It has almost been a year now since the excavations took place on the suspected Roman Fort at Budlake Farm on the Killerton Estate, Devon.
Over the last few months, Nancy has been sending the Budlake finds off to specialists and their reports will enable us to tell the story of the site.
Was it a Roman fort? When was it occupied and abandoned?
If there was a fort.. why was it built here? Was there a settlement here before the soldiers arrived?
What was the date of the far older prehistoric site we found unexpectedly beneath the playing card shaped triple-ditched enclosure?
The red Devon soil is very acidic and therefore corrosive. No bone (apart from cremated bone) or other organic matter like sea shells, leather or fabric could survive in it. Even the pottery has lost much of its finished surface.
We hoped for preserved pollen but none survived.
Fortunately, charcoal was found across the site and this was sent to Cathie. First to identify it and then to assess it for radiocarbon dating suitability.
Some chunks were from mature trees and therefore the date range for C14 would be too large. She selected the fragments of round wood and twig that would provide a closer date.
Our samples were from young oak, ash, gorse, blackthorn and hazel collected from sealed contexts across the site. We sent ten samples off, and eventually, last week, the C14 dates came back.
It turned out they represented a huge date range….from 8537-8297BC to AD 16-124.
Quite a spread… but the dates can be looked at in clusters.
You may remember that our earliest feature at Budlake was a small pit which contained tiny fragments of flint.
These are known as microliths and typical of the middle stone age or Mesolithic period. Our earliest date of c.8500 BC would be good for this but this date, from a fragment of hazel nut shell, was the earliest of four dates from the pit filling. The others were all from ash twigs, a sample from the top, the middle and the bottom of the pit … all dated from between 4700-4500 BC at 95% probability.
I phoned Olaf, our flint specialist. Yes, he felt that the flints from the pit were mixed and though most of the flints seem be typical of the Mesolithic material, there was Neolithic lithic technology evident.
The other three C14 dates from the pit were exciting because they date the soil to a period when the last hunter gatherer communities and the earliest farmers were in contact.
We can now compare the pottery finds and there were 8 small sherds mixed in the Roman deposits which were of fabrics likely to date from the Neolithic and Bronze Age.
Last year’s Trench I cut across the three ditches of the enclosure and found them to ‘V’-shaped in profile and averaging 3m wide and 1.75m deep. They looked typically Roman but contained next to nothing to date them. Fragments of blackthorn charcoal were found in the filling of the middle and inner ditch and both gave dates. From the middle ditch 201BC to 53 BC at 90% probability and from the inner ditch 47BC to AD66 at 95% probability.
Ed, the pot specialist dated the single rim fragment I’d found in the inner ditch as Late Iron Age…so was the triple-ditched enclosure Roman at all?
The other four C14 dates were very similar. Three came from Trench II, at the south entrance into the playing card shaped enclosure, and one more came from Trench III.
In Trench II, the enclosure’s middle and inner ditches had pottery and charcoal found within the debris filling them. From the inner ditch, a fragment of gorse charcoal and one of oak.. together with an fragment of oak charcoal from the middle ditch, all gave dates within a band AD 6 to AD 124 at 90% probability. The last date came from hazel charcoal from the filling of a large ditch found outside the enclosure AD 16-124 at 95% probability.
The mid range of all four would be AD 64-70. This would fit a Devon conquest period occupation of a Roman fort….though tempting, this way of averaging the dates is considered to be statistically incorrect and too precise for radiocarbon dating.
Still, the consistency of these four broad date ranges provides reasonable grounds to conclude that this is indeed a mid-late first century Roman fort and Ed’s pottery report provides back-up from the Trench II gateway ditch terminals.
From the middle ditch, a fragment of decorated South Gaulish samian bowl and part of a North Gaulish gritted mixing bowl together with a white ware butt beaker were all distinctively mid-late 1st century. The sherds of amphorae found in the gate-way post-pit had been imported from Spain and were also of a typically early Roman type.
Archaeology can be annoying because it rarely gives definitive answers. The best it can do is to use all available evidence to nudge us closer to the truth.
The case is scientifically well evidenced now to argue for a mid to late 1st century Roman fort at Budlake… but other finds give us the scent of an earlier presence there and perhaps the displaced mid-late Iron Age C14 dates and fragments of pottery suggest that the Romans occupied an existing settlement… fragments of which were scooped up and thrown into its ditches just as the Mesolithic was thrown into an earliest Neolithic pit.
The photo taken in 1984 shows the whole field full of archaeology and. it is hoped that further geophysical survey and excavation by Exeter University and Killerton’s HART volunteers will start to further unravel the palimpsest of time of this fascinating place.
Apologies for this unorthodox method, but we are urgently trying to reach you regarding the talk you are due to deliver for us on Thursday 23rd June. Can you please contact me on mikeATromanroadsDOTorg, or phone me on the number I supplied on my email to your National Trust email address.