The turn of the millennium saw the publication of two timely little books on screws and the tools to tighten or remove them.
Both Jenkinson and Rybczynski claimed recognition of the earliest known fixing screw and its adjusting device from a diagram of the late 15th. century Hausbuch lathe.3 The tool, in German, rejoiced in the name schraubendreher, in French tournevis. These descriptions transliterate into English as turnscrew. This name still survives in many parts of Britain from the Midlands to Scotland.
Tapered Screws
Curiously, Goodman4 mentions screwdrivers once only, and turnscrews not at all, in his already classic 1964 'History of Woodworking Tools'. It could be that he didn't consider this tool as a specific woodworking device. Perhaps more surprising is his date of after the beginning of the 19th century for the invention of the screwdriver. However, within ten years he revised5 that estimate (or corrected a rare mistake) to the 16th century. He captured the essential features of early slotted head woodscrews and the tournevis; in about 250 words, and a diagram, he traced the tapered screw back to its use by gunsmiths and armourers almost 500 years ago. Oddly, the picture of the tapered screw seems to be taken from that displayed in De Re Metallica for fixing leather to bellows boards – not for the above uses.
Parallel Shank–Parallel Thread Screws
Goodman was correct in assuming that it was in Birmingham that the first semi-automatic machine-made (but blunt) screws were produced by the late 18th century. Three brothers, John, Job, and William Wyatt achieved various contributions between 1751 and 1760. Several patent applications were made. Job and William Wyatt, GB751 (1760), are credited with developing the first successful automated manufacturing process. John Wyatt produced the world's first parallel shank–parallel thread iron wood screws in 1751. The parallel thread proved to be the breakthrough in holding ability of metal wood screws.1 The name self tapping screws is usually applied to tapered metal screws to be driven into thin metal sheet. Of course, a wood screw is also self tapping; it makes its own nut when driven into softwood or pilot-drilled hardwood. The 'one-way' screws, often used on window locks today, were actually patented in Britain in 1796; they were designed to frustrate access to coffins by body snatchers!
Japy, then the largest French screw manufacturing company, invented the first cold worked fly press to make screw head blanks (1806). Apparently, this could only be achieved by employing better quality iron than was used in Britain. It appears that the metal feedstock for machine made screws in Britain was similar to the blacksmith made wrought iron of the screw girdlers. These early 19th century screws were tapered, like Roman screws over a millennium earlier. They were made by the rounding and heading of square 'rod', cutting a slot in the head, and turning the blank into a split die. Machine manufacture displaced this employment (mainly in the 'Black Country' and Scotland) by about 1860.
A forerunner to Nettlefolds, E. Wooley, patented a method of forging headed screw blanks in 1818; production from this source began some twelve years after Japy. Jenkinson reports that Japy made 3.7 billion screws in 1806. Many of these apparently ended up as boot sole protectors for Napoleonic armies. Nettlefolds set up their first screw manufacturing business in about 1819 at a factory in Sunbury-on-Thames. The process was bought from a patent agent, Colbert, handling an application GB4117 (1817) for an overseas inventor. Jenkinson reasons that this was one J. Andrews who held a U.S. Patent for a 'Screw Cutting Machine Wooden and Metal', unnumbered, 1817. However, the U.S. Patent Office copies were lost in the devastating fire of 1836. Nettlefolds made considerable improvements, both in machinery and design. One major advance, ca. 1826, was in the development of the deep cut buttress thread. Their more fundamental approach was to achieve the cutting of a tapered core whilst maintaining a strong parallel thread.
This progressive development (1825–1850) is fundamental to wood screw design to this day; it allows ease of driving into softwood, and pre-bored hardwood, together with fast-holding characteristics.
Pointed Wood Screws
The first successful process to manufacture pointed screws was developed by J. T. Sloan in New York. Goodman4 was in error in attributing this invention to Nettlefolds. Sloan was astute in patenting the screw design (thus any manufacturing process, during the life of his patent, for this structure would be subject to a selection patent).
Sloan specified a process to produce this design in 1846 (USP4864). This was licensed (1849) to both the American Screw Company and the Eagle Screw Company. However this was only after their rival, the New England Screw Company, started manufacturing a pointed screw developed by C. Whipple (unpatented).
The beauty of the Sloan screw was that the thread at the point had the same pitch as the body of the screw. Many previous points were simply conical (unthreaded) or filed (very expensive). Sloan demanded £30,000 (about £2.5 million today) for manufacturing rights in Britain. They were covered by a British patent (GB11,791, 1847) taken out by an agent, Newton, on Sloan's behalf. In fact, to acquire this technology (1854) Nettlefolds took a capital injection from Joseph Chamberlain (of London). The company, now Nettlefolds and Chamberlain, acquired extensive intellectual property. They actually made machinery for themselves and exported models, mainly to Europe. However, Japy apparently acquired a direct licence and manufactured in France (1856) what Jenkinson describes as the best screw of its time and with the fastest throughput of any screw cutting machine in the world. An interesting historical point is that Joseph Chamberlain's son (of the same name) was working in his father's business when, at 18, he was despatched to Birmingham to manage the family interest in Nettlefold and Chamberlain. Apparently he proved to be a very successful businessman. His second son, Neville, became prime minister of Great Britain, 1937–1940.
Ancient Bolts and Screws
Stephanie Dalley8 clearly demonstrated that the first known helical screw was used to lift water in the Hanging Gardens of Nineveh (not Babylon). Dalley forwards persuasive evidence, from the cuneiform, that Sennacherib developed the device from specific logs ca. 700 B.C. Thus the so-called Archimedean screw was invented over 400 years before he was born (ca. 287 B.C.)! The widespread use of this system to lift water, by human and draught animal power, is known throughout the Middle East and Europe. The main applications were for agricultural irrigation and pumping water from mines. In case of deep lifts the latter is achieved by a number of helical screw pumps in series. Even today the technology is used for fluids in the form of screw conveyors; these are often used for moving grain and other easily fluidised solids. Thus the technology has been employed for circa 2,700 years.
As with so many phenomena the central principle can be reversed. The classical example of this is the ancient water wheel (and modern turbines). Reversed, energy is used to generate motion as in the paddle steamer (and propeller). In the case of the helical screw, its use to act to move a load, press, or tighten, required the development of the nut. This was achieved during the time of the ancient Greek empire. It is widely held that Archimedes developed the 'endless screw', the bolt, as a method of applying mechanical advantage. It is known that he originated a series of inventions relying on leverage and the inclined plane in military hardware. There is positive documentary evidence that Hero of Alexandria (1st century AD) devised the direct screw press; this is used today in a wide range of forms e.g. the letterpress. His outstanding development of ways to make the nut, or socket thread, paved the way to widespread use of the nut and bolt in the infinite forms that have been developed. Hero also used small bronze helical screws (machine screws virtually identical with those of today) in his instrument, the Dioptra, an ancient forerunner to the theodolite. These machine screws could have been filed; a suitable lathe was not developed for another 1,500 years. They could also have been made by use of a harder nut like iron. A truly inventive advance on his screw press was his solution to the problem of the screw thread binding when used in a beam press. As the beam was tightened it, of course, moved in an arc. Hero simply tapered the thread sufficiently to accommodate the beam 'nut' at its greater angle away from perpendicular. Similarly, for presses with threaded timber posts, he had them slightly tapered towards the top to reduce binding. A number of Hero's type presses are known. In particular, finds in Pompeii from 79 A.D. support the written evidence.
Whereas Rybczynski2 claims that the Romans "never made the connection between bolts and screws," Jenkinson unearthed a number of examples of tapered screws that Roman craftsmen made. Although scattered, they could be considered as an extension of Hero's earlier tapering of helical threads for presses. Indeed, Jenkinson shows that the Romans advanced metal working and screw making techniques at a greater rate than the Greeks. Of course, finding external wooden artefacts would demand highly unusual conditions. Some smaller metal items would have corroded to the extent of being unrecognisable in wet climates. Impressive survivals include a stand for a roller found in a Saalburg site near a water 'draw' well. It looks like a crutch threaded to fix into the ground with a "U" top to house an axle. It is dated ca.250 A.D. An amazing variant is a sectional assault ladder (4th century A.D.) comprising conical threaded screwed posts with through-rungs. These are shown assembled in pairs forming a ladder where the stiles are screwed into the ground. An earth borer and various small metal screws are other examples that have been recovered from Roman campsites. A most impressive screw – the Dalheim – iron screw (240 A.D.) is threaded the length of the axis of the screw. It has variously been de- scribed as a metal screw and an item of jewellery. Nonetheless, it was designed to thread, and hold, into some material. During the same period there were also a number of machine screws of impressive craftsmanship.
Turnscrews and Screwdrivers
A major advance from the tournevis is, of course, in the length of the blade. The stubby turnscrew ancestor would offer little control with a tight screw. A small deviation from the axis of the screw will cause a large angular deflection. Much improved contact between the slot and the blade is achieved by removing the handle as far as is practical from the tip; i.e. providing the tool a long handle is the solution to stripping the slot.
The illustration below shows some specific developments after the Hausbuch tool during half a millennium.
12” Turnscrew, A. Ridge; 18” Turnscrew, Clay.
These three items 20th century, author’s collection.
The evolution of the turnscrew, from ca. 1480 to date, completed a full revolution by a return to the socket head drive. The problems of the slotted-head screw had bedevilled all but the super-skilled tradesman for four hundred years. After much activity, mainly in North America, in the late 19th century P.L. Robertson, of Montreal, (1907) patented a square socket driving head. His design differed little from the Hausbuch screw head. The major departure was that the socket of Robertson's head had a slight taper. This had two advantages. Firstly, it allowed the tight positioning of the screw on the tool to leave a hand free to hold the item being fixed. Secondly, it aided in the manufacture by cold punching. There have been myriad designs of socket to aid driving and unscrewing in the twentieth century. Phillips, Posidriv, Torx, are a few examples. Some ingenious shapes are simply an attempt to deter unqualified hands tampering with (mainly electrical) products, but automated manufacturing plants have benefited infinitely from the development of socket headed screws.
Perhaps the most used head for bolts is the hexagonal driving socket. This was developed by Unbrako in Scotland and by Allen in the USA. The socket drive lends itself to the use of a rigid right angled key which can be used to obtain much more leverage than a screwdriver in the same axis as the screw. This key has the advantages of a spanner. In fact, the original Hausbuch screw could be regarded as a socketed coach screw; it could be adjusted with either a turnscrew or a spanner. Other extant 13th-15th century wood screws – non fixing type – i.e. for hanging pots, clothes, etc. all had heads which allowed turning with a tool held at right angles to the axis of the hardware.1
Historical Comment
With the fall of the Roman Empire in the fifth century AD, artistic and technical development suffered a major decline in Europe. However, it did not cease; 20th century historians are showing more interest in the so-called Dark Ages than their predecessors. Two points are worth making; firstly reading and writing were not widespread, thus books on technology were not written for the best part of a millennium. Secondly, fighting was widespread throughout Europe and, as always, the victor determined the future structure. The outstanding influence was Charlemagne (742-814), "saviour of civilisation". He used scholars from religious orders and craftsmen from the east to build his chapel at Aachen, the first outstanding, post Greek and Roman, building in Europe outside Constantinople. It seems incredible that screws and screwdrivers have not been found (except for suspected 9th century gold and silver ecclesiastical pieces, and a 13th century iron screw). It may just be due to the lack of literacy that we have no evidence beyond rare archaeological finds. Charlemagne himself could not write, an achievement of only a few in religious orders and the occasional monarch.
It is noteworthy that Gutenberg's printing press was built around 1440 – a mere 40 years or so before the Hausbuch. The printed word spread exponentially, and the reproduction of books freed thousands of clerics from their repetitious work. The Hausbuch has a number of illustrations of screws from adjusting devices on canon to unmistakable items of restraint and torture
The identification of the turnscrew proposed by Jenkinson and by Rybczynski arose from matching sizes of the lathe parts. The items in the illustration above include spanners and an interesting socket screw. This has a turning ring juxtaposed, to suggest fitting by hand and tightening with a rod. The smaller screws look as if they are suitable for fixing the manacles. There is currently no information on how long the Hausbuch screws and turnscrews existed before ca. 1480, but a good look at some of the iron studs in European cathedral doors could be very rewarding.
References
1) G.G. Jenkinson, Metal Wood Screws, pub. by author, Adelaide, S.A., 1999.
2) W. Rybczynski, One Good Turn, Simon & Schuster, London, 2000.
3) Hausbach Munchen ca. 1480. English version C.G. zu Waldburg Wolfegg, Venus & Mars: the World of the Medieval Housebook, (ca.1480), translated by A Seebohm, Prestel, Munich, 1998.
4) W.L. Goodman, The History of Woodworking Tools, Bell, London, 1964, p.9.
5) W.L. Goodman, Eureka! (ed) E. de Bono, Thames & Hudson, London, 1974, pp.226-7.
6) G. Agricola, De Re Metallica, 1556; translated H.C. & L.H. Hoover, The Mining Magazine, London, 1912, p. 365. Also published by Dover.
7) A. Félibien, Des principes de l'architecture, de la sculpture, de la peinture, et des autres arts. Babtiste fils, Paris, 1676, p. 453.
8) S. Dalley, Nineveh, Babylon, & the Hanging Gardens, IRAQ, Vol.LV1, British School of Archaeology in Iraq, 1994.
9) A. Smith, (reprint of) A Catalogue of Tools for Watch & Clock Makers, John Wyke of Liverpool, (ca. 1760), Winterthur, University of Virginia, 1978.
10) P. Walker, The Victorian Catalogue of Tools for Trades & Crafts, Studio Editions, London, 1994. (Richard Timmins ca. 1820).
Acknowledgements
I thank Rebecca Keary for tracking down the exact reference to Félibien mentioned by Goodman.5 I am also grateful to Lee Bidstrupp for obtaining the images used for the figures.
This article by Warren Hewertson first appeared in Newsletter 90, Autumn 2005.