steam Mill Mad
By the latter half of the eighteenth century, nearly all the available sites for waterpower around the major British industrial towns – Birmingham, Nottingham, Manchester, and Sheffield – were occupied. Legal battles raged over claims to the fall of water on each stream – claimants expended 10,000 pounds a year on water-related cases in York and Lancaster counties alone in the late 1700s.
Cotton only added more stress to this already difficult to situation. A typical, traditional water mill for grinding grain or fulling cloth used perhaps five horsepower, versus ten to twenty for an Arkwright-style factory. Cotton concentrated particularly in the Lancashire region northwest of Manchester, with easy access to the steady streams of water pouring off the Pennines to the east, the port of Liverpool to the west, and the transport hub and industrial trades of Manchester itself. But competition for good locations proved so fierce that by the 1780s, some spinners had to relocate to other parts of the country while others resorted to literal horsepower – machinery driven by horse wheels.
In 1781, Boulton, eagerly eyeing these developments, wrote to his partner Watt to prod him to renew his inventive powers in order to take advantage of the growing demand for power. “The people in London, Manchester & Birmingham are Steam Mill Mad, & therefore let us be wise and take ye advantage,” he wrote. In a separate letter he warned Watt that the market for mining pumps had nearly been saturated, and “…there is no other Cornwall to be found.”
For the steam engine to enter this new market, the mechanical structure of the engine had to be adapted accordingly. Instead of the reciprocating motion of a pump, it had to produce smooth, even rotation – any variation in speed would result in “wobbly” thread that got thicker and thinner along its length. The great John Smeaton himself expressed skepticism that it could be done: “I apprehend,” he wrote, “that no motion communicated from the reciprocating lever of a fire-engine can ever produce a perfect circular motion, like the regular efflux of water in turning a water-wheel…” He bet instead on hybrid steam/waterwheel systems. Mills using Newcomen engines to pump water uphill to supply their wheel dated back decades, but the concept reached its perfection with Joshua Wrigley’s fusion of a Savery-style steam pump with an overshot waterwheel. The action of the wheel itself worked the valves of the engine. The wheel thus effectively supplied itself with water. Wrigley’s contraption required no actual natural stream, simply a reservoir to draw water from, and so it provided the same geographical freedom from waterways that the steam engine did. But this roundabout approach could not compete in efficiency with a direct rotary steam engine – if it could be made to work.
Thought it required prompting from Boulton to get Watt moving on the rotative engine, he did not lack for ideas on the subject. His interest in rotary motion went all the way back to his initial impetus for thinking about steam engines in the first place – his friend James Robison’s notion that a steam engine might be used to drive a carriage, rather than horses. The idea of solving the problem of rotary motion remained in his mind, and in parallel with the reciprocating steam pump design that made him famous, Watt had designed a very different kind of engine specifically to do so. He called it the steam wheel.
The steam wheel consisted of a hollow circular tube, divided into three chambers separate by one-way valves. Central spokes supplied each chamber with steam. When steam entered one of the chambers, it forced a bolus of mercury through the rightward valve into the next chamber, creating, per Newton, an equal and opposite force on the closed valve to the left, and thus pushing the whole wheel around in a clockwise fashion.
Watt had, in effect, designed an elaboration of Heron’s whirling engine, with the steam forced to stay inside a tube where it could do real work, rather than spewing straight into the atmosphere. By using a “piston” of liquid metal, he had also neatly solved the problem that had bedeviled him with his steam pump, of sealing the piston against the walls of the cylinder. But despite years of tinkering, Watt never got this design to work effectively (a lucky thing for the health of the British public, considering the hazards of mercury vapor).
It was, it turned out, far simpler to transform the rocking of a beam into a rotary motion than to get steam to create rotation directly (success in that endeavor would come much later, with steam turbines). So much simpler that two other inventors beat Watt to the patent office with the idea. Over the previous decade, Boulton and Watt had developed a whole stable of skilled mechanics to supply them with engine parts and help them erect engines on site. Among them was Matthew Wasborough, a mechanic and clock-maker from Bristol who had for a time supplied parts to Watt for his Cornish engines. Wasborough patented a new engine design in 1779 that used pulleys and ratchets to create a one-way rotary pull, and then a flywheel to smooth the rotation.
John Pickard, a Birmingham button-maker, worked with Wasborough to improve this design further, by substituting a simple crank for the difficult pulleys and ratchets. Pickard patented this improvement in 1780. After the fact, both pro- and anti-Watt partisans claimed angrily that the crank was stolen from their team. A nosy party in either camp could certainly have easily spied on the other, since Wasborough and Pickard had erected an engine in Birmingham, not far from the Soho works where Watt did his experimentation. But it is equally likely that the idea came to each independently – that a crank could turn reciprocating motion into circular motion or vice versa had been known for centuries.
Wasborough died of a fever in October 1781, aged just twenty-five. But the Pickard crank patent still stood in Watt’s way. Watt set his mechanics to work to find a way around it. William Murdoch, the most skilled of the mechanics in Watt’s employ, came up with the solution. A Scotsman and the son of a tenant farmer, Murdoch had walked 300 miles from his hometown of Auchinleck to ask for a job of Watt in Birmingham in 1777, when he was twenty-three years old. Later he would go on to invent an improved steam-engine valve, a model steam carriage, and a pneumatic message system.
Murdoch’s solution to the crank problem was the so-called “sun-and-planet” gear – a planet cogwheel attached to the steam engine beam orbited around a sun cog attached to the drive shaft of the machinery. Mechanically, Murdoch had reinvented the crank by way of Rube Goldberg, but it was sufficiently different in form to legally bypass Pickard. As for the flywheel, Watt apparently assumed that this was too well known a principle to really be patented, and went ahead and included one on in his design. To further smooth out the stroke of the engine and therefore the rotation of the machinery, Watt also introduced “double-action” to the design – that is to say, the use of steam to push the piston both up and down. A valve controlled the supply of steam, connecting each side in turn first to the boiler to accept fresh steam, then to the condenser to exhaust it.
But as happens so often in engineering, pushing down a problem in one place forced another up somewhere else in the design. The moving end of the beam traced a curve, while the engine’s piston rod moved straight up and down. Since Newcomen, this discrepancy had been solved by attaching the piston to the working beam with a chain that curled around an arc as the piston pulled down. But the piston of the double-acting engine also had to push the beam up, and pushing a chain would get you nowhere. To solve this Watt devised a little metal parallelogram that he called the “parallel motion.” It translated between the two motions, curved and straight.
To further smooth the engine’s motion and confute Smeaton’s skepticism, Watt added a flyball governor. This consisted of two weighted arms that controlled the throttle valve that admitted steam to the cylinder. If the engine ran faster, the governor would spin faster, lifting the arms, constricting the valve, and so reducing the speed again. The reverse happened when the engine slowed down, admitting more steam to speed it up. This kept the engine operating at a constant rate of speed.
Watt’s design for a rotative engine, though not first, was best. The double-action and other refinements made it more efficient, smooth, and powerful than its competitors. Per tradition, Boulton and Watt provided an early working model to Wilkinson in 1783, to operate a hammer at his ironworks. They followed up with two fifty-horsepower beasts for Albion Mills on the Thames, which would drive a total of twenty pairs of millstones for grinding flour. The partners sold 278 total rotative engines before the fundamental Watt patent expired in 1800, surpassing the 171 pumping engines sold to that date. The purchasers included “lead works, rope works, malt distilleries, sugar, tobacco and snuff manufactories,” and more, but by far the most eager customers were the cotton spinners of Lancashire, who acquired forty-four of Boulton and Watt’s engines.
The Steam Revolution?
There is a certain school of historical thought, one that became particularly popular in the latter decades of the last century, that pooh-poohs any claims about abrupt, transformative change. According to this way of thinking, Rome never fell, it merely aged gracefully into Late Antiquity. Modern science did not suddenly spring into being in the 1600s, scholars fitfully adopt a variety of newish knowledge-making strategies in a piecemeal fashion over several centuries. And the supposed industrial revolution of the late eighteenth and early nineteenth centuries was an illusion, full of noisy engines and furious machinery, but signifying nothing in terms of economic statistics.
All the worse for the economists and their statistics, if they fail to reveal the great qualitative changes to British life that were obvious to everyone at the time. Aside from the aforementioned effects on the culture of clothing, the explosion of British cotton shattered an entire productive sector in India, a major event in the ongoing world-historical shift of economic power from the southeast of the Eurasian landmass to the Atlantic coasts of Europe and North America. Soon the landed gentry of the American South would reorient its entire existence around providing cotton fibers for British spindles, binding it irrevocably to the cause of slavery and setting the stage for the Civil War. And this is considering only one industry, albeit by far the most dynamic one in the early stages of industrialization.
What, then, can we say about the steam engine? Was there a steam engine revolution? It certainly did not instantly displace the waterwheel. Even in Britain, water-power continued to overshadow steam-power in both number of mills and operating horsepower well into the nineteenth century. In other countries with more abundant water sources, less industrial know-how, and more expensive coal, water power continued to dominate steam even into the second half of that century. Waterwheel design also underwent its own technical improvements, including the replacement of wooden wheels with cast iron ones, which were lighter and more efficient. But, of course, this substitution itself relied on cheap iron, which in turn depended on cheap coal, which depended on steam.
As we have seen, steam power arrived just in time to resolve a power crisis in Britain, due to the near-exhaustion of available waterways for traditional mills. But this purely quantitative role, as a new source of power too tap into, did not constitute a steam engine revolution, either. Two other factors did, however: its consistency across time and its indifference to geographic location.
Pre-industrial society moved to the rhythms of the day and the season. Almost all work happened on farms, where the needs of plants and animals drove the hours and season of work, and winter provided a natural lull between harvest time and the next year’s planting. Waterwheels did little to change this. A water-powered mill was subject to the yearly cycle of wet and dry seasons, winter frost, and random interruption by the chance events of flood and drought. The steam engine, however, drawing on a store of eons-old solar energy banked deep beneath the earth, was subject to neither the cycles nor the whims of nature. Industrial time, measured in equal increments and decoupled from the turning of the seasons, was aborning – with the help of another product of the same mines, artificial light from coal gas.
Even as it loosened the bonds of time, the steam engine did the same for those of geography. A water-powered mill had to go wherever water power could be found. A steam-powered mill, on the other hand, could be sited wherever was most convenient. The ideal site was a transportation hub where fuel and raw materials could arrive (and finished goods could depart) quickly and cheaply. This nearly always meant a city.
The cotton mills of Lancashire before the steam engine were scattered across the countryside, wherever an available stream could be found. But in 1782, Arkwright built Shudehill Mill, a steam-powered factory, directly in the city center of Manchester. At first it used a pumping engine to supply water to a water wheel, then later several of Watt’s rotating engines were installed. It was the first of many such enterprises. Fifty years later, in 1835, Alexis de Tocqueville visited Manchester and described the grim yet sublime complexion the city had acquired:
A sort of black smoke covers the city. The sun seen through it is a disc without rays. Under this half daylight 300,000 human beings are ceaselessly at work. A thousand noises disturb this damp, dark labyrinth, but they are not at all the ordinary sounds one hears in great cities. The footsteps of a busy crowd, the crunching of wheels of machinery, the shriek of steam from boilers, the regular beat of the looms, the heavy rumble of carts, these are the noises from which you can never escape in the sombre half-light of these streets.
… From this foul drain the greatest stream of human industry flows out to fertilise the whole world. From this filthy sewer pure gold flows. Here humanity attains its most complete development and its most brutish; here civilisation works its miracles, and civilised man is turned back almost into a savage.
This kind of industrial city, belching smoke and steam, bustling with Gradgrinds and Bounderbys, was a product of the steam engine, and the fruit of the steam revolution.
 Phyllis Deane, “The Output of the British Woolen Industry in the Eighteenth Century,” The Journal of Economic History 17, 2 (June 1957), 213.
Until 1774, however, when the ban on selling cotton cloth was fully lifted, most British cotton yarn was blended in the finished fabric with other fibers, such as the flax-cotton blend known as fustian. This avoided any potential legal trouble.
 There was precedent in the silk industry for large-scale spinning mills in Italy and England, such as the Lombes mill, also on the Derwent, which employed several hundred women to tend its machinery. But the problem of spinning silk was a simpler one, since it involved only a single motion to twist the fibers around one another. Robert Friedel, A Culture of Improvement (Cambridge, MA: MIT Press 2010), 222.
 100-count means that 100 hanks of thread weigh one pound. A hank is 840 yards long. So, 8400 yards of 100 count thread weighs one pound. The finer the thread, the greater the length in a pound, and so the higher the count. Friedel, 231.
 Friedel, 232. Dare I say it became “the fabric of our lives”? Only in a footnote, I guess.
 Terry S. Reynolds, Stronger than A Hundred Men: A History of the Vertical Water Wheel (Baltimore: Johns Hopkins University Press, 1983), 267-68.
 Quoted in Russell, 149.
 Quoted in Richard L. Hills, Power from Steam: A History of the Stationary Steam Engine (Cambridge: Cambridge University Press, 1993), 49. Smeaton wrote these words as part of a report to the Commissioners of the Navy, which helped convinced them not to use a Wasborough engine (see below) to grind corn for their victualling yards.
 Reynolds, 323-24.
 Douglas Self, “Steam Wheels,” The Museum of Retro Technology (http://www.douglas-self.com/MUSEUM/POWER/steamwheel/steamwheel.htm, accessed October 1, 2021). The steam wheel is described as the fifth item in Watt’s patent, https://upload.wikimedia.org/wikipedia/commons/0/0d/James_Watt_Patent_1769_No_913.pdf (accessed October 1, 2021).
 Information about Pickard and Wasborough (whose name is spelled variously in the sources) is relatively scanty, and most sources report only Watt’s point-of-view. Grace’s Guide to British Industrial History gathers some pro-Wasborough sources: “Matthew Wasbrough”, Grace’s Guide to British Industrial History (https://www.gracesguide.co.uk/Matthew_Wasbrough), retrieved October 2, 2021. An account of the events from a Watt partisan can be found in Samuel Smiles, Lives of Boulton and Watt (London: William Clowes and Sons, 1865), 289-293, though it somewhat muddles the rolls of Wasborough and Pickard.
 Watt’s new design is discussed in Russell, 149-152 and Hills, 60-69. The parallel motion did not actually produce a perfectly straight vertical motion, but it was close enough.
 Richard Woollard, “Albion Mill,” The Vauxhall Society (https://web.archive.org/web/20141018073918/http://www.vauxhallcivicsociety.org.uk/history/albion-mill/).
 Russell, 152.
 Peter Brown, The World of Late Antiquity: AD 150-750 (New York: W.W. Norton & Company, 1987); Steven Shapin, The Scientific Revolution (Chicago: The University of Chicago Press, 1996); N. F. R. Crafts, “British Economic Growth, 1700-1831: A Review of the Evidence,” The Economic History Review 36, 2 (May 1983), 177-199. More recently, the computer revolution, too, has proved a supposed flop, failing to make any dent in productivity statistics in the 1980s and 90s, a conundrum known as the Solow Paradox. Robert Solow, “We’d Better Watch Out”, New York Times Book Review, July 12, 1987.
 Reynolds, 325-330.
 Horses could also provide more dependable power, turning a horse wheel, but at much greater cost.
 Alexis de Tocqueville, Journeys to England and Ireland (New York: Arno Press, 1979), 107-108.