Exactly. Simon Winchester

their mutual commercial benefit, commingled. Wilkinson’s “New Method of Casting and Boring Iron Guns or Cannon” was married to Watt’s “New Invented Method of Lessening the Consumption of Steam and Fuel in Fire-Engines.” It was a marriage, it turned out, of both convenience and necessity.

      James Watt, a Scotsman renowned for being pessimistic in outlook, pedantic in manner, scrupulous in affect, and Calvinist in calling, was obsessed with getting his machinery as right as it could possibly be. While he was making and repairing and improving the scientific instruments in his workshop in Glasgow, he became well-nigh immured by his passion for exactitude, to much the same degree as had John Harrison in his clock-making workshop in Lincolnshire. Watt was quite familiar with the early dividing engines and screw thread cutters and lathes and other instruments that were then helping engineers take their first tentative steps toward machine perfection. He was accustomed to instruments that were carefully built and properly maintained, and that worked as they were intended to. He was mortally offended, then, when things went wrong, when inefficiencies were compounded, and when the monster iron engines he was now trying to build in the giant Boulton and Watt factory in Soho performed less well than the brass-and-glass models on which he had experimented back up in Scotland.

      His first prototype large engines were spectacular behemoths: thirty feet tall, with a main steam cylinder four feet in diameter and six feet long, a coal-fired boiler, and a separate steam condenser, all massive. All the working parts were connected by a convoluted spiderweb of brass pipes and well-oiled valves and levers, with a spinning two-ball governor that prevented runaways. Above it all was a heavy wooden beam that rocked back and forth with metronomic regularity, turning a huge iron flywheel that in turn worked a pump that gushed water or compressed air or performed other tasks fifteen times a minute. Once at full power, the engine produced a concatenation of noise and heat and a juddering, thudding, stomach-churning intensity that somehow seemed an impossible consequence of merely heating water up to its natural boiling point.

      Yet everywhere, perpetually enveloping his engine in a damp, hot, opaque gray fog, were billowing clouds of steam. It was this, this scorching miasma of invisibility, that incensed the scrupulous and pedantic James Watt. Try as he might, do as he could, steam always seemed to be leaking, and doing so not stealthily but in prodigious gushes, and most impudently of all, it was doing so from the engine’s enormous main cylinder.

      He tried blocking the leak with all kinds of devices, things, and substances. The gap between the piston’s outer surface and the cylinder’s inner wall should, in theory, have been minimal, and more or less the same wherever it was measured. But because the cylinders were made of iron sheets hammered and forged into a circle, and their edges then sealed together, the gap actually varied enormously from place to place. In some places, piston and cylinder touched, causing friction and wear. In other places, as much as half an inch separated them, and each injection of steam was followed by an immediate eruption from the gap. This is where the blocking came in: Watt tried tucking in pieces of linseed oil–soaked leather; stuffing the gap with a paste made from soaked paper and flour; hammering in corkboard shims, pieces of rubber, even dollops of half-dried horse dung. A solution of sorts came when he decided to wrap the piston with a rope and tighten what he called a “junk ring” around the compressible rope.

      Then, by the purest accident, John Wilkinson, in Bersham, asked for an engine to be built for him, to act as a bellows for one of his iron forges—and in an instant, he saw and recognized Watt’s steam-leaking problem, and in an equal instant, he knew he had the solution: he would apply his cannon-boring technique to the making of cylinders for steam engines.

      So, without taking the precautionary step of filing a new patent for this entirely new application of his method, he proceeded to do with the Watt cylinders exactly what he had done with the naval guns. He had Watt’s workmen haul a solid iron cylinder blank the seventy miles across to Bersham. He then strapped the blank (in this case, for the very engine that he, as customer, eventually wanted, so six feet long and thirty-eight inches in diameter) onto a firmly fixed stage, and then secured it with heavy chains to make certain it did not move by so much as a fraction of an inch. He then fashioned a massive cutting tool of ultrahard iron that was three feet across (which should in theory have produced a cut that left a thirty-eight-inch-diameter cylinder with one-inch-thick walls) and bolted it securely to the end of a stiff iron rod eight feet long. This he supported at both ends and mounted onto a heavy iron sleigh that could be ratcheted slowly and steadily into the huge iron workpiece.

      As soon as he was ready to begin working the piece, he directed, through a hose, a water-and-vegetable-oil mixture both to cool the thrashing metals and to wash away any fragments of cut iron; opened the water valve for the millrace and wheel that would set the rod and its cutting tool turning; and slowly and steadily, notch by notch by notch, set the rod moving forward until its cutting edge began chewing away at the face of the iron billet.

      After a full day of searing heat and grinding din, the cylinder was cut. The tool, hot but barely blunted, was withdrawn. The hole, three feet in diameter, looked smooth and clean, straight and true. Using a set of chains and blocks, he placed the heavy cylinder (now rather less heavy, as so much iron had been bored away) upward, on its end. The piston, fractionally less than three feet in diameter itself and smeared with lubricating grease, was carefully lifted up and over the lip of the cylinder and down into its depths.

      There was, I like to think, a round of cheers, for the piston slipped noiselessly and snugly into the cylinder and could be lifted up and down with ease and without any apparent leakage of air, of grease, of anything. It then took Watt just a few days, once the disassembled pieces were brought back to his works in Soho, to mount the cylinder in pride of place in what would now be his, and the world’s, first working full-scale single-action engine. He and his engineers then added all the supplementary parts (the pipes, the second condenser, the boiler, the rocking arm, the governor, the water tank, the flywheel) and then loaded the firebox with coal, added a primer, lit the fire, and, once the water was hot enough to set steam pouring from the safety line, opened the main valve.

      With an enormous chuff-chuff-chuff, the piston began to move up and down, up and down, out of the newly machined cylinder. The rocking beam above then began to oscillate up and back; the connecting rod on the far side started to move up and down, up and down; the set of eccentric sun-and-moon gears on the flywheel started to move; and then the huge wheel itself, tons of solid iron that would in effect store the engine’s power, started to turn.

      Within moments, with the governor’s shiny couplet of balls spinning merrily to keep matters in check, the engine was roaring along at full power, thumping and thudding and whirring and chuffing—and now all perfectly visibly because, for the first time since Watt had begun his experiments, there was no leaking steam. The engine was working at maximum efficiency: it was fast, it was powerful, and it was doing just what was demanded of it. Watt beamed with delight. Wilkinson had solved his problem, and the Industrial Revolution—we can say now what those two never imagined—could now formally begin.

      And so came the number, the crucial number, the figure that is central to this story, that which appears at the head of this chapter and which will be refined in its exactitude in all the remaining parts of this story. This is the figure of 0.1—one-tenth of an inch. For, as James Watt later put it, “Mr. Wilkinson has bored us several cylinders almost without error, that of 50 inch diameter … does not err the thickness of an old shilling at any part.” An old English shilling had a thickness of a tenth of an inch. This was the tolerance to which John Wilkinson had ground out his first cylinder.

      He might in fact have done even better than that. In another letter, written rather later—by which time Wilkinson had bored no fewer than five hundred cylinders for Watt’s engines, which were being snapped up by factories and mills and mines all over the country and beyond—the Scotsman boasted that Wilkinson had “improved the art of boring cylinders so that I promise upon a seventy two inch cylinder being not farther distant from absolute truth than the thickness of an old sixpence at the worst part.” An old English sixpence was even slighter: half of a tenth of an inch, or 0.05 inches.

      Yet this is a quibble. Whether the thickness of a shilling coin or the thinness of an old sixpence, it does not really matter. The fact is that a whole