O n P a p e r a n d P a p e r m a k i n g

Transcription

1 O n P a p e r a n d P a p e r m a k i n g Before launching into an analysis of the physical and chemical properties of the papers composing the samples, it is beneficial to give a brief history of paper in Europe, a description of paper, a synopsis of hand papermaking in general, and a study of papermaking in a typical 18 th to 19 th century paper mill in Sweden. The short description of papermaking that follows is based on the author s experience as a papermaker and paper conservation scientist, 45 as well as upon close reading of the works of Hunter, Barrett, and others (some of which were in translated or abridged form). 46 A Brief History of Paper Papermaking is an ancient, and guarded craft, 47 which was initiated in China around the year It flourished throughout Asia for a thousand years before it was introduced to Europe via trade with Arabs some time during the late 11 th or early 12 th centuries. 49 Due to various political, social, and commercial reasons, there was some initial resistance to the introduction of paper in Europe. Prior to the introduction of paper, most of the (preserved) written documents were produced on parchment. Early paper competed with parchment; likewise the nascent profession of papermaking was in competition with the established profession of parchment production. This competition led to the production of high quality, extremely permanent papers that resembled parchment in many characteristics. 45 I have had the immense pleasure of studying papermaking at the University of Iowa Center for the Book (UICB) under the instruction of Mr. Barrett and others. In addition to papermaking, I have performed research into paper permanence for the UICB, again under the supervision of Mr. Barrett. 46 For scholars of the history and methods of papermaking, Dard Hunter is the key source. In addition to his scholarly work, he revived the tradition of hand papermaking in Lime Rock, Connecticut. Another significant figure in hand papermaking is Timothy Barrett, at the UICB. 47 The guarded nature of the profession is such that there is a paucity of first-hand, contemporary documents related to the papermaking industry prior to the 19th century. As such, the craft has had to be rediscovered by trial and error, and there is still much that is not understood or reproducible. Hunter has compiled a good account of early writings and delineations of the craft of papermaking (Hunter 1930, 3-106). 48 The art of papermaking was announced by privy councillor Ts ai Lun to the Emperor HoTi in the year 105 (Hunter 1943, 312). 49 The first verified use of true paper in Sicily is The first paper was produced in Europe in in Spain, from thence paper production spread throughout Europe, reaching Norway as late as 1690 (Hunter 1943, , ). 28

2 Despite the high quality product, 50 papermaking did not receive its major impetus until the 15 th century. 51 There was, and still is, variation between regions and even individual mills, however, the core methods for preparing a hand sheet today differ little from those used throughout the centuries. This conservation of the technique has allowed for historians and modern craftspersons to attempt to reconstruct the practices and products of the ancient mills. While the technique of producing handsheets has on the whole been conserved, there have been several innovations that have drastically changed the industry, the most notable are, of course, the various papermaking machines, which all but killed the hand papermaking industry in the mid to late 19 th century, but innovation has not been limited to modern times. Since its introduction to Europe, paper production has grown from the cottage industry of the East to the modern, industrial behemoth of today. Some notable innovations are: stamper beaters-1151, gelatine size-1337, movable type printing press , the glazinghammer-1540, metal screws for driving printing and paper presses-1550, the Hollander beater -1680, glazing-rolls-1720, rag shortages in the 18 th century spurred research into alternative fibres for papermaking, the wove paper mould-1757, chlorine bleach-1774, Nicholas-Louis Robert s first paper machine-1798, Koops s experimentation with ground wood as a papermaking fibre, rosin size , the Fourdrinier paper machine-patented 1806, the cylinder washer for the Hollander beater-1835, a patent granted for a wood-grinding machine -1840, ground wood paper produced on a commercial scale-1846, Burgess Watt process for treating wood pulp-1851, and the sulphite process (Hunter 1943, ). 52 What is paper? A simple question that is easy to answer incorrectly. For example, according to Bookbinding and the Conservation of Books, A Dictionary of Descriptive Terminology, paper is a 50 Which, to this day, has remained non-reproducible. 51 The history of the European paper industry is intimately linked to that of printing; as the demand for printed material increased, so to did the demand for paper (Hunter 1943, ). 52 While the industry has continued to evolve, the 1870s serve as the terminal date for this study. 29

3 matted or felted sheet or web of fibre formed on a fine screen from a water suspension Paper may be produced from animal fibres, e.g., wool, fur, hair, silk; mineral fibres, e.g., asbestos; synthetic fibres, e.g., rayon or nylon; and even ceramics, metals, glass, and other materials. Most paper, however, is produced from cellulosic plant fibres, (Etherington and Roberts 2005, paper). Etherington and Roberts have provided an extremely good definition of paper, while simultaneously demonstrating a common misconception about paper. Paper is not a matted or felted material, nor is it interwoven, or a web; rather it is a layered material. 53 In a felted or matted material, fibres can be seen to transverse the z-axis 54 of the material, while in a layered material they do not. In paper the fibres essentially lie in the plain of the sheet in a nearly random arrangement. Consequently paper can be treated as a 2D structure as long as the vertical position of the fibre is not important. Furthermore crosssectional analysis of a sheet of paper shows that a statistically non-appreciable number of fibres transverse the thickness of the sheet, this is despite the fact that a fibre can be 10 to 20 times the thickness of the sheet in length (Corte 1982, ). Image 9 Cross section of a sheet of paper showing the layered structure (Corte 1982, 225). The number of transverse fibres depends on the formation technique, the fibre type, fluid viscosity, and several other factors, but generally, the number of transverse fibres, and thus felting, is increased with higher consistency of pulp in suspension The author believes that this common misconception is due to a Western attempt to understand papermaking and paper structure from the notion of the felt making industry which was invented (in Europe) in Greece by at least 900 B.C.E. or from the notion of weaving, which dates to the lithic period. 54 If the x and y-axes define the plane of a sheet of paper, then the z-axis is that which is normal to the plane. 55 A felt like structure is desirable for some applications, for instance absorbent tissue; it is undesirable for high quality sheets to be used for writing or printing. 30

4 Furthermore, paper must be considered and treated as a composite material. A composite material, although itself made up of other materials, can be considered to be a new material having characteristic properties which are derived from its constituents, from its processing, and from its microstructure (Prudhomme, 1976). As a composite material, paper provides many difficulties for analysis and understanding of its behaviour. Chemical tests provide information on the molecular level of the component materials, yet they do not adequately explain the complex behaviour of the composite material, while physical tests provide information about the integrated paper network, but not the isolated contributing components, and the results of physical tests are subject to question (Dwan 1987, 1-17). Image 10 Organization of cellulose polymers into fibrils and in plant cell walls (Tímár-Balázsy and Eastop 1998, 24) The next point of contention is the production of paper from non-cellulosic fibres, most practitioners agree that true paper is limited to properly prepared cellulosic material such that inter-fibre bonding is optimised. 56 That is to say, the fibres must be fermented, beaten, and/or chemically/enzymatically treated to break down their macrostructure and increase their fractal properties, thereby the surface area of the individual fibres. These changes result in the production of macro and microfibrils. It is these microfibrils, not the fibres, that overlap and interact on a microscopic and molecular level to form various bonds that give the sheet its structural integrity and prevent the sheet from delaminating (Peel 1999, 45-51). 56 For instance tappa, Polynesian bark cloth, is made from nearly the same materials as washi, Japanese paper, yet tappa is not paper because it has not been processed to the appropriate degree such that inter-fibre bonding is prevalent. Likewise fibre glass insulation, while superficially resembling paper, cannot be paper because the fibre material does not allow for inter-fibre bonding. 31

5 While non-cellulosic materials can be added to the paper pulp, they do not undergo the same chemical and physical changes as cellulose based pulp. These atypical fibres tend to show non-homogenous distributions or increased Image 11 Molecular structure of cellulose and amylose, the profusion of hydroxyl groups (-OH) allow for the cellulose polymers to form hydrogen bonds (and other bond types) between the microfibrils and the microfibrils and water. felting such that the resulting material rapidly loses its defining characteristics as paper. The Papermaking Process Hand papermaking is an incredibly labour intensive process, it can be seen that a sheet must pass through a workman s hands more than thirty times and approximately ten times under the presses (LaLande 1761, by Barrett 1989, 25). The first step in the process is to acquire the fibre source. Traditionally European papers were made from linen and hemp rags (Laitinen 1996; Piette 1990; Hunter 1930; Hunter 1943; Barrett 1989). 57 Rags became increasingly scarce as the demand for paper increased, this became especially significant in much of Europe and America during the late 18 th through the 19 th centuries (Hunter 1943; Hunter 1930; and Hills 1988). While it is possible to use textile cuttings, spinning waste, and unbleached cloth to make paper; it was not used in finer grades of paper (LaLande 1761 by Barrett 1989, 9). In the 18 th century, European papermakers and scientists began searching for alternative fibres from which paper could be made. 58 As early as 1716, the Society of Gentleman in England proposed in their Essays a method of preparing hemp, by washing, boiling, pressing and drying the raw plant material thus rendering it a substitute for flax and cotton as a papermaking fibre (Hunter 1943, 233). More famously, in a November 15, 1719 treatise, 57 Prior to the industrial period most cloth was home spun and woven, garments were worn until they were too weak to be serviceable. When a rag finally entered a mill, it had exhausted all other usefulness (Barrett 1989, 8). 58 It should be noted that this search for non-rag stuff was limited to Europe, as Eastern papermakers had been producing non-rag paper for well over a thousand years by the 18th century (Hunter 1943). 32

6 René Antoine Ferchault proposed the use of wood as a fibre source for papermaking. This was based on his observations of wasps (Hunter 1943, 233). Perhaps the greatest contribution to the search for alternative paper fibres came from Jacob Christian Schäffer. Between 1765, and 1771, he published a six volume treatise. This work was based on his personal experiments in fibre preparation and sheet forming. His treatises contained samples of the prepared papers, 59 (Hunter 1943, ). Now extremely rare, Schäffer's treatise was well known and widely circulated during the 18 th and 19 th centuries (Hunter 1943, 247). Once a fibre has been sourced it must be prepared accordingly. If rags are used they must be sorted, trimmed, washed (repeatedly), bleached, 60 retted, 61 washed (repeatedly), cooked (with or without lime 62 ), treated with lime (again), then they were ready to be beaten (Barrett 1989; Hunter 1943, ; Hunter 1930, ). 63 During the period in question two types of beaters were in use, the stamper, which dates from the Middle ages and the Hollander beater, invented (roughly) 1680, but not widely adopted until the late 18 th early 19 th centuries (Barrett 1989, 11-12; Hunter 1943, ; Hunter 1930, ). Stampers, as their name implies, used falling hammers to work on all of the rags in the trough at the same time; they required careful sorting and preparation of the rags, as beating was the bottleneck in the papermaking process. Hollander beaters use an improved beating action through a rotating cylinder fitted with blades, and the tolerance between the cylinder and the bedplate can be adjusted to vary the beating action. In general Hollander beaters are much faster than stampers, they do not require retting, and they can easily process tougher fibres. Stampers make better paper because they leave longer fibres than Hollander beaters. 64 During beating, the pulp can be 59 These often contained a fifth-part cotton fibres acting as binder. 60 Early bleaching processes did not make use of strong chemicals; rather sunlight, sour milk, sour malt, or wood ash was used (Barrett 1989, 7-9). It was not until 1774 that Karl Wilhelm Scheele invented chlorine bleach (Hunter 1943, 237). 61 That is, fermented in a humid, but not saturated environment (Barrett 1989, 9-12; Hunter 1943, ; Hunter 1930, ). 62 This term is necessarily vague as it is not known if quick lime, CaO, slaked lime Ca(OH) 2, or CaCO 3, limestone, chalk, or precipitated carbonate, was used. Undoubtedly different papermakers in different locations used different forms of lime for different types of paper (Barrett 1989, 10-12; Hunter 1943, ). 63 There can be considerable variation in the order of the steps according to locale, technology, and desired product. Dutch papermakers sorted rags into a greater number of grades than others (Barrett 1989; Voorn 1963, Voorn 1960). Dutch papermakers generally did not use fermentation (Barrett 1989, 12; Voorn 1963). There are also accounts of papermakers freezing beaten pulp and formed sheets (Renker 1961; LaLande 1761 by Barrett 1989, 22, 27). 64 It is possible to examine paper fibres and identify which beating technology was used, and in doing so roughly date the paper. 33

7 washed, and/or pigments or dyes added to brighten or colour the pulp; of the possible additives, CaCO3, 65 in the form of ground limestone, chalk, or seashells, was the most common during the period in question. It is finally time to address sheet forming. 66 In sheet forming there are four essential pieces of equipment: the vat, 67 the moulds and deckle, 68 the felts, 69 and the press. Sheet forming is normally executed by a team of three persons per vat: the vatman, the coucher, and the layman. The vatman dips the mould into the vat of pulp suspended in water, and pulls a sheet of pulp on the mould surface. The key step is when the vatman deftly shakes the laden mould back and forth in two dimensions. This shaking is what aligns the fibres in the plane of the sheet and gives paper its layered characteristic. The vatman then removes the deckle and passes the first mould to the coucher. While the vatman pulls the next sheet on the second mould, the coucher couches (deposits) the newly formed sheet on a felt. This is repeated until a post, a stack of 144 sheets of paper-alternating felt-paper-felt, is formed. The post is then taken to the press where the water is expelled under mechanical pressure from the paper and felts. After pressing, the layman separates the delicate, newly-formed sheets from the felts and places them in a pile. The felts are then returned to the coucher for the next post. Next the layman represses the stack of sheets and shuffles them to be pressed yet again. This shuffling, called exchanging, is repeated several times, and it gives the paper greater strength and a smoother surface. After exchanging the sheets have been formed, but the paper is far from finished In addition to whitening the pulp, it also added bulk and an alkali reservoir. 66 Which for most persons is the act of papermaking. 67 The vat is a large tub in which the pulp/water suspension is placed and into which the vatman dips the mould to pull the sheet. 68 These superficially resemble a window screen fitted with a picture frame. 69 Felt is a type of textile consisting of matted, and later woven, fibers generally from an animal hair fibre source. 70 Interested readers are directed to consult (Hunter 1943) or (Barrett 1989) for a more detailed account of sheet forming. Better yet, they should visit a paper mill and try their hand in the vat. 34

8 The paper is now ready to be dried, to do so the paper is pressed into stacks of four to eight sheets called spurs. These spurs are hung over horse or cow hair ropes and allowed to dry. Spur drying helps to keep the paper from becoming overly wrinkled or misshapen during the drying process. After drying the spurs are separated and the individual sheets are often coated or sized to prepare it for use (Hunter 1943; Barrett 1989). Gelatine or animal glue was the most common sizing agent in Europe from 1300 until the 19 th century when rosin size 71 became available (Voorn 1963a; Brückle 1993b; Gess 1989). During the intervening centuries there were changes in sizing practice: alum 72 was used with increasing frequency and concentration. Alum was first used as an additive to gelatine to allow it to be stored for longer periods of time. Later, alum was put to multiple uses: water clarification, filler retention, defloculation of pulp, 73 and for rosin sizing (Brückle 1993b; Gess 1989). The increased use of alum in papermaking has been linked to increasingly acidic profiles of paper and decreases in paper permanence. After sizing the paper is again dried. After which it can either be used as is, or it can be finished by coating, 74 glazing, 75 or some other surface treating process. Papermaking at Ösjöfors Pappersbruk Ösjöfors Pappersbruk was selected as a typical small-scale paper mill operating within the southern bonader producing region during the 18 th and 19 th centuries. The mill is located too far to the east to have likely supplied the Sunnerbo painters, but as paper bonader can be found throughout Southern Sweden, it is still possible that paper from this mill was used by bonader artists. 71 An internal sizing agent based on modified diterpinoid constituents of coniferous tree sap. 72 Alum can be AlK(SO 4 ) 2 or Al 2 (SO 4 ) 3 or a mixture with various contaminants. 73 That is, reducing the clumping tendencies of the fibres. 74 Application of a surface treatment such as pigment in a binding media. 75 Rubbing or polishing the surface with a stone, glazing hammer, or roll. 35

9 Ösjöfors Pappersbruk, Rumskulla, was founded 10 June 1777, by Lars Dristig. The mill operated under the names Ösjöfors, Ödesjöfors, and Ödsjöfors Pappersbruk (Malmros 1944). 76 The mill remained within the Dristig family until 1845, when it passed to O. Pehrsson. It operated continuously until 1926, when it became part of the collection of the Tekniskamuseet (Malmros 1944). The mill was featured in a film from 1920 about the Swedish paper industry for the 1923 Jubilee Exhibition at Göteborgs Stadsmuseum. Having been featured in the exhibition and being part of the Tekniskamuseet s collection, the mill is one of the most studied in Sweden. Ösjöfors Pappersbruk is considered to be typical of Swedish mills from the 18 th and early 19 th centuries. The mill was destroyed by fire on 27 August 2005 (Tekniska Museet 2005). Based on archival records and still images from the 1920 film it is possible to reconstruct the papermaking practice within the mill during its operation. What follows is an analysis of paper production at the Ösjöfors based on a production record from 1831, and a series of photographs 77 of the mill operation in In 1831, the mill employed six individuals (Malmros 1944). The mill was water powered and used a Hollander beater. There was a single vat and press and the mill produced both skrivpapper (writing paper) 79 and karduspapper (lower grade paper/paperboard) (Malmros 1944). 80 In 1831, Ösjöfors Pappersbruk produced 488 Ris 81 of skrivpapper valued at I have relied upon transcriptions and translations of the archival information used for this study; this adds two degrees of removal but resolves access and legibility issues. 77 Readers please note that Images 12 through 25 are from the Göteborgs Stadsmuseum Arkivet, they are subject to copyright and have been used with permission. Please consult the bibliography for information about locating this images within the museum s archives. 78 Because the mill was never renovated in the late 19th or early 20th centuries it is valid to draw conclusions about paper production from these images. 79 In 1831 in Sweden, the term skrivpapper encompassed a large variety of paper qualities and formats: printing and writing paper of various formats, drafting paper, letter paper, bank note and stamp paper, among others. No single mill produced every type of paper, in fact for certain types, bank note and stamp paper in particular, a mill must have had a special charter to make the paper. 80 As with skrivpapper, karduspapper included several grades and formats: cardboard, newsprint, grey paper, wallpaper, tobacco paper, brown wrapping paper, sugar paper, wax paper, etc. As with skrivpapper no single mill, or at least, no small scale mill, would have produced every grade and format of karduspapper. 81 A Ris is a unit of paper roughly equivalent to a ream. A ream formerly consisted of 480 sheets, now consists of 500 sheets, and a printer s ream is 516 sheets. 36

10 Riksdaler banko and 540 Ris of karduspapper valued at 502 Riksdaler banko (Malmros 1944). This production rate seems to be congruent with mills with a similar work force. Images of Ösjöfors Pappersbruk Ösjöfors Pappersbruk, 1920 In Image 12 it is possible to see the mill race that provided power for the waterwheel in the mill. Note that the windows in the loft as well as the back side of the mill were not glazed, but were fitted with louvers. This allowed for ventilation. The loft is where paper would have been spur dried; while the water wheel, Hollander, and vats would have been located on the lower floor near the back. Image 12 Ösjöfors Pappersbruk from below the millrace, Ösjöfors Pappersbruk, 1920 Image 13 depicts the front of the mill with glazed windows on the ground and first floors. The rooms with glazed windows would have been used for paper sorting, grading, finishing, weighing, dry storage of finished sheets, and rag storage. The rooms would have also doubled as office space if needed, and possibly as living quarters for workers. Image 13 Ösjöfors Pappersbruk from the front,

11 Rag sorting and cutting at Ösjöfors Note the dark coloured rags and the careless way that they are cut and mixed together. This lack of care in sorting and cutting is not typical for making high grade paper. It can then be assumed that the paper to be next produced will be a low grade paper. Also note that Ösjöfors Pappersbruk did not have the necessary facilities for using chemical bleach, so discoloured or coloured rags would have produced grey paper. Image 14 Rag cutting at Ösjöfors Pappersbruk, Rag sorting table at Grycksbo Pappersbruk This image is of a rag sorting room at a different mill from the same period. Note the white rags and the well lit and clean environment in which they were stored and processed. The bright white colour of these rags hints the Grycksbo Pappersbruk made use of chemical bleaching agents. These rags would have been used for producing high grade writing papers. Image 15 Rag sorting and cutting table at Grycksbo Pappersbruk, Hollander Beater at Ösjöfors The mill is powered by a water wheel which is located just behind the drive train visible in the centre of the image. The beater is just to the right of the train. The wheel and gearing are such that a second Hollander, and by implication, vat and press, could have been Image 16 Hollander beater at Ösjöfors Pappersbruk, accommodated to the left of the train effectively doubling the output of the mill. The mill employed six persons in This is excessive for a single vat mill. It is possible that in 1831, there was a second beater, vat, and press such that a workforce of six persons is justified. 38

12 Sheet forming at Ösjöfors Only the vatman is pictured in this image, but there would have been a coucher working with the vatman, as in Image 18. Note the position of the window, during daylight hours the vatman would have a raking light from the rear which would accentuate any formation irregularities. Also note that the new sheet being formed is between A4 and A3 in format, and it is thick and grey coloured. Image 17 Sheet forming at Ösjöfors Pappersbruk, Sheet forming at Mariedals Pappersbruk Based on the clothing and the obviously long exposure, this image is from the mid to late 19 th century. Within the length of exposure, the team has begun a new post: the coucher has laid the first felt and is waiting for the vatman to pass the mould with the fresh sheet; meanwhile the vatman has dipped the mould and deckle and pulled the sheet. Image 18 Sheet forming at Mariedals Pappersbruk, mid to late 19 th century. Pressing a post at Ösjöfors The press used at Ösjöfors Pappersbruk is a simple screw press with a wooden frame and a metal screw. The press does not appear to have a samson and therefore would have only generated pressures of roughly tons (Barrett 1989). The lower pressures would have produced thicker paper with a rougher finish and a less pronounced rattle. Using such a press, it is possible to produce finer sheets by pack-pressing with exchanging. Image 19 Screw press at Ösjöfors Pappersbruk,

13 Drying Paper at Ösjöfors Drying paper outside is irregular (in Europe during this period) and was likely only practised on lower grades of paper as the risk of discolouring fine, white paper is too great. A benefit of such practice would have been a slight whitening of the sheets due to exposure to sunlight. However, Ösjöfors was equipped with a drying loft and there is evidence (Image 25) that it was used to spur dry paper. Image Drying paper outdoors at Ösjöfors Pappersbruk, Spur drying at Katrinedals Pappersbruk Typically the spurs (5-10 sheets) were hung to dry in a loft. This method helped to keep the sheets clean and prevented undue shrinkage and warping during drying. Note the position of the louvers (visible in the lower right of the image) which would have directed air against the edges, and between, the spurs. Also note that the drying lofts were unheated thus allowing the paper to freeze during the winter. Image 21 Pappersbruk, Spur drying paper in a loft at Katrinedals Inspecting and counting sheets at Ösjöfors After drying, the sheets would be inspected, counted, and sorted according to grade. The fine sheets would go on for finishing while the broke sheets would either be reprocessed or sold as a lower- grade. Note that the sheets in this image are not the same as those being formed in Image 17, nor those made from the rags in Image 14. Image 22 Inspecting and counting sheets at Ösjöfors Pappersbruk,

14 C h e m i c a l a n d P h y s i c a l C h a r a c t e r i s a t i o n General Methods 82 Sampling The samples were excised from the terminal edges of the bonader by Martin Ericson at Studio Västsvensk konservering. The samples were individually placed in folded circles of acid-free Whatman filter paper and labelled with the museum registry number. After collection, the samples were always manipulated with forceps and the researcher took care to not wear wool, or other loose-fibred clothing, while working with the samples. Preconditions The researcher took care to perform the chemical and fibre analyses under blind conditions. That is, the samples were only identified by registry number and no further information regarding date, painter, region, or motif was collected until after the wet testing was complete. The blind was used to protect the researcher from making conjectures about the paper based on a known date, and it allowed the researcher to be more objective in the analyses. Surface Examination The front and reverse surfaces of the samples were examined with a dissection microscope and imaged with a digital camera. For the front surface, the presence of paint and/or ground, its general condition, and any peculiarities in the surface were noted. For the reverse side, the general condition, colour, formation quality, fibre distribution, and any atypical details were noted. Fibre Testing Examples of the various fibre types were selected from the larger and more populated samples. Slides of these fibres were prepared using water as the medium. The fibres were examined using light field, dark field, and UV-fluorescent optical microscopy. The fibres were identified, to the best of the researcher s ability, through the use of fibre atlases and by consultation of experts at Studio Västsvensk Konservering. The general conditions of the various fibres were noted and images taken with a digital camera. 82 Please consult Appendix D for for more information about the experimental methodologies. 42

15 Historical and Archival Research After the initial fibre analysis, the researcher suspended all wet tests and began to consult published sources on papermaking in general, hand papermaking, and the history of papermaking in Sweden, as well as archival information from the Swedish National Archives and the Göteborg Stadsmuseum Arkivet (Gothenburg City Museum Archive). Additionally the researcher began to research bonader painting, display, and collection and to visit exhibitions of bonader. The consultation of published sources, archive information, and museums exhibitions helped to place the samples within a general cultural/historical construct, rather than a hyper-abstract laboratory space. Surface ph Testing The researcher proposed that the samples could be sorted by ph profile. Based on earlier research at the University of Iowa Center for the Book, the researcher used a modification of the TAPPI standard for surface ph procedure that allowed enough time for incorporated salts to reach equilibrium in solution. This modification provides a more accurate ph measurement, and gives insights into the techniques and chemicals used during the production of the paper. Spot Testing The researcher now used a series of spot tests to qualitatively identify different additives, sizing agents, fibres, and chemical processing of fibre. These tests were conducted on both distributed fibres and samples of the intact paper. For most tests, the results were observed by dark field optical microscopy and the results recorded with a digital camera. The exceptions are Ehrlich s reagent test for 4-hydroxyproline and Raspail test for rosin; for these tests the solution colour was noted. Elemental Analysis The researcher selected one sample, HM 9996, for analysis in an Environmental Scanning Electron Microscope. The choice was based on amount of remaining sample, thickness of paper, and general good condition of the sample. Three micro-samples were excised by scalpel and placed on the mount such that frontal, reverse, and cross section could be imaged and elemental analysis performed without opening the chamber. The sampling allowed for 43

16 localised determination of elements in a semi-quantitative fashion, which gave further insights into paper formation, additives, and surface treatments. General Observations and Discussion 83 Surface and Physical Properties The paper samples are, on the whole, greyish white or a light brown and rather thick, with the exception of the sample HM 4587 which is extremely thin and wrinkled. The samples more closely resemble paperboard than writing paper, and they are representative of lower-quality products of a European paper mill active in the 19 th century. Again with the exception of HM 4587, the samples are stiff, yet plastic and not brittle. Exposed fibres at the edges of the sample and the paint surface can be brittle, therefore care must be taken during micro-sampling. The formation quality is slightly clumpy and uneven, which is typical of cheaper, thick papers or paperboard. That the samples are not excessively yellowed or brittle indicates that they do not have a highly acidic ph. The samples show considerable amounts of surface dirt. There are spots of iron oxide residue on several samples, these are artefacts of the display of the bonader. 84 Some samples have spots of a dark resinous substance on the reverse side. Samples of these deposits were taken and flame tests indicate that it is a coniferous resin or pitch. 85 The presence of this pitch on the reverse is likely an artefact of the display process. It indicates that that particular bonader was attached to a wall or ceiling space, and not left freely hanging as depicted in some images of a Dragstuga (decorated cottage), see Image 8. Fibres Close examination of the reverse of the samples reveals that the fibres are predominantly white and cellulosic. There are, however, many dark fibres embedded within the paper 83 Please consult Appendix A for itemised observations, images, and detailed results. 84 This can be surmised due to their localisation around small holes through the paper that show surface disruption patterns typical of a tack being thrust from front to back. 85 The samples flared when placed in a flame. 44

17 matrix. These fibres are predominantly blue, but some red, green, black, and brown fibres can be found. Microscopic examination of these fibres indicates that most are animal hair fibres, predominantly wool, though some could be cow or horse, and some of the coloured fibres are dyed flax, hemp, and/or cotton. Most of the coloured fibres are distributed throughout the paper matrix. This indicates that they were present throughout the pulp and not wild fibres floating on the surface of the vat. 86 The presence of these coloured fibres in the pulp indicates that strong bleaching agents were not used in the fibre preparation process. The ratio of coloured fibres to white fibres is low, this indicates that the fibre source was predominantly light coloured. The presence of the coloured fibres is likely incidental, and due to poor rag sorting or the addition of coloured rags, or even decorated sections (embroidery, seams, etc.) of white rags to bulk out a thin vat. There is a remote possibility that blue fibres were added to whiten the paper (Brückle 1993a), but this practice is more common for writing paper and not paperboard. There are also some dark, thicker animal fibres on the surface of the sheet and embedded in the paper. These fibres tend to transverse the sheet and are atypical in fibre type and orientation. Under microscopic analysis, the fibres show a scaling pattern different from wool, and closer to what one expects for cow, horse, or the leg/face region of a sheep. These fibres would have been used in felt and/or rope preparation, and they are likely artefacts of the couching felts or of the ropes used for loft drying. The fibres conditions and deposition patterns further support this conclusion. Examination of the bulk fibres shows that the majority are shortened and highly disrupted with many microfibrils present. The fibre length and degree of disruption is typical of aggressive use of a Hollander beater or over retting of the rag stock. Since Hollander beaters were common in Sweden during the 19 th century, it is likely the former. These highlydisrupted fibres are difficult to identify with any confidence, but based on archive research, they are likely linen or hemp from rag sources. The large population of short, highly-disrupted 86 Wild fibres can result from improperly prepared pulp, or from aerial deposits, for instance, lint from the clothing worn by the vatman or coucher or from the rag sorting and cutting process if it is in an adjacent room. 45

18 white fibres indicates that the dregs from several vats for making finer quality paper could have been combined with pulp made from coarser fibre sources. These coarser fibres have been disrupted to a lesser degree, and have been identified as being linen, hemp, cotton, shrive portion of bast fibres (unidentified, but probably jute), and, surprisingly, partially processed wood fibre. The presence of these fibres indicates that the mill(s) were using alternative fibre sources. The cotton fibres came from imported, industrially produced cloth or raw fibre, the hemp and linen could be remnants from the indigenous textile production industry or from the use of raw fibres, the unidentified bast fibre could be jute from ropes and nets, and the wood fibre is a bit of an enigma. 87 Fibre Staining, Spot Tests and Microscopy To facilitate fibre identification, a series of chemical stains and tests were used: Graff C stain, fluoroglucinol, and neocarmine. 88 While other tests, I3 -, 89 Aluminon, Raspail, Ehrlich s, flame, and acid drop tests identified additives. The results of the fibre identification and the chemical tests are outlined in the itemised results section in the appendices. 90 There were difficulties with the positive interference with the Ehrlich s test for 4-hydroxyproline. 91 This was due to the presence of glue paint on the bonader. Care was taken to remove the paint layers and prevent contamination of the samples, but a single speck of glue is enough to generate a false positive. There were also difficulties with the Raspail test for rosin Some of the wood fibres present in the samples stain as if they were partially processed by cooking in a possibly alkali environment. The presence of these fibres in the paper samples predates the accepted date of first use of the such fibres in Sweden. 88 The researcher also attempted to use the Herzberg stain, but suspended using the test on the samples because the prepared reagent did not behave as expected on known standards. 89 As the Graff C stain has an I 3 - component, the separate starch test was suspended, and the Graff C stain doubled for fibre identification and testing for the presence of starch. 90 For detailed results please consult Appendix A Hydroxyproline is a modified amino acid that is found in animal connective tissue. Since glue and gelatine are both derived from animal connective tissue, there is a high risk of sample contamination and false positive tests when working with bonader samples. 92 These difficulties resulted in unclear results. It is noted in the literature that the Raspail test for rosin can give a false negative result if the samples are old, and weak (off colour) results due to quasi-positive interference from resins in wood pulps. 46

19 In an effort to circumvent the glue interference, with the Ehrlich s test for 4-hydroxyproline, micro-samples from HM 7986, HM 13698, and HM 9996 were collected and the outer surfaces cleaved off. 93 The central portion of the sheet was then tested for 4-hydroxyproline. If the paper were vat sized with gelatine, then one would expect a positive result, due to the gelatine being absorbed into the sheet. While if the sheet were only coated with glue or a glue/pigment mixture then one would expect a negative result, since the glue would remain on the surface. HM 7986 and HM both yielded negative results for the presence of 4- hydroxyproline, and thus for gelatine size, while HM 9996 yielded a pinkish yellow solution which can be interpreted as a weakly positive result. When the test was repeated the same colour evolved, but this can be a false positive since even a small amount of glue paint residue would generate this result. The Raspail test for rosin size was conducted according to standard protocol and a reddish orange colour evolved for each sample with the exception of HM 8320 which remained yellow indicating a definite negative. A true positive result would be the evolution of a raspberry red colour. While the reddish orange is different from the negative result of yellow to brown, it still does not match the expected positive result. This can be explained by assuming an aged and degraded rosin or by interference from other terpinoid compounds within the paper (Barrett and Mosier 1992). ph testing The most conclusive test was the surface ph determination of the samples. Noting previous difficulties in determining the ph of alkali papers and recent publications discussing the issue (Strlic et al 2004), an alternative protocol to that given by TAPPI was employed. This protocol allows for the adsorbed compounds to reach equilibrium by regularly sampling the ph over a five minute interval, thus giving a better picture of the ph of the paper, but as noted by Strlic et al (2004) there are still inherent problems in determining the true ph of an alkali paper. When sampling for this test, care was taken to choose regions with little to no paint, and to scrape the surfaces to remove any remaining paint and to increase the available surface area. The results are summarised in Graph 1, see page These samples were selected because of their exceptional thickness and ease in splitting the paper. 47

20 With the exception of HM 9996 (which stabilised at 7.4) and HM 7986 (which stabilised at 4.95) all of the papers are mildly acidic. This is likely due to the presence of alum, in varying concentration, in all of the papers. Alum was added at many stages during the paper making process in increasingly superfluous quantities during the 19 th century. On the other hand, these papers are not nearly as acidic or brittle as one might expect for 19 th century papers. A possible reason would be a large alkali reservoir either in the form of adsorbed calcium and magnesium, or in a chalk/carbonate based white pigment used for the ground. The curves can be sorted into five different behaviours, and from these behaviours, it should be possible to sort the papers by manufacturer. Elemental analysis An Environmental Scanning Electron Microscope with an X-ray detector was used for elemental microanalysis of sample HM Three samples were prepared such that the front and reverse surfaces as well as the cross section could be analysed. The technique is semi-quantitative and it was possible to note that there was much more calcium on the front and reverse surfaces than in the cross section. 94 This indicates that both surfaces were treated with a calcium rich substance, likely glue paint with chalk. While there was some calcium in the cross section, it was of much lower concentration and likely due to residual amounts in the water used during the paper making process. The other elements, with the exception of silicon, are fairly evenly distributed through the sheet indicating that they were present before the sheet forming stage. 94 Please see Appendix B for copies of the XRF data. 48

21 C o n c l u s i o n s The Paper From the chemical and physical analyses and the historical and archival research it is possible to reconstruct the techniques used to make the paper that the bonader are painted upon. This sample set is quite interesting in that it represents a cross section of papers made during a period of intense change within Sweden in general and the paper industry in particular. Fibre Sources The paper is predominantly from rag sources, but there is evidence of early use of wood fibre, and wood fibre processing. The papermakers were showing some ingenuity in dealing with an increased demand for finished product with a decreased supply of raw materials. They experimented in using cheaper, alternative fibres for lower grades of paper, including that which bonader are painted upon. Some of their work with non-rag fibres predates the accepted period of commercial use in Sweden. Their use of wood fibres and of processed wood fibres indicates a familiarity with publications from Koops and Schäffer among others. Furthermore, the Graff C stain indicates that some of the wood pulp has been cooked, possibly in a basic environment. If so, then papermakers in Sweden were also working towards what is now called the Burgess-Watt process. Beaters The use of alternative fibres and the degree of disruption of all the fibres in the papers indicates that the paper mills used Hollander beaters. 95 Swedish mills adopted this technology, as early as 1759 at Tumba (Voorn 1960), if not earlier at other mills. By the mid 19 th century Hollander beaters could be found in most paper mills in Sweden. Alum The papermakers made extensive use of alum during the papermaking process. It is unclear if the alum was used in combination with a sizing agent such as gelatine or rosin, but aluminium can be found distributed throughout the paper matrices. Aside from use as a preservative and/or complexing agent for size, alum was used for a multitude of purposes as 95 While it is possible to prepare alternative fibres using a stamper beater, it is not economically feasible as each fibre type would have to be beaten separately and wood fibres would require too much time in the beater for them to be a viable substitute or even additive to a rag pulp. 49

22 noted by Brückle (1993b). The percent of alum added, its state, and the presence of other adsorbed compounds creates a unique ph curve which can be used to sort the paper, and thereby the bonader, independent of stylistic analysis. ph Profile Using the surface ph, a ph profile can be generated allowing the papers to be sorted into five families, neutral rising ph: HM 9996; slightly acidic falling ph: HM 4838; slightly acidic rising ph: HM 4587 and HM 9995; slightly acidic constant ph: HM 8320, HM 13698, and HM 9454; and acidic falling ph: HM The bonader were then sorted by stylistic analysis yielding four families, Sunnerbo School: HM 9996, HM 9995, and HM 4587; Västergötland: HM 4838; Paintings by Carl Reinic Rosenberg: HM 8320, HM 13698, and HM 9454; and an Unattributed Earlier Artist: HM Since HM 9996 was painted by a different artist than the other two bonader from the Sunnerbo School it is permissible to consider that the stylistic sorting yields five distinct groups, Suunerbo A: HM 9996; Sunnerbo B: HM 9995 and HM 4587; Västergötland: HM4838; Carl Reinic Rosenberg: HM 8320, HM 13698, and HM 9454; and Unattributed Earlier Artist: HM Surface ph as a function of time (in seconds) HM 8320 HM HM 4587 HM 9995 HM 9996 HM 9454 HM 7986 HM Graph 1 It is possible to identify five separate ph behaviors and to group the samples by the behavior. There is a high correspondence between ph behavior grouping and stylistic grouping. 96 The author was not able to place this artist stylistically, but the bonader is dated to

23 The correlation between the ph based sorting and the stylistic sorting is quite close. This is encouraging as surface ph testing is inexpensive, fast, non-sampling and quasi-non-destructive; if further tests confirm the above results, then ph profiles can be used as an additional tool to help place unattributed/misattributed bonader and to identify manipulated bonader (Nyström 2002, Nyström 2003). Sizing at the Mill It is unclear as to whether the paper was sized at the paper mills or not. Since glue and gelatine are derived from the same source, there is too much interference from the glue paint to determine if gelatine is present. The Raspail test for rosin is unreliable for older or degraded rosin size. The Graff C stain is I3 - based and should have identified the presence of starch, but none was found in the sheets. 97 Since the papers were lower quality, grey paper, it is unlikely that the mills went to the expense of sizing them. Inter-Fibre Substance In several sheets there seems to be an inter-fibre substance. This could be a filler, residue from a degraded internal sizing agent, or detached microfibrils. The scanning electron micrograph of HM 9996 shows the fibres are coated with some substance. This could be the observed inter-fibre material. This material has not been identified and further work is needed. Surface Treatment Elemental analysis indicates that calcium and silicon are in higher population on the paper surfaces, with only residual calcium and silicon in the cross section. From this it can be conjectured that a calcium based surface coating was applied to the paper after sheet formation. The increased presence of silicon indicates that the calcium source was from a ground pigment: chalk, limestone, or other source, rather than a precipitated pigment such as CaCO3 (ppt). Despite the acidic ph and lignin containing fibres, the papers are in relatively good condition: supple, little yellowing, and only slight brittleness of the fibres. This could be due in part to 97 Starch size was not common in Northern Europe during the period, but according to a recipe from Bruzaholms Pappersbruk, Småland, dating from the period in question, starch was added in conjunction with the alum for internal sizing of paper with rosin soap. With that in mind it would not have been surprising to find some starch residue in association with rosin sized papers. 51