History Plays a Role

The Endpaper Project began in 1989 when a group of rare book conservators at the Library recognized the need for a wider selection of strong and compatible papers to use with historical books. The initial phase of the project was to support the development of one style of Italian printing paper from the late sixteenth century because of its prevalence and fine qualities. Interested papermakers were asked to create a paper that matched the “Genoese type” book papers. A similar search for paper was made in 1756 by the Society of Arts in England, which offered a premium for “making one Rheam of Paper, which upon Trial will be judged to come nearest in all its qualities to the French Paper, proper receiving the best Impressions from Copper Plates.”3 Library of Congress Matching Workshop, July 2004. Left to right: Otoyo Yonekura, Min Ah Song, and Yasmeen Khan comparing new paper samples with old books.
The first Endpaper Project orders were placed with an internationally representative group of six papermills. Each mill was provided with specification guidelines for permanence and durability, a color swatch, and an original paper sample borrowed from the book Spirituali, imprint Venice, 1574. The resulting papers were beautiful and useful, as well as very different from one another and from the historical sample.
These results echoed those of the Society of Arts’ request: “Although the Society of Arts did award several second prizes, it was not until 1787 that they awarded a first prize,” thirty-one years later!4 The rich variety of endpapers submitted for the initial Endpaper Project demonstrated that papermaking is just as complex and subtle now as it was in the eighteenth century.
Another lesson from the initial Project request was how complicated it is to describe a specific paper in a language that is mutually understood by both conservators and papermakers. In an effort to facilitate collaboration, a meeting of the papermakers and conservators was held at Center for the Book at the University of Iowa (UICB), where one of the sessions focused on revising a draft of papermaking terms and definitions compiled from sources available at the Library (see http://www.loc.gov/preserv).5
Looking at Endpapers Today
This new phase of the Endpaper Project, in Spring 2004, focused on looking at available papers and conservators’ priorities for choosing endpapers for books from various centuries. The Library requested that the papermills send samples that best matched sixteenthto eighteenth-century, western book papers and that they choose two papers to be tested. They were also asked to describe the reasons for their choices, as well as to provide a general manufacturing description (http://www.loc.gov/preserv/). A significant effort was made to identify all mills producing endpapertype papers and to be as inclusive as possible. Many of the mills also sent sample books, color swatches, and other papers they thought we might find useful, some of which were included in the matching session. We also added papers to the matching sessions from mills that were unable to send samples but that we had in stock.6
Book conservators consider four main categories when choosing endpapers: the paper must not degrade quickly (permanence), it must be strong enough for the use (durability), it must have compatible working and handling properties, and ideally it would aesthetically reflect the rare book paper. The quality of permanence addresses the need for the endpaper to protect the book pages and not cause degradation.
Durability is necessary because paper works as a hinge when used as an endpaper. Compatibility of working properties is important because the weight, drape, and handle of a paper are the qualities that support both the text and the book structure. The qualities needed to match the aesthetic characteristics can be as varied as the wide range of papers that have been used to print books for the past 500 years. This project was an endeavor to locate and evaluate endpapers using these categories. Evaluation of Permanence and Durability
Testing was performed under the direction of the Research and Testing Laboratory of the Library of Congress using the Technical Association of the Pulp and Paper Industry (TAPPI) testing (T) methods. All of the papermakers identified in this article have given permission to use their mill’s name.
The papers were evaluated for pH, brightness, and folding endurance both before and after artificial aging. While the unique furnish and manufacture of each paper prevents comparison of the “before” test results of one paper to those of another, retention values (derived from the differences between the before and after aging test results) can be compared across the papers when evaluating permanence.
Artificial Aging: TAPPI T402 and LC Humid Oven Aging Procedure
Artificial aging is exactly what it says it is: artificial. The test method is an approximation of aging, which uses heat and humidity to encourage “aging.” Because there are so many elements within a paper that will react differently to these conditions, the method is not considered to be absolute but rather a tool to provide a relative basis for comparison. There is no direct correlation of number of years of natural aging to number of days of artificial aging. Fourteen days of aging will produce enough change in papers to be able to approximately predict if a paper will degrade significantl over time. This sampling of papers was aged for 14 days at 90°C and 50% RH.
pH: TAPPI T50
The pH was measured by the cold extraction method. The retention of pH by a paper before and after aging can be an indicator of degradation if there is a significant change. All of these papers had minimal change in pH as shown in Table 1
What should be noted is that all papers fall within an acceptable pH range, including those without an added alkaline material and one with alum added to the gelatin size (9a, 9b).7
Brightness Evaluation: TAPPI T452
The brightness evaluation of a paper before aging is related to the color of the paper and does not indicate quality. Measuring brightness before and after aging is a method to determine how much color change occurs after artificial aging, see Table 2.
Significant color change (greater than 10 points) is an indicator of the production of degradation materials. For example, groundwood paper can change approximately 20 points when artificially aged for 28 days. This is usually attributed to the lignin content. A color change of 5 points or more is visible, so paper that changes less than 5 points is considered a high quality paper. A change up to 10 points can be considered acceptable depending on the cause of discoloration. For example, a slight discoloration due to gelatin sizing is more acceptable than discoloration due to bleach residue.
BEFORE AFTER PAPERMAKER PAPER AGING AGING CHANGE
Anonymous 1a 8.33 8.11 0.22
Anonymous 1b 8.16 7.61 0.55
Anonymous 2a 7.73 6.82 0.91
Anonymous 2b 7.83 6.81 1.02
MacGregor 4a 7.08 6.66 0.42
MacGregor 4b 6.67 6.72 0.00
Anonymous 5a 6.73 6.53 0.20
Anonymous 5b 7.07 6.61 0.46
Peter & Donna Thomas 6a 7.15 6.76 0.39
Anonymous 7a 7.98 7.38 0.60
Anonymous 7b 8.48 8.30 0.13
Twinrocker 8a 8.15 7.46 0.69
Twinrocker 8b 7.34 6.69 0.65
Twinrocker 8c 6.76 6.72 0.04
University of Iowa 9a 6.80 6.99 0.00
University of Iowa 9b 7.02 7.33 0.00
University of Iowa 9c 7.62 7.81 0.00
Werkstatt für Papier 10a 7.48 7.68 0.00
Werkstatt für Papier 10b 7.43 7.58 0.00
TABLE 1. pH
Folding Endurance: TAPPI T511 & T402
The intention of this research was two-fold: to evaluate  the durability of the paper samples by evaluating their physical strength with a folding endurance test and to predict their “permanence” or lasting quality by comparing any decrease in physical strength after subjecting these samples to accelerated aging. Folding endurance is used to estimate the durability of a paper by its ability to withstand repeated bending and folding under tension provided by a 1.0 kilogram weight. The sample paper was cut into ten strips of a standard size in the same grain direction. Each test strip was subjected to constant folding, in this case, across the grain, until it broke. The number of double folds registered until the breaking points were recorded for each of the ten test strips, and the average value and the standard deviation (SD) were calculated.
SD is a measure of the variability of the individual measurements; the higher the SD value, the more variability in the sample.
The fold endurance test has long been preferred in our testing laboratories because of its higher sensitivity in comparison with other measures of physical strength. However, this higher sensitivity can also be a drawback if the papers of interest do not have the highly uniform formation characteristics that papers made from short fibers normally do.

BEFORE AFTER PAPER AGING AGING CHANGE
1a 81.03 71.24 9.79
1b 65.40 58.06 7.34
2a 76.03 64.78 11.25
2b 74.87 66.12 8.75
4a 41.20 39.36 1.84
4b 44.77 42.66 2.11
5a 61.27 43.05 18.22
5b 59.90 45.29 14.61
6a 86.00 77.11 8.89
7a 68.27 67.55 0.72
7b 48.20 47.00 1.20
8a 68.67 64.70 3.97
8b 68.29 62.82 5.47
8c 63.11 54.35 8.76
9a 57.23 54.25 2.98
9b 57.80 54.62 3.18
9c 67.30 64.16 3.14
10a 56.90 49.54 7.36
10b 48.27 46.90 1.37
TABLE 2. BRIGHTNESS EVALUATION

Durability
To meet Library specifications, machine-made, text weight paper must be able to withstand at least 100 folds under the above testing conditions. One would expect quality handmade papers to be as strong or stronger than similar weight machine-made papers.
To interpret our results, one must take into account the weight, thickness, furnish, and sizing of each papers.8 An examination of the data in Table 3 shows that six of the nineteen papers did not meet the minimum requirement of 100 folds before aging. The low fold endurance values for these papers are an indication of their poor fiber and sheet strength; the low SD values indicate they are homogeneous. The remaining papers are considered quite strong even when factoring in high SD values, the latter indicating greater variability in the samples.
Permanence
There was no significant loss of strength after aging in any of the initially durable papers that retained strength well beyond the minimum of 100 folds. The results of the before- and after-aging fold endurance tests of several papers show high variability and have a high standard of deviation. This variability can be due to any of the following: the greater number of fibers that settle on either side of the ribs compared to the thinner areas in-between typical of antique-laid mould design, fiber clumping, or inclusions in the sheets.
The comparison of before- and after-aging fold endurance numbers with such high variability prevents a precise prediction of permanence. Readers are strongly recommended to consider those papers as “high quality” that maintain a high fold endurance after aging even if there is a significant drop in fold endurance, if there is an accompanying high SD. For example, even though test paper 10a, an antique laid paper, exhibits a large drop in fold endurance after aging, if SD calculations are applied, the fold endurance retention could be as high as 100%.9
Summary of Test Results
The results suggest that most of these papers would be considered appropriate for conservation work with the exception of those with fold endurance values under 100 and those that darkened significantly. The remaining papers all tested well for durability as well as retention of pH and brightness readings after aging, which are indications of permanence.
The pH of all of the papers were within the near neutral to alkaline range both before and after aging indicating that most papers will not become acidic very soon. Some loss of brightness is inevitable after aging, and papers that changed more than 10 points of brightness after aging should be not be considered for use in conservation.
A few papers new to conservators did change over 10 points, which suggests that the furnish, the water used in production, and/or some other element of the paper or pulp (such as bleach used in the preparation of the fiber) are contributing to its discoloration. The fact that the pH of the same papers did not drop significantly suggests that further investigation into cause of discoloration is necessary. A few papers submitted by those papermakers previously unknown to us were too heavy for use as typical conservation endpapers (over 130 g/m2). These papers also did not have the durability to attain the required minimum 100 folds.
Many of the papers met and exceeded the minimum requirement of 100 folds tested with a 1.0 kg weight for machine-made papers both before and after aging, which means the papers are both durable and have the quality of “permanence.” It is helpful to keep in mind that once a paper has met a minimum strength requirement, there are many other factors to take into account, such as flexibility and surface strength when evaluating a paper. The best choice is not always the strongest paper.
Matching Workshops Reveal Priorities
To evaluate the aesthetic, working, and handling qualities of handmade endpapers, five matching workshops have been held since the inception of the Endpaper Project.10 Matching workshops help conservators identify which papers work best, what paper types are needed, and to refine eye, hand, and selection criteria. Several observations about conservators’ priorities have been drawn from the results of these sessions, the most recent held in July 2004 at the Library of Congress. In preparation for the session, the authors examined more than 200 early European books from the Library of Congress Law Library’s Rare Book Collection. This survey began to show that regional traits could be more distinctive than period of manufacture. There were many instances when two papers were very similar—mould, formation quality, color, and surface texture—except that the dates of manufacture spanned more than 100 years. Consequently, the twenty-eight books selected for the matching workshop embodied characteristics specific to  region and date, and were sorted by imprint country, which is not necessarily where the paper was made, and then by period. The two predominant paper types found in our survey have been referred to as “fine to coarse white” papers and “brown” papers. This does not refer particularly to color, but rather to the furnish and character of the paper. John Krill describes these categories in his book, English Artists’ Papers: With a variety in qualities of white rags, it is not surprising that white paper was available in different grades; the most common were “super-fine,” “fine” and “coarse.”...”White” then referred simply to the quality of the furnish, and the rags themselves were usually a cream or off-white color.11
However, there is that marvelous group of “fine white” papers, particularly from Spain and Italy that are very white in color.
The brown papers common to the collection of sixteenth- to eighteenth-century European books in the Law Library tended to have flecks, shives, and colored fibers. According to Krill, Brown paper came from the use of soiled, coarser and poorer papermaking fibers and from less refined or even inferior papermaking procedures, such as incomplete beating....Wool and silk appear in a rag list of 1682 along with hemp, flax, and cotton cordage....Like the white, brown paper was not of necessity brown in color. Today, we might describe one of these brown papers as being a light or dark buff, beige, gray or brown depending upon the color of its furnish.12
All of the matching workshops were similarly designed. Participants were asked to first describe several historical paper samples using a given list of categories. Then the participants paired contemporary handmade papers with old books and prioritized their reasons for choosing that particular paper. The choice categories were the same as the description, except that color was considered separately due to the overwhelming tendency for it to dominate. Color of the sheet can be changed through toning by the conservato if it is light enough and without too strong a Imprint 1783. Page in transmitted light showing "fine white" qualities: free of inclusions with even fiber dispersion and formation. color bias. External sizing is the only other paper quality that, theoretically, can be modified by the conservator.
When evaluating the aesthetic and handling qualities of the papers, the overwhelming response was that the top three priorities were hard to separate because they are interrelated. Weight and surface texture competed for the first priority, but look-through factored in as well. The preferences of conservators are listed in the following order based on their comments after the workshop.
Basic Requirement
Weight & Drape: Both are considered a basic conservation requirement because the purpose of endpapers is to provide protection in the transition from cover to textblock. If a paper is too heavy or stiff, it will impose upon the adjacent pages affecting the function of the text block. Endpapers that flow with the text block will support the text well.
Top Priorities
Surface texture: This is a significant aesthetic factor because it reflects historical evidence pertaining to the character of the mould and felts. A strong impression of a modern pressing board can render a paper incompatible. Look-through: Seen in transmitted light, this is a particular consideration when it reveals antique-laid and chain patterns and the fiber dispersion characteristic of early papers.
Secondary and Desirable Priorities
Opacity: Historical book papers often reveal the image of text from preceding pages due to varying degrees of translucency. Some of the sample papers were too opaque to consider for use with the earlier dated books. Sizing: The feel of the paper surface often influences this choice: when too soft, strength and abrasion resistance is compromised, and when too hard, the flow and drape is affected. Although external sizing is an important strength characteristic, it can be adjusted by a conservator either by adding gelatin sizing to make the sheet harder or softening a sheet by washing out gelatin sizing; no type of internal sizing, however, is adjustable.
Color: Considered Separately
Color: This characteristic is considered separately due to the overwhelming tendency for it to dominate selection priorities. If a paper is light enough to tone, it can be considered for a wider range of books. In the matching workshops, it was observed that the papers which tend more toward the pinkish, warm direction were chosen more often than the papers with a cooler, greener tone. It was agreed that shives and colored fiber characteristics could not be achieved through toning, and therefore there is a definite need for more papers with those characteristics. In the 2004 paper samples that had flecks, it was felt that the “inclusions” were too large and distinctive.
Conclusion
The goal of this project was to look at the papers available today and evaluate them for conservation using the four basic categories of permanence, durability, working properties, and aesthetic characteristics, as well as to identify conservators’ preferred paper qualities. We are very happy to report that most of the submitted papers meet permanence and durability requirements, that the list of hand papermakers making conservation-quality, western-style endpapers has grown, and that their products reflect their responsiveness and dedication to the craft. Examination of the aesthetic characteristics of these papers revealed a marvelous array of surface characteristics, look-through, and look-down (seen in raking light).
The papers tended to group themselves by century based on these aesthetic characteristics. Some papermakers tend to specialize in papers with a more uniform furnish, which is very useful for nineteenth- and early twentieth-century books. Some papermakers produce paper with more prominent fiber characteristics, and other papermak- ers make a wide variety of papers. This reflects a similar tendency we noted during the survey: regional traits span years. This tendency is a welcomed echo of the past and provides conservators with an array of papers that protect and support old books.
We learned that the first priority of conservators is the weight and drape, closely supported by the aesthetics of surface characteristics.
Having a paper that is easy to tone is extremely important. To satisfy the general needs of conservators, two predominant types of paper would be very useful to have available: fine-to-coarse, white “Italian printing papers” and “brown” papers. The brown papers tended to have either shives only, textile fibers only, or a mixture. For the earlier paper types, the challenge appears to be that of maintaining lively characteristics and strength of fiber while achieving a look-down and look-through featuring prominent, fine chain and laid lines, and felt texture. Although we have not yet seen a paper that matches the look-down, look-through, and handling characteristics of the Italian printing papers of the sixteenth to eighteenth centuries, many more papers exhibited aspects of these papers. After all, it took thirty-one years for the Society of Arts in the eighteenth century to award first prize to a papermaker who could reproduce the qualities of “French” copper-plate paper. We hope the efforts of the dedicated and talented papermakers continue their pursuit of creating papers with the character of old papers while using young materials.
We want to express our appreciation for the support from the papermaking mills and distributor Anton Glaser for their faith and willingness to provide paper samples at their expense. Some sent us many full sheets and sample books willingly and without remuneration. Sample books are extremely valuable to bookworkers, particularly when they are at least 5 x 4 inches in size. We also benefited from our increased awareness of the range of mills making quality handmade papers. As a result of the combined efforts of everyone involved, there has been a significant increase in the production of historically compatible handmade papers in the past two decades. We hope that the list of participants (including suppliers) and contact information will help bring this information to conservators and paper users, as well as encourage other papermakers have a “go.” Our collections will benefit, and we are very excited about the increasing range of papers available to the conservation community. Seventeenth-century book with  original endsheet protecting textblock from deteriorating cover materials while maintaining good joint strength and drape.
Notes
1. John Krill, English Artists’ Paper: Renaissance to Regency, 2d ed. (New Castle, Del.: Oak Knoll Press, 2002), 71–72.
2. Mary Wootton, Jesse Munn, and Terry Boone Wallis, “Observations Concerning the Characteristics of Handmade Paper: The Library of Congress Endpaper Project 1996,” Book and Paper Group Annual 15 (1996): 179–197.
3. Krill, English Artists’ Paper, 87.
4. Ibid., 88.
5. For a detailed account about the UICB Meeting, see “Quality Materials: Handmade Paper, Taking a Closer Look,” Book and Paper Group Annual 12 (1993): 61–65. Beginning in 1994, a follow- up round of paper orders was placed. Because the permanence and durability specifications were easily met by the majority of papermakers, the succeeding orders focused more on defining the aesthetic qualities and working properties. The papers produced from the orders showed that just as the Italian and French papermakers of old had reputations for specific types of papers, modern hand papermakers tend toward specialties as well. Some of the project papers (whose predominant sheet characteristics tended toward an opaque and uniform formation, regularly and widely spaced chain lines and relatively smooth surfaces) were well suited for use with nineteenth-century books. Others paired better with the earlier papers, whose surfaces typically show more fiber and felt characteristics, increased transparency, and evidence of variations of mould manufacture.
6. Those mills are Dieu Donné, Griffen Mill, Moulin de Verger, and Vinzani Hand Papermaking.
7. For more information on gelatin sizing, see Anne-Laurence Dupont, Gelatine Sizing of Paper and its Impact on the Degradation of Cellulose During Aging: A Study Using Size-exclusion Chromatography (London: Archetype, 2003).
8. As you review these numbers, understand that some of the papers requested for testing were too thick and stiff to use as conservation endpapers. We included them in the testing because of the effort the papermaker made to participate and the potential usefulness the paper might have for other types of projects.
9. When calculating to achieve a 95% confidence level, the high standard deviation for the fold endurance value after aging suggests that this value (799) could in fact be as high as 799 plus two times the standards deviation of 408. This would give a fold endurance value after aging of 1615, as compared to its initial average fold value of 1233 plus 2 times 218, the standard deviation, or 1669; that is, a fold endurance retention of 100%.
10.At the Iowa meeting, papermakers and conservators matched papers from Moulin de Verger, MacGregor and Vinzani, Sea Pen Press, UICB, Twinrocker, and Dieu Donné and then identified the first priority as surface texture; second, drape/flexibility, followed by weight. For participants in the Guild of BookWorkers Standards Seminar in Chicago 1999 and at the Folger in 2000 for a number of professional and amateur book conservators, fine binders, book artists, the first priority again was surface texture, followed by handle/drape, and third, look-through. At the July 2004 session, Library book conservation staff selected weight as the first priority, followed by surface texture and look-through.
11. Krill, English Artists’ Paper, 50, 52.
12.Ibid., 55–56 passim. [The English] are...still obliged to [France, Holland, and Genoa] for our best Papers....The French excel us in Writing-Paper, and the Genoese in Printing-Paper, from whom we take annually a great many Thousand Pounds worth of that Commodity.
—London Tradesman, 17471

‍