Collection Care blog

Behind the scenes with our conservators and scientists


Discover how we care for the British Library’s Collections by following our expert team of conservators and scientists. We take you behind the scenes into the Centre for Conservation and the Scientific Research Lab to share some of the projects we are working on. Read more

17 October 2016

From West to East: Conservation of the Chinese novel ‚ÄėDream of the Red Chamber‚Äô

By Heather Marshall

Six volumes arrived in the Conservation studio at the British Library. They are all bound in Western style bindings, quarter bound in dark blue leather and marbled paper, and they are very much in the format and style with which you will be familiar when looking at a traditional book.

One of the six volumes, open, showing the restricted opening, detached boards and spine piece.

I discovered that these volumes contain one of the greatest Chinese novels: the Dream of the Red Chamber (in Chinese: Hong lou meng), written by Cao Xueqin in the 18th century. This edition (British Library shelfmark: 15333.a.1) was made during the Qing dynasty, and it is dated 1811. It is a printed item.

The Dream of the Red Chamber is also called ‚ÄėThe Story of the Stone‚Äô (in Chinese: Shi tou ji) and it is one of China‚Äôs greatest classic novels. Written during the Qing dynasty, this work is widely recognised as one of the pinnacles of Chinese fiction. It narrates the story of two branches of a wealthy aristocratic family (the Jia family) and gives us a vivid account of the Chinese culture and society at that time, portraying funerals, rites, cuisine and medicine aspects.

However, it is relatively unknown in the West, despite the fact that the first translation into English was produced in 1812, just one year after the production of this Chinese copy. It is curious that the original item has at some point been rebound in Western style. The size of the novel is remarkable and when bound in Western style it arrived at 23 fascicules in 6 volumes! The current Penguin books edition is longer than War and Peace!

Books and other collection items in libraries often go through several transformations by being put into a number of different bindings in their several hundred year lives. This can vary from complete rebinding from one style to the next or an amalgamation of styles, through different repairs deemed necessary over the years.

Until recent years, when an item from East Asia was ingested in a Western library‚Äôs collection, it was usually rebound with Western style technique, including a hard cover with a spine, which usually contained the title in English. It was in fact believed that in this way the items would have been better stored on the shelves and the paper would have been more protected from dust or humidity. The same happened for the British Museum Library‚Äôs items from China (now the British Library‚Äôs Chinese collection).

The British Library Lead Curator for East Asian Collections, Sara Chiesura, has informed us that more recent and contemporary acquisitions are left with the original binding and put in a custom made box for shelving. This approach goes along with the current conservation trends, which tends to intervene as little as possible on the original items, and to adopt non-invasive techniques to stabilise them.

As a Book Conservator making decisions about the type of repairs needed to a fragile or damaged binding can be very complex and you will always need to consider the evidence which gives the story of the past. Do you keep a book in its current format and make repairs? Sometimes, in fact, even if the Western style binding is not ideal for East Asian items, the binding itself becomes part of the history of the item, and can sometimes reveal information about previous owners or collectors.

Would a new binding be better able to protect the book for the future and honour its history too instead?

The repair work on these Chinese items became therefore a rare case at the British Library. After a discussion with the curators, we decided that the ‚ÄėDream of the Red Chamber‚Äô would have been taken out of its current western style bindings and been put into a Chinese style thread bindings, housed in wrap around ‚ÄĚTao‚ÄĚ cases. Using strong, archival materials, the 6 volumes (23 fascicules in their original format) deserved the sympathy of a return to their origins and a non-Western approach was the best option for the original fragile, thin paper and the need for gentle opening and a secure binding.

As with many Chinese books that have been rebound in Western style, the tight hard binding in contrast to the fragile and this Chinese paper often does not work. In this case, the solid animal glue at the spine created a tight opening and did not allow the bindings to open far enough to see all the text. The detached boards and spines caused much damage as well.

Above: one volume after the animal glue was removed from the spine, clearly showing the divisions into fascicules of the original Chinese book and leaving the evidence of the more recent Western style binding.

Working with the item, I immediately noticed some original holes (in the paper text block), which gave me evidence of where the original Chinese thread sewing was. I could therefore reuse these original holes to recreate the sewing for the Chinese binding, tracing back the binding‚Äôs original state.

The so-called ‚ÄúThread Binding‚ÄĚ was the usual Chinese binding format in 1811. It is strong and the sewing rarely breaks. Thread bound fascicules in wrap around cases (known as ‚ÄúTao‚ÄĚ cases) would last indefinitely. They allow the flexible pages of each fascicule (which are double sheets, with the fold where the pages are turned) to open flat under their own weight and not be restricted by the spine.

The treatment stages:

Open fascicule showing repairs to be made with Japanese tissue.

Repairs were made (with wheat starch paste, a neutral adhesive and Japanese tissue) to the delicate spine edge which is still partly glued together by the animal glue used on the previous Western binding.

Detail showing remains of the original paper twists and their holes.

The paper twists are used to bind the loose leaves of each fascicule before it is sewn. They are made from a twist of paper pointed at one end and in this case inserted in two places into each fascicule. The Chinese style binding is very strong and the paper twists are often thought of as an ‚Äėinner binding‚Äô. They hold each fascicule together before sewing and can act as a back up to the sewing if it breaks.


Paper twists inserted.

Trimmed paper twists.

One stitched fascicule.

Each of the 23 fascicules were re-stitched using the 4 existing sewing holes, using a strong linen thin thread. Extra care was taken to re-use these holes by sewing with a very blunt needle. The needle then finds the hole and eases gently through without causing any damage to the fragile paper.


Traditionally thin silk (lined with paper) is folded around the corner of each fascicule to protect the corners and give another source of strength. In this particular case aero cotton was lined with Japanese paper.

Re-sewn fascicules (with corners and covers).


Flexible opening.


Cases (Tao style cases) were custom made to ensure the fascicules have a snug fit and give very good protection to the book. It is a big advantage of the Chinese style of binding that the elements work together but can also be separated. The wrap around case has to be exactly the correct size. If it is too tight the fascicules can bend, if it is too loose fascicules may fall out! These cases were often remade several times in the book‚Äôs life. In this case strong, archival materials have been used so correct storage and handling will ensure good protection for the fascicules and cases in the future.

The wrap-around case is made of board covered in cloth, lined with paper, and then fastened with bone pegs.


‚ÄúChinese Book‚ÄĚ is the expression used for the whole object. For this copy of the ‚ÄėDream of the Red Chamber‚Äô 23 individually bound fascicules were encased in 6 wrap around cases (keeping with the format of the 6 western style volumes which arrived in the studio).


With thanks to Sara Chiesura for input on this post.


Atwood, Catherine, ‚ÄėJapanese folded sheet books: Construction, materials and conservation‚Äô in The Paper Conservator, papers from the 10th anniversary conference, 14-18 April 1986, part 2

Helliwell, David ‚ÄėThe Repair and Binding of Old Chinese Books‚Äô in The East Asian Library Journal, Spring 1998, Vol. VIII, no. 1

Ikegami, Kojiro, ‚ÄėJapanese Bookbinding‚Äô adapted by Barbara B. Stephan, Wetherhill, London, 11th edition 2007

Wood, Michael, ‚ÄėWhy is China‚Äôs greatest novel virtually unknown in the west?, in The Guardian, 12th February 2016

04 October 2016

Conservation code cracking: finding meaning in hidden symbols

Everyone loves a good puzzle - and this was definitely found to be the case when Flavio Marzo, conservation team lead for the Qatar/British Library Project, sent an email around to colleagues with a series of mysterious symbols attached. What did they mean? The ancient doodles were uncovered by Flavio during conservation work of a manuscript from the Delhi Collection. Could the code be deciphered by British Library experts? Flavio reports.

Digitisation processes can be quite repetitive. Here at the British Library Qatar Digitisation Project we try to achieve the best we can and this means a lot of quality checks to ensure high levels of efficiency through standardised processes.

There is no difference in our approach to the conservation strand of these digitisation projects. Standardised treatments are applied daily to a great number of items that are processed and prepared to enable a good final result and to ensure safe handling of library material. Unsurprisingly, these items are unique and their content extremely fascinating.

Recently, as part of the material scoped for the second phase of the project, I had to repair a manuscript before imaging and uploading onto the Qatar Digital Library.

This manuscript contains two mathematical treaties bound together dating back to the beginning of the XVIII century. It was in need of conservation treatment because the sewing of a previous restoration attempt was impairing the opening making some of the text inaccessible and impossible to be imaged.

The manuscript is one of 36 scoped for the project belonging to the Delhi Collection. The Delhi Collection encompasses more than 2900 manuscripts stored across the British Library. The manuscripts are all that is left of the Imperial Mughal Library that was acquired by the English run Government in Delhi after the final destruction of the Delhi Red Fortress.

Those texts are now finally becoming available to readers for the first time thanks to the surrogates that we are uploading onto the Qatar Digital Library website. They all are in very poor condition and for this reason many of them have never been made available to readers in our reading rooms.

The prime concern for conservation when treating items is to find the right balance between the level of intervention necessary to make a book strong enough to be safely handled while still preserving the unique and invaluable physical features related to its history and use. These concerns are even more apparent for the Delhi Collection manuscripts since their history and the vicissitudes relating to their move to London are still quite confused.

This manuscript and the treatments carried out to conserve it are a very good example of how challenging it can be to decide what to do and where to stop, but also a very unique case of a fascinating discovery. When the little manuscript (measuring just 182 mm high, 120 mm long and only 7 mm thick) was brought to the studio its book block was detached from its cover.

Delhi Arabic 1925
At a certain point of its life the manuscript was restored and a new over-casted sewing was made to keep the loose, badly damaged pages together.

Construction of the book block
Diagram of the construction of the book block and the full leather cover.

This new sewing, even if achieving its purpose, was badly impairing the opening of the book making some of the marginal notes illegible.

Hidden annotation
Annotation disappearing into the gutter before (left) and after (right) the removal of the over-casted sewing passages.

We know from historical sources that these manuscripts were moved from Delhi to Calcutta for evaluation in the view to be then transferred to London; it was during this time that they were left neglected and befell extensive damage. It was most likely around the same time that the manuscripts were crudely restored and the present cover was applied to the text.

Changes are unavoidable during restoration processes, but conservation is committed to keeping this to a minimal level and always trying to preserve evidence of past treatments while keeping detailed treatment documentation.

After consultation with the curators it was decided to remove the over-casted sewing to improve the opening. The passage holes of the sewing thread were left undisturbed and even the passages of thread on the first and last sheets, not causing any harm to the book, were left and secured in place with wheat starch paste.

Unfortunately most of the pages of the two small manuscripts, repaired even before this last restoration campaign, became loose with no clear evidence of the original construction of the sections.

Many of the sheets were attached to each other at the inner joint and it was decided, after discussion with the curators, to keep this arrangement since no other evidence of thread passages was found. New joints were made with Japanese paper to create the bifolia for the quires.

Book block construction after conservation

Diagram of the construction of the manuscripts after conservation.

The three sections were sewn together with an unsupported sewing using the holes found in only three conjoint bifolia. This is represented in the previous diagram by continuous black lines.

The different layers of original spine lining were re-adhered as they were originally. The end leaves were re-connected to the book-block by gluing them along the spine edges to the first and last leaves of the book, as they were previously.

Only the front right paste down, originally attached to the inner face of the board was left detached and this was due to a very interesting discovery. During the conservation treatment of the end-leaves some hidden manuscript annotations came to light.

Right board
The right board of the cover was made from reused manuscript material. A couple of signatures (now under investigation) appeared, accompanied by what looked like a series of squiggles almost entirely hidden by the leather cover.

After a more careful examination it became clear that these symbols were actually much more than simple doodles. I decided to figure out how to decode them.

The code
The line of symbols emerged partially obscured by the turn in of the leather cover on the fore edge of the inner right board.

An email was written with images attached and it was sent to all colleagues working here at the 6th floor within the British Library/Qatar Partnership: an open invitation to participate in the decoding. Less than an hour later the mystery was solved. The squiggles were in fact a rebus - a puzzle where words are represented by pictures and letters, and its translation came out as: I see you but you cannot see me

The breakdown is shown below:


What an incredible and exciting discovery!

This really is the most appropriate motto to what I am always saying about conservation and the challenges in preserving evidence of historical clues: they are there, they look at you, but we are not necessarily able to see them.

The curator of the Arabic manuscript strand of the project, Bink Hallum, was the person who cracked most of the code. This demonstrates how tasks can be resolved through collaboration and sharing of expertise.

So many invisible pieces of information, during our careers, look at us from the items we handle everyday. We don't always have the necessary knowledge to see them, but surely we have the responsibility to preserve and convey them for posterity.

Flavio Marzo

26 September 2016

Fingerprints & their potential impact in relation to handling library collections

Back in early 2016, Terry Kent, a consultant specialising in forensic fingerprint analysis, contacted British Library Conservation to learn more about how we assess the impact of handling on our collections with reference to our use (or not) of gloves in the reading room. This was pertinent timing for us since we were on the cusp of refilming and updating our videos that provide instructions to library users about handling collection items. We invited Terry to the British Library to discuss the issue with us in more depth as part of our Continuous Improvement Programme.

In June, Terry Kent gave a presentation about the potential effect of fingerprints on paper artefacts at the ICON (Institute of Conservation) Conference ‚ÄėTurn and Face the Change‚Äô in Birmingham. Lively debate ensued. It became clear that there is some perception that the British Library has a blanket policy of no gloves - regardless. Not so, and in this blog post we would like to give brief insight, with Terry‚Äôs contribution, into how we assess and mitigate risks to collection items to enable access to and use of a vast and varied collection in a working research library (and how this then helps us form a handling policy).


Humanities reading room in the British Library.

By way of background:

  • The British Library has 12 reading rooms; 11 at St Pancras, London and 1 in Boston Spa, West Yorkshire.
  • These have 1200 reader desks and accommodate 400,000 reading room visits per year.
  • Reading rooms are divided into general and special collections, and focus on different subject areas (e.g., Humanities, Maps, Rare Books & Music and Science).
  • To request items readers need to register for a reader pass and sign the conditions of use.

Given this level of use the challenge is to balance the need to make items available to users while at the same time protecting them from further degradation and potential damage in order to ensure their longevity. Collection items are assigned different reading categories, based on factors including their age, condition, and value (historical, religious, cultural, etc.) which affects how and when they can be used, for example:

  • Which reading rooms they can be read in.
  • Whether there is a digital copy (or other surrogate) which should be referred to instead.
  • Whether readers need to provide additional information about why they need particular items before they can be issued.
  • Whether or not the items can be copied.
  • Whether readers need to sit at invigilated desks when they use the items or meet other conditions of use in order to use them. 

Where readers are using original items we encourage them to handle items as little as possible and with care as we know that even with careful handling collection items face risks.

Reading Room Placemat 2

‚ÄėHandling instructions‚Äô place mat on a desk in a reading room.

A range of different factors can damage collections and lead to loss - these are summarised in the figure below.

Of these ten categories, any risks presented by fingerprints due to sweat transfer would be covered by ‚ÄėContamination‚Äô (which also includes aggressive volatiles, pollutants and other damaging chemicals). Any potential risk to an item must be considered in light of a number of factors - the likelihood of it occurring, the extent and nature of damage it will cause if it does occur, the degree to which it will limit how the item can be used, and the measures that can be taken to limit or prevent it.

Risks do not exist in isolation, so responses to risks - such as the use, or not, of gloves - must be based on a comprehensive understanding of the nature of an item, its vulnerabilities and the requirements for its use by staff and readers. Furthermore, solutions to any such problems must not exacerbate other risks or introduce new ones.

Preventive Conservation

Risk factors.

Terry Kent writes,

A widely referenced paper, in the conservation field, and several forensic references, refer to fingerprint deposits consisting of 'over 98% water'. Recent analytical and theoretical studies of latent fingerprints, demonstrate that this figure is substantially in error. The deposit from a single human finger touch, whilst varying widely between individuals, is likely to contain less than 20% water and on average be about four micrograms of a mixture of amino acids, salts, primarily sodium and potassium chloride, fatty acids, squalene and many other trace compounds.

What is less well researched is the effects such deposits may have over time on substrates such as papers and textiles. We know that body soiling of fabrics will lead to yellow-brown staining, and fingerprint deposits on some papers will darken when heated (accelerated ageing using elevated temperatures); although it is unclear whether this will occur at lower temperatures over longer time periods.

There are other potentially negative effects of fingerprint deposits from a conservation standpoint; again not well researched, these include the effects of microbial or bacteriological activity on such deposits. There is also the potential of the deposit to attract and retain dust and other material from the environment.

The protective effect of hand washing, standard practice for many institutions and effective for the removal of transferred dirt, is less effective for the secretions which lead to fingerprints - it has been shown recently to be negated by natural replenishment of secretions in as little as five to ten minutes. So we need to consider the likely impact of these deposits on various substrates.

Reading Room use

Rare item being used, open access item being handled on shelf.


We are always looking at new evidence to challenge or support our current practices. Clearly fingerprints do have an impact on library and archive materials, although the extent of this is not yet clearly understood. The impact must be considered in light of other risks to the collection items given the context in which we work. Our policy is tailored to the requirements of individual items and the risks they face and the way they can be accessed and consulted. There is no one size fits all. Fragile, rare and significant items are subject to much tighter access and handling controls to minimise risks (including fingerprints) compared with items on open access. A core purpose of the British Library is to allow access to the national collection and our role in conservation is to manage that process as effectively and pragmatically as possible. We hope this blog post generates some thought and debate on the subject of handling and the impact of fingerprints. The collective authors plan to present their thoughts in a longer article in a future ‚ÄėICON News‚Äô.

Cordelia Rogerson, Paul Garside, Sarah Hamlyn with thanks to Terry Kent for co-writing this post.