Every organic economy runs on the annual output of its land, and the way to see how tight that constraint was is to work out how many acres of woodland a ton of iron used to cost.
The previous lesson left cotton spinning growing at 7 per cent a year against water power that could not grow at all. The way out ran through two materials. Iron had to become cheap enough to build machines from, and a fuel had to be found that did not compete with food for land. Both problems were solved in the same coalfields by the same people, and the second solution is the more fundamental of the two.
The acreage of an iron industry
Iron is made by reducing its ore with carbon at high temperature. Before the eighteenth century the carbon was charcoal, and charcoal is made by burning wood slowly under cover, which converts about four tons of wood into one ton of charcoal.
Take the numbers through. Smelting a ton of pig iron in an eighteenth century blast furnace consumed roughly two tons of charcoal, so roughly eight tons of wood. A managed coppice woodland, cut on rotation as English woods were, yields something like two tons of dry wood per acre per year. Therefore each ton of pig iron produced annually required about acres of woodland kept permanently in production.
That ratio is the whole story of pre-industrial metallurgy. British pig iron output in 1806 was about 244,000 tons, which on the same arithmetic would have needed nearly a million acres of dedicated coppice. Output in 1850 was about 2.25 million tons, needing 9 million acres, which is 28 per cent of the surface of England including its cities, mountains and arable. A charcoal iron industry of the size Britain actually had by 1850 is not merely expensive. It is geometrically impossible.
Britain had been feeling the constraint for two centuries. Iron making had migrated to wherever wood remained, the Weald and then the Forest of Dean and then Scotland and Wales, and by 1700 Britain was importing much of its bar iron from Sweden and Russia, where the forests were.
Coke
Abraham Darby, a Quaker brassfounder who had learned about coke in the malting trade, smelted iron with coke at Coalbrookdale in Shropshire in 1709. Coke is coal baked in the absence of air, which drives off the volatile matter and leaves a hard porous carbon, and it does what charcoal does without the woodland.
The story is usually told as though 1709 changed everything, and it did not, which is instructive. Coke smelting spread slowly for forty years, and the reason is chemical. Coal contains sulphur, sulphur passes into the iron, and sulphurous iron is brittle when worked hot, so early coke iron was fit for casting into pots, cylinders and rails but hard to forge into the bar iron that smiths and toolmakers needed. Darby's local coal happened to be unusually low in sulphur, which is one reason it worked for him first.
What made coke win in the end was a property charcoal lacks. Charcoal is soft and crushes under its own weight, which limits how tall a blast furnace can be; coke is hard, so the furnace can be built taller and blown harder, and a taller furnace with a stronger blast is more efficient and produces far more per day. Coke did not merely replace charcoal, it removed the ceiling on furnace size, and by the 1790s almost all British pig iron was coke-smelted.
Example. Compare the coke and charcoal routes on land use alone, for Britain's 1850 output of 2.25 million tons of pig iron.
The charcoal route needs about 4 acres of permanent coppice per annual ton, so 9 million acres. The coke route needs coal: roughly 3 tons of coal per ton of pig iron by the 1840s, so about 6.75 million tons of coal a year, which came out of pits occupying a few thousand acres of surface between them. Expressed as land, the ratio is on the order of a thousand to one. That comparison is the reason it is a mistake to treat coal as simply a cheaper fuel. Charcoal is a flow, limited by the rate at which trees grow on a fixed area; coal is a stock, limited by how fast it can be dug. An economy running on a flow has a ceiling set by its acreage, and an economy running on a stock does not, at least not on any timescale the people involved can see.
Now you. Britain's coal output was about 15 million tons in 1800 and about 62 million in 1850. Express each as the acreage of coppice woodland that would have been needed to supply the same energy, and compare with the land area of Great Britain, 51.7 million acres.
Answer
Take a ton of coal at about 29 gigajoules and dry wood at about 15, with a coppice yielding 2 tons an acre each year, so an acre-year supplies about 30 gigajoules, almost exactly one ton of coal. The 1800 output is therefore worth about 14.5 million acres of woodland, and the 1850 output about 60 million acres. The second figure exceeds the entire land area of Great Britain, mountains, cities, roads and farmland included. By 1850 Britain was consuming more energy from its coalfields than the whole island could have grown if every acre of it had been planted with trees and nothing else, which is Edward Wrigley's central point and the sharpest single statement of what coal did. Treat the arithmetic as an order of magnitude rather than a measurement: energy contents vary by coal rank and wood species, coppice yields vary with soil, and neither route converts fuel to useful work at the same efficiency.
From pig to bar
Smelting is only the first half. Pig iron straight from the furnace holds three or four per cent carbon, which makes it hard and brittle, good for casting and useless for anything that must bend, be hammered or take a thread. Turning it into malleable wrought iron means burning most of that carbon out, and until the 1780s that too was done with charcoal in a finery forge.
Henry Cort patented the two processes that fixed it, rolling in 1783 and puddling in 1784. Puddling melted pig iron in a reverberatory furnace, where the flame passes over the metal and the coal never touches it, so the sulphur problem disappears and coal can be used directly. A workman stirred the melt with a long bar until the carbon burned off and the purified iron gathered into a pasty ball, which was then hauled out and worked. Rolling replaced the hammer: passing the hot ball through grooved rollers squeezed out slag and drew it to shape maybe fifteen times faster than hammering.
With Cort's processes the whole iron industry could run on mineral fuel from ore to finished bar. British pig iron output rose from about 25,000 tons in 1720 to about 6 million in 1870, a factor of 240, at an average of 3.7 per cent a year for a century and a half.
The human cost of puddling deserves a sentence, since it is easy to lose in the output series. It was done by hand in front of an open furnace, stirring several hundredweight of molten metal for twelve hours, and puddlers were famously strong, famously well paid and famously short-lived.
The hot blast
One further improvement is worth stating because it shows how much slack remained in a process everyone assumed was mature. James Beaumont Neilson, manager of the Glasgow gas works, patented in 1828 the idea of heating the air blown into a blast furnace instead of blowing it cold.
The intuition of every ironmaster was the opposite, since cold air is denser and it seemed obvious that the furnace wanted cold. Neilson's furnaces cut coal consumption in Scotland from roughly 8 tons per ton of iron to roughly 2.5, a saving of about 69 per cent, and they could use raw coal rather than coke and work the previously useless blackband ironstone of Lanarkshire. Scottish pig iron production, negligible before, became a major share of British output within twenty years.
The point to take is not the number but the pattern. Neilson's change required no new material, no new machine and no new science; it required someone to test an assumption that had never been tested. That describes a large fraction of the improvements in this course.
Example. Put the hot blast in money. Coal at a Scottish pithead cost on the order of 5 shillings a ton and pig iron sold for around £5. What does the saving do to the cost of a ton of iron?
The saving is tons of coal per ton of iron, worth shillings, which is £1 7s 6d. Against a selling price near £5 that is about 28 per cent of the revenue of a ton of iron, turning up as pure margin for whoever adopted it first and as a price cut once everyone had. So adoption was not really a choice: an ironmaster who declined the hot blast was going to be undersold within a few years, which is the mechanism by which a good process becomes universal without anybody planning it. Neilson then spent much of the 1830s and 1840s in court enforcing his patent against Scottish ironmasters who had adopted it without paying, and won.
Now you. Why did the hot blast transform Scotland in particular rather than Shropshire or south Wales?
Answer
Because it changed what counted as ore and what counted as usable fuel. Lanarkshire sat on blackband ironstone, a mineral holding both iron and enough coal within it to help fire its own smelting, which had been ignored because working it with the old cold blast was uneconomic. The hot blast also allowed raw splint coal to be used directly instead of being coked first, and Scotland had splint coal in quantity. So a region with poor conventional endowments was suddenly holding first-rate ones, and Scottish pig iron output went from negligible to a large share of Britain's within twenty years. The general lesson is that a resource is not a fact about geology alone: it is a fact about geology and technique together, and a change in technique can create a resource where there was none.
Getting the coal out
None of this works unless the coal keeps coming, and British coal output rose from about 2.6 million tons in 1700 to about 110 million in 1870, a factor of 42 at an average of 2.2 per cent a year. Britain was producing the great majority of the world's coal for most of that century.
The binding constraint on a coal pit is water. Shallow seams near an outcrop drain themselves; every metre deeper adds water that has to be lifted continuously, and by the late seventeenth century the profitable shallow coal in the northeast and the Midlands was worked out. Drainage was the reason the first steam engines were built, and the first place they were economic was at the pithead, where the small coal they burned had almost no market value. Coal paid for the engine that made deeper coal possible, which is the tightest feedback loop in the whole of industrialisation.
The other constraints were human. Deep pits are hot, dark and full of explosive gas, and the Felling colliery explosion of 1812 near Gateshead, which killed 92 men and boys, was the disaster that prompted Humphry Davy's safety lamp of 1815. The lamp made deeper and gassier seams workable, which meant more men underground in more dangerous places, and the total death toll went up rather than down. That is a recurring shape in this subject and worth naming: a safety improvement that removes a constraint on production usually raises exposure faster than it lowers risk.
Example. Britain produced about 15 million tons of coal in 1800 for a population of about 10.5 million, and about 62 million tons in 1850 for about 20.8 million. Work out coal per head, and turn the 1850 figure into something a person can picture.
Per head, 1800 gives tons and 1850 gives tons, so output per person roughly doubled while the population itself doubled. Taking a ton of coal at about 29 gigajoules, the 1850 figure is about 86 gigajoules a year for every man, woman and child in Britain. A human labourer sustains something like 100 watts of useful mechanical output for perhaps 2,000 hours a year, which is 0.72 gigajoules. The comparison is therefore about 120 to one in raw energy terms. That ratio is deliberately unfair, because heat engines of the 1850s converted only a few per cent of the heat into work, so the honest figure for equivalent workers is more like a handful per person rather than 120. Even the honest figure is unprecedented: no previous society had ever had more mechanical energy at its disposal than its own muscles could supply.
Now you. Why is coal at a pithead a fundamentally different economic good from coal delivered to London?
Answer
Because for most of this period the cost of coal is overwhelmingly the cost of moving it. At the pithead the price was a few shillings a ton and the small coal unfit for the London market was nearly worthless, which is why fuel-hungry engines and fuel-hungry furnaces were built at pit mouths and nowhere else. In London, delivered by sea from Newcastle, the same coal cost several times as much, and inland away from navigable water it could cost several times more again, which is why so much early industry sat on coalfields or on canals rather than near its customers. This single fact explains a large part of the industrial geography of Britain, and it is why the transport lessons and the fuel lessons in this course are the same argument seen from two ends.
What cheap iron let people build
The consequences run in two directions. The first is structural: iron cheap enough to use in quantity made possible the Iron Bridge over the Severn at Coalbrookdale, cast by Abraham Darby's grandson and opened in 1781, spanning just over 30 metres with some 378 tons of cast iron in it. It was assembled with carpenter's joints, dovetails and wedges, because nobody yet knew how to design in iron, and it is standing.
The second is precision, and it is the one that unlocks the next lesson. John Wilkinson patented a boring machine in 1774 that held the cutting bar rigidly at both ends, so it bored a true cylinder rather than following the wanderings of the hole it started from. James Watt said of a cylinder Wilkinson bored for him that it erred by no more than the thickness of an old shilling across a diameter of several feet. That is roughly a millimetre and a half, and it is the difference between a steam engine and a leaky curiosity. Henry Maudslay's screw-cutting lathe around 1800 and Joseph Whitworth's standard screw threads of 1841 continued the same line, which is the substitution of measured, repeatable machine work for a craftsman's judgement.
Cheap iron and accurate holes are the preconditions for everything mechanical that follows. What the coalfields still lacked was a way of turning their fuel into motion at a price worth paying, and the engine that did it is the next lesson.