One Good Turn: A Natural History of the Screwdriver and the Screw

by Rybczynski, Witold

  —

  The matchlocks at the Metropolitan Museum are displayed in a small room that is part of a large area devoted to arms and armor. After examining the guns I decide to take a look at the armor. This is not research—I simply have fond boyhood memories of reading Ivanhoe and seeing the Knights of the Round Table at the movies. The centerpiece of the main gallery is a group of knights mounted on armored steeds. The armor, which was tinned to prevent rusting, is shiny. There are banners and colorful pennants, which give the display a jaunty, festive air; it is easy to forget that much of this is killing dress. The day I visit, the place is full of noisy, excited schoolchildren. I stop at a display case containing a utilitarian outfit, painted entirely black—not the Black Knight, just a cheap method of preventing rust. The beak-shaped helm has only a narrow slit for the eyes. “Neat!” the boy beside me exclaims to his companion. “It’s just like Darth Vader.”

  The display is German armor from Dresden, dated between 1580 and 1590. This is slightly later than what is generally considered to have been the golden age of armor, which lasted from about 1450 to 1550. Contrary to the movies of my boyhood, King Arthur’s knights, who lived in the sixth century, would have worn chain mail, not steel armor. Protective steel plates came into use only at the end of the thirteenth century. First the knees and shins were covered, then the arms, and by about 1400, the entire body was encased. The common method of connecting the steel plates was with iron, brass, or copper rivets. When a small amount of movement was required between two plates, the rivet was set in a slot instead of a hole. Removable pieces of armor were fastened with cotter pins, turning catches, and pivot hooks; major pieces, such as the breastplate and backplate, were buckled together with leather straps.

  The Dresden suit is identified as jousting armor. Jousting, or tilting, originated in martial tournaments in which groups of mounted knights fought with lance, sword, and mace. By the sixteenth century, this rude free-for-all had evolved into a highly regulated sport. Two knights, each carrying a twelve-foot-long blunted wooden lance, rode at each other on either side of a low wooden barricade called the tilt. The aim was to unseat the opponent, have him shatter his lance, or score points by hitting different parts of the body. To protect the wearer, jousting armor was heavily reinforced and weighed more than a hundred pounds (field armor was lighter, weighing between forty and sixty pounds).

  The black Dresden armor was for the scharfrennen, a particularly deadly German form of joust fought with sharpened lances and particularly popular with young men. Such combat required additional protection. The helm, called a rennhut, covered only the head and upper part of the face. The lower part of the face and the neck were protected by the renntartsche, a large molded plate that extended down to cover the left shoulder and was attached to the breastplate. A small shield, called a tilt targe, was fastened to the breastplate. Such “target” pieces were designed to fall off when struck; sometimes they were fitted with springs that caused them to fly dramatically into the air to the delight of the wildly applauding spectators.

  Like most of the armor in the gallery, the steel plates of the Dresden suit are held together by rivets and buckled straps. Then I notice something: the renntartsche is screwed to the breastplate—the slotted heads, about half an inch in diameter, are plainly visible. Armorers, too, used screwdrivers! Since armor plate is relatively thin, these screws are probably mated with nuts, although I can’t see them since they are hidden inside the suit. The Greenwich Armory outside London employed a dozen or more general armorers as well as a variety of specialists such as platers, millmen, helmsmiths, mail-makers, and locksmiths. It was probably the latter who fabricated the screws (medieval locks sometimes used threaded turning mechanisms).

  We can be fairly sure how these screws and nuts were fabricated. In Mechanick Exercises, Moxon includes a section titled “The Making of Screws and Nuts,” a process that could not have changed much since the Middle Ages. He describes how, after the head and shank are hammered out of a forged blank, the “screw-pin,” that is, the thread, is cut with a die called a screw plate. The screw plate, made of tempered steel, has several threaded holes of different diameters. The blank is placed in a vise, and the screw plate is forced down hard and turned to cut the threads. (The corresponding nut is threaded with a tap, a tapered screw fitted with a handle.) “Screw the Nut in the Vise directly flat, that the hole may stand upright, and put the Screw-tap upright in the hole; then if your Screw-tap have a handle, turn it by the handle hard round in the Hole, so will the Screw-tap work it self into the Hole, and make Grooves in it to fit the Threds [sic] of the Screw-pin.”13 Moxon’s complicated instructions underline the combination of delicacy and brute strength that was needed to make a screw in this fashion.

  Looking more closely at the Dresden armor, I see that the helm is attached to the backplate by large wing nuts. Since the highest points in a joust were accorded to a hit to the helm, special precautions had to be taken to protect the head. Field helms were close-fitting and worn over a coif of chain mail; the heavy jousting helm, on the other hand, did not touch the head. It was supported on the shoulders like a modern deep-sea diver’s helmet and attached to the breastplate and backplate with leather straps to keep from getting knocked off. “In suits for the joust or tourney these adjustable fastenings could not always be depended upon,” observes Charles Ffoulkes in a 1912 book on armor, “and the great helm . . . [was] often screwed on to the suit.”14 Wing nuts, such as the ones on the Dresden armor, were a later refinement that allowed the exact angle of the helm to be closely adjusted. This was important. The so-called frog-mouth helm had a narrow, beaklike viewing slit, designed so that the knight could see out as he leaned forward in the saddle, riding toward his adversary. At the last minute, just before the moment of impact, he would straighten up and the lower part of the helm would protect his eyes from stray splinters. It required nerve: galloping down the list, aiming the heavy lance at one’s opponent who was barely visible through the helm’s shaking, narrow slot, then sudden darkness followed by the jarring crash of wood against steel.

  Bracket for jousting helm and protective renntartsche, Dresden, sixteenth century.

  Multipurpose armorer’s tool, sixteenth century.

  It is unclear exactly when screws were substituted for straps. Ffoulkes refers to a French military manual, written in 1446, that provides a detailed description of jousting armor. The text refers to most attachments as cloué (literally “nailed,” as rivets were called arming nails), but in one place describes a piece as being rivez en dedens (fixed from the inside), which sounds like a screw and nut. I came across references to helms being screwed to breastplates as early as 1480.15 The oldest screw in the Metropolitan Museum is part of a steel breastplate that is identified as German or Austrian and dated 1480–90. If screws were used in the 1480s, that would make them the same age as the screws in the matchlock in Pollard’s History of Firearms and the metal screws in the Housebook. Ffoulkes describes the heads of the screws as square or polygonal. However, all the screws I saw at the Metropolitan were slotted.

  I look through Ffoulkes’s chapter on “Tools, Appliances, Etc.” According to the author, few armorer’s tools have survived. He describes a display in the British Museum: “In the same case is a pair of armourer’s pincers, which resemble the multum in parvo tools of today, for they include hammer, wire-cutter, nail-drawer, and turnscrew.”16 He refers to a photograph. Excitedly, I turn to plate V.17 I had missed it earlier. Upon closer examination I can make out what looks like a pick at the end of one handle, and at the end of the other—a flat screwdriver blade. The caption beneath the photograph gives the date as the sixteenth century.

  Another combination tool. I am disappointed that the oldest screwdriver resembles the kind of gimcrack household gadget that is sold by Hammacher Schlemmer. Although Ffoulkes calls this a turnscrew, like the screwdriver blade that was part of the arquebusier’s spanner, it probably didn’t have a special name. With so
few screws, all that was needed was a part-time tool.

  CHAPTER FOUR

  The Biggest Little Invention

  IN SEARCHING FOR the first screwdriver I have become interested in screws. When Agricola compared the screw to the nail as a way of constructing bellows, he observed that “there is no doubt that it [the screw] surpasses it in excellence.”1 In fact, the wrought-iron nail is a remarkable fastener. It bears little resemblance to the modern steel nail. The modern nail is round and pointed and forces itself between the wood fibers. Such nails are reasonably effective when driven into softwood (spruce, pine, fir), but will usually split hardwood (maple, birch, oak). Moreover, even in softwood the holding power of a round nail is weak, since it is kept in place only by the pressure of the fibers along two sides. The wrought-iron nail, on the other hand, is square or rectangular in cross-section with a hand-filed chisel point. The chisel point, driven across the grain, cuts through the wood fibers rather than forcing its way between them, just like a modern railroad spike. Such nails can be driven into the hardest wood without splitting it, and they are almost impossible to remove, as I discovered when I nailed a replica wrought-iron ship’s nail into a board as an experiment.I

  Wrought-iron nails have limitations, however. If they are driven into a thin piece of wood, such as a door, their holding power is greatly reduced and their protruding ends must be clenched—bent over—to keep them fast. Wrought-iron nails are most effective—and easier to fabricate—when they are relatively large (at least an inch or two long). That is why the earliest screws replaced nails in small-scale applications such as fixing leather to a bellows board, or attaching a matchlock to a gunstock. Even a short screw has great holding power. Unlike a nail or a spike, a screw is not held by friction but by a mechanical bond: the interpenetration of the sharp spiral thread and the wood fibers. This bond is so strong that a well-set screw can be removed only by destroying the surrounding wood.

  The problem with screws in the sixteenth century was that, compared to nails, they were expensive. A blacksmith could turn out nails relatively quickly. Taking a red-hot rod of forged iron, he squared, drew, and tapered the rod to a point, pushed the reheated nail through a heading tool, then with a heavy hammer formed the head. The whole procedure, which had been invented by the Romans and was still used in the 1800s (Thomas Jefferson’s slaves produced nails this way at Monticello), took less than a minute, especially for an experienced “nailsmith.” Making a screw was more complicated. A blank was forged, pointed, and headed, much like a nail, but round instead of square. Then a slot was cut into the head with a hacksaw. Finally, the thread was laboriously filed by hand.

  Gunsmiths manufactured their own screws, just as armorers made their own bolts and wing nuts. What about clockmakers? Turret clocks appeared in Europe as early as the fourteenth century. The oldest clock of which we have detailed knowledge was built by an Italian, Giovanni De’Dondi. It is an astronomical clock of extraordinary complexity. The seven faces show the position of the ancient planets: the Sun, Moon, Mercury, Venus, Mars, Jupiter, and Saturn; in addition, one rotating dial indicates religious feast days, and another displays the number of daylight hours in the day. De’Dondi fashioned the bronze, brass, and copper parts by hand. It took him sixteen years to build the clock, which he finished in 1362. Although the original was destroyed by fire in the sixteenth century, the inventor left detailed instructions, and two working replicas were built in London in 1962. One of these now belongs to the National Museum of American History, and I catch up with it in Montreal, where it is part of a temporary exhibit. The exquisite seven-sided machine stands about four feet tall; the gearwheels are driven by suspended weights. I examine the mechanism. As far as I can see, all the connections are pegged mortises and tenons, a detail adapted from carpentry. The projecting tenons have slot-holes into which a wedge is driven. These wedges vary in size from tiny needlelike pins to one inch long. There must be several hundred such attachments, but I can’t see a single screw.

  According to Britten’s Old Clocks and Watches and Their Makers, the standard work of horological history originally published in 1899, “screws were entirely unknown in clocks before 1550.”2 Their introduction was a result of the demand for smaller and lighter domestic clocks, especially watches. According to Britten’s, “Even the earliest watches generally possess at least one screw. These screws have dome-shaped heads and the slots are V-shaped. The thread is coarse and irregular.”3

  —

  By the mid-sixteenth century, applications for screws had grown to include miniature screws and bolts in watches, larger screws in guns, and heavy bolts in armor. Yet it was another two hundred years before demand grew enough that a screw industry developed. The Encyclopédie mentions that the region of Forez, near Lyon, specialized in screws, which were available in a variety of lengths—one-half inch to four or five inches. These screws were still so expensive that they were sold individually. According to the Encyclopédie, heads were either slotted or square.

  In England, screw-making was concentrated in the Midlands. It was organized as a cottage industry. Forged-steel blanks with formed heads were made in large quantities by local blacksmiths and delivered to the so-called girder, who, with his family and an assistant or two, worked at home. The first step was to cut the slot, or “nick,” into the head with a hacksaw. That was the easy part. Next the thread, or “worm,” had to be filed by hand. Some girders used a spindle—a crude lathe—turning a crank with one hand and guiding a heavy cutter with the other, back and forth, back and forth. Whichever method was used, the work was slow and laborious, and since the worm was cut by eye, the result was a screw with imperfect, shallow threads. According to one contemporary observer, who had seen screw-girders at work, “The expensive and tedious character of these processes rendered it impossible for the screws to compete with nails, and consequently the sale was very small. The quality was also exceedingly bad, it being impossible to produce a well-cut thread by such means.”4

  Both Moxon and the Encyclopédie mention that screws are used by locksmiths to fasten locks to doors. I also come across references to eighteenth-century carpenters using screws to attach hinges, particularly the novel garnet hinge. A garnet hinge resembles a |—, the vertical part being fastened to the doorjamb and the horizontal to the door. Garnet hinges, used with light cupboard doors and shutters, were screwed rather than nailed to the frame. Heavy doors, on the other hand, were hung on traditional strap hinges that extended the full width of the door and were nailed and clenched.

  Strap and garnet hinges are still used today, but by far the most popular modern door hinge is the butt hinge, which is not mounted on the surface but mortised into the thick end—the butt—of the door. Butt hinges are aesthetically pleasing, being almost entirely hidden when the door is closed. They were used in France as early as the sixteenth century (butt hinges are illustrated by Ramelli), but were luxury objects, crafted by hand of brass or steel. In 1775, two Englishmen patented a design for mass-producing cast-iron butt hinges.5 Cast-iron butt hinges, cheaper than strap hinges, had one drawback: they could not be nailed. Nails worked themselves loose as the door was repeatedly opened and closed, and since the nails were in the butt of the door, they could not be clenched. Butt hinges had to be screwed.

  By coincidence, at the very moment that butt hinges were being popularized, a technique for manufacturing good-quality, inexpensive screws was being perfected. Years earlier, Job and William Wyatt, two brothers from Staffordshire in the English Midlands, had set out to improve screw-making. In 1760, they patented a “method of cutting screws of iron commonly called wood-screws in a better manner than had been heretofore practiced.”6 Their method involved three separate operations. First, while the forged blank of wrought iron was held in a rotating spindle, the countersunk head was shaped with a file. Next, with the spindle stopped, a revolving saw-blade cut a slot into the head. Finally, the blank was placed in a second spindle and the thread was cut. This was the m
ost original part of the process. Instead of being guided by hand, the cutter was connected to a pin that tracked a lead screw. In other words, the operation was automatic. Now, instead of taking several minutes, a girder could turn out a screw—a much better screw—in six or seven seconds.

  It took the Wyatt brothers sixteen years to raise the capital required to convert a disused water corn-mill north of Birmingham into the world’s first screw factory. Then, for unexplained reasons, their enterprise failed. Maybe the brothers were poor businessmen, or maybe they were simply ahead of their time. A few years later, the factory’s new owners, capitalizing on the new demand for screws created by the popularity of butt hinges, turned screw manufacturing into a phenomenal success. Their thirty employees produced sixteen thousand screws a day.7

  Machine-made screws were not simply produced more quickly, they were much better screws. Better and cheaper. In 1800, British screws cost less than tuppence a dozen. Eventually, steam power replaced waterpower in the screw factories, and a series of improvements further refined the manufacturing process. Over the next fifty years, the price dropped by almost half; in the following two decades, it dropped by half again. Inexpensive screws found a ready market. They proved useful not only for fastening butt hinges but for any application where pieces of thin wood needed to be firmly attached, which included boatbuilding, furniture-making, cabinetwork, and coachwork. Demand increased and production soared. British screw factories, which had annually produced less than one hundred thousand gross in 1800, sixty years later produced almost 7 million gross.8