Creating the Twentieth Century

by Vaclav Smil

  Alternatives to frequently jamming type bars appeared even before the first visible typing: James Hammond’s revolving cylinder in 1881, and George Blickensfelder’s revolving wheel (the precursor of daisywheel design) in 1893. The first electric typewriter, with a typewheel printhead, was sold in 1902, but IBM’s Selectric became the best-selling model only during the 1970s, just before the typewriters gave way to PCs. And the QWERTY arrangement? Sholes chose this irrational order deliberately in order to slow down the speed of typing and hence to minimize the jamming of type bars. Rational keyboard designs—for English language typewriters August Dvorak’s famous statistics-based rearrangement that speeded up typing by more than a third (Dvorak et al. 1936)—have been available for decades, but the universal design was not only maintained by virtually all 20th-century typewriters but also has been transferred, without any second thoughts, to computer keyboards.

  Papermaking

  By the beginning of the 19th century, steadily rising demand for paper was putting greater strain on the supply of rags. In different places and at different times, this feedstock included cotton, linen, flax, hemp, and even silk fibers. Small runs of artisanal handmade sheets are still made from rags, while cotton gin trash and flax waste are used to make paper for the U.S. currency (Brennan 2000). Demand for larger volumes of fibrous feedstock further increased after the introduction of a new papermaking technique that produced long strips rather than single sheets of paper. The machine for this semicontinuous production was invented in 1798 by Nicolas Louis Robert, a clerk at the Essonne Paper Mills in France, and the English rights for its use were bought by Henry Fourdrinier (1766-1854), the owner of a mill in Kent.

  By 1803 the Fourdrinier brothers and Bryan Donkin had constructed the first practical machine at a mill in Frognore in Hertfordshire (Wyatt 1911). This machine produced strips of paper up to 3.6 m wide in lengths of 15 m, and they had to be taken off wet and hung to dry. The basic principle remains two centuries later: pulp is laid on a continuous wire mesh at the wet end of the machine, most of water is expelled in the felt press section, and the process is finished by passing paper over a series of heated cylinders. Performance is a different matter: by 1910 the largest fourdriniers were 5 m wide and could produce rolls of paper nearly 13 km long at the speed of 240 m/min. Today’s large fourdriniers can produce rolls of up to 18 km long at speeds surpassing 400 m/min.

  During the first half of the 19th century, paper recycling became more common, and the search for new feedstocks led to the use of jute sacking, cereal straws, Manila hemp, and esparto grass, but wood was obviously the most abundantly available source of fiber. Mechanical pulping was developed for practical use by Heinrich Volter in Bautzen by 1845, but it has been used commercially on a larger scale only since the early 1870s. The process had high yield (up to 90% of wood becomes pulp), but this also means that it removed hardly any lignin. This polymer normally constitutes between 30% and 40% of wood and imparts to it strength, but its presence in the pulp makes for an inferior paper. Furthermore, wood resins in mechanical pulp contribute to paper’s rapid deterioration, betrayed by distinct yellowing with age. These drawbacks make little difference for ephemeral paper uses, and that is why mechanical wood pulp still accounts for about 20% of all pulp produced worldwide (FAO 2003). Its most important use is for newsprint, and it also goes into toilet paper, towels, cardboard, and building board.

  Experiments to remove lignin, resins, and other intercellular matter and to produce nearly pure cellulose began during the 1850s when the soda (or alkaline) process, patented by Charles Watt and Hugh Burgess in 1854, was introduced in the United States for the treatment of deciduous woods, mainly for poplars, aspen, beech, birch, maple, and basswood. The first American mill using the process was in Pennsylvania in 1863. Chipped wood was boiled with sodium hydroxide (NaOH) in large boilers for up to nine hours at pressures of up to 1 MPa, and the bleached pulp (total yield about 50%) was used for book and magazine papers. The soda process was gradually displaced by the sulfite (or acid) process that eventually produced most of the world’s paper during the first third of the 20th century.

  Benjamin Chew Tilghman (1821-1901), better known for his patenting of sandblasting, began his work on this treatment during the 1850s. His U.S. Patent 70,485 for treating wood with acid, boiling it under pressure, and adding a suitable base substance for easier bleaching (Tilghman 1867) was followed by many improvements contributed mostly by engineers in Europe (Steenberg 1995; Edge 1949). An indirect boiling process, with steam at 125-135°C circulating through copper or lead coils to boil the mixture for at least 20-30 hours, was patented in Sweden in 1871 by Carl Daniel Ekman (1845-1904) and in Germany in 1874 by Alexander Mitscherlich (1836-1918). With its longer boiling at lower temperature, this process yields stronger pulps than does the direct boiling method that uses hot steam (140-160°C) for 8-10 hours. In Europe this new method became generally known as Ritter-Kellner process. Carl Kellner (1851-1905) developed it in 1873 when he worked for Baron Hector von Ritter-Zahony’s paper factory in Gorz, and had it patented in 1882 (Dvorak 1998). America’s sulfite pulp production began only in 1883 at the Richmond Paper Co.’s plant in Rumford, Rhode Island (Haynes 1954). Figure 5.6 shows an early design of American pressurized tanks used for pulp production.

  Feedstock for both methods is ground coniferous wood that is boiled under pressure of 480-550 kPa in a solution of bisulfites of calcium or magnesium until all nonfibrous constituents of wood (resins, gums) are dissolved and cellulose fibers can be recovered, washed, and bleached to remove all traces of the liquor (Edge 1949). Just before WWI, sulfite pulp accounted for about half of all wood pulp produced in the United States (Keenan 1913). The sulfite process produces fairly strong pulp, but its acidity eventually embrittles the paper, whose disintegration is further accelerated by its susceptibility to air pollution and high humidity. Most matter that was printed between 1875 and 1914 has already entered the stage of fairly advanced disintegration: even when carefully handled, brittle pages of many 100-125-year-old publications that I read in preparation for writing this book tear easily, and some literally crumble. And the same fate awaits most of the pre-1990 publications as the first official paper standard for permanent papers was adopted in the United States only in 1984, and the international norm, ISO 9706, came a decade later.

  The second pulpmaking innovation, and the one that has dominated global papermaking since the late 1930s, came shortly afterward with the introduction of the sulfate process. In 1879 Swedish chemist Carl F. Dahl invented a pulping process that used sulfates rather than sulfites to boil coniferous wood in large upright boilers at pressures between 690 and 890 kPa for about four hours (Biermann 1996). The sulfate process produces much stronger paper, hence the use of the Swedish (or identical German) word for strength to label the process kraft pulping. The residual lignin colors paper brown, and so bleaching is required to produce white paper, and additives (acids, bases, sizing agents, adhesives, resins, fillers) are used to control pH, strength, smoothness, and brightness.

  FIGURE 5.6. Apparatus for treating wood with bisulfite of lime. Reproduced from Scientific American, March 29, 1890.

  While the kraft process yields superior pulp at a lower price, it also produces highly offensive hydrogen sulfide that used to escape uncontrolled from the mills, and sulfate pulps are also more difficult to bleach. The first American kraft mill was built only in 1911 in Pensacola, Florida, but eventually the advantages of sulfate pulping—the only chemical is inexpensive sodium sulfate, a large amount of energy is produced during the recovery process, and pine trees, abundant worldwide, are well suited for it—made it the leading form of papermaking in the United States and around the world. By the year 2000, the sulfate process accounted for nearly 70% of the world’s wood pulp production; mechanical pulping produced about 20% of the total, and the rest comes from the sulfite and semimechanical treatment (FAO 2003).

  Importance of these relatively simple chemical processes—both
involving nothing else but pressurized boiling of ground wood in acid solutions—that were commercialized before WWI is not generally appreciated. Yet sulfite and sulfate pulping made it possible to convert the world’s largest stores of cellulose sequestered in tree trunks into high-quality paper and ushered in the age of mass paper consumption. In 1850, near the end of the long epoch of rag paper, there were fewer than 1,000 papermaking machines in Europe and about 480 in the United States, and the annual per capita output of paper in Europe and North America amounted to less than 5 kg. By 1900 Germany alone had 1,300 machines and the worldwide total surpassed 5,000. U.S. paper production doubled during the 1890s to 2.5 Mt, and by 1900 it surpassed 30 kg per capita. Paper became one of the cheapest mass-produced commodities of modern industrial civilization.

  Consequences of this change could be seen everywhere. Picture postcards, invented in Austria in 1869 and available in France and the United States by 1873, became very cheap and as popular as today’s e-mail. In 1867 Margaret Knight (1838-1914) designed the first flat-bottom paper bag and soon afterward also a machine to make such bags; better machines, by Luther Childs Crowell in 1872 and by Charles Stillwell in 1883, made their production even cheaper. This seemingly simple innovation endured throughout the 20th century, and its dominance began to recede only during the 1980s with the rise of plastic grocery bags. By the mid-1890s the largest mail order business in the United States, Montgomery Ward & Co., whose first catalog in 1872 was a single page, was distributing seasonal catalogs containing more than 600 pages (Montgomery Ward 1895). Writing slates disappeared even from the poorest schools: the just cited catalog was selling 160 cream-colored pages of its School Spelling Tablet for three cents.

  The combination of cheaper paper and halftone reproduction (see the next section) led to an enormous expansion of the publishing industry, particularly in the United States with its new 10-cent illustrated magazines (Phillips 1996). In addition to the rise of mass journalism, the printing of large editions of books, ranging from bibles and historical novels to how-to-do manuals, became affordable. As books became cheaper, the total number of public libraries also increased rapidly; in the United Kingdom the Local Government Act of 1894 made it possible to have them even in remote rural parishes (Fleck 1958).

  Consumption of paper is still going up even in countries where its per capita use already amounts to more than 200 kg a year (FAO 2003). By the year 2000 the U.S. annual paper consumption surpassed 300 kg/capita, nearly 15% above the 1990 mean, and similar increases were recorded in Japan (now at 250 kg/capita) and in the European Union (now averaging about 220 kg/capita). The link between the use of printing and writing paper and affluence is even stronger than between the consumption of primary energy and gross domestic product: the correlation coefficient for the world’s 35 most populous countries is 0.98! The world’s affluent economies, whose population accounts for less than 15% of global population, now consume nearly 80% of all printing and writing paper. These realities expose the myth of the paperless office that should have become, according to the forecasts of electronic technoenthusiasts, the norm by now.

  As Sellen and Harper (2002) demonstrate, most people prefer reading, evaluating, and summarizing information printed on paper spread around on their desks rather than displayed on multiple overlapping windows on their computer screens. I belong to this majority: although I have used a succession of IBM machines since the mid-1980s, I maintain only a small number of electronic files but I treasure and intensively use my archive of printed materials that now amounts to hundreds of thousand of pages. Of course, tree-saving electronic paper would be a truly revolutionary departure, and several research teams are at work to come with a practical, thin, and erasable display (Chen et al. 2003).

  Accessible Images

  There was no shortage of excellent images in the pre-industrial world, and some of them were reproduced in relatively large number of copies; they ranged from exquisite copper engravings of Renaissance and Baroque masters to fascinating multicolored ukiyo-e prints that depicted everyday life of Tokugawa Japan. But all of these images had two things in common: they were produced in artisanal fashion, one at a time, and at a relatively high cost. Invention of lithography—by Alois Senefelder (1771-1834)—made the reproduction process much cheaper, but prints still had to be pulled off one by one in single sheets. Introduction of steam-driven flat bed and, later, rotary presses speeded up the printing process, and electric motors made it even more flexible—but none of these innovations changed the painstaking process of cutting, engraving, or lithographing the originals.

  And, of course, both engravings and woodcuts were inherently labor-intensive interior techniques entirely unsuited for any rapid capture of images in the open or for recording moving people, animals, or objects. Only the development of photography changed that but, despite many impressive improvements, the scope of this new art was limited for decades by cumbersome and inconvenient procedures as well as by the absence of simple techniques to reproduce these images in newspapers and books and to make inexpensive copies for personal use and distribution.

  In 1826 Joseph Nicephore Niepce (1765-1833) produced the world’s first photograph, a very crude image of nearby buildings seen from an upper window of his home that was captured on light-sensitive bitumen and fixed on a pewter plate. During the late 1830s the first nature morts and photographs of buildings and people appeared on polished silver-plated copper plates made by Louis Jacques Mandeé Daguerre (1789-1851). And William Henry Fox Talbot (1800-1877) and John Herschel (1792-1871) invented independently the process of photography on sensitized paper (Frizot 1998; Schaaf 1992). In 1839 Herschel used for the first time the terms “negative” and “positive,” and in 1847 Claude Niepce, Joseph’s cousin, invented the glass plate. Four years later Frederick Scott Archer introduced a cumbersome wet collodion process that became the standard procedure and has left behind a remarkable record of portraits, documents of daily life, and landscape images.

  Wet plates were dominant for about three decades. The first dry, silver-bromide—sensitized, gelatin-emulsion plates were produced in 1864 by W. B. Bolton, and B. J. Seyce and many inventors, including Joseph Wilson Swan of the incandescent light fame, improved them during the 1870s (Abbott 1934). An easily produced and highly sensitive kind, prepared by Charles Bennett, reached the market only in 1878. Still, taking pictures required patience, skill, and considerable expense due to relatively long exposures and heavy, bulky, and fragile plates. Moreover, darkrooms were needed to develop the images. The three kinds of innovations that revolutionized photography were (1) the reduction of exposure time that allowed for capturing an expanding range of natural settings and movements; (2) development of convenient photographic film to replace heavy glass plates; and (3) availability of affordable handheld cameras that could take fairly good snapshots without any special manipulation, and access to inexpensive custom film development that obviated the need for an in-house darkroom and the knowledge of photographic chemistry.

  Cameras, Films, Photographs

  The earliest photographs required very long exposure times. Niepce’s faint view from his window took eight hours to capture; two decades later typical da-guerrotypes required 30 seconds when the plate was exposed in a studio. Talbot took the first high-speed flash image in 1851 by using a spark from the discharge of a Leyden jar, but that was clearly an impractical arrangement for casual use. Progressive improvements that began during the 1850s reduced the typical exposure times, and by 1870 Talbot was able to expose his plates for just 1/100 of a second. The most revealing breakthrough in high-speed photography came when Eadweard Muybridge (1830-1904), an Englishman who started his American photographic career with the images of Yosemite in 1867, developed a technique to capture motion in rapid sequences with a camera capable of shutter speed of as much 1/1,000 of a second (Hendricks 2001).

  This work was prompted by Leland Stanford—at that time a former governor of California, the president of the Central Paci
fic, the future founder of the eponymous university, and an avid horse breeder—who wanted to confirm his belief that a galloping horse will, at one point, have all of its legs off the ground simultaneously. In 1878 Muybridge proved this to be correct: perhaps his most famous photo sequence confirmed that all four hooves of a trotting horse are momentarily off the ground (figure 5.7). After his return from a European tour, where he showed his Zoopraxiscope, a viewbox that displayed a rapid series of stills that conveyed motion (the principle used later in animated cartoons), Muybridge made an agreement with the University of Pennsylvania to prepare a comprehensive series of sequences of animal and human locomotion.

  This work began in 1884, and it used a fixed battery of 24 cameras positioned parallel to the 36-m-long track with a marked background, and two portable batteries of 12 cameras each that were sited at both ends of the track. These cameras captured the motion of animals and people as they broke strings stretched across the track and activated the shutters in turn. The complete set of 781 sequences was published in 1887 (Muybridge 1887), and it includes such predictable images as running at full speed and throwing a baseball as well as such oddities as a woman pouring a bucket of water over another woman. But more widely practicable and truly instantaneous photography was made possible only when highly sensitive dry plates, which reduced exposures to fractions of a second in well-lit environments, became readily available. Even amateur photographers could now capture movement, and by 1888 their task was made easier by new devices that measured the time of required exposure.