Tag Archives: conservation treatment

Conservation of Piranesi bound etchings: treatment

by Michele Brown

In my previous post, I described the overall condition of Cornell’s set of bound Piranesi engravings. Now, let’s look at the treatment of these volumes.

Objectives:

The usual objectives for rare book restoration are compatibility, strength, usability, reversibility and longevity. Repairs should be sympathetic with the original binding, and also strong and flexible so the volume may once again be used. Since nothing lasts forever, reversibility is important for future repairs. For example, once the original spine has been removed, lining the back with tissue and paste provides a barrier against succeeding lining materials applied with pva, and enables removing them, if necessary. Materials should be archival and of high quality. When possible, original components should be saved or reused. I thought of these objectives when considering treatment for the Piranesi volumes.

Treatment:

Rare Books and Manuscripts curators Katherine Reagan (Ernest L. Stern Curator of Rare Books & Manuscripts) and Laurent Ferri (Curator of the pre-1800 Collections) agreed that due to their size and condition, the Piranesi volumes should be fully rebacked with leather. The original spines would be saved. Missing endbands would be replaced, but weak endbands would be retained and reinforced. Original sewing would be strengthened or repaired if necessary.

Mid brown book calf from J Hewit & Sons worked well for the new spines, and needed only mild toning to blend in with the original leather. I used Fiebing’s Pro Oil Leather Dye, now called Pro Dye, to mimic the mottled calf found on the bindings. (1)

The steps for rebacking were essentially the same for all of the volumes that were treated:

Photography and assessment. Due to their size, Simon Ingall of DCAPS (Digital Consulting and Production Service) photographed them “before” and “after” in the DCAPS studio.
Photographing Piranesi volume.
Removing the original spine. The original spine was lifted using a single-bevel knife. In most cases the spine was already peeling so this process was fairly easy.
Volume 4 spine before treatment.
Cleaning the back with a wheat starch paste poultice.Stiff wheat starch paste (Aytex P and filtered water) was applied to the spine, allowed to sit, and then scraped off with a blunt knife.
Removing spine linings.
Repairing/reinforcing weak sewing. In some cases, the sewing at the beginning and/or last sections was weak or broken. If so, the section was repaired and then resewn around the original sewing support using unbleached linen thread.
Resewn sections in volume 4.
Resewing on new cords. Volumes Vols 15 and 16 had extensive structural damage, including broken cords. Consequently, both volumes were pulled and then resewn with new cords before rebacking.
Lining the back with Hanji 1101 using wheat starch paste. This served to protect the spine from adhesives used later in the repair process, and helped consolidate weak sewing.
Volume 2 lined with tissue.
Lining the back with a strip of unbleached cotton stretch cloth, followed by by hand-made paper. Unbleached cotton was cut with slots to accommodate raised cords, and glued to the back using pva (Jade 403). After lifting the board leather, the cotton lining was extended across the joints and glued under the lifted leather using pva. This strengthened the joint and board attachment. For additional strength, panels of hand-made paper were glued between the cords

Back lined with unbleached cotton stretch cloth.

.Reinforcing existing, or replacing missing, endbands. If necessary, new endbands were sewn over cords (Clarkson BC cord) using silk twist.
Replacement endband in process.
Lifting the sides and paste downs to accommodate a new leather spine. Using a “lifting” knife, the sides were lifted a few millimeters away from the joints.
Leather being lifted at the joint.
Attaching the new spine. The new calfskin spine was pared and dyed and then attached using stiff wheat starch paste.
Putting down the lifted leather on the outside and inside of the boards. A mix of 50/50 wheat starch paste and pva (Jade 403) worked well.
Covering the inner joints with compatible, tinted tissue. Color Kozo from Hiromi Paper was attached using paste.
Lifted areas of the inner joints covered with Color Kozo.
Replacing the original spine. The edges of the original spine pieces were beveled using a Dremel tool and then attached using the 50/50 paste/pva mix.)
Consolidating deteriorated leather. Dry and deteriorated leather was treated with Cellugel (hydroxpropyl cellulose in isopropanol) and SC6000 leather conditioner.
Finished rebacking, volume 4.

The corners of most volumes were worn and damaged. These were repaired with 100% kozo Moriki (2). The leather of the original corners was lifted, the new corner material was attached using a mix of 50/50 wheat starch paste and pva, and then toned using Golden acrylics. This was usually done before starting the rebacking process.

As a result of the above repairs, the bindings were strengthened and made functional while their original components were retained. Cornell’s Piranesi is now more accessible to scholars.

A 2014 post from the Cornell in Rome’s blog provides an excellent overview of Piranesi’s work.

Notes:

(1) Traditional Fiebing’s dye is alcohol-based. Although Pro Dye is supposed to retain its color longer, it is known to fade. The Leather Conservation Centre in London has developed water-based dyes that when used with a fixative, have been shown to be more stable than other leather dyes. Unfortunately, these have been unavailable to US conservators for a while. Hewit’s also markets water-based leather dyes. I prefer Fiebing’s dyes, but use them sparingly. Skin Deep, Volume 30 offers a good overview of leather dyes.

(2) Moriki tissue is a convenient leather substitute for corner repair. It may be difficult to find 100% kozo Moriki, but we have a supply in the Cornell conservation lab.

Suppliers:

Gane Bros. (unbleached cotton stretch cloth)

J Hewit & Sons (calfskin)

Hiromi Paper (CK Color Kozo)

Polistini Conservation Material (Hanji)

Talas (Aytex P, Jade 403, SC6000, unbleached linen thread, silk twist)

University Products (Cellugel)

Conservation of Piranesi’s bound etchings

Leave a reply

by Michele Brown

The 18-volume set of bound etchings by the Italian artist, architect and print maker, Giovanni Battista Piranesi (1720-1778) is one of Cornell’s treasures. The Cornell set contains volumes presented in 1774 by Pope Clement XIV (1704-1774) to Henry Frederick, Duke of Cumberland (1745-1790) and brother of the King of England, during his residence in Rome; the volumes that contain works by other artists were added later in the 18^th century.

Beautifully bound in full calf (1) in the rococo style, the volumes have matching full gilt spines. Of interest is the difference between the decoration and sewing on Volumes 1 and 2, and the rest of the set. The first 2 volumes are decorated with an elaborate roll around the perimeters of the boards but are otherwise relatively plain; whereas the boards of the remaining volumes are heavily decorated. The first 2 volumes are sewn with recessed sewing and have fake bands on the spines. The remaining volumes are sewn on raised cords

The front board of volume 1, before treatment.

Front board of volume 3, showing elaborate decoration.

The front board of volume 3, before treatment.

Portions of the fake bands are missing on volume 2.

The raised cords are clearly visible under the peeling spine on volume 3. Note the similarities in spine tooling on both volumes.

When Willard Fiske acquired the volumes in 1873, he wrote the following to Ezra Cornell:

“Ithaca, New York, Feb. 19, 1873. Wednesday morning. [To] the Hon. Ezra Cornell. Dear Sir, I am rejoiced to tell you that I last evening received a note from our agent in London, saying that he had secured the Piranesi. All the parts are of the first and best edition — much superior to the second edition published by Piranesi’s son in the early part of this century. Appleton’s Cyclopaedia speaks of these works as “unique in art”. […] The work fills two cases, which arrived in New York [the] day before yesterday [and] we shall probably have them here in ten or twelve days. Respectfully yours, W. Fiske.” (2)

For many years, they were readily accessible to architecture students and were heavily used. Now, due to their fragile condition, access is by appointment only. Restoring them to a usable, working condition is the final project of my career at Cornell.

Their size and condition provide some interesting challenges for restoration and conservation.

Size:

The volumes are atlas folios. Volumes 11-18 measure 21.75” wide by 31” long. Volumes 1-10 are a more manageable 16.5” wide by 21.5” long, but still pose challenges for handling.

Condition:

Most of the bindings are in poor condition. In addition to cracked joints, worn corners, missing endbands, and losses to the covering material, the sewing components (cords and thread) on most of the bindings are extremely brittle.The cords pulled completely away from the spine of volume 16 when we opened the front board for photo documentation. The spine panels of several of the books are in various stages of peeling away from the backs, showing the backs have minimal or no linings. The spine photo of volume 3 (above) shows this clearly.

The sewing was weak or broken in some volumes, resulting in sections becoming detached.

Loose sewing thread indicates a weakness or break in the sewing.

Volumes 15 and 16 have significant water and mold damage. Two etchings had been removed from volume 15 and mounted separately. We decided not to reverse this treatment and reinsert them into the volume.

Water stains on volume 15, before treatment.

Previous repair

At some point, the joints and losses of all of the volumes had been treated with a plastic-like substance that matches the description of Liquick Leather, a pva product developed in the 1950s and used to repair leather bindings. (3) Unfortunately, this sort of quick “repair” was common in the 1950s and is seen on books in many libraries . It is impossible to remove without causing damage to the area it has covered. Although, some conservators have experimented with using solvents, it is often removed mechanically. Either way, the original leather surface is usually damaged.

Detail of the front board of volume 3, showing Liquick residue.

The spines and corners of Volumes 11, 12, 13 and 14 have been skillfully repaired with goatskin, but unfortunately, there is no record of who repaired these volumes or when. Until the 1980’s many of Cornell’s rare books were sent to contract binders for restoration. As was the custom during earlier periods of book restoration, missing tooling was replaced as part of the restoration process. Again, this was expertly done, apparently by someone trained in bookbinding and gold tooling.

Look for upcoming posts about the history of bookbinding and conservation at Cornell.

Proposed Treatment:

These volumes will be rebacked with calfskin; corners will be repaired with Moriki tissue. The missing endbands will be replaced; original spines will be retained, but missing tooling will not be replaced.

My next post will describe treatment of the volumes in more detail.

Thank you to Laurent Ferri, Curator of Pre-1800 Collections, Rare and Manuscript Collections, for his information and suggestions for this post.

Notes:

(1) The books are bound in smooth leather, which looks like calfskin. However, the leather separates and “peels” in a way that’s similar to sheepskin.

(2) In the Ezra Cornell Papers, 1-1-1, box 34, folder 11. Transcribed by Laurent Ferri.

(3) This pdf describing the benefits of Liquick Leather was posted in the discussion section in a blog post by Rita Udina: https://ritaudina.com/en/2015/02/17/bibliopaths-lacquer-binding-book-restoration/

Conservation treatment of Shakespeare’s Fourth Folio

Leave a reply

by Michele Brown

Cornell University Library is commemorating the 400th anniversary of Shakespeare’s death with a series of digital vignettes highlighting Cornell’s Shakespeare collection. Earlier this year, the Fourth Folio came to the conservation lab for treatment and evaluation.

The title page, before repairing the binding.

The binding on Cornell’s Fourth Folio appears to be contemporary with its 1685 printing date. It is a full calfskin binding with a gilt spine.The leather is tightly adhered to the back of the book (this is known as a tight back). When it came into the lab, the front board was detached, the back board was weakly attached, and a large piece of covering was missing from the back board. The tailband was mostly missing, with the core being held on with a few loops. The corners were worn with some losses. It had been previously repaired.

The front board of Shakespeare's Fourth folio, before treatment. The tailband is hanging off.

The front board of Shakespeare’s Fourth Folio, before treatment. Notice how the tailband is hanging off.

The tight bond between the original leather spine and the back of the book created difficulties for repairing the binding. Usually, when a book is rebacked the original spine is removed, the back is lined with cloth over tissue for strength and reversibility, and a new spine made from material similar to the original spine is applied. The original spine would then be stuck on over the new spine. The result is a book with reinforced sewing, new materials in all the areas that take the most stress, and yet with all of its original components retained.

The previous repair leather was applied only to the joints and head and tail caps, indicating perhaps, that the conservator had experienced difficulty removing the original spine.

The spine of the Fourth Folio, before treatment. The repair leather is visible at the edges and head of the spine.

Since the spine of a tight back binding is adhered directly to the pages of the book, it is not always possible to remove the spine without damaging it or the pages. When a book is historically significant with a contemporary binding, it is important to retain as many of its original components as possible. That is the case with our copy of the 4th Folio. We can tell our copy was previously repaired not only by the “new” leather on the outside joints, but also by the treatment of the first few pages and the inner joints. We don’t know when the earlier repair was done, but the repair leather had deteriorated and needed to be replaced. The front board had become detached and needed to be reattached.

I tried lifting the original spine after first facing it with Hanji paper using methyl cellulose paste as the adhesive. I was able to remove the second panel (containing the title label), and the tail panel, but could only lift the edges of the rest of the spine. The edges of the boards, both inside and outside, were easily lifted.

Once the leather was removed from the second and last panels, these areas were first lined with usu mino tissue using wheat starch paste, and then with unbleached cotton stretch cloth (from Gane Bros) using pva. The stretch cloth was cut wide enough to stretch over onto the boards under the lifted leather. A lining of Conservation Wove paper (from Katie Macgregor), using pva, made for a nice smooth surface on these 2 panels.

Ideally, all of the panels would receive this kind of reinforcement, but the spine adhesion was tenacious. How could I add strength to my repair? I decided to use joint tacketing on the front board. Joint tacketing consists of drilling several holes into the joint, angled so they come out on the shoulder of the spine. Two corresponding holes for each tacket are drilled into the board. Linen thread is passed through the joint to the spine and secured with a loop. The tails pass through the holes in the board. A square knot is tied to hold the threads in place. This holds the board to the spine in a way that is similar to its original attachment. The joint tacketing link above provides a detailed description of the procedure.

Before drilling the holes, a strip of Hanji paper was attached to the inside shoulder, one edge aligning with the shoulder, the rest extending towards the fore edge. This would be folded up over the linen thread loops towards the end of the treatment. Then, holes at 5 stations were drilled into the joint and 2 per each station were drilled into the board. At each station, the thread was passed through the shoulder, looped, and then passed through the board and tied in a square knot.

The front joint, with tackets and the strip of Hanji tissue.

This is a closer view of one of the tackets, showing the loop and square knot.

A new tailband was woven using silk twist over the original tailband core.

The finished tailband and cap.

New calfskin (from Hewit’s) was pared, dyed and attached to the second and bottom panels, extending across the spine and onto the boards. New calfskin strips were also added to the headcap and joints under the lifted spine edges and onto the boards under the lifted sides. The lifted board material was put down with wheat starch paste.

New leather attached to the joints and 2 panels.

This is a closer view of the tacket and new leather. The spine pieces have not been replaced at this point.

This is a closer view of the tacket and new leather. The spine pieces have not been replaced yet.

The corners were repaired with dyed calfskin or colored tissue. The Hanji strip in the front inner joint was pulled across the joint and adhered under the lifted paste-down with paste. Then, it was covered with CK color kozo. The back joint was also covered with CK color kozo. Wheat starch paste was used for these steps.

The finished front inner joint. The first few pages also had minor repairs.

The finished front inner joint. Losses on the fly leaf were replaced with color kozo from Hiromi, using paste.

The lifted spine panels and edges of the spine were re-attached using wheat starch paste. This hid the tacketing threads on the shoulder. The corners were repaired with dyed calfskin and toned kozo. The loss on the back board was patched with dyed calfskin.

The front board of the finished repair.

The finished spine, with the original pieces reattached.

As a result of this treatment, the book is much stronger and more easily handled. If the repair leather fails, the tackets should help keep the front board attached. The cloth liners across the 2 panels that were lifted will also provide extra support.

One of our goals in the conservation lab is to make the physical collections accessible for study and analysis. Here’s hoping that readers of Shakespeare will be able to enjoy Cornell’s copy of the Fourth Folio for another 400 years!

Mystery solved

2 Replies

by Michele Brown

Recently, “Ask a Librarian” received the following query:

“Hi, I have a question about the 1579 edition of Hill’s “profitable instruction of the perfite ordering of bees”, which is your “SF523 .H996 Phillips Beekeeping Collection”.

Over at Distributed Proofreaders we are trying to create a free e-book of this for Project Gutenberg.

We started with the scans from the Internet Archive / Biodiversity Heritage Library, which are of your copy. Your catalog entry says “56 [i.e.57] p. 19 cm.” My problem is that on the last page (the second 56) it has a catchword, and you can see the bleed through from the printing on the verso! So there have to be more pages. Plus, the table of contents says the second treatise has 8 (viij) chapters, and the last page is the end of chapter four.

If the whole of signature Kk was missing I could understand it, but the recto of at least the first leaf is there, so the verso has to be there too!

So I am asking for two things. Firstly, can someone look at the second page 56, signature Kk, and confirm that it continues on from there (I would assume for a full quarto signature).

Secondly, is there any way I can get images of the remaining pages, either by you rescanning it for the Biodiversity Heritage Library or by you sending me images directly?

I think you have the only copy of this that is complete (the British Library copy being shorter than yours) so I do not have a plan B here. Thanks in advance, Neil. “

Since this book is held in Mann Library Special Collections, the query went to Linda Stewart, Life Sciences Bibliographer and Special Collections Librarian. Linda then contacted the Conservation Lab.

Although the book in question was printed in 1579, it had a library-style binding typical of the 20^th century. The verso of the last page had a blank leaf stuck to it that appeared to be covering up text. This blank leaf matched the paper throughout the book, not the endpapers of the new binding.

Cloth binding added by a commercial binder, probably in the 20th century.

The title page.

The last page has a catchword, implying there is more to follow.

A blank page has been pasted to the verso of the previous page.

But, there is no more text. A blank page has been pasted to the verso of the last page.

We tried to see the obscured text by placing a light sheet behind the laminated page, but this was unsuccessful.

Since we were unable to read what might be behind the pasted blank sheet, we decided to disbind the book and soak the last page in filtered water in order to separate the layers.

First we removed the commercial binding; then we cleaned the spine with a wheat starch paste poultice. This gave us an opportunity to see the steps involved in the previous binding process.

Binders often used waste-paper in the bindery to construct spine linings.

Binders often used bindery waste to construct spine linings.

We used a wheat starch paste poultice to remove the old linings.

Under the paper linings, we discovered a cloth lining.

Spines are often lined in coarsely woven cloth for strength.

We could see that the new binding was put over the leather spine of an earlier binding. The book was oversewn before rebinding it.

We can see the remnants of the spine of the earlier binding under the new oversewing added by the commercial binder.

Remnants of an earlier leather spine with a gold tooled title.

The book had originally been sewn on three cords.

Removing all of the linings reveals the original sewing structure.

The sections are held together by thread passing through the center of each section.

Once the spine was clean, we separated the book into sections.

The sewing thread was cut and the sections were carefully separated.

The book, separated into sections.

Finally, we were ready to separate the last page from the blank page that was pasted to it.

The last page was soaked in a bath of filtered water.

The blank page was teased off gently while both pages were still wet.

The blank page came off easily and we discovered that the printing on the verso has nothing to do with the rest of the book. We can also see the instructions “Paste over this” written with what appears to be iron gall ink at the top of the page.

The printing on the verso is revealed.

How did this happen?

Thomas Hyll died around 1576; his books were reprinted several times. Could there have been some confusion in the print shop when this book was re- printed in 1579? Perhaps an examination of all of his work would show us exactly which book the verso printing belongs to.

Meanwhile, what is the next chapter for this book? The sections will be repaired. It will then be resewn and rebound in binding more sympathetic to its time. It will also receive a new custom-made clamshell box and then it will be returned to the E.F. Phillips Beekeeping Collection where it will remain available for scholars for decades to come.

Treatment of Caesalpini’s De Plantis Libri XVI, 1583

4 Replies

Original condition of the book. Here the text block is separated from the boards

The treatment Caesalpini’s De Plantis Libri XVI, 1583, stemmed from concern about the presence of mold spores. The book’s condition when it arrived was actually quite good. The sewing was still intact throughout the text block and it was still attached to the boards. Unfortunately, the fly leaves had separated from the text-block and were only just attached to the book’s cover. Someone had repaired the inner hinge in the front and back of the book and it was this repair that was keeping the fly leaf attached. Finally, after it was determined that the “mold” was not actually mold but soot and bacteria colonies, treatment could begin.

First, I began with dry cleaning each page with an Absorene Dirt Eraser sponge, also known as a dry cleaning or soot sponge, which we purchased from Gaylord. This sponge is designed to attract any surface dirt that may be present, though not necessarily visible. As treatment progressed, I began to find that the sponge was not picking up much dirt. This seems to indicate that the pages are actually quite clean on the surface, and that the smudges of soot have become embedded in the paper over time.

Example of Worm Damage

The next step was to reattach the fly leaves to the text block. I used usumino, which is a Japanese paper (from Hiromi) and wheat starch paste (from Talas) to attach the fly leaf to the text block. In order to repair the inner hinge in the front and back of the book, I used a thicker Japanese tissue called sekishu (also from Hiromi). This will help protect the the board attachment. As mentioned in the previous post, there was worm damage throughout the book. This was apparent on the fly leaves so I used a colored kozo Japanese tissue to repair any worm damage and to consolidate some areas where the damage was significant.

Reattached boards and repaired worm damage

Finally, we wanted to try to clean the vellum covers with something a little stronger than a dry cleaning sponge but not so invasive that it damaged the integrity of the materials. After consulting with Michele Brown, I used cheese cloth dampened with ethyl alcohol to gently rub across the cover. Both boards and spine were cleaned in this manner. Though the change was not visible on the surface of the book, the white cheese cloth did become gray with dirt.

Front cover after alcohol treatment.

With the treatment of this book complete, I constructed a custom-fit clam-shell box with Velcro tabs for this item to live in for the foreseeable future.

New custom made clam-shell box

Rebinding Newton’s Principia

Leave a reply

By Michele Brown

This year, the curator of the History of Science Collections requested a new binding for Cornell’s copy of the first edition (1687) of Philosophiæ Naturalis Principia Mathematica by Isaac Newton, generally referred to as Newton’s Principia.

It had been rebound in the 1970’s in a way that was unsympathetic with the date of its printing. This by itself would not normally be a reason to rebind it, but the sewing had broken, there were water stains throughout, and mold growth had damaged the edges of many pages. As a whole, the volume was fragile.

It is used frequently for teaching, and its deteriorated condition made this difficult.

The curator was anxious for this work to have a more appropriate binding. Since the book needed to be resewn, he agreed that washing the pages could improve their flexibility and increase their longevity.

Washing and paper repair

The pages were washed using filtered water alkalinized to a pH of 8.5. They were deacidified using a calcium hydroxide solution, and then re-sized with 1% gelatin solution. We were concerned with lightening the paper excessively and losing any of the manuscript notations. Washing and resizing removed much of the staining and returned flexibility to the pages. In the photo above, the pages on the left have been washed.

Each folio was reinforced with thin usumino (from Hiromi) using wheat starch paste (Aytex P from Talas and filtered water). Paper edges damaged by mold were reinforced with thin usumino or tengujo tissue (also from Hiromi). It was resewn onto 5 raised cords, reusing the original sewing holes.

Binding

Unfortunately, there is no record of the original binding. The original endbands were also gone—replaced with machine-made endbands in the 1970s—but since a fragment of green thread remained I sewed on new green and white endbands.

It was covered as a tight back using fair calf (from Hewit’s) that was dyed using Hewit’s water-soluble aniline dye and Fiebing’s leather dyes.

We decided a Cambridge panel binding was appropriate for this time period. We found two valuable resources for deciding on the design. The Massachusetts Center for Renaissance Studies had several images of bindings for this period. English Bookbinding Styles 1450-1800 by David Pearson (British Library and Oak Knoll Press, New Castle: 2005) was also a good source.

Terry Buckley’s 1996 Guild of Book Workers presentation on leather staining and dyeing provided valuable guidance for dyeing the leather for the Cambridge panel design.

The new binding is more attractive and opens easily, making it more useful as a teaching tool. It is also in a better condition for digitizing, should that option be considered in the future.

Cornell University Library Conservation

Cornell University Library Conservation is a part of Digitization and Conservation Services. We provide preservation assistance for all Cornell Library collections, including consulting, conservation treatment, and collections maintenance and management strategy.

Tag Archives: conservation treatment

Conservation of Piranesi bound etchings: treatment

Conservation of Piranesi’s bound etchings

Conservation treatment of Shakespeare’s Fourth Folio

Mystery solved

Treatment of Caesalpini’s De Plantis Libri XVI, 1583

Rebinding Newton’s Principia