The dry summer of 2012 revealed a number of enigmatic crop marks, which we were keen to investigate further. In the spring of 2015, we conducted a geophysical survey on the area, which displayed a large number of pits, ditches and irregular-shaped features spreading over an area of at least 20 hectares, and indicating a sizeable settlement as well as potential areas of industrial activity.

Armed with this information, we selected two sections of the site for sample excavation. Trench A was positioned to expose two potential roundhouses, while Trench B was established to examine a large and distinct roundhouse, measuring at least 15m in diameter.

The results were far better than we could have hoped for.

While we had only been able to see the outline of three roundhouses on the geophysical survey, our excavations revealed the footprints of at least 16 discrete buildings, indicating that the area had originally been far more densely populated than we suspected.

We would expect somewhere between 150 and 200 roundhouses across the area of the geophysical survey if the number of ring-gullies revealed across our two trenches continued throughout the rest of the site. This is particularly unusual as the settlement lacks of any form of enclosure - such as a rampart and ditch in the form of a hillfort.

As a result, Duropolis - as we have called it - is of a size unheard of in Dorset and the south-west outside the security of a hillfort.

Returning to the roundhouses themselves, one of their key features was the ‘ring gully’ - a shallow cut that appeared to represent the foundation slot for a cob or wattle-and-daub wall. All the gaps we identified faced in a south-easterly direction, away from the prevailing wind, in the manner of many other, later prehistoric roundhouses.

In addition to the roundhouses, we also examined a total of 18 cylindrical pits, typically referred to as storage pits. Interestingly we discovered special, placed deposits in these pits, which were backfilled when no longer in use. These depositions varied from pit to pit, and included the fully articulated remains of a dog, quern stones, upended and perforated pots, and the inverted skulls of a cow and a horse, among other items. The inference is that the inhabitants of the settlement were making offerings to their gods to help assure the continued success and fertility of their crops, herds and community.

Elsewhere on the site, we uncovered evidence of quarrying and at least one possible furnace bottom. We also believe it is possible that the processing and preparation of food occurred here, as well as pottery, ceramic, weaving and textile manufacture.

The artefacts we recovered, and particularly the ceramics we unearthed, indicate that the major phase of activity at this site was between 200-50 BC, making it a settlement of the middle to late Iron Age.