How to make an Iron Age shield (out of bark)

Two thousand three hundred years ago a shield was made from the bark of a willow tree. It ended up in the bottom of a wet pit in Enderby, Leicestershire, where it remained until 2015, when it was discovered during excavations by the University of Leicester Archaeological Services (ULAS). This fragile artefact has since been preserved with Polyethylene Glycol, known as PEG, and is now part of the British Museum collection thanks to the generosity of the landowner, Everards of Leicestershire. It is on display in Room 50, Britain and Europe 800 BC–AD 43.

The shield is the only known example of its kind in Europe. Before 2015, pre-Roman shields had been found made of wood and animal skins, but it was not known that they were also made of bark. Most of the evidence for shields comes from the metal details added to organic backings: bronze fronts, rims, bosses (the protruding central part), handles and decorative details. To find an ancient shield made entirely of plant materials is incredibly special and important for understanding weaponry and craft skills in the first millennium BC. Ancient wood specialist Mike Bamforth shares his thoughts on the use of bark…

The survival of ancient bark artefacts in Britain is scant at best, being limited to a few stitched bark vessels and the odd well lining. However, internationally there is abundant evidence for a wealth of uses from beautiful, folded bark containers to full-on bark canoes.

The shield is almost complete, but missing parts of the body, the edge and half of the handle. Most of the boss, a small part of the rim and a large part of the body do survive. Examining this object raised questions about how it was made, whether parts of the design had a practical function, and how effective it was in protecting a person.  

To test some theories, the team decided to reproduce the shield using the same materials, keeping as close to original techniques as possible. This research was led by Matt Beamish of ULAS and the first attempt in 2018 raised more questions than answers. So with a revised design in mind, we tried further experiments in 2021 and in 2022.

Vital to the whole project were three people: Matt Beamish, Diederik Pomstra and Paul Windridge. Matt is a Project Manager at ULAS. His archaeological knowledge of, and fascination with, the shield has driven this project forward since it was discovered in 2015. Diederik joined us from the Netherlands, bringing his incredible skill, knowledge in working with plant materials and calm dedication. Paul is a highly skilled woodsman, maker and brilliant storyteller. Paul and Matt also contributed their knowledge of the local ecology and arranged access to materials and a place to make the shields.

I had the pleasure of becoming part of this dedicated and enthusiastic team when I became a curator at the British Museum in September 2021. We were supported on site by Mike Bamforth, a specialist in Ancient Wood who examined and interpreted the original find alongside specialists at York Archaeological Trust. Nick Harris and Andrew Parry from the production team at the British Museum kindly joined us to record the process with videos and sound, encouraging us all to share our thoughts and theories on camera.  

With us in spirit were a team of highly skilled and knowledgeable scientists, conservators and archaeologists from the British Museum, York Archaeological Trust and ULAS, including Caroline Cartwright who identified the bark was indeed willow, and Barbara Wills, who picked out the finer details of how the basketry boss was made. 

The team gathered in Leicestershire, only a few miles from Enderby where the shield was discovered, armed with high-definition photos, videos and scaled drawings of the original shield. Thirty-six hours later we had made a viable, physically and visually appealing recreation of the shield. 

First, supplies of willow bark, lime bast, crab apple wood and hazel sticks used in the original shield had to be gathered and prepared. Late spring to early summer is the best time for harvesting the willow bark, as later in the year the tree holds onto its protective covering so it cannot be peeled away easily. In late spring 2021, pandemic restrictions meant the team couldn't gather to make the shield so, instead, the bark was harvested and frozen to preserve it. Not an authentic process but a necessary adaptation. For the final attempt in June 2022 the bark was harvested straight from the tree the day before making the shield.

Close study of the basketry boss showed it was made with willow foundation rods wound into a domed shape, starting from the middle, with lime bast (Tilia species) stitched over the willow rods. Diederik wove the boss at home in the Netherlands, using notes, photos and high-definition videos of the original shield boss. It was such a thrill to see the recreated version reflecting all the details that could be seen in the original.

The harvested bark was kept wet in a nearby lake ready for working. A template was used as a guide for cutting the right shape, which was based on the dimensions of the surviving parts of the shield and comparison to other contemporary shields in wood and metal.

The bark was used inside-out so that the rough outer bark forms the inside of the shield against the body of the bearer, and the smooth inner bark forms the exterior of the shield facing their opponent. This intentional design may have helped weapons glance off the shield rather than snag on the rough, fissured outer bark. The bark had to be encouraged to bend against its natural curve around the tree. In the original shield, thin wood batons referred to as laths were found inserted into the bark board. We had wondered if these were to help the bark to bend against its natural curvature and this proved to be the case. When all the laths were inserted into the reproduction, the bark lay flat instead of in its natural curve, seemingly increasing its rigidity.

Before cutting out the hole for the boss, Diederik and Paul began the tricky process of carving the strips of crab apple wood to form the laths and inserting these between the layers in the bark. In 2021 we thought the design used long thin strips of wood passed through loops to create a woven effect. These proved too weak to be of any use. Looking again at the original shield, we realised that the laths were actually short, thick and pushed into narrow tunnels cut into the bark.

With the laths successfully inserted, the bark around the rim was attached, then the basketry boss was added and finally the round wood handle. Only a small section of the rim had survived on the original shield, so a good deal of time was spent working out how it was attached. It is made of split, roundwood hazel. Single stitches proved inadequate at holding the rim to the edge of the shield. Scrutiny of the original revealed small groups of about four slits at intervals around the edge of the shield. Flat lime bast stitches were still present in some of the slits, which must be the remnants of the stitches that held the rim on. We were still uncertain of the efficacy of this design so added some pins made from hawthorn thorns for extra security. As the shield dried, we found that our plan was still not entirely successful because the rim had pulled away from the edge of the board.

The boss was stitched to the shield with twisted lime bast cord. Diederik replicated this process using a bone needle similar to Iron Age examples to sew a running stitch through holes made with a pointed tool called an 'awl'.  

Next came the handle. Diederik removed the bark from a poplar branch and swiftly ran it through a flame to char it. It was then carved with notched and sloped ends and stitched to the shield around the notches. The sloped ends meant the attached handle pulled the bark into a shallow curve towards the holder, giving it more strength. 

Ta da! A bark shield is made. At this stage it was quite heavy with moisture, which made it difficult for the shield-bearer to hold close to the body, but once dry it became light and manageable. Drying proved the biggest challenge in the cool, damp environs of the East Midlands. The shield needed to be in a relatively warm and dry place with good airflow around it to dry without distorting or going mouldy. Perhaps the original shield was hung from the rafters of a roundhouse to dry in the ambient warmth from the hearth. This process has been added to the to-do list for future attempts.

Finally, we decorated the shield like the original. Fine scored lines and traces of red colouring on the front were analysed during the early stages of conservation. The design was reproduced with paint made from ochre and boiled linseed oil, materials that would have been available 2,300 years ago in Britain. Paul and Matt scored and painted the dried shield to achieve the sharp lines seen on the original. The result is a geometric pattern quite different to the swirling designs found on contemporary bronze-fronted shields like those from Battersea and Witham, also on display in Room 50.

Plans are afoot to test the capabilities of the reproductions with support from the Engineering department at the University of Leicester to confirm whether any of the holes in the original are weapon damage. And we might just try making another one. 

Everards of Leicestershire, University of Leicester Archaeological Services, The British Museum Research Grant and Mike Winterton and family.

Matt Beamish, Diederik Pomstra, Paul Windridge, Mike Bamforth, Nick Harris and Andrew Parry.

Julia Farley, Barbara Wills, Caroline Cartwright, JD Hill, Melanie Giles, Matt Hitchcock, Steven Allen, Gareth Beale, Michael Biggs, Konstantinos Chatzipanagis, Derek Hamilton, Claire Robinson, Luke Spindler, Chloe Watson, Heidi Addison, Mags Felter, Chloe Watson, Penelope Walton Rogers and Rachel Crellin.