Industrial Book Production
Industrial Book Production
Abstract and Keywords
Chambers's Edinburgh Journal described the changes in the book trade in the first half of the nineteenth century as a “great revolution in the business of the printer.” Machine printing and stereotyping helped to save the labor costs of press work and composition. Such journals defined stereotyping as a “means of keeping up fictitious types to answer future demands, at an expense infinitely inferior to that of keeping the actual pages standing.” William Savage did not deny that their ability to print faster and on larger sheets of paper offered the machines great merit in terms of speed, which was essential to newspapers and certain periodicals. The factors that can alter the structure and organization of the book trade were publishers' decisions to use machine-made paper, stereotyping, and machine printing.
Chambers's Edinburgh Journal described the changes in the book trade in the first half of the nineteenth century as a “great revolution in the business of the printer.” Its own proprietors' early experiences with hand printing were represented as a period of continuous trial and vexation, ended only by the arrival of the steam-printing machine. It wholeheartedly celebrated the “stupendous spectacle of moral power” that was revealed in the application of the “inert mechanism” of the steam-printing machine to “knowledge and human improvement.”1 The Journal's editors were so excited that they devoted a long article in 1835 to describing the publication's own production processes, from papermaking and typesetting through stereotyping to machine printing. Its competitor, the Penny Magazine produced by Charles Knight and the SDUK, had done the same thing in 1833.2 Machinery was a key feature of these descriptions, and stereotyping was routinely included as one of the recent technical advances, despite not actually involving a machine.
Despite the flood of innovations in the early nineteenth century, certain aspects of the print production process remained resolutely unmechanized and unmodernized. The monotype and linotype machines, which were the eventual solution to mechanizing the composition process in the 1890s, would fundamentally reconceive the task by combining it with typecasting. Until then, compositors retained their position as well-paid, highly skilled craftsmen.3 Similarly, many of the decorative parts of the binding and illustration processes remained in the hands of skilled craftsmen throughout the century. Folding the printed sheets of paper into pamphlets and sewing them together into (p.32) book blocks was also difficult to mechanize, but publishers felt less pressure to do so since these tasks were usually carried out by unskilled female workers for low wages.
Even before the industrialization of printing, the few publishers and charities who specialized in cheap print had developed a variety of tricks and techniques for reducing the costs of as many aspects of production as possible. For instance, they often reprinted public domain works as cheap sources of texts. They did not pay high fees to authors of new works. Illustrations were kept simple and few. Bindings were plain. Paper was the most significant component of physical book production costs in Britain between 1836 and 1856, accounting for 40 to 45 percent of the total cost. To save on paper, the smallest typefaces possible were used, and page margins were reduced. The other major costs were the highly skilled task of composition, which accounted for 30 to 35 percent, and the press work or machining that comprised the remaining 20 to 30 percent.4 Machine printing and stereotyping helped to save the labor costs of press work and composition, which is why they were adopted so quickly by those involved in the cause of cheap print and popular education.
In Britain, the cost of paper was artificially raised by government taxation (see chapter 1). But even without taxation, paper was an expensive commodity. For most of the nineteenth century, paper was made from pulped linen or cotton rags. Once the rags had been soaked and macerated to a suitable consistency, a wooden frame criss-crossed with wires was lowered into it, to catch a thin layer of pulp. The paper maker transferred this carefully to a sheet of felt, squeezed the water out, and then hung it up to dry for a month or so. In earlier times, the scarcity of rags had been a problem, but the industrialization of cotton manufacturing encouraged people to buy new clothes rather than mending old ones, and it also generated cotton waste that could be used to supplement the rags. By the 1830s, supply was no longer such a problem. Machine production helped reduce prices, but it was only in the last two decades of the century that prices fell dramatically, after the introduction of esparto grass and wood pulp as alternative raw materials and the development of next-generation machinery.5
The first commercially successful papermaking machine was built near London in 1803 by Henry and Sealy Fourdrinier. The brothers were not terribly effective at defending their patent and subsequently went bankrupt, but machines based on their designs became very successful. By 1830, two-thirds of the paper produced in Britain was machine-made.6 Printers often (p.33) formed long-term relationships with particular paper merchants, preferring certain manufacturers for covers and endpapers, and others for book paper. Papermaking factories were usually located outside the cities, near sources of plentiful clean water, and the arrival of the railways made it easier both for paper makers to dispatch their wares to the city and for printers to seek tenders more widely than before. With paper forming such a high fraction of the physical production costs of a printed item, finding the best price was essential for producing cheap books and magazines. In the 1850s, W. & R. Chambers bought paper for their Edinburgh print works from mills in nearby Colinton and Penicuik, but also from Airdrie (near Glasgow) and as far away as Aberdeen.7 Samples and prices were generally sought by post, but in 1854, David Chambers made plans to visit the paper mills of Buckinghamshire personally, in the hope of negotiating better prices.8
Because the Fourdrinier machine dripped paper pulp onto a continuous moving web of wires, it could make very large sheets of paper (or even continuous rolls). Although a handmade sheet of paper is limited to the size of frame that a man can hold in his outstretched arms, machine-made paper by the 1830s could be twice that big, and by the end of the century, it would be four times that size. Being able to make more paper on each cycle of the manufacturing process helped reduce costs in itself. The size of the platen in a hand press limited the size of paper it could print, so hand-press printers who bought machine-made paper would have had the sheets cut down to one of the traditional sizes, such as post, crown, or demy. Printing machines could apply pressure across a larger area, which meant they could take larger sheets of paper. This did not have to mean larger book pages; rather, it meant more pages printed per cycle. In 1820, William Chambers had hand-printed eight pages on each side of a sheet for his Songs of Robert Burns, a common layout known as octavo. (Four pages per side would be quarto; twelve pages per side, duodecimo.) With a sheet of paper twice as large, and a machine capable of printing it, he could have printed two octavo sets of pages on one sheet. For William Clowes & Sons, one of the big steam-printing establishments in London from the 1830s and the printers of the Penny Magazine, this technique “effected a complete revolution in our business” because it was now possible to print a book “in half the time we used to do.”9
Composition and Stereotyping
Early nineteenth-century attempts to mechanize composition typically involved mechanical arms to pick out the letters of type, controlled by some form of keyboard, usually resembling a piano; the typewriter was not yet invented. None was satisfactory, and the major development in this area still (p.34) relied on the compositor's skilled work as a starting point. The stereotyping process involved using a mold to cast an exact replica of the surface of the block of composed type. The resulting stereotype plate (or plates, since a separate plate was usually cast for each page on the sheet) could be placed in the bed of a printing press or machine. The idea of stereotyping had been around since the late eighteenth century, but it was first made practicable in the 1810s. The process did not become commonplace in Britain until the 1850s, even though American printers were using it from the 1820s.10
The great advantage was that these stereotype plates could be stored and reused almost indefinitely (until they wore out), long after the original type had been dispersed. Thus, Chambers's Journal defined stereotyping as a “means of keeping up fictitious types to answer future demands, at an expense infinitely inferior to that of keeping the actual pages standing.”11 For most printers, type was far too valuable an asset to be locked up doing nothing, so it was usually dispersed after the initial print run. Supplying back issues or printing subsequent editions thus required the type to be reset, which is why there are often textual variants between first and later editions of texts prior to the mid-nineteenth century. A publisher therefore had to balance the cost of keeping type locked up, or the cost of storing a long print run that might not sell, against the cost of resetting the type. Chambers's Journal calculated that £100 of type were locked up in each of its issues, whereas the cost of making a set of plates was only 45s.12
The ability to reprint at no additional expense was extremely useful to publishers issuing steady-selling books, such as Bibles, tracts, and schoolbooks. The British and Foreign Bible Society and the American Bible Society were early adopters, using stereotyping from the 1810s.13 Subsequent users found other ways of using plates. Some publishers made multiple sets and ran several presses simultaneously, as William Clowes did with the Penny Magazine in London in the 1830s.14 Plates could also allow periodical publishers to respond to an unexpected demand for back issues, or additional sets of plates could be sent to an agent in another city to print a local edition. Plates even became property that could be sold across the Atlantic in the absence of copyright. Plates were most useful for texts that did not need to be revised, but revisions could be made by cutting out a piece and soldering in a replacement, so plates did not “freeze” a text forever.
Stereotyping in the first half of the nineteenth century was difficult to do well, and most publishers believed the effort was worthwhile only for frequently reprinted publications such as Bibles. Making a stereotype plate involved covering the block of composed type with a thin layer of grease and then with a half-inch layer of plaster of Paris. The plaster had to be removed (p.35) without breaking and baked until completely dry. It was then placed in a perforated iron box, which was lowered into a vat of molten lead alloy. If all went well, on opening the box the foundryman would find a plate about a sixth of an inch thick, “exhibiting the perfect appearance of the faces of the types on which the stucco was plastered.”15 However, the high temperature of the molten metal often shattered the mold, and achieving the required thin and uniform layer of lead alloy was tricky. Even after three years of regular weekly practice, the man who stereotyped Chambers's Journal sometimes took six attempts to produce a usable plate.16
The wider adoption of stereotyping in Britain in the late 1840s was partly due to an increasing awareness of its usefulness, but also to the development of a different way of making the mold, using flexible papier-mâché. This was easier to make and, unlike the plaster mold, could be reused. In the United States at the same time, publishers were starting to turn to electrotyping, where electricity was used to deposit a fine layer of copper on a wax mold. Wax was easier to use than plaster, and the copper surface (which was backed with type metal once it was removed from the wax) retained finer details. These new processes made plates easier to produce, while larger print runs and multiple editions for the mass market made them worth having.
In a hand press, sheets of paper were laid on top of a flat printing surface, and pressure was exerted from above by a screw or lever pressing down on a flat surface called a platen. Hand printing was a stop-and-start process, as the type had to be inked and a new sheet of paper inserted before each pull of the lever. Two experienced men might manage to print 250 single-sided sheets in an hour's work, but rates were often slower. The challenge for anyone trying to invent a printing machine was how the continuous power provided by a water wheel or steam engine could be applied to this discontinuous process.
In the early 1810s, a Prussian inventor called Friedrich König came to London, where he hoped to find supportive investors for his proposed powered printing machines. Among his designs was one that was completely different from the hand press: it involved a cylindrical platen that rolled the paper over the flat printing surface. His sponsors introduced him to John Walters, the owner of the Times. Walters commissioned two of König's machines, and the Times was first printed by steam on November 28, 1814 (which conveniently helped Walters break a strike by his pressmen).17
The newspaper industry was at the forefront of developments in printing (p.36) machinery, simply because other printers found the hand press quite adequate. Most books were printed in runs of no more than 1,000 copies, and if they happened to be a day or two late in the shops, it did not really matter. The exception to this rule was, of course, the daily or thrice-weekly newspapers, where timing was critical and the circulation was often up against the maximum possible by hand printing. Some of the most successful British weekly papers and magazines did begin to use steam in the 1820s, and so did a very few book printers, including the Edinburgh firm of Ballantyne, which was struggling to meet the apparently endless demand for Scott's latest novels. But with high installation costs, very few printers could justify investing in steam printing, and there were still only six printing machines in London in 1820 (in addition to the pair at the Times).18 Even among newspapers, only a few really needed the power. The Times was the best-selling newspaper of the early nineteenth century, with a circulation of 11,000 copies in 1830. Its nearest competitor was the Morning Chronicle, with about 6,200 copies.19
König's first machine produced around a thousand single-sided sheets per hour, but by the late 1810s he and his successors Augustus Applegath and Edward Cowper had produced a “perfecting” machine that could print the same number of double-sided sheets. Later designs introduced more cylinders to feed paper more quickly, and mounted the type on a revolving drum (curved stereotype plates, which could be mounted on a true cylinder, were not developed until the 1860s). In 1851, the Times was being printed on an eight-feeder cylinder perfector, designed by Applegath, which produced 10,000 sheets per hour. Applegath's machines mounted the revolving type on a vertical drum, but Robert Hoe of New York produced a competing design with a horizontal drum: his ten-feeder machines could print about 20,000 sheets an hour. Hoe would later develop web-fed rotary machines (in which the curved printing surface prints onto a continuous web of machine-made paper), and by the 1870s, his machines would dominate the newspaper trade on both sides of the Atlantic.20
The newspaper printing machines were highly specialized monsters that provided striking statistics for all those who wished to appreciate the great achievements that had been made in the provision of cheap literature, but if they had been the only option, powered printing might well have remained of limited relevance to the rest of the book trade. Those machines were too big and too expensive for most book publishers. In the United States, machine printing was made more accessible by the 1821 invention of a flat platen printing machine by Daniel Treadwell in Boston. Treadwell's machine was smaller and cheaper than the König-based cylinder machines, yet at (p.37) 500 sheets an hour, its output was still a significant improvement over hand printing. It was thus potentially useful to printers who worked with substantial print runs but did not face the tight time constraints of newspapers. The availability of these platen machines (and those developed by Bostonian Isaac Adams in the 1830s) enabled a variety of American printers to adopt machine printing in the 1820s and 1830s.
In Britain, however, Treadwell and Adams machines were not generally available until the later 1840s. Thus, any British printers considering machine printing in the 1820s had to look at a variant of the newspaper machines. One of the biggest printing establishments in London was that owned by William Clowes, who had bought Applegath's printing workshop in 1826: by 1843, he owned twenty-four perfector machines as well as twenty-four hand presses, and could produce 1,500 printed reams per week.21 When Applegath & Cowper's patent expired in the early 1830s, generic versions of their machine became available at cheaper prices and could be more widely used.22 Among the producers of these generic machines was Thomas Middleton, who had worked with Applegath & Cowper and would supply W. & R. Chambers with machines in the late 1830s and 1840s (see chapter 7). The Applegath & Cowper cylinder machines (and generic imitations) were faster and more expensive than Treadwell platen machines. Thus, the main reason for acquiring such a machine remained the printing of a successful newspaper or weekly periodical.
However, machine printing was also attractive to philanthropically motivated publishers, most of whom subcontracted their printing and thus were more concerned by running costs than purchase price. Machines could finish a run more quickly, so that fewer hours of labor had to be paid for. They could also be operated by less skilled, cheaper labor (two boys to feed in the paper, and a man to oversee several machines).23 These lower labor costs meant that, per unit printed, machine printing was cheaper than hand printing. However, since the preparation of the machine was more complicated (and thus expensive), printers needed to be planning long runs to benefit from lower unit costs. For the Religious Tract Society, which produced tracts in runs of five or ten thousand, or for penny magazine publishers with circulations of twenty or thirty thousand, this was not a problem. Charles Knight estimated that the Penny Magazine's costs would have increased by 40 percent if it had been hand-printed.24
The widespread perception was, however, that the quality of print produced by machines was not as good as that done by hand. It was difficult to get both sides of a page lined up properly on the early machines, though “perfecting” machines solved that particular problem. William Savage, author (p.38) of a Dictionary of Printing who had followed the fortunes of printing machines since the early test runs at the Times, was convinced that machines were incapable of “producing the finest work, and printing the finest impressions from highly finished engravings on wood.” The finest workmanship, he wrote, required the best ink, evenly spread and firmly pressed into the paper. The machines required a weaker, thinner ink, and their speed meant that the inking was “done in an imperfect manner with an inferior ink.” Printing a page incorporating images and text resulted in “crude” and “muddy” images, as the machines had trouble applying equal pressure to the engraved wooden blocks surrounded by metal type. Nonetheless, Savage did not deny that their ability to print faster and on larger sheets of paper gave the machines great merit in terms of speed, which was essential to newspapers and certain periodicals.25
Publishers' decisions to use machine-made paper, stereotyping, and machine printing changed the structure and organization of the book trade. As in other industries, skilled craftsmen were replaced with semiskilled machine attendants, though compositors retained their position, and stereotyping added a new craft process to the trade. For the consumer, the result was greater choice and increased availability of cheap books and magazines.
It is all too easy to assume that cheap prices were the inevitable consequence of great inventions, so it is worth remembering that the prices of reprinted out-of-copyright books had been falling since the late eighteenth century;26 and that religious charities and political radicals had been producing cheap print (legally and illegally) since the 1810s. Cheap print did not require steam printing or stereotyping—but those new techniques made things much easier for publishers of cheap print. Still, it is significant that, in Britain, such publishers were in a minority in the 1830s and 1840s. Papermaking machines spread quickly, but printing machines were a more complicated story. Publishers had to be able to afford them, but they also had to need them. This is why the different ways that British and American publishers perceived their audiences in the 1830s had such an impact on the uptake of machine printing. American publishers, almost all of whom aimed for a cheap product for a wide audience, could see a clear use for printing machines, and the availability of the cheaper Treadwell machine made it an easier decision. In Britain, the only option was an expensive cylinder machine, and since most publishers targeted a limited circle of affluent customers, the decision against was equally straightforward.
In the 1830s, only a few British publishers used steam printing and stereotyping. The tract and Bible societies, Charles Knight, and W. & R. Chambers (p.39) were among the small number outside the newspaper industry using printing machines. And Chambers stood out from their peers for two reasons: rather than contract out their printing like most publishers of the era, they decided to do their own, thus becoming intimately involved with the machinery. Second, they decided to use their new machinery for more than just periodicals. (p.40)
(1.) “Mechanism of Chambers's Journal,” CEJ, June 6, 1835, 151.
(2.) “Commercial History of the Penny Magazine,” Penny Magazine 2 (1833): 377–84,
(3.) Twyman, Printing, 61–63. On wages, see Weedon, Victorian Publishing, 76–79.
(4.) Weedon, Victorian Publishing, 87.
(5.) Weedon, Victorian Publishing, 66–67
(6.) McKitterick, 1830–1914, 92–96
(7.) See entries in WRC 161 (LB 1851), passim.
(8.) David Chambers to Robert Chambers, May 6, 1854, in WRC 126 (Misc. Letters).
(9.) Weedon, Victorian Publishing, 65
(10.) Winship, “Printing with Plates.”
(11.) “Printing and Stereotyping,” CEJ, September 29, 1832, 278.
(14.) Knight, Old Printer, 255–56
(15.) “Printing and Stereotyping,” 278.
(16.) “Mechanism of Chambers's Journal,” CEJ, June 6, 1835, 150.
(17.) Many histories of printing and printing machines are available. The account in Gaskell, Bibliography, is straightforward; that in Twyman, Printing, is well illustrated; those who are particularly interested in the presses and machines may wish to consult Moran, Printing Presses. For a comparison of the relative claims of Nicholson and König as inventors, see Savage, Dictionary of Printing, 449–62.
(18.) Twyman, Printing, 52
(19.) Twyman, Printing, 51; McKitterick, 1830–1914, 76–77.
(20.) Moran, Printing Presses
(21.) Gaskell, Bibliography, 289
(22.) Gaskell, Bibliography, 251–63; Twyman, Printing, 51–52.
(23.) Gaskell, Bibliography, 290
(24.) Knight, Old Printer, 255–56
(25.) Savage, Dictionary of Printing, 465–67.
(26.) St. Clair, Reading Nation, 19–20