James Watt: A Life in Five Chapters

Chapter 1 · The Instrument Maker to the University

Greenock, London and a Glasgow workshop

1736 – 1763 · Greenock · London · Glasgow

James Watt was born at Greenock on the Clyde in January 1736. His father was a shipwright, merchant and dealer in ships' instruments, and Watt grew up in a workshop.

He was a sickly child with severe headaches — migraines that troubled him all his life — and was largely taught at home before attending grammar school.

In 1755 he went to *London* to learn instrument-making, and served an intense one-year apprenticeship under John Morgan in Finch Lane, working long hours in poor conditions. He came back with the skills of a first-rate maker and with damaged health.

In Glasgow he tried to set up as an instrument maker and was blocked: the *Hammermen's Guild* refused him permission to trade, on the grounds that he had not served a full seven-year apprenticeship in the burgh.

The *University of Glasgow* rescued him. Because the university was outside the guild's jurisdiction, it could appoint him *mathematical instrument maker to the university* and give him a workshop within its precincts.

That is how a tradesman came to be among the professors. He became friendly with *Joseph Black*, who was then working out latent heat, and with *John Robison*.

It matters that Watt was an *instrument maker* rather than a millwright or an engineer. Instrument work demands precision to a small fraction of an inch and a habit of thinking about what an apparatus is actually doing. He brought both to a field where a cylinder bored anywhere near round was considered good work.

“I can think of nothing else but this machine.”

— James Watt, in correspondence during the development of the separate condenser

Why this matters

Watt was trained as a precision instrument maker, not a millwright, and brought a standard of accuracy to heavy engineering that did not exist in it.

You have the guild's refusal and the university's workshop. What would you ask him?

Ask Watt

• “Why did the guild refuse to let you trade?”
• “What does instrument-making teach that engineering does not?”
• “What was the London apprenticeship like?”
• “What did Joseph Black talk to you about?”
• “How did a tradesman end up among professors?”

Chapter 2 · The Model That Would Not Run

Finding the waste in a Newcomen engine

1763 – 1765 · Glasgow · Glasgow Green

In 1763 the university asked Watt to repair a *model Newcomen atmospheric engine* used for demonstrations. It would run only a few strokes before the boiler was exhausted.

The Newcomen engine works like this. Steam is admitted to a cylinder under a piston. Cold water is then sprayed *into the cylinder*, condensing the steam and creating a partial vacuum, so that *atmospheric pressure* pushes the piston down. That is the working stroke. Then the cycle repeats.

It had been in use since 1712, mostly for pumping water out of mines, and it was extremely wasteful — but coal at a pit head is nearly free, so nobody minded much.

Watt was a measurer. Rather than tinkering, he investigated: how much steam does a stroke actually use? How much water does condensing it require? What is the volume of steam from a given volume of water?

And he found the problem. The cold water spray cools *the cylinder itself* on every stroke. The next charge of steam must then *reheat the whole cylinder* back to boiling before it can fill it. Most of the steam — and therefore most of the fuel — is spent warming metal that was deliberately cooled a moment earlier.

Joseph Black's *latent heat* explains why the waste is so enormous: condensing steam releases a very large quantity of heat, all of which is dumped into the cylinder and the injection water and must be supplied again.

Watt saw the requirement: the cylinder must be kept *always hot*, and the condensation must happen *always cold*. Those are contradictory in one vessel.

On a Sunday walk on *Glasgow Green* in May 1765 the answer came: use *two* vessels. Condense the steam in a *separate condenser*, kept permanently cold and connected to the cylinder by a valve. The cylinder stays hot; the condenser stays cold; neither is cycled.

The fuel saving is about three-quarters.

Why this matters

The Newcomen engine cooled its cylinder on every stroke and then reheated it, and Watt's separate condenser removed that waste to cut fuel use by about three quarters.

You have the cooled cylinder and the walk on Glasgow Green. What is your question?

Ask Watt

• “Where exactly was the Newcomen engine wasting fuel?”
• “Why can't one vessel be hot and cold at once?”
• “How did latent heat help you see the size of the loss?”
• “Why did nobody mind the waste before?”
• “What made you measure rather than tinker?”

Chapter 3 · Eleven Years Between Idea and Engine

Roebuck's bankruptcy, Boulton's capital and Wilkinson's boring machine

1765 – 1776 · Glasgow · Kinneil · Birmingham · Bersham

Watt had the idea in 1765. A working commercial engine did not exist until 1776.

The eleven years are the real story, and they are the part usually skipped.

*Money*. Watt had none. He worked for years as a *surveyor* — laying out canals across Scotland, including the Forth and Clyde and the Caledonian — because it paid, and it took him away from the engine.

*The first backer*. John Roebuck, of the Carron ironworks, funded him in exchange for two-thirds of the invention. Roebuck went *bankrupt* in 1773. Watt's share was, at that point, worth nothing; Roebuck's creditors valued the engine interest at zero.

*Manufacturing*. This is the deepest problem. The separate condenser requires a cylinder in which the piston makes a reasonably tight seal, because leaking steam destroys the vacuum. Cylinders of the period were cast and then bored inaccurately; a large cylinder could be out of true by a considerable fraction of an inch. It did not matter for a Newcomen engine, where the seal was made with a layer of water on top of the piston. It mattered enormously for Watt.

In 1774 *John Wilkinson* patented a boring machine, developed for cannon, in which the boring bar is supported at *both ends* rather than cantilevered. It could bore a large cylinder true to a small fraction of an inch.

Watt said it made his engine possible. The theory had been waiting for the machine tool.

*The partner*. Matthew Boulton, of the Soho Manufactory in Birmingham, took over Roebuck's interest in settlement of a debt, and became Watt's partner in 1775. Boulton supplied capital, a factory, a workforce, salesmanship and relentless commercial drive — including getting Watt's patent *extended by Act of Parliament* to 1800.

The first commercial engines were installed in 1776.

Why this matters

Watt's engine had to wait nine years for Wilkinson's boring machine, because the theory required a cylinder nobody could yet make accurately enough.

You have eleven years, a bankruptcy and a boring machine. What would you ask?

Ask Watt

• “Why did it take eleven years?”
• “Why does the cylinder have to be bored so accurately?”
• “What did Boulton bring that you did not have?”
• “What was it like surveying canals instead of building engines?”
• “Can an invention be ahead of the tools to make it?”

Chapter 4 · Turning a Shaft

Rotary motion, double action, and an engine that governs itself

1781 – 1788 · Birmingham · Soho

The first Watt engines were pumping engines: a beam rocking up and down, driving a pump rod down a mine shaft. That is a large market and a limited one.

The transformation came from producing *rotary motion* — a turning shaft, which can drive machinery of any kind.

A crank is the obvious way to convert reciprocating to rotary motion, and Watt could not use one: *James Pickard* had patented the crank in 1780. So Boulton and Watt worked round it with the *sun and planet gear*, patented in 1781, in which a gear fixed to the connecting rod runs round a gear on the output shaft.

It works, and it has a side benefit — it turns the shaft twice per stroke — and it exists because of a patent, which is a fair illustration of what patents do.

In 1782 came the *double-acting* engine: steam admitted alternately to *both sides* of the piston, so that the engine drives on the upstroke as well as the down. That roughly doubles the power from a given cylinder.

In 1782 he also patented *expansive working*: cut off the steam supply part-way through the stroke and let the trapped steam *expand* to complete it, doing work as it expands. Less steam, nearly the same work — a further large gain in efficiency, and a direct anticipation of thermodynamic reasoning.

And in 1788 came the *centrifugal governor*, adapted from a device used in windmills. Two weighted arms spin on a spindle driven by the engine. As the engine speeds up, the weights fly outward, and their movement closes the steam valve. As it slows, they fall in and open it.

The engine regulates its own speed, automatically, with no attendant.

That is a *feedback control system*, and it is one of the earliest deliberate ones. Maxwell wrote a mathematical analysis of it in 1868 that is a founding paper of control theory. Every thermostat, cruise control and automatic regulator descends from it.

Why this matters

The centrifugal governor lets an engine regulate its own speed with no attendant, and it is one of the earliest deliberate feedback control systems.

You have the spinning weights and the self-regulating engine. What is your question?

Ask Watt

• “How does the governor control the engine?”
• “Why did you need a sun and planet gear instead of a crank?”
• “What is expansive working, and why does it save steam?”
• “What changed when the engine could turn a shaft?”
• “Did you know you had built a control system?”

Chapter 5 · The Patents

Twenty-five years of monopoly, and what it cost

1775 – 1819 and after · Birmingham · Cornwall · London

Watt's 1769 patent, extended by Act of Parliament in 1775, ran until *1800* — twenty-five years — and it was drawn very broadly.

Boulton and Watt enforced it aggressively. They did not usually sell engines outright; they charged a *royalty* of one-third of the fuel *saved* compared with a Newcomen engine, calculated annually, which required access to the customer's books. Cornish mine owners in particular resented it bitterly and there was persistent litigation.

And they used the patent to *block rivals*.

*Jonathan Hornblower* developed a compound engine, using steam expansively through two cylinders in series, and was sued into bankruptcy. Compounding is a genuine improvement and it was suppressed for the life of the patent.

*Richard Trevithick* developed *high-pressure* steam, which allows a much smaller and lighter engine — the essential requirement for a locomotive or a steamboat. Watt opposed it strenuously, believing high-pressure boilers so dangerous they should be prohibited by law, and Boulton and Watt's patent position discouraged its development.

Watt was not simply being obstructive about high pressure; boilers of the period *did* explode and kill people. But he was wrong, and the entire future of steam transport was in the direction he opposed. The first successful steam locomotive ran in 1804, four years after the patent expired.

The fair summary is that Watt's improvements made steam power economic and that his patent then held the field static for a quarter of a century.

He also treated his employee *William Murdoch* poorly, discouraging his experiments with steam locomotion and taking his gas-lighting work into the firm without proper credit.

He retired in 1800 when the patent expired, wealthy, and spent his last years in his attic workshop at Heathfield, developing a machine for copying sculptures. He died in August 1819, aged eighty-three.

The SI unit of *power* is named after him, and *horsepower* — his own unit, defined by measuring what a horse could actually do so that an engine could be priced against the animals it replaced — is still in use.

Why this matters

Watt's improvements made steam power economic and his patent then blocked compound and high-pressure engines for twenty-five years, which is a real cost to set against the benefit.

You have the royalty, the lawsuits and the blocked rivals. What would you ask?

Ask Watt

• “Was it right to charge a share of the fuel saved?”
• “What happened to Jonathan Hornblower?”
• “Why were you so opposed to high-pressure steam?”
• “Did the patent hold engineering back?”
• “How did you decide what a horse could do?”

What Watt changed

The separate condenser made steam power economic, and with it the factory system and much of the industrial revolution. Watt's habit of measuring an engine's duty — work obtained for fuel burned — is the ancestor of every efficiency calculation, and the centrifugal governor is one of the earliest deliberate feedback control systems. Horsepower and the watt are both still in daily use.

A debate that continues

Watt's patent, extended to 1800, was used to block Hornblower's compound engine and to discourage Trevithick's high-pressure steam — the direction in which the whole future of steam transport lay — and historians continue to argue about whether it advanced or retarded engineering on balance.

Keep exploring — ask Watt

• “Would you enforce the patent the same way again?”
• “What did you owe William Murdoch?”
• “Where else is a machine wasting most of its fuel?”

Related lives

• Joseph Black — Whose Thermometer Would Not Rise
• James Prescott Joule — Who Weighed Heat Against Work
• Vladimir Shukhov — Who Broke the Long Molecules
• Lord Kelvin — Who Found The Bottom Of The Scale

Related themes

Energy transfer and efficiency · Heat engines · Feedback and control