Gelatin completely permeated the paper, and, when properly dried, made the paper very resistant to water and nearly doubled the strength by most measures. In particular it would stand up to erasing with a knife. According to Clapperton2 and others, gelatin sizing was invented in Fabriano, Italy in about 1300 CE, attributable to the widespread availability of skins and bones from the large local tanning industry. Tanning and papermaking were often located near each other, both dependent on fresh water and a stream to take away the filthy effluent.3 Partly because of its location on the navigable River Medway (an estuarine tributary of the River Thames), Maidstone was a center for trade for many industries including leather.4 Tanning may not have been one of the largest industries locally but it was carried out in the area and, interestingly, the family of James Whatman I set up tanning in the village of Loose just over a mile upstream from Hayle Mill. Tanneries produced a variety of by-products including unusable parts of the hide which contain collagen. The Greens would have found it easy to procure materials locally for making gelatin at the Mill. The process started by soaking these in water which was done, in later years at least, in a drum rotating in the millrace (the channel to the mill wheel). The vast numbers of fresh-water shrimp and small fish in the stream would have helped to remove the meatier and fattier parts of the hide and to wash away the blood and dirt. To refer to the water in the Loose Stream at that time as "fresh" is misleading. The combined effluents of the tanneries, fulling mills, and paper mills would have made for a foul liquid. By the early twentieth century, for economics and reliability, Hayle Mill began to buy in gelatin as a granular material from large factories elsewhere. The granules were dissolved in warm water accompanied by extensive stirring. We had two stainless-steel size "melters." The first was used to dissolve the size in water and mix with a solution of alum. We used potash alum (potassium aluminum sulfate) but there are many other varieties of alum available, mainly papermakers' alum (aluminum sulfate) which is often contaminated with iron and other impurities. The second melter stored the warm solution (technically, a colloid) used to top up the size machine where the bath was maintained at about 105 degrees Fahrenheit (approximately 40 degrees Celsius). While the warm temperature kept the solution fluid, it also sped up the degradation of the size, partly due to microorganisms in the solution. To prevent the growth of bacteria, we added Formalin to the bath, which also had the added benefit of hardening the gelatin and making it virtually insoluble once dried.5 For centuries paper was sized with gelatin by hand dipping into the warm solution followed by pressing and loft drying. This continued to some extent into the twentieth century in Europe, but most mills, including Hayle Mill, adopted sizing machines in the late nineteenth century. In this process, the sheets were carried through a long bath of warm size solution between two continuous felts and pressed between metal rollers before being removed. A common problem with gelatin sizing was staining; indeed I have a thick file on this problem. When Robert Henderson Clapperton worked at Hayle Mill in the 1930s, he observed that this problem could usually be avoided by cooling the paper soon after sizing. A surplus sizing machine was adapted so that the sized sheets could be carried slowly on a continuous felt exposed to the air for a few minutes, resulting in a great reduction of staining. Cooling the sized sheets also helped to lower the incidence of under- sizing or uneven sizing.6 After sizing, the damp, sized paper was taken to the loft for drying. Although ledger and printing papers could be hung over a cow-hair rope (because they would in any case be folded before binding), watercolor paper had to be dried flat on a treble (a hessian screen made of jute fabric). The paper was laid up at room temperature with the louvers open on the sides of the loft. Before we had a steam boiler, the paper remained in the unheated loft for days or weeks. The location of most paper mill lofts was across narrow valleys. In our case I could always detect a diurnal breeze in the loft either up or down the valley, however calm the day seemed. The introduction of steam for powering engines meant that it could also be used to heat the loft. However it could be easily overheated, softening the gelatin, and this could lead, we believed, to sizing failure. In the 1960s I fitted each loft with electric thermostats which may have been of some help. At Hayle Mill we laid up the paper on the first day, closed the louvers, and turned on the heat for a day or two. On day 4, we turned off the heat and opened the louvers to allow the breeze to rebalance the moisture content on day 5. On day 6 the paper was taken down and stacked, and on day 7 the workers, and maybe even the master, rested. In 1807 Moritz Friedrich Illig publicized in Germany his invention of rosin–alum sizing, whereby rosin–alum size is added to the beater, avoiding the need for a secondary sizing process after the paper is dried.7 This process only came into common use in the 1840s.8 It seems inevitable that Hayle Mill was using this form of sizing to a limited extent in the nineteenth century but we have no records of its introduction. In the twentieth century we were using rosin–alum sizing in our extensive range of note-paper for domestic letter writing, but again I have not as yet found documentation regarding formulation or procedures. From the early 1950s our popular machine-made watercolor paper, Bockingford,9 was rosin sized. While our records on rosin are thin, most likely we bought in the rosin as a solution, diluted it to a quarter of its original strength, and added an alum solution. Rosin sizing became increasingly associated with discoloration and weakening of the paper over time with disastrous results, especially if it is combined with mechanical wood pulp. The rosin is not entirely to blame as it is generally considered to be its combination with acidic papermakers' alum (aluminum sulfate) that is harmful. Even though gelatin sizing also includes alum, usually it was potash alum (potassium aluminum sulfate) which does not appear to cause harm due to some protective effect of the gelatin. During the 1970s I had vigorous discussions with paper conservators who considered any pH of less than 7.00 to be unacceptable, and as gelatin sizing was ineffective without the addition of alum, gelatinsized papers always tested below pH 7.00. I argued that pH levels do not tell the whole story; gelatin-sized papers are often in perfect condition after hundreds of years. I believe the conservation profession has come to share my view.10 Despite being familiar with the problems caused by rosin– alum sizing, I was shocked to find how quickly Bockingford darkened in sunlight when I first started working at Hayle Mill in 1968. Within a few months I had persuaded my father to switch over to Aquapel sizing. To my knowledge Bockingford was the first acid-free watercolor paper to be put into production. Aquapel was introduced by the Hercules Powder Company in 1955.11 At about this time William J. Barrow's research into the reasons for paper impermanence and ways to avoid it was becoming well known, including his recommendation to discontinue use of rosin–alum sizing.12 However until the late 1970s, the paper industry considered making alkaline or even neutral-pH paper commercially impossible as it was devoted to alum as a cure-all for many problems. Even a small amount of alum in the paper-machine system would wreck Aquapel sizing. In 1965, the great bookbinder and conservator Sydney M. Cockerell asked my father Rémy John Barcham Green to make his handmade marbling paper with alkaline sizing. Our first attempt at using Aquapel was a disaster as the paper repelled the marbling inks! However it was almost certainly the first handmade paper to be sized in this way. By the late 1960s I was making a number of handmade and machine-made papers with Aquapel sizing which had a pH of just under 8 due to the hard nature of our water. By varying the addition rate, the degree of sizing could be adjusted from slight resistance through hard sizing to impermeability. To preserve our trade secret, I referred to these papers as neutral sized for a number of years without mentioning Aquapel, thus frustrating a number of would-be competitors. Aquapel sizing transformed our business because it freed us from the huge cost of gelatin sizing. However, like starch and rosin, it did not add to the strength of the paper as gelatin did. This led me to research stronger fibers and increasingly we used abaca and then flax pulp. The latter produced excellent results and was almost as strong as gelatin sized papers although not quite as resistant to the use of a scalpel. There are many other aspects to sizing to consider. Gelatin-sized papers are less permeable to both air and oil than paper sized with other agents or none at all. The oil resistance can affect various methods of printing as well as oil painting. The air resistance seems to seal out to some extent the effects of air pollution. The diverse effects of sizing have led to a wide variety of test methods. There are no doubt many other natural and synthetic materials that can be used for sizing and the ways in which fiber is cooked, beaten, and formed into sheets all interact with sizing in complex ways. Sizing is quite a subtle process; it is not just a case of being absorbent or repellent. Different uses require different standards. Juice and milk carton boards must be fully repellent, watercolor papers should be hard sized but allow a wash to be evenly applied, and most traditional printing papers are soft sized. The standard test is the Cobb Test which essentially measures the increase in weight of a 100-square-centimeter piece of paper when covered in 1 centimeter of water for 60 seconds, expressed in grams per square centimeter (g/m2).13 Watercolor paper would typically have a Cobb of just under 20 g/m2 which is fairly easy to control. A softsized paper could be just over 30 g/m2 and this can be very difficult to maintain consistently. But the Cobb tester does not indicate the character of the sizing. Most papermakers use a steel pen to mark curved lines on the paper to observe the rate of drying and to check for any feathering. For watercolor papers we would apply a wash with a one-inch bristle brush, then scrub the surface counting the strokes until it started breaking up. When I first started working at Hayle Mill, one of the roles I first learned was undertaking the "wet out" test for gelatin-sized papers with my grandfather Jack Barcham Green who was 83 years old at the time. We would set out a large shallow tank of water in a cold, damp room in the heart of the mill. We selected six sheets of paper at random from each dryworker's batch, and placed the sheets in a pile near the tank. Water from the springwater supply ran slowly into the tank at about 11 degrees Celsius (52 degrees Fahrenheit). We floated the first sheet on the surface of the tank and ran our hands over the surface of the paper to let the water work into the sheet. We placed the next sheet on top and continued to set and stack until all six sheets were in the tank. Then, we turned over the sheets and gradually restacked them onto the water surface with more hand spreading of the water. This went on for several minutes until all the sheets were thoroughly wet. Then we held up each sheet in front of a fluorescent tube fastened to the wall behind and closely scrutinized. Any slight irregularity of the sizing could be observed in this way. We checked for overall absorption (a well-sized sheet was not as heavy or floppy as a soft-sized one), evenness (i.e., no spots or areas that were more or less translucent than others), and loss of general integrity (a wellsized sheet would still be strong when wet). Unless all six sheets were evenly sized, the entire sizing batch would have to be resized. We typically tested paper for twenty to thirty minutes a day during which I learned a lot from my grandfather, including how cold 11 degrees Celsius really is! I have never heard of this rigorous test being used in any other mill but I am sure many had their own special methods. notes 1. Maureen P. Green, "Hayle Mill—How a Small Papermaking Company Thrived in the Nineteenth Century Using Traditional Techniques Which Were Being Superseded by New Technology in the Mainstream Paper Industry" (unpublished PhD thesis, Institute of English Studies, School of Advanced Study, University of London, 2011). A revised version will be published by The Legacy Press, Ann Arbor, in 2016, covering the first hundred years of Hayle Mill including the purposes for which it was built, the types of paper made, and the techniques involved. For more, see http://www.thelegacypress.com/hayle-mill.html (accessed August 9, 2015). 2. Robert H. Clapperton, Paper: An Historical Account of Its Making by Hand from the Earliest Times Down to the Present Day (Oxford: The Shakespeare Head Press, 1934). 3. Simon B. Green, "An Outline History of Sizing Methods with Special Reference to Practices at Hayle Mill," in The Institute of Paper Conservation Conference Papers, Manchester (London: Institute of Paper Conservation, 1992), 197–200. 4. Jane Andrewes, "Industries in Kent, c. 1500–1640," in Early Modern Kent, 1540–1640, edited by Michael Zell (Suffolk: Boydell & Brewer, 2000), 135. Tanning was a domestic industry, often undertaken by two or three men on farms with Maidstone as a distribution center by the river along the Medway and the Thames to London. The vast oak forests of Kent supplied the bark trade of which Maidstone was also a center. 5. By the 1970s we stopped using Formalin when it became recognized as a carcinogen. 6. Despite best practices, gelatin sizing typically has a very high failure rate (often 30 percent). 7. William J. Barrow Research Laboratory, Permanence/Durability of the Book—VII: Physical and Chemical Properties of Book Papers, 1507–1949 Sizing notes from 1944 on Moreleign note paper, indicating that it had size-stained edges and had been passed through the sizing machine with hot water at 147 ˚F (64˚C) to remove the stain. The stain is no longer visible on the sample, so either the rinsing worked, or the paper may have been resized. Even at the end of the war, we still needed to make nice notepaper! (Richmond: W.J. Barrow Research Laboratory, 1974). More information on Barrow research and publications can be found at https://www.loc.gov/preservation/ scientists/projects/barrow_books.html (accessed August 9, 2015). 8. See Jim Thurn's 2003 paper "History, Chemistry, and Long-Term Effects of Alum-Rosin Size in Paper" in The Cochineal, https://www.ischool.utexas .edu/~cochinea/html-paper/j-thurn-03-alum.html (accessed August 9, 2015). 9. Bockingford was developed at Hayle Mill but it is now made at St. Cuthberts Mill in Somerset, England. 10. The leading research in the field was done by Irene Bruckle. See Irene Bruckle, "The Role of Alum in Historical Papermaking," The Abbey Newsletter, vol. 17, no. 4 (September 1993), http://cool.conservation-us.org/byorg/abbey/an/an17/ an17-4/an17-407.html (accessed August 9, 2015). 11. Aquapel® is the registered trademark of Hercules Incorporated. See Herbert H. Espy, "The Genesis of Alkaline Sizing and Alkaline-Curing Wet- Strength Resins," The Abbey Newsletter, vol. 3, no. 3 (August 1990), http://cool.conservation-us.org/byorg/abbey/ap/ap03/ap03-3/ap03-307.html (accessed August 9, 2015). 12. See Wikipedia, http://en.wikipedia.org/wiki/William_Barrow_%28chemist%29 (accessed August 9, 2015). 13. Go to http://www.tappi.org/content/tag/sarg/t441.pdf to see the latest proposed version of the test developed by Cobb R.M & Lowe D.V. in 1934. Paper Sample: Crisbrook Imperial 72-lb. Rough, Gelatin Sized text by simon barcham green This vintage Hayle Mill paper sample, gelatin sized as described in the preceding article, was probably made in the late 1960s. Crisbrook was originally made out of white cotton rags but from at least the early 1950s it was made out of cotton linters pulp. In the 1970s we reduced the range of Crisbrook to just Imperial 140- lb. in Rough, NOT (CP, or "cold pressed") and HP ("hot pressed"). Shortly after, we stopped selling the sized version as the unsized, waterleaf paper was the more popular variety, by then primarily used for printmaking. Crisbrook was originally promoted as being "that essential bit softer" as not everyone liked the much harder RWS (a paper we named after the then-referred-to-as Royal Watercolour Society) made from nearly 100-percent linen rags.