The Dawn of Innovation (9 page)

The long evolution brought forth a new culture. Or perhaps it was the other way round. The great-grandfather of the British empiricist tradition, after all, was a Briton: William of Ockham, a medieval monk. Whatever the reason, the British were different from other Europeans. The eager emigrants fleeing to the American colonies may have viewed their mother country as stifling and class-ridden, but no other established nation was as free and democratic or gave as much scope to the individual. Great Britain was nominally ruled by a languid upper class, but unlike a country like France or Italy, it allowed room for energetic climbers in the middle; indeed, the top strata admitted the most successful strivers to their own ranks. Joseph Whitworth came from a middle-class family but was made a baronet in recognition of his great contributions to British technology.

England was commercial to its toes. Money talked—even dominated the conversation. Contracts were honored, patents generally respected. The country was empirical, swayed by what worked, disposed to clinch an argument with numbers. Its banking and monetary practices were the best in the world, its currency reliable. Honor was paid to the new and the better. The powers that be were more willing than elsewhere to disrupt old rhythms and break old patterns. Turnpikes, canals, and then railroads proliferated sooner than anywhere else. In short, Great Britain was the perfect petri dish for the viruses of industrial revolution, the benign and the noxious alike.
¹¹

All complex developments are in some degree path-dependent. Early choices and random cornerings may dominate outcomes far down the road. Because of the importance of the navy to national survival, a powerful stream of British technology was driven by naval priorities, which imparted a particular twist toward the ultraprecise. A century later, that bias, perhaps interacting with a certain upper-class intellectual style, may have disadvantaged Great Britain in its inevitable industrial confrontation with the United States.

Eighteenth-century British admirals grumbled about keeping their ships out past August. The navigational apparatus for taking latitude readings—the north/south position—was quite accurate. The noon sighting—to fix the latitude, recalibrate the ship's clock, and turn the calendar—was inviolable ritual on British warships. A captain could readily find a line on the same latitude as the mouth of the English Channel and ride it home. Without obvious landmarks, however, it was much harder to divine how far east or west you were. That was the longitude problem, and to a seafaring nation it was of first importance. Without accurate longitude readings, ships could lose all sense of location in open oceans. Even when familiar trade routes were known to harbor pirates, merchants dared not vary from them for fear of getting lost. Muddled positioning extended voyages far beyond expectations: men got scurvy; missing a small island with fresh water could be a death sentence.

Almost all voyages home were by way of the Channel, so making the entrance was a routine but dangerous part of any sea captain's job. The Channel entrance is wide: the distance from the Isles of Scilly off the south coast of Cornwall to Ushant in Brittany is about 120 miles. It is an area of swirling currents and treacherous, relatively featureless rocky coasts, and the setting for Alfred Hitchcock's
Jamaica Inn
, in which locals live off the pickings from shipwrecks and sometimes engineer the wrecks. In fog or other difficult weather, any captain could lose his bearings.

A military convoy returning home in 1704 mistook the looming Cornwall coast for Guernsey, which is off Brittany. Believing they were in the channel, they turned north and sailed some sixty miles up the Atlantic coast before realizing their mistake. More tragically, in 1707, a returning war fleet, under the impression they were well into the channel mouth, made the turn north and ran directly on the rocks of the Scillies, losing four warships, a popular admiral, and 2,000 crewmen. The shock led to the passage of the Longitude Act, which offered a series of prizes for full or partial solutions to the challenge of accurately positioning a ship at sea on the east-west axis.
¹²

There were two potential solutions. One involved time. If you set a clock at Greenwich time before leaving England, you could accurately calculate longitude simply by taking the difference between Greenwich time and sun time at your location. But the clock had to be
extremely
accurate, off by less than three seconds a day. If the Greenwich time readings drifted by even very modest amounts, they would add up to disablingly large variations over the weeks or months of a typical sea voyage. Just as challenging, the clock would need to be utterly impervious to the extremities of a sailing-era sea voyage: the sharp temperature changes, the storm batterings, the salt everywhere. In 1721, no less an authority than Isaac Newton declared that it would be all but impossible for a solution to the longitude problem to come from the “Watchmakers.”

The “Astronomical” solution, which Newton preferred, was at least as difficult. Sailors long ago learned to fix latitudes because the apparent path of the sun was so readily observed and easily measured. The so-called fixed stars also had a regular apparent path around the earth, but it was far too slow to be useful. Then there was the moon, which does have a regular pattern but an extremely complicated one that varies with the seasons, local variations in the earth's magnetic fields, and much else. In principle, however, it was possible to precisely chart the moon's position with reference to the fixed stars. If you looked up the moon's position in a moon chart, it would tell you the exact time that pattern occurred over Greenwich. If you also knew your local time, you could calculate, again in principle, your east-west position.
¹³

In principle. But observational instruments were not nearly accurate enough to track anything but the grossest positional changes of the moon with reference to the fixed stars. Even if they had the requisite accuracy, it would be very difficult to take such readings from the deck of a rolling ship. There were also some nasty mathematical complications to correct both your position and the reading from Greenwich to that of an observer at the center of the earth.

Newton, for once, was wrong, and the watchmakers won. A self-taught genius named John Harrison built four candidate clocks over thirty years. They were highly innovative but extremely complex, and there were serious questions about their reproducibility. Nevertheless, all of them met the requirements for the prize, although it took the intervention of the king to secure Harrison his award, in part because of opposition from the astronomers who dominated the awards committee.
^l

In the event, roughly a quarter century after Harrison's death in 1777, watchmakers in both France and England were turning out affordable and reliable pocket-sized marine chronometers that enabled longitudinal calculations satisfactory for most purposes. Only a few of Harrison's innovations were retained. Most chronometer makers chose to stick with traditional forms and mechanisms but learned how to execute them at new orders of precision.

The astronomers got there too, by dint of an informal fifty-year international collaboration to build the necessary tables of lunar motion, along with the development of the sextant, the first instrument with the precision required for useful positional readings on the stars. The great Swiss mathematician Leonard Euler contributed practical methods for correcting the data, but they still took an expert some four hours of calculation.
Their practicality, that is, was hardly better than that of Harrison's strange instruments. But the lunar charts were maintained and promulgated, and the correction math simplified, so by the first decades of the nineteenth century the two approaches were coexisting comfortably, with practical navigators frequently checking one against the other.

It is a remarkable episode. A century elapsed between the 1707 Scilly Isles tragedy and achievement of a stable solution, but it was pursued consistently and diligently over that entire span. Although there was an international flavor to the longitude project throughout, it was driven primarily by the British.

Its lasting stamp on British technology was something of an obsession for absolute mechanical precision, or what British machinists came to call “the truth.”

Until well into the nineteenth century, machinists' tools typically lacked graduated measurement markings. In fact it was hard to do. All draftsmen knew how to make accurate divisions by geometric methods, using a compass and square edge, but beyond fairly crude resolutions, any method of marking by hand was apt to be greatly inaccurate. The solution was leverage. Releasing a pin could drop a trip-hammer: a small motion was converted into a much larger one. But leverage could be reversed, and a gross movement converted into a much finer one. And that was the path of truth.
¹⁴

The illustrations on pages 52–53 show various methods of achieving greater precision from imprecise measuring tools, most commonly by exploiting the leverage gained from screws and gears. Assume a tool or workpiece held by a chuck that is moved by a screw with twenty threads to the inch. Rotate the screw one full rotation, and the chuck advances by a twentieth of an inch. A gear train with a net twenty-to-one gear-tooth ratio would accomplish the same result. Such solutions are easy to envision, but they just relocate the problem—from making accurate measurements
to making accurate screw threads and gear teeth. Clockmakers had small machines for cutting gear teeth early in the eighteenth century, but they weren't especially precise. Individual prodigies like John Harrison could work marvels of precision by hand, but that was not a solution either. The challenge was to embed the required level of precision in machinery that could make other machines, so those machines could pass on their precision to generations of new tools and instruments. That took the better part of a hundred years.

Why did it take so long? Because as a practical matter, it is not possible to make an accurate screw thread without an accurate screw-cutting lathe, which is impossible to make without tools with accurate screw threads. In other words, accuracy could be achieved only by a process of successive approximation. And that is a tedious path, with many byways, involving better metals, better bearings, even better magnifying lenses. The work took place primarily in England because naval and other high-end engineering applications had created a market for high-precision scientific instruments for astronomy, surveying, and a host of industrial uses. Brilliant scientists in other countries were not as successful. For example, a French nobleman, the Duc de Chaulnes, made several important advances in gear-cutting machinery, but he was working with his own money and lacked the thick network of machine users, designers, and craftsmen that existed in England.

Jesse Ramsden usually gets credit for inventing the first industrial-scale dividing engine. It didn't cut the gear teeth but marked their placement, which was the essential task. “Inventing,” in this context, is not quite the right word, for all such machines were developments of others' work. Several important features of the Ramsden dividing engine, like screw-based motion controls, were inspired by predecessors like de Chaulnes, as Ramsden freely acknowledged. One of his best-known instruments representing “the best design of the time”
¹⁵
(from the late 1780s) was a large theodolite, or surveying telescope, which fixes locations by taking the intersection of horizontal and vertical circles. Measurements were read from verniers (see illustration) and viewed through microscopes. Ramsden built two of the instruments, which were used for a complete survey of Great Britain. At a distance of ten miles, the instrument was accurate to one arc second, or about three inches.

 

A.
Astronomer Tycho Brahe (1546–1601) popularizes the use of linear transversals to achieve greater precision. If the marks on the two axes at the limits of the day's technlogy for acurate gradiation, simply making the grid and drawing the transversals as shown improves the accuracy by a factor of ten. The heavy vertical line intersects at the. 04horizontal mark.
B.
By yhe eighteenth century astronomers learned to improve astral measurements with are transversals. The numbered ring is marked in 5-degree intervals. The six inner concentrics subdivide to an accuracy of 50 minutes. The right-hand heavy line from the origin intersects at the 150-minute mark, so the angle measures 5 degrees plus 150 minutes = 7,5 degrees.
C.
The Vernier Caliper uses a second sliding rule to mark out very small distances. In the example, the caliper marking is beyond the.30 position on the fixed scale. The additional distance is read from the vernier scale at the point where it lines up
most closely
with a marking on the fixed scale, as shown. 5Humans are very good at recognizing when two moving lines line up—"vernier acuity.")