—
Take a close look at a modern screw. It is a remarkable little object. The thread begins at a gimlet point, sharp as a pin. This point gently tapers into the body of the screw, whose core is cylindrical. At the top, the core tapers into a smooth shank, the thread running out to nothing. The running-out is important since an abrupt termination of the thread would weaken the screw.
The first factory-made screws were not like this at all. For one thing, although handmade screws were pointed, manufactured screws had blunt ends and were not self-starting—it was always necessary first to drill a lead hole. The problem lay in the manufacturing process. Blunt screws could not simply be filed to a point—the thread itself had to come to a point, too. But lathes were incapable of cutting a tapering thread. Screw manufacturers tried angling the cutters, which produced screws that tapered along their entire length. Such screws had poor holding power, however, and carpenters refused to use them. What was needed was a machine that could cut a continuous thread in the body of the screw (a cylinder) and also in the gimlet point (a cone).
An inventive American mechanic found the solution. The first American screw factories had been established in Rhode Island in 1810, using adapted English machines. Providence became the center of the American screw industry, which by the mid-1830s was experiencing a boom in demand for its products. Beginning in 1837, a series of patents addressed the problem of manufacturing gimlet-pointed screws, but it took more than a decade of trial and error to get it right. In 1842, Cullen Whipple, a mechanic from Providence who worked for the New England Screw Company, invented a method of manufacturing screws on a machine that was entirely automatic. Seven years later he made a breakthrough and successfully patented a method of producing pointed screws. A slightly different technique was devised by Thomas J. Sloan, whose patent became the mainstay of the giant American Screw Company. Another New Englander, Charles D. Rogers, solved the problem of tapering the threaded core into the smooth shank. Such advances put American screw manufacturers firmly in the lead, and by the turn of the century, when the screw had achieved its final form, American methods of production dominated the globe.
—
Ever since the fifteenth century, screws had had either square or octagonal heads, or slots. The former were turned by a wrench, the latter by a screwdriver. There is no mystery as to the origin of the slot. A square head had to be accurate to fit the wrench; a slot was a shape that could be roughly filed or cut by hand. Screws with slotted heads could also be countersunk so they would not protrude beyond the surface—which was necessary to attach butt hinges. Once countersunk screws came into common use in the early 1800s, slotted heads—and flat-bladed screwdrivers—became standard. So, even as screws were entirely made by machine, the traditional slot remained. Yet slotted screws have several drawbacks. It is easy to “cam out,” that is, to push the screwdriver out of the slot; the result is often damage to the material that is being fastened or injury to one’s fingers—or both. The slot offers a tenuous purchase on the screw, and it is not uncommon to strip the slot when trying to tighten a new screw or loosen an old one. Finally, there are awkward situations—balancing on a stepladder, for example, or working in confined quarters—when one has to drive the screw with one hand. This is almost impossible to do with a slotted screw. The screw wobbles, the screwdriver slips, the screw falls to the ground and rolls away, the handyman curses—not for the first time—the inventor of this maddening device.
American screw manufacturers were well aware of these shortcomings. Between 1860 and 1890, there was a flurry of patents for magnetic screwdrivers, screw-holding gadgets, slots that did not extend across the face of the screw, double slots, and a variety of square, triangular, and hexagonal sockets or recesses. The latter held the most promise. Replacing the slot by a socket held the screwdriver snugly and prevented cam-out. The difficulty—once more—lay in manufacturing. Screw heads are formed by mechanically stamping a cold steel rod; punching a socket sufficiently deep to hold the screwdriver tended to either weaken the screw or deform the head.
The solution was discovered by a twenty-seven-year-old Canadian, Peter L. Robertson. Robertson was a so-called high-pitch man for a Philadelphia tool company, a traveling salesman who plied his wares on street corners and at country fairs in eastern Canada. He spent his spare time in his workshop, dabbling in mechanical inventions. He invented and promoted “Robertson’s 20th Century Wrench-Brace,” a combination tool that could be used as a brace, a monkey wrench, a screwdriver, a bench vise, and a rivet maker. He vainly patented an improved corkscrew, a new type of cuff links, even a better mousetrap. Then, in 1907, he received a patent for a socket-head screw.
Peter L. Robertson’s 1907 patent for a socket-head screw.
Robertson later said that he got the idea for the socket head while demonstrating a spring-loaded screwdriver to a group of sidewalk gawkers in Montreal—the blade slipped out of the slot and injured his hand. The secret of his invention was the exact shape of the recess, which was square with chamfered edges, slightly tapering sides, and a pyramidal bottom. “It was early discovered that by the use of this form of punch, constructed with the exact angles indicated, cold metal would flow to the sides, and not be driven ahead of the tools, resulting beneficially in knitting the atoms into greater strength, and also assisting in the work of lateral extension, and without a waste or cutting away of any of the metal so treated, as is the case in the manufacture of the ordinary slotted head screw,” he rather grandly explained.9
An enthusiastic promoter, Robertson found financial backers, talked a small Ontario town, Milton, into giving him a tax-free loan and other concessions, and established his own screw factory. “The big fortunes are in the small inventions,” he trumpeted to prospective investors. “This is considered by many as the biggest little invention of the 20th century so far.”10 In truth, the square socket really was a big improvement. The special square-headed screwdriver fit snuggly—Robertson claimed an accuracy within one one-thousandth of an inch—and never cammed out. Craftsmen, especially furniture-makers and boatbuilders, appreciated the convenience of screws that were self-centering and could be driven with one hand. Industry liked socket-head screws, too, since they reduced product damage and speeded up production. The Fisher Body Company, which made wood bodies in Canada for Ford cars, became a large Robertson customer; so did the new Ford Model T plant in Windsor, Ontario, which soon accounted for a third of Robertson’s output. Within five years of starting, Robertson built his own wire-drawing plant and powerhouse and employed seventy-five workers.
In 1913, Robertson decided to expand his business outside Canada. His father had been a Scottish immigrant, so Robertson set his sights on Britain. He established an independent English company to serve as a base for exporting to Germany and Russia. The venture was not a success. He was thwarted by a combination of undercapitalization, the First World War, the defeat of Germany, and the Russian Revolution. Moreover, it proved difficult to run businesses on two continents. After seven years, unhappy English shareholders replaced Robertson as managing director. The English company struggled along until it was liquidated in 1926. Meanwhile, Robertson turned to the United States. Negotiations with a large screw manufacturer in Buffalo broke down after it became clear that Robertson was unwilling to share control over production decisions. Henry Ford was interested, since his Canadian plants were reputedly saving as much as $2.60 per car using Robertson screws. However, Ford, too, wanted a measure of control that the stubborn Robertson was unwilling to grant. They met but no deal was struck. It was Robertson’s last attempt to export his product. A lifelong bachelor, he spent the rest of his life in Milton, a big fish in a decidedly small pond.
Meanwhile, American automobile manufacturers followed Ford’s lead and stuck to slotted screws. Yet the success of the new Robertson screw did not go unnoticed. In 1936 alone, there were more than twenty American patents for improved screws and screwdrivers. Several of these were granted
to Henry F. Phillips, a forty-six-year-old businessman from Portland, Oregon. Like Robertson, Phillips had been a traveling salesman. He was also a promoter of new inventions, and acquired patents from a Portland inventor, John P. Thompson, for a socket screw. Thompson’s socket was too deep to be practicable, but Phillips incorporated its distinctive shape—a cruciform—into an improved design of his own. Like Robertson, Phillips claimed that the socket was “particularly adapted for firm engagement with a correspondingly shaped driving tool or screwdriver, and in such a way that there will be no tendency of the driver to cam out of the recess.”11 Unlike Robertson, however, Phillips did not start his own company but planned to license his patent to screw manufacturers.
All the major screw companies turned him down. “The manufacture and marketing of these articles do not promise sufficient commercial success” was a typical response.12 Phillips did not give up. Several years later a newly appointed president of the giant American Screw Company, which had prospered on the basis of Sloan’s patent for manufacturing pointed screws, agreed to undertake the industrial development of the innovative socket screw. In his patents, Phillips emphasized that the screw was particularly suited to power-driven operations, which at the time chiefly meant automobile assembly lines. The American Screw Company convinced General Motors to test the new screw; it was used first in the 1936 Cadillac. The trial proved so effective that within two years all automobile companies save one had switched to socket screws, and by 1939 most screw manufacturers produced what were now called Phillips screws.
The Phillips screw has many of the same benefits as the Robertson screw (and the added advantage that it can be driven with a conventional screwdriver if necessary). “We estimate that our operators save between 30 and 60 percent of their time by using Phillips screws,” wrote a satisfied builder of boats and gliders.13 “Our men claim they can accomplish at least 75 percent more work than with the old-fashioned type,” maintained a manufacturer of garden furniture.14 Phillips screws—and the familiar cross-tipped screwdrivers—were now everywhere. The First World War had stymied Robertson; the Second World War ensured that the Phillips screw became an industry standard as it was widely adopted by wartime manufacturers. By the mid-1960s, when Phillips’s patents expired, there were more than 160 domestic, and 80 foreign licensees.15
The Phillips screw became the international socket screw; the Robertson screw is used only in Canada and by a select number of American woodworkers.II A few years ago, Consumer Reports tested Robertson and Phillips screwdrivers. “After driving hundreds of screws by hand and with a cordless drill fitted with a Robertson tip, we’re convinced. Compared with slotted and Phillips-head screwdrivers, the Robertson worked faster, with less cam-out.”16 The explanation is simple. Although Phillips designed his screw to have “firm engagement” with the screwdriver, in fact a cruciform recess is a less perfect fit than a square socket. Paradoxically, this very quality is what attracted automobile manufacturers to the Phillips screw. The point of an automated driver turning the screw with increasing force popped out of the recess when the screw was fully set, preventing overscrewing. Thus, a certain degree of cam-out was incorporated into the design from the beginning. However, what worked on the assembly line has bedeviled handymen ever since. Phillips screws are notorious for slippage, cam-out, and stripped sockets (especially if the screw or the screwdriver are improperly made). Here I must confess myself to be a confirmed Robertson user. The square-headed screwdriver sits snugly in the socket: you can shake a Robertson screwdriver, and the screw on the end will not fall off; drive a Robertson screw with a power drill, and the fully set screw simply stops the drill dead; no matter how old, rusty, or painted over, a Robertson screw can always be unscrewed. The “biggest little invention of the twentieth century”? Why not.
* * *
I. At the beginning of the nineteenth century, handmade nails were replaced by cut nails, stamped out of sheets of wrought iron (later steel), with a similar rectangular cross-section. Cut nails are sharpened by hand with a file.
II. Starting in the 1950s, Robertson screws began to be used by some American furniture manufacturers, by the mobile-home industry, and eventually by a growing number of craftsmen and hobbyists. The Robertson company itself was purchased by an American conglomerate in 1968.
CHAPTER FIVE
Delicate Adjustments
IN READING ABOUT the Wyatt brothers’ factory in Staffordshire, I had been struck by the statement that their screw-making machines were operated by children. During the eighteenth century, children commonly worked in coal mines, workshops, and factories, but were usually given only menial tasks. Even a machine as simple as a screw girder’s spindle required an experienced—not to say strong—operator. The Wyatt machines were obviously different. I had stumbled on a landmark of industrialization.1 At a remarkably early date—the industrial revolution would not get fully under way for another hundred years—the Wyatt brothers not only pioneered the use of multipurpose machines to achieve mass production, they were the first to put into place the guiding principle of industrialization. Their factory was the earliest example of an industrial process designed specifically to shift control over the quality of what was being produced from the skilled artisan to the machine itself.
The screw girder’s spindle and the Wyatt brothers’ screw-making machines are both examples of simple turning-lathes. In a lathe, the blank, or workpiece, is rotated around an axis, somewhat like a potter’s wheel. However, while a potter creates a shape by building up clay, the turner removes material. As the workpiece turns, a sharp cutter is applied to the surface and, depending on the desired shape, removes inequalities until every part is equidistant from the axis. The lathe is an ancient tool that appears to have been invented in Europe, since the earliest surviving pieces of lathe work are an eighth-century B.C. Etruscan bowl, and a sixth-century B.C. bowl found in Upper Bavaria.2 Although these wooden objects were definitely turned, nothing is known of the lathes themselves. Turning technology eventually spread to the rest of the Mediterranean world, including Egypt, where the oldest depiction of a lathe, dating from the third century B.C., has been found in a bas-relief on a grave wall. The piece being turned, which appears to be a furniture leg, is held vertically. The turner’s cutting tool resembles a chisel; his assistant rotates the piece by pulling a cord looped around the rotating axle, or mandrel. Since the workpiece rotates in alternate directions, the turner cuts only on every other turn.
The Egyptian bas-relief shows the turner and his assistant kneeling on the ground. It reminds me of my first visit to India, when I saw a carpenter at work squatting on the floor. Just as the world is divided into those who wrap and those who button up, or those who eat with their fingers and those who eat with utensils, it is divided into craftsmen who work kneeling, squatting, or sitting on the ground, and those who work erect—or sitting—at a bench. The ancient Egyptians belonged to the former category; the Romans, to the latter. Since the Romans invented the plane, they needed a flat surface to which the workpiece could be fastened, and the result was the first carpenter’s bench.
Although Europeans in the Middle Ages often relaxed by sitting on cushions on the floor in the Oriental manner, they worked erect. This habit probably prompted the thirteenth-century European invention of the so-called pole lathe. The turner works standing up at a pole lathe. The workpiece rotates not vertically but horizontally. A cord is looped around the mandrel with one end attached to a hinged treadle, and the other fastened to a flexible pole, resembling a bowed fishing rod, that keeps the cord taut. The turner, alternately pressing and releasing the treadle with his foot, now has both hands free to guide the long-handled cutter, which he braces under his arm or over his shoulder for added stability. Like the Egyptian lathe, the pole lathe turns back and forth.
Screw-cutting lathe, from The Medieval Housebook of Wolfegg Castle, c. 1475–90.
The simple pole lathe was used by wood turners for a long time—working examples s
urvived in England until the early 1900s. For turning metal, however, a more effective machine was required. Here the screw again plays a vital role, for the ancestor of the modern lathe is in fact a machine for cutting screws. It was invented almost three hundred years before the Wyatt brothers’ screw-making lathes and appears in the Medieval Housebook, the fifteenth-century manuscript that I had consulted in the Frick Collection. The beautiful drawing is precise. The lathe, a radical departure from the pole lathe, consists of a heavy frame mounted on a solid workbench. The blank is held horizontally between two adjustable supports and rotated by turning a hand crank. One end of the blank is attached to a lead screw. As the blank turns, the lead screw advances through a threaded hole in one of the supports and pushes the blank through a box containing a sharp cutter that incises the thread. The operator has only to set up the blank in the jig, wedge the threaded support and the cutter-box in place, adjust the depth of the cutter, and turn the crank.
The Housebook lathe is made of wood, but it is a true machine tool; that is, it is a tool in which the machine—not the craftsman—controls the cutter.3 It anticipates many features of the modern bench lathe: the two supports (today called a headstock and a tailstock); the frame (ancestor of the modern slide-rest) that allows flexibility in the location of cutter-box and stocks; a continuous drive that can be connected by a belt drive to an external power source such as a waterwheel; a rotating lead screw that advances the blank by tiny increments; a design that integrates the lathe with the workbench; and heavy construction that assures rigidity and a relatively high degree of precision.
One Good Turn: A Natural History of the Screwdriver and the Screw Page 5