Readers of Vidimus will no doubt be aware of the recent exciting news that glass panels from the “Ancestors Series” at Canterbury Cathedral may be the oldest existing stained glass windows in England. This series was created and installed in the Cathedral, beginning in the late 1170s (after a fire in 1174) and continuing through to 1220. It has recently been discovered that a panel from this series is much older than originally thought, dating back to the period 1130-1160. This discovery was made by a team of scientists from UCL and conservators from Canterbury Cathedral using a technique known as portable x-ray fluorescence (pXRF), with an approach developed for the purpose by Dr Laura Ware Adlington, who was then a PhD student at the UCL Institute of Archaeology. Their results confirm research carried out in the 1980’s by the art-historian Professor Madeline Caviness who suggested that four of the panels installed in the 13th century were stylistically much older. Full details of the press release issued by UCL can be accessed here and the detailed report on the work may be freely downloaded from the journal Heritage.
Readers may be interested to know how pXRF can help date glass. It is a method of analysis which determines the chemical elements and their concentrations in a material. Medieval glass was composed of two main ingredients, a flux (typically the ash of inland trees or bracken) and silica, usually in the form of sand. The compositions of ash and sand vary depending on where the glass was made, and the proportions used in glass making changed over time and differed between workshops. This opens up the possibility of dating or sourcing glass from its chemical composition. While a chronology of glass chemical compositions has been established for the post-medieval period in England (see works by Dr David Dungworth), similar work is still on-going for medieval glass.
It is not straightforward, however. Handheld pXRF works by directing a beam of x-rays at the glass of interest. The x-rays interact with the glass surface and send out a spectrum of new fluorescent x-rays which are diagnostic of the elemental make-up of the glass. Because pXRF is non-destructive and portable, it is well suited for windows but they present a number of difficulties: the presence of weathering or dirt on the glass surface, the difficulties in positioning the instrument against the glass surface because of the protruding lead cames, and the fact that air absorbs the X-rays of light materials.
The authors overcame these issues by devising a special 3D-printed nose which allowed the positioning of the pXRF instrument between the cames and by concentrating on a few diagnostic elements which were characteristic of the sand and the flux, which can travel through air and which were generated below the surface of the glass, minimising the problems due to weathering. This approach allows the identification of glass groups that are related to different recipes, sources, and/or date. The authors were therefore able to analyse and compare panels from different stages of the Canterbury construction programme and concluded that the Nathan panel, which was installed in the early thirteenth century, has a chemical composition resembling the earlier panels, and is therefore consistent with Caviness’s interpretation.
With scaffolding and a suitable attachment, a pXRF instrument, which in looks and weight is rather like a hand-held electric drill (without the drill bit), can be held against a stained glass window. With appropriate protection for the operator, a number of readings can be safely taken from all parts of the window and compared with analyses from other windows to establish the relationships between them.