When we set out at the start of the season we were fairly confident that this would be the last year of excavation in Trench 1. There seemed only a limited number of questions left to be answered and since we had exposed boulder clay and bedrock over the full area of the trench at the end of last season, it’s not as if there was much more to dig. Trench 1 always seems to have more secrets to reveal though, so the appearance of more post-holes and small features, due to slow weathering, and the difficulty of answering one of our last research goals, frustrated us in the end. I can’t say I am sad to have a little more to do next year.
We started the year with two outstanding research goals. The first was to complete the investigation of the corn-drying kiln (a kiln or oven used to dry damp cereal for use or storage) in the north-west corner of the trench and the second was to further investigate the timber defensive rampart leading to St Oswald’s gate and also to see if we could also trace evidence of the rampart along the northern limit of the perimeter of the site. It presence here had long been speculated on but never proved.
The kiln has been a fascinating feature, constructed from fired clay set around laths of timber that would have at one point supported a domed top, fragments of which we find broken and forming a substantial part of the fill of the bowl. It also contained a considerable quantity of charred cereal grain that will help us learn a little more about the diet of the inhabitants. The kiln was likely used to dry damp grain on its way into the fortress for long term storage. This would explain its location close to the entrance at St Oswald’s Gate. At the moment we have a series of potential dates for the kiln based on archaeomagnetic samples. The most likely period of its use is the late Anglo-Saxon period, but a radiocarbon date will be needed to confirm this.
One aspect that has intrigued us about the kiln for a while is that it underlies the earliest phase of high medieval wall, which is probably 12th century in date. The edge of the rock plateau is close by so when the kiln was in use there must either have been a much thinner stone wall or no defensive feature at all, given that a timber rampart would have been at a severe risk of burning down if right next to a kiln. We know that we have at least two phases of timber defence on the west side and that the kiln cuts and post-dates the latest of these. This summer we have further investigated this, excavating more of the later rubble foundation for a timber sill beam to reveal a series of post-holes beneath it. On the same alignment. The post-holes are spaced about 0.5m apart and, like the sill beam, almost certainly represent the inner face of a box rampart.
We have long speculated that the rampart likely continued around the fortress perimeter on the north side and have even seen what appeared to be a linear formed from a stony clay extending in the right place and on the right alignment for this. We began the excavation of a slop through the boulder clay across the trench to try and investigate this idea. We assumed that if the spread was to represent the former fill of a timber rampart then we would have re-deposited boulder clay on a boulder clay surface that had not been disturbed. The theory seemed reasonable but proving it has turned out to be much harder than thought. The boulder clay turns out to be more variable and mixed than we thought from its surface and, not at all surprisingly, hard to dig! By the end of the season we had a number of potential post-holes, a series of enigmatic patches of apparently burnt clay and no clear buried surface. Unable to make firm conclusion just what this evidence means, we will be conducting a limited amount of further excavation next year. Hopefully we will be able to get some secure answers, but if not then at least we will have done everything we could to reveal this last secret.