Apart from the time, equipment, space, and expertise needed, tub sizing could result in the loss of salable paper: tears, uneven sizing, sheets that stick (or "block") together and cannot be separated, and mildewing or mold growth. Tub sizing was so onerous that well before the discovery of a viable internal sizing agent (rosin–alum), papermakers unsuccessfully tried all kinds of sizing materials that were added to the vat, such as gelatin or oil. Although announced in 1807 by Moritz Friedrich Illig, rosin–alum sizing was not common until the 1830s. To make the water-soluble rosin soap, powdered tree resin was boiled in a caustic soda solution. Typical sizing recipes called for rosin soap, starch (usually potato or farina), and alum (acting as a mordant to precipitate the rosin and starch onto the fibers) in different proportions based on the weight of dry pulp. The starch was mixed in a small amount of warm water before being added to the beater.3 Typical steps in rosin–alum sizing were: 1) in beater, thoroughly wash (without beating) processed pulp; 2) test pH; if above 9, add alum to neutralize processing residues, e.g., lime, hard water; 3) add rosin soap, starch, and alum; 4) beat for several hours. As the paper traveled through the dryer section of the papermaking machine, fourdrinier or cylinder, this internal size was melted onto the fibers. It is a misconception that rosin–alum sizing was confined to use with machine-made papers. It was also used in hand-papermaking mills that employed special drying apparatuses. The common use of rosin–alum in most papers did not preclude external sizing, however. Internally sized paper, both machine- and handmade, could easily be surface sized by sending it through sizing machines installed in between dryer sections or through machines that stood alone. Internally sized paper enhanced external sizing because it was not as absorbent as unsized paper, and a thinner, more even film of size formed on one or both surfaces, depending on how the sizing machine was set up. Gelatin, starch, and later casein were surface sizes applied by machine. When letterpress printing on the hand- or common press began in the West in the fifteenth century, soft- or slack-sized, or even unsized, rag paper was the most appropriate for printing. These relatively absorbent yet strong papers were easily dampened before printing and were less prone to mildewing compared to more heavily gelatin-sized paper. Mildewing could occur in packs of damp paper run through the press once and then stored covered so that the sheets would not dry out until the second side could be printed. Dampening yielded a more pliable paper, which eliminated the necessity of changing the tympan packing to accommodate thick and thin sheets within batches of paper and the irregular heights of type forms and blocks. Uniform impressions resulted, even though not much pressure was generated from the handpress, compared to the later cylinder press. Another advantage to printing on damp paper was that less ink was consumed. When lithography was introduced at the beginning of the nineteenth century, the pressure generated by the first presses was also relatively weak, and dampening was commonly performed on weakly sized papers. Soft- or unsized book papers were not suitable for writing on, however, and it may be that patrons who wanted to write in their books (beyond the well surface-sized endpapers) may have instructed their bespoke binders to prepare the printed sheets for writing. I doubt that gelatin tub sizing was performed in pre-nineteenth- century binderies because of the problems obtaining fresh gelatin solutions and the expertise required. It is possible that the printed sheets were either sent to a papermill for sizing, or if done in the bindery, sheets could have been dipped in or brushed with alum water to harden any gelatin already in the sheet and then hung to dry.4 Conversely, intaglio (plate) papers required "strong" or hard sizing because once the paper was dampened, it had to be strong enough to stretch without breaking under the greater pressure generated by the cylinder press. There was less concern about plate papers mildewing because only one side was usually printed on in one run, and the prints could be dried soon after printing. Because the inked lines in intaglio printing are above the surface of the paper (as opposed to below the surface in relief printing), embedding the ink in a distinct film of gelatin helped make it more resistant to abrasion. During the nineteenth century when literacy (letter-writing), leisure time (artistic endeavors), and paper-based business practices (bookkeeping) burgeoned, papers were hard-surface-sized in order to withstand the application of water-based inks. This sizing also allowed mistakes to be erased and those areas to be written over without the ink feathering. The so-named plate, drawing, writing, ledger, currency, and bond papers often had the same furnish with the degree of surface sizing plus plating or calendering being the significant differences between them. The popularity of chromolithography by mid-century required a new type of paper that would ensure the perfect registration of colors because sheets were printed off many stones with ink-drying periods in between. Paper made primarily from cotton fiber from rags, which yielded a softer paper compared to linen rag, with minimal or no surface sizing meant that it could be printed dry without expanding or contracting between impressions. Additionally the risk of the paper mildewing over the course of many press runs was eliminated. Many people think, incorrectly, that once paper was made by machine, its quality decreased to the point of utter destruction or unusableness. This is simply not true as evidenced by the millions, if not billions, of nineteenth-century books in the stacks of research libraries in the US alone. Reality is that in terms of its permanence and use-suitability and durability, the manner of sheet formation is less important than the care taken by the papermaker to prepare the fiber before formation and to finish the sheet. Unfortunately excessive amounts of alum result in acidic papers, no matter why it is introduced. Much of the gray-brown discoloration and overall weakness, but not necessarily embrittlement, of rag and later chemical-wood-pulp paper can be attributed to excess alum. Generally the extreme brittleness in very cheap grades of book paper and news paper dating from the late 1880s well into the twentieth century is the result of groundwood-pulp papers sized with rosin–alum. This condition is due to the large proportion of lignin in groundwood pulp, which when combined with excess alum, contributes to the severe degradation of cellulose, resulting in the so-called "brittle book" problem. Over the past decades, the industrywide practice of alkaline-sizing copying and printing papers has eliminated this problem and the now-routine addition of calcium carbonate as a buffer against acidity ensures permanent and durable papers. ___________ notes 1. Paper is sized in order to fill the interstices throughout its structure and to seal the surfaces. Once sized, paper is more able to resist the absorption of liquids, pollutant gases, and airborne particulates, as well as soiling from handling. Sizing also increases the strength of the sheet by "gluing" the fibers together. Dictionaries suggest that "size" is derived from the Middle English "assize," and it may be that the paper required for recording assizes or ordinances—to "fix" legalities in writing—was described as assized and then sized paper. 2. Research indicates that while a small amount of alum and gelatin can hard size paper, adding increasing proportions of alum actually decreases the effectiveness of the gelatin. 3. See for example A. Proteaux, "Further Remarks on Sizing," chap. V in Practical Guide for the Manufacture of Paper and Boards (Philadelphia: Henry Carey Baird, 1866). In 2008 a research project was conducted to determine if nineteenthcentury American book papers were sized with gelatin. The results indicate a trend to replace gelatin sizing with starch (in association with rosin–alum) as the century progressed. Gelatin sizing was most prominent in the early part of the century from 1790 to 1839, but the last gelatin-sized samples were from two 1866 books. This anomaly may be due to the Civil War and its aftermath when starch would have been reserved for food, while otherwise unusable animal products were available for glue and gelatin. The earliest positive test for starch was in an 1837 volume, and most of the books published after 1870 tested positive for starch. For a full description of this research, see Cathleen A. Baker, Morgan S. Adams, and Shannon Zachary, "Sizing in Nineteenth-Century Book Papers," Book and Paper Group Annual 28 (2009): 1–6. 4. Treating a too-absorbent paper by applying alum water prior to drawing or painting was often recommended in nineteenth-century artist manuals.