concerning a Cornish Giant, and a trip up Camborne Hill; the triangular relationship between power, weight, and pressure; George Washington’s flour mill and the dredging of the Schuylkill River; the long trip from Cornwall to Peru; and the most important railroad race in history

THE CAUSE OF THE ACCIDENT at Poldory Mine in January 1784 is not precisely known, nor is its date; some accounts have it occurring on the sixteenth, others on the nineteenth. Its effect on the mining families of Cornwall’s Gwennap Valley, about two miles from the town of Redruth, was depressingly routine: tragic for the six families that lost their breadwinners, a reminder to everyone else of the dangers of mining—cave-ins, floods, fires, and suffocation.

The risk of each hazard increased in direct proportion to the mine’s depth, and by 1784, Cornwall’s copper was being carved from seams several hundred feet below the surface. This made Poldory, and its neighbor, the Ale-and-Cakes,* utterly dependent on the Boulton & Watt engine pumping the water out of the two mines and into the Great County Adit that drained a good portion of central Cornwall. If the engine didn’t work, neither did the miners.

The reliance on steam power, however, had introduced a new danger: catastrophic failure of the engine itself. Not all steam engine failures are catastrophic—complete and sudden. Valves can stick, or beams crack, harming only the engine itself. A failure of the engine’s cylinder housing or boiler is different. The sudden release of steam under pressure is literally explosive, sending shards of metal flying outward at several hundred feet per second, driven by jets of scalding steam. It was precisely this sort of accident that killed three miners, and maimed several more, at Poldory, and when the news reached the engine’s designer and builder, his reaction was predictable. On January 24, James Watt wrote to Thomas Wilson, a mine owner living in the town of Truro in Cornwall,

I am exceedingly shocked1 at the account of the accident at Poldory and should have been Glad to have had some particulars. They must certainly have had a very strong steam otherwise, the people would have had time to escape. Please also to advise who the people were and how so many came to be about the boiler; Copper tubes must be entirely given up without men can be found more carefull [sic] in the management of them. If any of the families of the deceased or the surviving persons who were scalded are in distressed circumstances, I am sure that Mr. B[oulton] will Join me in being pleased that you should give a small matter for their immediate relief as if of your own accord without mentioning our names …”

The phrase “strong steam” is telling. By 1784, Watt was decades removed from his first experience with Newcomen engines, but those decades had done nothing to ease his fear of the caged power of steam under pressure. Some of the concern dates to his earliest experiments: two years of sealing materials that failed to seal, tubing that leaked, and cylinders with seams that burst. The larger part, however, was an analytical blind spot, one that he shared with the most sophisticated scientists of the eighteenth century.

That blind spot, about the nature of the relationship between heat and motion, was no longer the belief that phlogiston was released whenever anything was burned. The failure of phlogiston theory to account for any number of observed phenomena had opened the door for another, more useful though still flawed, to take its place. This was the idea that heat remained a substance, but a weightless one, called caloric. The Scottish philosopher William Cleghorn, another protégé of Joseph Black, theorized that this “subtle fluid” (in the words of Lavoisier, who also coined the word “caloric”) was a gas whose properties included varying levels of attraction to different types of matter, thus explaining why the heat capacity of coal was different from that of glass. The theory further held that caloric could be neither created nor destroyed, but only changed from Black’s latent heat—the potential locked up in a combustible substance—to sensible heat, and then back again, with the total amount of caloric in the universe staying constant.

Caloric theory remained the conventional wisdom for sixty years or more for a simple reason: It explained, better than any alternative, dozens of physical phenomena. Hot fluids cool, in caloric theory, because caloric repels itself, thus diffusing from an area of high concentration to a lower one. It explains heat radiation and Boyle’s Law, and even formed the basis, forty years after Poldory, for Sadi Carnot’s Réflexions, and the first working theory of steam engines: that their capacity depended only on the difference between high temperature and low temperature, which, in Watt’s steam engine, was the difference between the temperature of the boiler and that of the condenser.

This didn’t mean that no one was thinking outside the caloric box. There was, for example, the thoroughly remarkable Benjamin Thompson of Massachusetts, a loyalist American who, after backing the losing side in the Revolutionary War, moved, first to England (where in 1779 he was made a Fellow of the Royal Society), and four years later, apparently on a whim, to Bavaria. There he found himself, on behalf of the Prince-Elector2 Karl-Theodor, running an espionage network that stole design sketches from the Soho Manufactory and spirited them out of England. For this and other services (including the invention of Rumford Soup, a concoction of peas, barley, potato, and old beer intended to meet the nutritional needs of Europe’s poor) he was made a Count of the Holy Roman Empire in 1791.

Seven years later, Count Rumford (as he had styled himself, after the New Hampshire town where he had been a schoolmaster) performed an experiment investigating the nature of heat, inspired by his observation of Bavarian metalworkers boring out a brass cannon barrel, which generated enough heat from friction that the barrel was too hot to touch. Lavoisier’s theory predicted that the caloric associated with drilling should have melted the brass shavings into which it presumably had been transferred. Since it did not, Rumford tested it, rather cleverly, by boring a cannon barrel underwater. The borer’s friction produced enough heat to keep the water boiling—and the water continued boiling as long as the borer was spinning, thus disproving the idea that caloric was a property contained within matter, since it was never exhausted.

If caloric was not a fluid, it had to be something else. Rumford’s monograph, An Experimental Enquiry Concerning the Source of the Heat Which Is Excited by Friction, argued that the “something else” was actually motion: that heat and motion are essentially the same thing. This was critical, and surprisingly slow in coming. It is, after all, not hard to find places outside the laboratory where mechanical work creates heat: rubbing two Boy Scout–approved sticks together, for example. John Locke himself observed the heat produced by the mechanical energy of a wheel rubbing against its axle.

But even though Count Rumford shot a very large hole in the idea of caloric, theories are overturned only by better theories, and he didn’t really have one to offer. So long as caloric theory was still an effective way of approximating physical reality, it wasn’t going away. This was a real obstacle to engineers, since a key element of caloric theory confused the nature of work in steam engines. Caloric theory held that heat was latent3 within combustible materials and was simply converted from latent to sensible, which “proved” that there could be no advantage to a high-pressure engine, which would simply increase sensible heat at the expense of latent heat.

This was confounded by another blind spot: Until the middle of the nineteenth century, scientists and engineers alike widely believed that the source of a steam engine’s work was the pressure of expanding steam on a moving piston. More pressure, therefore, equaled more work. So far, so good. But the analogy they used as a model was waterpower, and that was not so good. A waterwheel produces the same amount of work when relatively little water flows over a long distance as it does when a lot of water flows over a short distance; a waterwheel will make as many revolutions when it catches a trickle that has fallen a thousand feet as when it is driven by a river falling a few inches. By the same token, they reasoned, a steam engine would be just as efficient using low pressure to drive a long beam as high pressure to drive a shorter one—and steam, as the families of the Poldory miners would certainly testify, is a lot safer at low pressure.

Reasoning by analogy is always suspect. What was missing in this example was an understanding that it was not the pressure that mattered, but the heat—heat energy and mechanical energy were just different forms of the same thing. The heat needed to raise one pound of water a single degree is equivalent to lifting 772 pounds one foot, and vice versa: lifting 772 pounds one foot generates an equivalent amount of heat.* The slow realization of this relationship is explained largely by the low efficiency of the early engines. If you’re converting only 2 percent of the heat energy into work, you’d be hard pressed to notice its absence. And even when they noticed it, the blind spot endured. No matter how many times engineers observed4 more work being done with more heat, they were unable to make any sense of the results.

This blinkered view didn’t change the need for steam engines to deliver more power, of course; water needed to be pumped from ever deeper mine shafts, factories needed more power to drive larger wheels. The response of Boulton & Watt was Archimedean: longer levers driven by ever larger condensing engines, using cylinders up to five feet in diameter, with strokes of nine or even ten feet. That these behemoths offered a pretty unattractive power-to-weight ratio didn’t seem at the time to be much of a problem; low pressure—about 8 to 10 psi—was reliable, safe, and affordable. A multi-ton condensing steam engine could pull a bucket from a mine shaft five hundred feet deep, or run a dozen shafts at a cotton spinning factory. But it couldn’t pull itself any distance at all.

It wasn’t that no one thought about the possibility of steam locomotion. In 1784, only four months after Poldory, Watt himself described, in the same patent that included the parallel linkage, a piston-driven steam carriage that used “the elastic force of steam5 to give motion.” In 1785, Boulton & Watt’s brilliant engineer William Murdock actually built a “steam carriage” and for the first time filed a patent in his own name. He even built a scale model of his carriage—a cylinder with a diameter of three-quarters of an inch and a stroke of an inch and a half—and decided to exhibit it in London, though there is no reason to believe that it would have been a success. Asking a steam engine to move itself—to say nothing of cargo or passengers—meant making it powerful andlightweight; and the only way to do that was increasing the pressure in the cylinder. With his fortune tied to the future of low-pressure steam power, Matthew Boulton himself intercepted Murdock6 on the way to London and persuaded him to return to building stationary Boulton & Watt engines in Cornwall.

Which was where, in any case, the next revolution in steam engines was going to occur. There, and in America.

BY THE MIDDLE OF August 1787, the fifty-five delegates to the world’s first, and most consequential, Constitutional Convention had been meeting in Philadelphia for three months of an extremely hot summer. They had debated judicial appointments, executive departments, and every conceivable duty of a national legislature. They had established the line of command for the new nation’s armed forces and proposed a national postal service. Given all that, it’s no great shock that the first acknowledgment that the newly proposed federal government had any place in protecting the activities of inventors didn’t come until August 18, when James Madison of Virginia proposed that the national legislature be empowered to “encourage knowledge and discoveries.” The same day, Charles Pinckney of South Carolina submitted a proposal that the government be able “to grant patents for useful inventions.”7

Four days later, on the twenty-second, the convention adjourned for the afternoon and headed to the banks of the Delaware River to see a demonstration of the power of such useful invention: a forty-five-foot-long boat that resembled an Iroquois war canoe, with six oars on either side. The motive force for those oars, however, was not muscle. It was steam.

The steamboat’s inventor, a onetime clockmaker and silversmith from Connecticut named John Fitch, had turned a traditional twelve-inch condensing cylinder on its side and used it to drive a piston with a three-foot stroke tied to an eighteen-inch axle. In his own words, “Each revolution of the axle tree8 moves twelve oars five and a half feet. As six oars come out of the water six more enter the water; which makes a stroke of about eleven feet each revolution. The oars work perpendicularly and make a stroke similar to the paddle of a canoe.”

Fitch’s steamboat was not, as many histories have it, the world’s first. In 1772, two ex–artillery officers in the French army, the Comte d’Auxiron and Charles Monnin de Follenai, received a fifteen-year exclusive license to run a steamboat along the Seine. Unfortunately, their first attempt, a marriage of a Newcomen engine to a Seine bâteau, was less than successful: the engine was so heavy it sank the boat. Slightly more successfully, in 1783, the Marquis de Jouffroy d’Abbans took a 140-foot boat mounting a Newcomen-style engine out on the Saône from Lyon. He did make it all the way back to the dock, where cheering crowds met it—just in time, before the engine’s vibrations destroyed the boat.*

The great importance of Fitch’s steamboat was not that it survived its inaugural trip; it was his audience, who were properly impressed with what turned out to be the steamboat’s maiden voyage. One delegate, William Samuel Johnson of Connecticut, wrote on the twenty-third, “the exhibition yesterday9 gave the gentlemen present much satisfaction and will always be happy to give him every countenance and encouragement in their power which his ingenuity and industry entitles him to.” Two weeks later, the Brearly Committee (named for its chairman, David Brearly of New Jersey, and also, and unfortunately, known as the “Committee of Leftovers”) reported on fourteen proposals to the convention; the last one was a recommendation to “provide limited patents10 to promote science and arts.” The patent clause was incorporated, without a single dissenting vote, into Article I, Section 8, paragraph 8 of the United States Constitution.

It’s easy to see why the American revolutionaries were so taken with the British attitude toward intellectual property. In almost every relevant way, they were British. The common law was as well known on the banks of the Potomac as along the Thames. Virginians signed on to seven-year apprenticeships as carpenters, millwrights, and glaziers exactly as their counterparts did in Yorkshire (sometimes an apprenticeship would begin in the latter and conclude in the former). The éminence grise of the American Revolution, Benjamin Franklin, was not only a Fellow of the Royal Society but also one of the most prolific inventors of the entire eighteenth century; Thomas Jefferson, the revolution’s intellectual soul, took enough time off from his writing and architecture to design revolving bookstands, copying machines, revolving chairs, and even a new and improved moldboard plow. If those models weren’t sufficient, eighteenth-century America showed even more enthusiasm than Britain itself for the intellectual forebears of patent law. Locke was considerably more influential among the American constitutionalists than he ever was to English parliamentarians, and had even drafted the first constitution for the Carolina colony. And not just Locke: the Mayflower had carried a set of Coke’s writings from the Old World to the New, and both Jefferson and Madison had gotten their legal training from reading them. In the eighty-fourth of the Federalist Papers, Alexander Hamilton compared Coke’s 1628 Petition of Right to the Magna Carta.

Fitch’s influence on the future of steamboats was less enduring.* He gave up on oars fairly quickly and experimented with circular paddles and even an early propeller screw, which he used for ferry service between Philadelphia and Burlington, New Jersey, but his failures outnumbered his successes many times over. Though he pursued a national patent, and de facto monopoly, on steam-powered water travel with monomaniacal zeal for years (much of his time exhausted in disputes over priority with another inventor, James Rumsey, who used a jet of water propelled by a steam-driven pump to drive a boat on the Potomac in 1787), the final award of patent, almost three years to the day after his demonstration for the delegates in Philadelphia, was so limited in its language as to be commercially useless.

The fact that a national patent was available at all, however, was the significant thing. One reason that those constitutionalists were even in Philadelphia that summer was the realization that the union originally established in 1777 under the Articles of Confederation and Perpetual Union was responsible for governing a nation that covered a territory bigger than France, Britain, Germany, and Spain combined, with fewer tools than a New England town meeting. The confederation—unable to tax its citizenry or even levy soldiers in wartime, powers that were reserved to the individual states—wasn’t deriving much benefit from its size.

And while national defense was obviously a more urgent deficiency, economic issues were a close second, including the recognition that some national authority needed to promote, and protect, invention. In 1800, there were only a few more Americans living in the New World than Dutch living in the old, and we’ve already seen that the Netherlands, despite its great wealth, was still too small to support British-style inventing; Massachusetts, or Virginia, wouldn’t stand a chance. Inadvertently, the constitution had stumbled on the fundamental issue of scale in intellectual property. The value of a bar of gold or a bushel of wheat—Romer’s rivalrous property—is no greater in a large country than a small one. On the other hand, the value of a nonrivalrous patent or copyright increases in direct proportion to the number of people one can sue to prevent its theft.

Not everyone agreed, even in eighteenth-century America. Thomas Jefferson, most notably, was reflexively offended by even the slightest odor of monopoly; in a much-quoted letter sent to his friend Isaac McPherson, Jefferson wrote:

If nature has made any one thing11 less susceptible than all others of exclusive property, it is the action of the thinking power called an idea, which an individual may exclusively possess as long as he keeps it to himself; but the moment it is divulged, it forces itself into the possession of everyone, and the receiver cannot dispossess himself of it. Its peculiar character, too, is that no one possesses the less, because every other possesses the whole of it. He who receives an idea from me, receives instruction himself without lessening mine; as he who lights his taper at mine, receives light without darkening me…. Inventions then cannot, in nature, be a subject of property.

Nonrivalrous property indeed; small wonder that Jefferson is regarded as the intellectual godfather of the twenty-first century’s “information wants to be free” movement. Partly because of his resistance, the First Congress, which opened for business in March 1789, failed to consider a system for granting patents for nearly a year. The first American patent statute was, however, eventually passed, and was signed by George Washington on April 10, becoming the law of the land—or of twelve-thirteenths of the land, since Rhode Island hadn’t yet ratified the Constitution.

The original procedure was fairly straightforward. Each patent application was sent to a committee of the United States Senate, who then referred it to the attorney general, who passed it to the president, who signed it and returned it to the secretary of state (in 1790, Thomas Jefferson). That was changed soon enough to the newly named Commissioners for the Promotion of Useful Arts: Jefferson, Secretary of War Henry Knox, and Attorney General Edmund Randolph.

The American system was simplicity itself compared to the contemporaneous British system, which was a thing of cartoonish complexity:

Step 1: Inventor prepares petition to the Crown, including an affidavit sworn before a “Master of Chancery”

Step 2: Master sends petition plus accompanying affidavit to the Home Office, who reads, endorses, and sends to the Attorney General and Solicitor General

Step 3: The Attorney General and Solicitor General review and, if they approve, return both petition and report to the Home Office

Step 4: Home Office prepares a warrant; sends to the King

Step 5: King signs, and Secretary of State countersigns, the warrant

Step 6: Secretary of State sends warrant back to Attorney General and Solicitor General; they prepare a bill describing the invention, and transcribe it onto the actual letters patent (written on parchment, in order to have the force of law)

Step 7: Attorney General and Solicitor General send parchment bill back to Secretary of State; King and Secretary again sign and countersign, thus making it, literally, a “King’s Bill”

Step 8: Secretary of State sends the King’s Bill to the Signet Office, which prepares an identical version, known as the “Signet Bill” (on parchment, of course; see Step 6)

Step 9: Signet Office sends Signet Bill to the Lord Privy Seal, who prepares a Writ of Privy Seal—yes, on parchment—and sends it to the Lord Chancellor with Signet Bill and Letters Patent

Step 10: Lord Chancellor “engrosses” Letters Patent on parchment with language identical to the Writ, dates it, and finally seals it

The system was not only absurdly complicated but outrageously expensive. In 1792, the official cost of a patent12 was £70 for England and Wales, but “gratuities” to every secretary, official, and even doorman standing along the way typically ran another £20; the tariff including Scotland and Ireland could easily exceed £300. The cost of a U.S. patent application,13 by comparison, was fifty cents; if the patent was awarded, the recipient owed the federal government two dollars, plus another dollar for affixing the Great Seal of the United States.

Even so, the American system was a little slow getting started. The United States issued only three patents during all of 1790. The first went to Samuel Hopkins for his method of making potash; the second, “for manufacturing candles,” was granted to a Boston chandler named Joseph Stacey Sampson. Both patents ended up generating far more wealth as rare documents, sold and resold to collectors avid for the signatures of Jefferson and Washington, than they ever did for their inventors. By number three, however, the system had identified an inventor who would do as much as any man alive to put steam on the move: a Philadelphian named Oliver Evans.

EVANS WAS BORN IN 1755 in the colony of Delaware and was apprenticed at the age of sixteen to a wheelwright in Pennsylvania. By the end of the customary seven years of training and toil, he had already built his first invention, a machine tool for making the leather “cards” used for removing unwanted material from wool and cotton. One of Evans’s most distinctive skills was the ability to see machines as a series of related steps, and his tool was an early, but telling, case in point: an assembly line in a box, it first bent the wire, then cut it to the proper length, and finally punched holes in precut pieces of leather, on which it mounted up to one thousand wire teeth per minute, with all the steps driven by a single rotating wheel. The hand-cranked card-making machine, for which Evans sought and received patent protection in Delaware, was an immediate success; so successful, in fact, that copies of it turned up, within months—and royalty-free—in Massachusetts, an early reminder to the inventor of the problematic character of intellectual property.

Evans’s next invention took the process of using one mechanical motion to drive a succeeding one a giant step further. Starting in 1788, he designed and built an automated mill that turned wheat into flour in a single continuous operation, illustrating, on a much larger scale, his gift for sequential mechanics. This time an elevator—actually wooden cups affixed to belts—lifted the unmilled grains of wheat onto a conveyor belt, which, in turn, was driven by a rod with lands cut into it: effectively a horizontal screw. The belt then pushed the wheat through the millstones into a hopper where a mechanical rake alternately stirred and sifted the flour. And once again a single rotating shaft synchronized all the operations.

Of all the components of the mill, only the hopper was truly original; but the real novelty, and value, of the invention came from the way it coordinated several existing mechanisms. Evans had once again exhibited a gift for seeing the big picture, literally, if uncomfortably: “I have in my bed viewed the whole operation14 with much mental anxiety.” In the event, it was the “whole operation” rather than any specific machine that earned Evans his first patents: first a fourteen-year grant in Maryland and Pennsylvania, a little later one of the same duration in Delaware, and one in New Hampshire to run for seven years.

His receipt of U.S. patent number 3 in 1790 generated prominent clients, and equally prominent critics. Though George Washington installed an Evans machine at his Dogue Run mill in 1791, his fellow Virginian Thomas Jefferson was nonplussed by Evans’s claim of novelty: “If the bringing together under the same roof15 various useful things before known entitled him [Evans] to an exclusive use of all these … every utensil of life might be taken for use by a patent…. I can conceive how a machine may improve the manufacturing of flour, but not how a principleabstracted from any machine can do it.”

Whatever Jefferson’s concerns about patents in general, and Evans’s in particular, he discharged his responsibility for them with diligence, and eventually—once he conceded the success of the law in promoting invention—enthusiasm. Evans himself was certainly spurred by the financial incentives of patent ownership, though their promise, in his case, frequently overran their actuality. His goal had been to earn his living16 selling licenses to other millers, but he found them resistant to automation, and in a classic pattern he spent more in legal fees suing patent infringers than he earned in licenses, despite renewing his patents nationally in 1808. By then, however, though his grist mills typically ran on waterpower, he was fully entangled in the next stage of the steam engine revolution.

In his own later recollection, Evans’s interest in using steam power for transportation dated back to his apprenticeship; in 1773, he had noticed a blacksmith’s apprentice using steam as the propellant in a gun. By 1783, he was already attempting to patent a steam locomotive, though his application to the Pennsylvania legislature was rejected for lack of a working model.

Using steam to propel vehicles over land was considerably more difficult than doing so on water, for reasons of simple physics. A wheeled vehicle needs to combat both gravity and inertia, plus any number of sources of friction, particularly the wheel against the axle rod. By contrast, relatively little power is needed to propel a lot of weight on water, primarily because of the lack of any real changes in elevation, close to perfectly even terrain, and negligible friction; this is why shipping large amounts of freight has always been cheaper by water than by land. As a result, a separate condensing engine could drive John Fitch’s steamboat. With the water supporting the engine’s weight, enough power could be produced to drive the craft, though slowly: four miles an hour, a brisk walking pace.

Locomotive engines, however, needed to put out more power with less weight; to be, in short, efficient. One of Evans’s great Usherian insights was that if one could dramatically increase the temperature of the steam, and therefore its pressure, the separate condenser could be dispensed with, at a huge weight savings. Another was that even though the steam, without a condenser, would simply disappear into the air once it had driven the piston, it would still use less fuel than was used up constantly reheating and recondensing. All he had to do was discover a way to increase the heat and pressure in the cylinder by an order of magnitude—but without turning it into a steam-powered hand grenade.

The key was the boiler. Low-pressure engines, from Newcomen to Watt, had boiled water in chambers whose basic design wasn’t much different from that of an oversized teakettle: a pot, or “haycock” shape, in which the furnace was placed below the water. Despite a number of improvements, it had remained essentially unchanged ever since.

Evans’s stroke of brilliance was to place his furnace inside a water-filled chamber: more surface area for the heated gas, more heat transfer to the liquid. With nowhere to go, the boiling water turned to steam whose pressure, now about fifty pounds per square inch, drove the piston, first in one direction, then in another.

In 1804, Evans applied for and received a federal patent for a high pressure “vibrating steam engine” that incorporated a “Boiler with the furnace in the center17 of the water enclosed in brick work.” To Evans, like every other inventor of the steam age, heat meant fuel, so increasing the first meant adding more of the second. But he also realized “the more the steam is confined18 … the greater will be the power obtained by the fuel. For every addition of 30 degrees of heat to the water doubles the power [and] doubling the heat of the water increases the power 100 times. Thus, the power of my engine rises in geometrical proportion, while the consumption of fuel has only an arithmetical ratio” (emphasis added). Evans had discovered how to produce the same power from a 500-pounder that Watt was getting from a two-ton monster; and as a bonus, it burned less fuel for every increase in horsepower—a huge advantage in an engine that needed not only to be mobile but also to carry its fuel with it. It was, by any definition, an act of genius.

It was not, however, a success. The value of high-pressure steam was greatest where its attractive power-to-weight ratio mattered most—steam locomotion—and America was still several decades away from railroad building. Evans was, like most inventors, forced to invest his time where he could find customers. In 1805, he built a thirty-foot-long fifteen-ton steam dredge for the Philadelphia Board of Health, which was the first steam land vehicle built in the United States; the same year, he created the first significant improvement over Watt’s linkage,19 using an isosceles triangle to connect two bars at a single pivot point, a design later known as the Russell linkage.*

However, also in 1805, he was the litigant in no fewer than four different suits, brought on both philosophical and technical grounds. One result was that his income was slashed, leaving Evans, in his own words, “in poverty at the age of 50,20 with a large family of children and an amiable wife to support.” Despite his many triumphs, Evans’s life seemed, to him at least, a succession of unsatisfying successes and paranoia-reinforcing failures. In 1795, he had published The Young Mill-wright and Miller’s Guide, which was effectively an entire millwright’s education between hard covers, went through dozens of printings, and was probably the bestselling “professional” book in the young republic. But he ran out of money before completing its sequel for steam engineers, and he published it, rather petulantly, under the title The Abortion of the Young Steam Engineer’s Guide.

Fig. 7: America’s first working steam “locomotive”: not only the first practical high-pressure steam-powered vehicle of any sort, but the first amphibious vehicle as well, though this was largely an accident. Evans’s workshop was more than a mile inland and fifteen miles from the dredging site on the Schuylkill River, so he added wheels to his paddle-driven dredge, which he had named Orukter Amphibolos, a drawing of which is in the upper right corner, and drove it proudly down Philadelphia’s Market Street. Image by permission of the Library of Congress

Sometimes his petulance verged on a persecution complex: “And it shall come to pass21 that the memory of those sordid and wicked wretches who opposed [my] improvements, will be execrated, by every good man, as they ought to be now….”

In 1806, however, Evans had opened the factory he called the Mars Works, where he built more than a hundred steam engines and boilers and thousands of components for mill machinery, enough that he ended up dying a rich man in 1819.

Evans was a visionary and a pioneer. But despite his prediction that “the time will come,22 when people will travel in stages moved by steam engines from one city to another almost as fast as birds can fly,” his greatest contribution to the history of steam locomotion was almost incidental: his decision to share the design of his boiler and high-pressure steam engine with his compatriots in Britain. As he wrote in the 1805 Abortion of the Young Steam Engineer’s Guide:

In 1794–95, I sent drawings,23 specifications, and explanations, to England to be shown to the steam engineers there, to induce them to put the principles into practice and take out a joint patent for the improvement, in their names … Mr. Joseph Stacey Sampson, of Boston [the same one who received the second patent issued in the U.S.] who carried the papers to England, died there, but the papers may have survived.

The timing is suggestive. In fact, it is a powerful bit of circumstantial evidence for the notion that the papers not only survived, but were read by a Cornish mine engineer named Richard Trevithick.

TREVITHICK’S ANCESTRY, ON BOTH sides, is dotted with members of Cornwall’s mining aristocracy. His father and uncle were both captains in some of the region’s largest and most profitable copper mines, including the legendary Dolcoath mine, where Richard Sr. built the deep adit in 1765 and constructed a Newcomen engine in 1775. Richard Jr.’s early life, partly in consequence, was slightly atypical, as he was neither the product of a formal apprenticeship, nor was he schooled to be a scientist or engineer; instead, by 1784, he was already working in a relatively senior position at Dolcoath, reporting to his father, and over the next ten years, he practiced his somewhat nomadic craft at half a dozen different mines all across Cornwall, including the Tincroft and Wheal Treasury (“wheal” is a Cornish term simply meaning “mine”).

By the 1790s, he was a local hero, partly because he was seen as an adversary of Boulton & Watt. The steam engines provided by the Birmingham firm had improved productivity, but their royalty system generated mostly resentment, since the more copper the mines produced, the more gold their owners owed to Boulton & Watt. So when Trevithick testified on behalf of Hornblower during the lawsuit, he endeared himself to most of Cornwall. Trevithick’s appearance didn’t hurt. He had grown to be a huge man, at least six feet two, and “Cap’n Dick,” as Trevithick was widely known, became the subject of Paul Bunyan–like mythmaking; one story has him throwing a sixteen-pound sledgehammer over the top of a twenty-foot-tall Newcomen engine.

Trevithick was his country’s champion not only because of his strength, but his cleverness; in 1797, he invented and built a system that improved upon, and so replaced, the chain of buckets that had been used since Newcomen to pump water out of mines. But the engines that drove them were still built by Boulton & Watt, or under license from them—at least until 1800, when Watt’s original patent finally expired. The immediate consequence of the expiration was a dramatic increase in the appeal of the Boulton & Watt designs now that they were available without the Boulton & Watt royalties.

Trevithick, however, wasn’t trying to imitate the Boulton & Watt engine, but to replace it. Most especially, like Evans, he wanted to dispense with Watt’s separate condenser, but unlike Evans, he didn’t know whether it was feasible. This was a basic scientific question, for which Trevithick, an indifferent student (his teachers, though they noted his intelligence, also called him “disobedient, slow, obstinate24 … frequently absent and very inattentive”), had no answer. However, he wasn’t shy about using the skills of those with more formal education, and during the Hornblower trial, he had struck up what became a lifelong friendship with Davies Gilbert, the future president of the Royal Society. In 1797, Trevithick asked him to calculate how much power would be lost if instead of capturing the steam in a separate condenser, the engine simply exhausted it into the air.

The answer was Trevithick’s real eureka moment. Gilbert explained that with each stroke,25 the cylinder would lose exactly as much pressure inside as the pressure outside: 14.7 pounds per square inch at sea level. This would obviously be disastrous for a Boulton & Watt separate-condensing engine, which generated only a little more than ten pounds per square inch inside the cylinder; exhausting the condensation would leave it with no pressure at all.

But what if the pressure inside the cylinder could be increased?

In 1800, five years after Joseph Stacey Sampson had brought Oliver Evans’s drawings to Britain, and three years after Davies Gilbert had shown that an engine operating at 60 psi would lose only a quarter of its pressure at each stroke, Richard Trevithick introduced his first high-pressure steam engine at the Wheal Hope copper mine. It used almost precisely the same method Evans used for increasing the steam pressure from the boiler. It is not known whether Trevithick saw the Evans design firsthand or—more plausibly—through William Murdock, his neighbor in the town of Redruth for six months in 1797 (and who, despite his employment by the despised firm of Boulton & Watt, was friendly with Trevithick). The sequence of events, however, is persuasive: Sampson carried Evans’s drawings to Britain in 1795, to show to “the steam engineers there.” The British steam engineer best known to be working on high-pressure steam for locomotion was William Murdock, who was Trevithick’s neighbor when the Cornishman asked Davies Gilbert to give a scientist’s opinion on what was, in essence, the practicality of Evans’s engine.

The Wheal Hope engine showed that a high-pressure engine was practical; a little more than a year later, Trevithick was ready to demonstrate that it was mobile as well. On Christmas Eve 1801, with apparently no warning, Trevithick appeared on the High Street of the town of Camborne aboard a carriage unlike anything anyone had ever seen before. No drawings survive, though a sketch dated a year later shows a four-wheeled flatbed truck with a vertical steam engine set over the front wheels, a twelve-foot-high boiler over the rear, and the driver (wearing what looks like a top hat) set behind. We do have the testimony of one of the blacksmiths who worked on the castings for the machine, “old Stephen Williams”:

Captain Dick got up steam,26 out in the high-road, just outside the shop at the Weath. When we get see’d that Captain Dick was agoing to turn on steam, we jumped up as many as could; maybe seven or eight of us. ’Twas a stiffish hill going from the Weight up to Camborne Beacon, but she went off like a little bird….

When she had gone about a quarter of a mile, there was a roughish piece of road covered with loose stones; she didn’t go quite so fast…. She was going faster than I could walk, and went on up the hill about a quarter or half a mile farther, when they turned her and came back again to the shop….

To the spectators at Camborne Hill wrapped up against the cold that Christmas Eve, the vision of a wheeled vehicle moving uphill without being either pushed or pulled must have seemed something like levitation. For millennia, the force driving wheeled vehicles had always been something external to the vehicle, which meant that the primary traction against the roadway was also generated externally: by a horse’s hooves, for example. It was by no means obvious that simply turning a wheel would generate enough traction to pull itself—the first bicycles were still decades in the future—and Trevithick apparently spent the week before his experiment hand-cranking the wheels of a model along cobblestone and dirt roads.

Just as startling to the audience lining the half-mile-long High Street, the engine on board the Camborne carriage made its journey belching smoke like something out of myth. Trevithick’s engines, having dispensed with a separate condenser, were thereafter known as “puffers” because their steam exhausted directly to the air. The distinctive clouds familiarly associated with steam engines* had finally been born.

It didn’t survive its own infancy; the engine that Stephen Williams rode up Camborne Hill on Christmas Eve was destroyed before New Year’s. Two of Trevithick’s drivers27 celebrating the season in a local pub left the boiler unattended, the water boiled away, and the chamber became hot enough to set fire to the engine, and to the shed in which it was housed. Which didn’t, in the end, matter all that much. In March 1802, Trevithick, with help from a fellow Cornishman, the scientist Humphry Davy, applied for and received a patent on the new locomotive.

The market for self-propelled steam engines was still a fraction of that for stationary ones, thousands of which were by then pumping water and operating machinery throughout Britain. One of the more avid users was the Coalbrookdale foundry,28 then run by the fourth generation of the Darby family, where the boiler for the Camborne Hill locomotive was built; by the end of 1802, the Darbys had hired Trevithick to build a new stationary engine operating at the seemingly impossible pressure of 145 psi.

Their faith in the Cornishman, and in high-pressure steam, was understandable, but it was controversial. The future of steam power was very clearly at stake, and the established powers of British manufacturing had everything riding on the existing Boulton & Watt low-pressure designs. The Birmingham firm’s founders had retired in 1800, but their sons were, if anything, more hostile both to high-pressure steam and to Trevithick than were their fathers. No one knew this better than Cap’n Dick; after the boiler of a high-pressure engine exploded in Greenwich in September 1803, Trevithick wrote of his belief that “Mr. B. & Watt29 is abot to do mee every engurey in their power for they have don their outemost to repoart the exploseion both in the newspapers and private letters.” But though the Boulton & Watt public relations campaign could slow the adoption of high-pressure steam engines—and infuriate Trevithick—it could not stop it. In 1803, a Welsh ironmaster named Samuel Homfray, jealous of his competitors at Coalbrookdale, invited Trevithick to bring his magic across the Bristol Channel from Cornwall to Wales.

Industrialization in Wales, as in England, had been partly a function of geology, and the great ironworks that had grown up around the town of Merthyr Tydfil in the 1760s were testimony to the generous local supplies of hematite, limestone, coal, and water, the key ingredients for a furnace-and-forge economy. Just as rich was the supply of ironmongers—four large ironworks, each competing with one another to meet the ever increasing demands of British industry. Competing, and also cooperating: the four ironworks each used the same route to transport their goods to Cardiff and the coast: the twenty-four-mile-long Glamorganshire Canal, which they had built, owned, and, in theory, shared.

They did not, however, share it happily. The owner of the Cyfarthfa ironworks, Richard Crawshay, was also the majority shareholder in the canal, and as a result, barges hauling Cyfarthfa iron were granted preferential treatment. His partners, the owners of the Dowlais, Plymouth, and Penydarren works, were angry enough to build a parallel route for their iron, a horse-drawn railway. One of them was Samuel Homfray of Penydarren, Richard Trevithick’s new employer.

Homfray’s original objective in hiring the Cornishman seems a bit unformed in retrospect. It is certain that he wanted a new steam engine to run the hammer in Penydarren’s forge, but unlikely that he had already thought about replacing the horses pulling his iron-filled carts. For one thing, the existing railway was built for horses rather than locomotives; the rails were set into concrete stones set four feet apart, and had no crossties in between to trip up the horses. Also, the grade was extremely gentle;30 on the way from Merthyr to the wharf at Abercynon, the railway dropped only one foot for every forty-five traveled, putting less strain on the horses both going and returning.

Fig. 8: The Penydarren engine, a replica of which is still on display at the National Maritime Museum at Swansea, was as important in its way as either Newcomen’s 1712 Dudley Castle pump, or Watt’s 1776 New Willey engine, or even the Stephensons’ Rocket. National Railway Museum / Science & Society Picture Library

In the event, something provoked Crawshay to bet Homfray five hundred guineas that no steam locomotive could do the job of the horses. To win, Trevithick needed to build an engine capable of hauling ten tons of iron ore from Merthyr to the wharf, nine and a half miles away.

The Penydarren locomotive is practically an encyclopedia of innovation. As protection against explosion, it used one of Trevithick’s cleverest inventions, the so-called “fusible plug,” a small lead cylinder inserted into a predrilled hole in the wall of the engine’s boiler—a hole that, in a properly operating engine, would always be underwater. If, however, the water level in the boiler were to fall low enough to become dangerous, the heat would melt the lead plug, allowing the steam to blow out the fire. The Penydarren engine also incorporated a U-shaped fire tube, a return flue that carried the air heated by the furnace from one end of the boiler and back again, which put at least twice as much surface area in contact with the water. Even more important, it didn’t just exhaust the spent steam into the air, but used a chimney that, in Trevithick’s own words, “makes the draft much stronger”31—that is, the exhaust steam, hotter than the surrounding atmosphere, rose. By doing so, it pulled more oxygen into the furnace, raising its temperature and increasing the efficiency of the heat engine itself.

In other respects, it was a bit of what a later generation of engineers would call a kludge. Trevithick was obliged to build an engine that would serve Homfray’s forge as a steam hammer, whether or not it worked to transport his ore, and it was therefore cobbled together from pieces intended for different functions. Its piston operated like a slide whistle, driving only the two wheels on the engine’s left side and conserving momentum with an enormous flywheel set behind them. And, on February 21, 1804, it worked—sort of:

… yesterday we proceeded32 on our journey with the engine, and we carried ten tons of iron in five wagons, and seventy men riding on them the whole of the journey … the engine, while working, went nearly five miles an hour; there was no water put into the boiler from the time we started until our journey’s end … the coal consumed was two hundredweight.

And, sort of, it didn’t. The problem lay less with the locomotive than with the rails, which cracked like twigs. The engine, whose five and a half tons were distributed over only four wheels, and with only two of them driving, put an unanticipated lateral strain on the railway. Though Trevithick would try again, on a coal railway in Newcastle in 1805, the rail problem remained unsolved; even in 1808, when Trevithick demonstrated his “Catch-Me-Who-Can” locomotive on a half-mile oval near Gower Street in London for a shilling a head, it was regarded more as a circus act than as any useful industrial advance.

Trevithick’s engine, the first driven by high-pressure steam, earned him a considerable claim on the title “father of railways,” but the birth of steam locomotion was still a decade or so in the future. More important, though less romantic, was another of Trevithick’s innovations, one that was nearly as large an improvement over the first high-pressure design as that had been over the Boulton & Watt separate condensing engine—indeed, as big an improvement as Watt’s separate condenser was over Newcomen’s original atmospheric engine.

For nearly a decade, Trevithick’s high-pressure engines had been making significant inroads into the dominant position of Boulton & Watt in Cornwall’s mines. By 1812, he determined to displace them once and for all. In a pump built for the Wheal Prosper mine in Cornwall, Trevithick modified his existing high-pressure steam design so that instead of exhausting the condensed steam directly to the atmosphere, as with the Penydarren and Camborne Hill engines, he allowed it to expand into a lower-pressure chamber first. In the new engine, the pressure on the piston came from both the expansive property of high-pressure steam on top of the piston and the atmospheric pressure on the chamber once the steam has been condensed. Steam flowed into the top half of the cylinder and pushed the piston down some distance, at which point a valve closed and the steam expanded to fill the now smaller volume. Trevithick, in comparing an early model to a Boulton & Watt atmospheric engine, discovered that he could produce 40 psi using one-third the coal that the atmospheric engine needed to produce 4 psi. Even more innovative, the new engine’s boiler lay horizontally, which allowed the fire tube to run through its middle, heating the water both efficiently and to high pressure. “My predecessors,” Trevithick said, “put their boilers in the fire;33 I have put the fire in the boiler.” The result, in 1812, was the first really successful “Cornish engine.”

It was certainly successful as measured by the still-in-use benchmark of “duty,” which measured the pounds of water raised one foot by a bushel of coal. A high-performing Newcomen-style engine typically performed in the neighborhood of 5,000 pounds; Smeaton’s many improvements nearly doubled that number—to 9,600 pounds—without changing the basic design, and a 1778 Watt engine, with separate condenser, achieved a duty of 18,900 pounds. By 1812, Trevithick was boasting34 of 40,000 pounds, which is likely an exaggeration, but an objective report of three Cornish engines at the Dolcoath mine reported 21,400, 26,800, and 32,000 pounds in 1814. By 1835, another Cornish engine achieved a duty of 100,000 pounds.

However, efficiency, as measured in duty, was not everything. The price of the Cornish engine’s dramatic achievement was that its multiple chambers and valves demanded an unforgiving level of both precision and maintenance. Without either, they were more subject to breakdowns—and to the purchaser of a steam engine trying to make delivery of a scheduled amount of cotton, produce a quantity of iron, or pump water, it mattered little to have the most efficient steam engine if it was out of commission for two days a week. As a result, it is the last advance in steam power with Trevithick’s name attached to it.

Instead, like a homing pigeon, he returned to his origins: precious metals mining. Trevithick became obsessed with reopening the silver mines of Cerro de Pasco in Peru, which had once been among the richest of Spain’s possessions in the New World. Trevithick convinced himself that he would be able to make the Peruvian mines profitable once again, and he left Britain planning to do so in October 1816, arriving in February of the following year.

His timing could have been better. By 1817, most of South America was in rebellion against Spain; the month before Trevithick arrived in Peru, the Argentine general José de San Martín had crossed the Andes into Chile and was preparing to head north. A month before that, Simón Bolívar had returned to Venezuela from Haiti. Though Peru would remain under Spanish control for another five years, Trevithick’s engines (he had shipped four pumping engines and four winding engines ahead of his arrival) were still in their crates when his romantic soul got the better of him and he joined the rebellion. While in Caxatambo, Peru, he even designed a new carbine for Bolívar’s army, but when the city was occupied by the Spaniards in 1818, Trevithick was forced to flee north35 to Costa Rica, leaving an estimated £5,000 in ore and uncounted more pounds’ worth of lost equipment.

Trevithick’s South American adventure carries an almost unwieldy tonnage of symbolism: a representative of the dominant world power of the nineteenth century caught in the collapse of the dominant one of the sixteenth. Even more pointedly, it offers a high-contrast picture comparing history’s two longest-lasting approaches to the very idea of wealth: wealth as technology versus wealth as precious metals. Whatever meaning is retrospectively poured into it, however, the experience as Trevithick lived it seems to be less about the metaphorical war between two notions of political economy than about the real thing. Evading Spanish patrols in the Nicaraguan jungles, which the Cornish inventor was forced to traverse on foot, destroyed whatever romance the rebellion still offered. Trevithick’s journey, which included a dozen hair’s-breadth escapes, the deliberate capsizing of his boat by an offended traveling companion, and bouts of illness too frequent to count, found him arriving at last in Cartagena, Colombia, exhausted, sick, and broke—in his own words, “half-drowned, half-dead, and the rest devoured by alligators.”36

There his story took an unlikely turn. In Cartagena, he met the son of an old friend, who lent him £50 for his fare home. When Trevithick finally returned in 1827, he had nothing on his person to show for a decade in South America but two compasses—one for drafting, the other for navigating—a pair of silver spurs, and his gold watch.

Aficionados of dramatic coincidences could, however, take some comfort in the name of the man who paid for Richard Trevithick’s ticket home. He was Robert Stephenson, of Newcastle, and along with his father, George, is Trevithick’s only serious competitor for the title of “father of railways.”

THE DEPICTION OF GEORGE STEPHENSON by Samuel Smiles, the prolific biographer and self-help author* who did more than anyone else to establish the heroic archetype for British inventors, is a textbook example of self-discipline and deferred gratification. His first job was as a picker: a laborer whose entire job was separating coal from the stones that accompanied it from mineshaft to colliery. Soon enough he was working as an assistant fireman, then as the “plugman” operating a set of valves on the steam-driven pump at another collier’s. When he turned twenty, he was appointed as the brakeman, responsible for maintaining the winding mechanism that pulled the coal-filled buckets from the work area of the mine.

None of this permitted much time for education, even of the practical sort, and in a previous century, Stephenson might have simply become a superbly reliable artisan. By the end of the eighteenth century, however, Britain—particularly outside London—had spent decades making heroes out of onetime laborers who had become wealthy by acquiring and producing useful knowledge. So inspired, Stephenson taught himself to read and write and hired someone to teach him the rudiments of arithmetic. By 1801, the ambitious twenty-two-year-old brakeman took on additional work moonlighting as a watch repairman; two years later, now a father, he took charge of the Boulton & Watt steam engine driving the wheels of a Scottish spinning factory that had grown prosperous making cloth for the military uniforms needed for the war against Napoleon. This did not exempt Stephenson from service himself, nor did the birth of his son, Robert, in 1803; the following year he was drafted for the militia, and went into debt paying for a substitute to serve in his place. The same year, he returned to the Killingworth pit mine, where he taught himself mechanical engineering by spending his one day off each week dismantling and reassembling the colliery’s steam engine.

In 1811, Stephenson, like Watt thirty years before, made his first real mark on the world repairing a Newcomen engine, this one an old model at one of the Killingworth pits. For the job he was paid £10, and far better, was hired to manage all the engines owned and operated by the collieries of the so-called “grand allies,” a group of aristocratic investors that owned the Killingworth pit, at the impressive salary of £100 per year—the equivalent of more than $100,000 in current dollars.37 Stephenson’s salary was an insurance policy for the stationary engines on which the colliers depended; it rapidly turned into an investment in an entirely new industry. In 1814, Stephenson built his first locomotive to transport coal at Killingworth: the Blucher, a giant step toward practical steam locomotion.

• • •

THE TWO GREAT PROBLEMS in harnessing steam power for transportation were, broadly speaking, both a function of weight. The first one—increasing the engine’s power-to-weight ratio—was addressed, if not solved, by the realization, by both Evans and Trevithick, that more heat meant more pressure and therefore more work. That notion, which is obvious in retrospect but was revolutionary at the time, practically demanded a whole series of “micro-inventions” intended to turn up the dial on steam boilers: return flues, for example, which not only “put the fire in the boiler” but increased exponentially the area heating the water. Even more important, exhausting the steam through a chimney located above the furnace created a draft—a “steamblast”—that raised the heat even further.

The other weight problem was the one that licked Trevithick at Penydarren: The tracks on which the locomotive ran were just not able to survive the tonnage traveling over them. Driving a five-ton steam locomotive over rails designed for horse-drawn carts was only slightly more sensible than driving a school bus over a bridge made of wet ice cubes. In both cases, it’s a close call whether the vehicle will skid before or after the surface collapses.

This is why all of the dozens of inventors attempting to put steam on the move were obsessed with the durability and traction of the surface holding their vehicles; for centuries, the rails originally designed for horse transport had been made of wood, occasionally reinforced with iron edging. Not until 1767 did the Darbys of Coalbrookdale begin casting iron rails for wagonways, which made them far stronger; within twenty years some unknown innovator had added an arched rim, or lip, to prevent wheels from slipping off.

The rims, or flanges, were fine for keeping the wheels from moving laterally, but they did nothing to increase traction—a real challenge for smooth iron wheels on smooth iron rails. In 1811, John Blenkinsop, an employee of another collier located in the city of Leeds, patented a Trevithick-style engine with a cogged driving wheel, and accompanied this with a new sort of track, this one made of cast iron with an edge rail carrying a toothed rack. The cog-and-rack not only eliminated any possibility of skidding, it transmitted five times the force of Trevithick’s original engine, and Blenkinsop-style engines remained popular through the 1820s, despite the enormous cost of producing miles of what were essentially horizontal iron gears.

Two years later, a civil engineer named William Chapman applied for and received a patent for an engine that propelled itself by hauling itself along a chain; even odder, William Brunton also managed to patent his “Brunton Mechanical Traveler” that “walked” the locomotive along by operating mechanical feet driven via a complicated series of levers and linkages. Slightly less eyebrow-raising,38 in 1815, William Hedley’s “Wylam Dylly” engine tried to solve the excess weight problem by doubling the axles from two to four, thus distributing the weight over a larger area.

Stephenson’s locomotive, which made its maiden journey on July 25, 1814, took a different approach. The engine driving the Blucher (named for the Prussian general who would pull the Duke of Wellington’s chestnuts out of the fire at Waterloo almost exactly a year hence) incorporated an early version of the blastpipe: a vertical tube with a narrowed exit that carried the exhaust into the chimney, creating a draft, just as Trevithick had more or less accidentally discovered a decade before. It also ran on a reversed version of the most popular track design, putting a flanged wheel on a smooth rail. Two years later, Stephenson, in collaboration with the ironmonger William Losh of Newcastle, produced, and in September 1816 jointly patented, a series of improvements in wheels, suspension, and—most important—the method by which the rails and “chairs” connected one piece of track to another. Stephenson’s rails seem mundane next to better-known “eureka” moments, but as much as any other innovation of the day they underline the importance of such micro-inventions in the making of a revolution. For it was the rails that finally made the entire network of devices—engine, linkage, wheel, and track—work.

Stephenson’s working life* marks the point in the development of steam technology when the value of what economists call “network effects” finally overtook the importance of any individual invention, however brilliant. Setting the distance between the smooth tracks on which the Blucher traveled at four feet eight and a half inches was arbitrary—that was the width of the Killingworth Colliery wagonway—but its specific width was irrelevant. The value of any standard is not its intrinsic superiority, but the number of people using it. Like the famous example of the QWERTY keyboard, the Stephenson gauge became the world standard, and it is still the width used on more than 60 percent of the world’s railroads.

Of course, simply laying rails a particular distance apart does not make for a monopoly unless others follow. And others weren’t about to follow Stephenson’s lead until they were persuaded that there was some advantage to it, in the form of either increased revenue or lower costs. To a proprietary line, such as the ones that connected coal mines with ports, the advantage of a standard wasn’t all that obvious; as at Killingworth, it was frequently more economical to use existing wagonways and roads than to redesign them to a new standard. The same didn’t apply to so-called “common carriers,” who needed, by definition, to accommodate rolling stock they didn’t own, or to travel on railways they didn’t build. The first common carrier to realize this,39 the Stockton and Darlington Railway, was, not at all coincidentally, one of George Stephenson’s employers. But by far the most important one was the one intended to connect the cities of Manchester and Liverpool.

It is almost indecently tempting to place the Liverpool & Manchester Railway at the climax of the entire history of British industrialization. The first temptation is posed by Manchester itself, which was, when George Stephenson took on the job as chief engineer of the proposed railway in 1825, the most “industrial” spot in all England, with all that implied: “a town of red brick,” in the words of Charles Dickens, “or of brick that would have been red if the smoke and ashes had allowed it.”

The reason, of course, that those bricks were covered by smoke and ash was that the city was home to the world’s largest textile manufacturers, factories that used coal to turn cotton into clothing. Richard Arkwright’s mills, which gave the city its nineteenth-century nickname of “Cottonopolis,” had become so successful that the choke point for the industry’s growth was no longer technological imbalance (the difference between efficient spinners and inefficient weavers, for example) but transportation. Manchester was making cotton faster than it could ship it, and Britain’s canal system, even with its sophisticated locks, was less and less able to handle the load, which was easily exceeding a thousand tons of cargo daily: raw cotton in, finished goods out. So much cloth was being made in Manchester, in fact, that by 1800, the port of Liverpool on the River Mersey was the world’s most important; less than eighty years old, it handled more than a third of all the world’s trade. The need for a railway to connect Manchester’s mills with the port city had become urgent.

It is metaphorically satisfying to talk about threads being woven together when talking about cotton, but the thread that mattered to the Liverpool & Manchester Railway was made of iron: thirty miles of it, smelted, forged, and wrought in ironworks like Coalbrookdale on the Severn, and laid down as rails between the two cities that were now producing, in their mundane way, more wealth in a year than the entire Roman Empire could in a century.

But while there were clearly massive financial incentives for building some kind of railway between the factory and the port, the railway’s directors were uncertain that a locomotive railway was the best option. Some of the investors and directors in the enterprise were promoting the use of rope cables to haul boxcars full of cotton the entire thirty miles, using stationary engines roughly every mile and a half. Others wanted different kinds of locomotives (though no one, happily, was arguing on behalf of Brunton’s mechanical “walker”). After much to-ing and fro-ing, it was decided to settle the problem with a contest.

On May 1, 1829, the Liverpool & Manchester Railway ran an advertisement in the Liverpool Mercury inviting “engineers and iron founders” to submit plans for locomotives to compete for the winning design. The offer of a £500 prize, the equivalent in average earnings of more than $500,000 in 2010, brought the crackpots out in force. The treasurer of the Liverpool & Manchester, Henry Booth, described the applications:

From professors of philosophy40 down to the humblest mechanic … [from] England, America, and Continental Europe. Every element and almost every substance were brought into requisition and made subservient to the great work. The friction of the carriages was to be reduced so low that a silk thread would draw them, and the power to be applied was to be so vast as to rend a cable asunder…. Every scheme which the restless ingenuity or prolific imagination of man could devise was liberally offered to the Company….

The oversupply of perfect vacuums and perpetual motion machines was in part a testimony to the utter transformation of British cultural attitudes toward innovation over the preceding century. By the 1820s, the Patent Office was approving nearly three hundred new inventions annually, and rejecting thousands. In the event, the Liverpool & Manchester had made the conditions for entry fairly strict: entries had to be mounted on springs, weighing no more than six tons including water (if on six wheels) or four and a half tons (if on four); they must operate at between 45 and 60 psi, while being prepared for a test at up to 150 psi; they must consume their own smoke (to keep the route as clear of ash as possible; this effectively required the engines to burn coke rather than coal); and they were required to pull a gross load of twenty tons at ten miles an hour back and forth along a mile-and-a-half course forty times, reproducing the sixty-mile round trip between Manchester and Liverpool.

The stipulations eventually weeded out all but five applicants, only three of which could be called serious. One of the others never actually made it to the starting line, and the other—the Cycloped, whose source of propulsion was a horse trotting on a treadmill and which was only allowed to compete because its designer was on the railway’s Board of Directors—proved good for nothing more than comic relief.

Two of the remaining three competitors were joint favorites to win the prize: the Sans Pareil, built by Timothy Hackworth, master mechanic of the Stockton and Darlington Railway (and therefore George Stephenson’s former employer), and the Novelty, the creation of a former Swedish army officer now living in London, John Ericsson.

The third, entered by Henry Booth and George Stephenson and to be built by his son Robert—Richard Trevithick’s rescuer, and an even more skilled engineer than his father—was Rocket.

Between May and September of 1829, Robert Stephenson—who had promised his father, “Rely upon it, locomotives41 shall not be cowardly given up. I will fight for them until the last. They are worthy of a conflict”—labored at his workshop in Newcastle-on-Tyne to construct the world-changing locomotive. While it incorporated a key design feature suggested by Booth (the multitube boiler, about which more below), every other innovation contained in the final entry was the work of Robert, who had explicitly identified four areas for potential improvement in the final design: transmission of the largest amount of power from the pistons to the wheels; preservation of the greatest amount of traction between wheels and track; minimizing the loss of heat between boiler and cylinder; and maximizing the amount of heat within the boiler itself.

Those innovations are the reason that any list of the most significant engines in locomotive history always includes the Stephensons’ entry at Rainhill, the site of the competition’s final trials. First were its mechanics: the way it transmitted the reciprocating motion of its pistons to its wheels. The “premium engine,” as the two Stephensons referred to it, used two pistons set at a 45-degree angle above the front axle, each one attached to a slip eccentric, which is a sort of linkage in which a disk is attached to an axle but offset “eccentrically” (essentially a simpler, and more efficient, version of the sun-and-planet gear). One set of slip eccentrics turned the reciprocating motion of the pistons into rotation, while another set worked in reverse, opening and closing the steam valves as the engine cycled.

To increase the amount of traction between wheels and track, Stephenson and his assistant William Hutchinson calculated the optimal arrangement of weight over the wheels and determined to use the engine to operate only the front wheels; it was far more efficient, both in tractive power and durability, to drive only two large (4′8″ diameter) wheels and use the back wheels (with a diameter of 2′6″) for balance.

Fig. 9: Little though it resembled the great locomotives of the nineteenth century, Rocket pioneered virtually all of their engineering innovations, from the high “blastpipe” chimney to the multitube firebox to the slip eccentric gears on the driving wheels. National Railway Museum / Science & Society Picture Library

But the truly revolutionary significance of the engine was its boiler design. Twenty years before Rainhill, Oliver Evans had demonstrated that raising the boiler’s heat by doubling the amount of fuel increased the engine’s power by at least ten times; Richard Trevithick had goosed up the heat in his boiler with a U-shaped return flue. The principle was, in retrospect, obvious: Since the water was heated by conduction with the chamber containing heated gas, increasing the surface area of the chamber would transmit more of that heat to the water surrounding it. Robert Stephenson was just about ready to take that principle to its logical conclusion.

Rocket’s boiler did not have a single flue, even a U-shaped one. Instead, as suggested by Henry Booth, twenty-five copper tubes, each three inches in diameter, were fitted into a firebox inside a water jacket, with somewhat wider copper tubes connecting them to the barrel of the six-foot-long, three-and-a-half-foot-diameter boiler. The cylinders exhausted their steam into two blast pipes inside the chimney, whose slightly narrowed openings guaranteed a powerful draft of air. Robert Stephenson spent the entire month of September testing to ensure that the boiler and cylinder were reliably steamtight to the point that they could handle up to 150 pounds of pressure per square inch. It finally passed Stephenson’s inspection only the day before it left his Newcastle workshop and was placed on a series of horse-drawn carts for the 120-mile journey to the Rainhill course, ten miles east of Liverpool.

The first day of the trials, October 6, was largely a day for demonstration, as each competitor tried the course without hauling the weight required by the contest’s rules. Novelty, at two and a half tons, the lightest of the three remaining entrants, was by far the fastest. Using two vertical cylinders to drive a crank attached to the leading axle, it was also, by general consent, the prettiest engine in the competition (painted royal blue, with its boiler and water tank covered in polished copper), and it was made the early favorite, a position it improved on the following day, when, hauling more than eleven tons, Novelty easily hit a speed of 20 mph.

On October 8, the final specifications for the contest were published: Each engine was no longer required to haul twenty tons, but a load of three times its weight, including the water in its boiler, with allowance made for the engines—Novelty and Sans Pareil—that hauled water in the locomotive rather than in a separate tender. Though the entrants were to have competed in the order of their “race cards”—Novelty, then Sans Pareil, then Rocket—the first two needed last-minute repairs, and Rocket went first.

Rather surprisingly, given the historical significance and number of spectators, no one knows who actually drove Rocket on its October 8 debut at Rainhill. Robert McCree, from the Killingworth Colliery, had driven it during testing, but at least one report suggests that he was, like Robert Stephenson, only a passenger (and possibly a fireman, loading fuel into the firebox). If so, the driver could only have been George Stephenson himself, and he, like Rocket, covered himself in glory, along with coal dust. It took only fifty minutes for the fuel (like the other competitors, Rocketused cleaner-burning coke, to “consume its own smoke”) to bring the pressure in the boiler up to the required 50 psi from a cold start, and by 10:00 A.M., the engine was on its way.

And so it continued. Aware that the rules mandated an average speed of 10 mph, the Stephensons kept their pressure well below its maximum for the first back-and-forth laps. It took a bit more than six minutes, at an average speed of around 15 mph, to complete the first mile and a half: just about the pace of a good twenty-first century fifteen-hundred-meter runner. By the tenth lap, the engine was moving closer to 20 mph, but not until the last lap did the Stephensons open up the steam regulator and let Rocket fly. When they passed the grandstand at the eastern end of the course, Rocket was pulling its twenty tons at more than 30 mph, all while consuming “only” a little more than 200 pounds of fuel an hour. Thousands of spectators rushed the finish line to cheer the Stephensons on their triumph.

The rest of the competition was something of an anticlimax. Novelty didn’t get a chance to compete until Saturday, October 10. The same high power-to-weight ratio that had made it such a fan favorite four days earlier allowed it to race off at what must have seemed magical speed, completing its first mile in less than two minutes. Before its second, however, a blowback from the engine’s furnace burst the bellows used to create chimney draft. The explosion ended Novelty’s day. On its next run, the favorite managed only one lap before another pipe exploded; since this was the pipe that fed the boiler, the resulting detonation ended with “the water flying in all directions.”42 When the boiler gave out, Ericsson gave up.

Sans Pareil did perform brilliantly. The heaviest of the three finalists, it pulled a full twenty-four tons at better than 15 mph. But not for long. After twenty-two and a half miles, its boiler, rather embarrassingly, ran dry, melting the fusible plug that stopped it cold. The reason was its enormous consumption of fuel: nearly 700 pounds per hour.* The victor, by acclamation, was the Stephensons’ Rocket.

IT’S NOT NECESSARILY OBVIOUS that the Rainhill Trials mark the moment in history when the steam revolution became finally, and utterly, inevitable. One year later, the Liverpool & Manchester Railway opened for business, with eight Stephenson-built locomotives traveling on Stephenson’s standard-gauge track before luminaries that included the then prime minister, the Duke of Wellington, but the conflicts over the proper use of steam power didn’t vanish. To the end of his life, Stephenson fought a running battle with an even more famous engineer, Isambard Kingdom Brunel, over the latter’s preference for an “atmospheric railway system” operated by stationary engines. Brunel, the son of Marc Brunel, of the Portsmouth Block Mills, even designed the Great Western Railway to run on a gauge nearly three feet wider than Stephenson’s (though he soon discovered the impossibility of overcoming an early monopoly advantage).

There is, after all, something as arbitrary about ending the story of the steam revolution at Rainhill in 1829 as beginning it in first-century Alexandria. Unfortunately for historians, if not for history, such convenient end points are as capricious as the textbook dates for the Industrial Revolution itself, which the careful reader will remember were originally matched as a lecture hall convenience to the regnal years of George III. One might just as well have decided that the story ended in 1819, the year that James Watt and Oliver Evans died—and, coincidentally, the year of the first steamship crossing of the Atlantic, by the American-built Savannah. Or 1824, when Sadi Carnot finally explained the thermodynamics of steam power. Or 1838, when I. K. Brunel’s Great Easternconnected a steam railroad with a true transatlantic steamer (the Savannah was really a three-masted sailer, with paddlewheels added).

The reason for ending with Stephenson’s triumph nonetheless seems persuasive. Rainhill was a victory not merely for George and Robert Stephenson, but for Thomas Savery and Thomas Newcomen, for James Watt and Matthew Boulton, for Oliver Evans and Richard Trevithick. It was a triumph for the ironmongers of the Severn Valley, the weavers of Lancashire, the colliers of Newcastle, and the miners of Cornwall. It was even a triumph for John Locke and Edward Coke, whose ideas ignited the Rocket just as much as its firebox did.

When the American transcendentalist Ralph Waldo Emerson met Stephenson in 1847, he remarked, “he had the lives of many men in him.”43

Perhaps that’s what he meant.

* The names of eighteenth-century Cornish mines are as personal, and as obscure, as the names given to thoroughbred racehorses and recreational sailboats.

* Or, indeed, any form of thermal or electromagnetic energy. This particular bit of equivalence, the British Thermal Unit, is an early nineteenth-century measurement that has been mostly replaced by a frighteningly large array of units, including calories (and kilocalories), joules (and kilojoules), electron volts, kilowatt-hours, and therms, each of which can be converted to the others.

* Some histories still insist that in 1543, a naval officer in the service of Charles V of Spain named Blasco da Garay used steam to propel a boat across Barcelona harbor, though the story has been thoroughly debunked for more than a century.

* In the east end of the North Corridor on the first floor of the Senate wing of the U.S. Capitol is a series of frescoes painted by the Italian émigré artist Constantino Brumidi, thematically coordinated with the specific duties of Senate committees; over the doors leading to Room S-116, where the Committee on Patents originally met, are three portraits. Two of the subjects—Benjamin Franklin and Robert Fulton—are as well known as any names in American history. The third is John Fitch.

* The Russell (or Scott Russell) linkage was actually invented and patented in 1803 by the watchmaker William Freemantle and only decades later named for the naval architect John Scott Russell.

* Confusingly so. Steam is actually invisible; the clouds are just evidence that steam has condensed back into water vapor.

* Literally; in addition to biographies of Watt, Smeaton, Maudslay, Dudley, Boulton, and dozens of other inventors and engineers, he also wrote, in 1859, the worldwide bestseller titled Self-Help: With Illustrations of Character and Conduct.

* Not all of Stephenson’s historically significant inventions were associated with railroads, or even steam. His invention of a safety lamp, one that placed a barrier such as metal gauze between the candle and surrounding gas practically saved the deep coal mining industry. Stephenson’s eponymously named “geordie” was virtually simultaneous with a similar one invented by the Cornish chemist, and onetime partner of Richard Trevithick, Humphry Davy; the dispute over primacy continues to this day.

* Hackworth never did accept his loss at Rainhill, and he and his supporters argued that the boiler failure was actually sabotage; perhaps imprudently, Hackworth had ordered it from Robert Stephenson’s workshop in Newcastle-on-Tyne. In the event, his accusation was dismissed, and the Liverpool & Manchester Railway ended up buying Sans Pareil.

 The inaugural day for the Liverpool & Manchester is famous for the death of Liverpool MP William Huskisson, who was run down by Rocket. Just as widely reported, and far more lauded, was the heroic dash George Stephenson made in Rocket to the nearest hospital, during which he averaged 36 mph for fifteen miles.

If you find an error or have any questions, please email us at Thank you!