Technology reveals hidden history

Advances in X-ray microtomography technology originally developed for dentistry could allow historians to peer into centuries-old parchment scrolls too damaged to unroll, unlocking the secrets of the past.

Prof Graham Davis, senior lecturer in the Institute of Dentistry at Queen Mary, University of London (QMUL) is leading the team developing scanning technology that will make such discoveries possible.

Over time, parchment breaks down into gelatine, which sticks layers of a rolled scroll together. This is exacerbated by the common use of iron gall ink, which accelerates the degradation of the substratum of collagen, tending to make the parchment more brittle and prone to gelatinisation. Many ancient documents in libraries and museums remain unread due to the threat of damage if they are unrolled.

Microtomography was originally developed for dentistry by Jim Elliott in the early 1980s to detect differences in mineral concentrations in three dimensions. It soon became apparent that it had a number of different applications, and has since been used for everything from examining other biological systems in hospitals to analysing aircraft materials.

Dental technology remains the most advanced in quantifying the grey-scale of images, making it the ideal base for reading the scrolls. Prof Davis worked alongside Elliott and is now leading a project to customise microtomographic technology for the purpose.

The project sprang from a chance encounter with Prof Tim Wess, head of the school of optometry and vision sciences at Cardiff University, who was interested in imaging ink on parchment before some of these precious documents become unreadable.

Work at Kentucky University had demonstrated the feasibility of the methodology on a small sample, but did not have the necessary contrast resolution to take it further.

‘The problem is, as you move from small samples to larger samples, you need much greater contrast resolution just to be able to see the same things,’ said Davis. ‘You tend to get a grainy, noisy image from them, and ring artefacts [concentric rings round objects]. On most commercial systems, they’re not visible as they have got ways to deal with them but when you boost the contrast to try and find these very small differences, you begin to see them.’

Davis’ team is developing a scanner with improved scanning geometry designed to be immune to these ring artefacts and to give the best possibly contrast resolution.

‘Because we want this high-contrast resolution, we have to use much longer X-ray exposures,’ said Davis. ‘I’ve just finished scanning a sample on our existing system that took six days, and we anticipate scrolls taking a lot longer than that.’

The team plans to build a scanner that is an upgraded version of existing designs using commercial components, then develop its own software and scanning strategies.

Davis said the main problem to be overcome is signal to noise. ‘For a given size of specimen, there’s a fourth-order relationship between resolution and required X-ray exposure, so if you want to double the resolution for the same size specimen, you need a four-fold increase in X-ray exposure, 10 times increase in resolution, and 10,000 times increase in exposure, which becomes absurd.’

The iron in the ink makes it visible during a scan, which would not work with carbon-based inks. ‘The irony, no pun intended, is that iron gall ink is one of the agents that is causing parchment to deteriorate,’ said Wess.

The QMUL group will produce a 3D volume dataset representing a whole scroll in 3D, then the computing department at Cardiff will take that dataset and virtually unroll it. Target parchments will come from important collections around the UK. ‘These parchments are so damaged that the archives that have them don’t know what is on them or how to recover any information from them — this is almost a last chance for them,’ said Wess.

‘We’re working with organisations such as the National Archive at the Bodleian, the Welsh National Library and the British Library.’

The EPSRC-funded project runs until the end of 2012, by the end of which the team hopes to have gathered some useful historical information from otherwise unreadable scrolls. It could also uncover erased writing on rolled palimpsest — reused parchment.

Initial samples will ideally be those thought to be of little historical relevance in case there is any element of risk with the technique, though the researchers will also carry out an extensive biochemical study to assess potential damage.

Wess said an additional benefit will be the unusual exchange of information between two fields of science. ‘I like the way medical science informs cultural heritage science in this project,’ he said. ‘We can then take the techniques and equipment we develop and offer high-resolution tomography back to other fields.’

Berenice Baker