In September 2017 and 2018, AOC delivered community excavations at the King’s Seat, Dunkeld, with Perth and Kinross Heritage Trust (PKHT) on behalf of Dunkeld and Birnam Historical Society.

King’s Seat hillfort is situated on a steep rocky outcrop to the west of Dunkeld overlooking the Tay River valley to the north and south. The site comprises a low wall enclosing a small upper citadel, a further series of 3-4 terraces or ramparts enclosing a mid-level terrace on the west side of the hill and a lower D-shaped enclosure enclosing an area to the east at the base of the hill. A massive glacial erratic (right) composed of the local schist stone dominates the summit. Four mysterious holes were discovered on the upper surface of the stone. It is not clear what these relate to, but they must have been bored into it in antiquity.

The two week seasons of excavation in 2017 and 2018 were preceded by a substantial exercise in vegetation clearance undertaken by a small but enthusiastic team of local volunteers from the society led by PKHT. The fieldwork represented the first recorded excavations on this spectacular hillfort site, which has long been assumed to be later prehistoric or early historic in date, perhaps both!

Trenches so far have targeted three key areas: the upper citadel wall; the interior of the upper citadel; and the series of 3-4 ramparts on the west side of the hill.

The ramparts were found to be substantial banks built of large stones and boulders with a turf or earth component which were placed onto the natural subsoil. One had a timber component, demonstrated by substantial postholes that would have held timbers at the back of the structure. Another had a front face comprising small but good building stone which was discovered underneath the slumped material from the structure collapsing off the steep hillside. The banks on the east were generally between 2m and 3m wide although heavily eroded and collapsing down the hill.

The upper citadel wall was a lower construction of stone and earth with evidence of a very rough stone inner kerb or face. The front face of this feature was not identified and is likely completely collapsed down the vertical cliffs that characterise the upper part of the hill.

In some places in the interior the bedrock was very close to the surface, more or less under a thin layer of forest loam. In some flat terraces there was a build up of material which had been very disturbed and mixed by the action of rhododendron and tree roots. Within this layer of hillwash or turf structure collapse a really unusual number of exciting artefacts were identified. Finds on the site included evidence of ferrous and non-ferrous metal working in the form of crucible fragments; slag (metal working waste); and stone and clay moulds. In addition to this there were a lot of iron objects including blades; a socketed tool or arrowhead; part of a set of shears; and a number of other objects that will need to be x-rayed before their original form can be identified. Other finds included rare early medieval pottery imported from the Continent and several glass and ceramic beads (one shown below, right) which are of an Anglo Saxon style. When compared with finds from other sites in Britain they likely date to the 6-7th centuries AD, demonstrating not only that the site was likely in use around then but that it was an important high status site of trade and production in the early historic period. Finds such as spindle whorls (below, left) demonstrate textile processing occurred on the site and the large number of whetstones for sharpening blades is notable. A piece of vessel glass from an Anglo-Saxon style glass beaker (below, centre) gives us a tantalising glimpse of a society drinking from fancy glasses and potentially living or feasting on the site.

The location of the site and the type of rocky outcrop with multiple terraces that the site sits on is similar to other important early historic sites in Scotland such as Dunadd and Dundurn. The metalworking assemblages we have found so far are comparable with sites such as Mote of Mark, Dunadd and Buiston crannog. The presence of continental pottery and Anglo-Saxon glass beads and vessels points towards a site fully integrated into extensive trading networks during this period.

Underneath this finds rich layer we finally started uncovering evidence of structures and more secure activity which will be very useful to put all of these finds into context. A charcoal and ash rich deposit filled with animal bone was identified and sampled around a substantial hearth setting (left). These rich deposits representing in situ burning will hopefully yield important radiocarbon dates to help us date the activity at the site as well as informing us about the economic indicators of the people who lived and worked here.

Although we have lots of information and lots of finds to analyse we still have lots of questions about the site so we will return for a final season in 2019. The last trenches will aim to continue investigating the upper citadel interior; the interior of the middle terrace and the wall of the lowest annexe/enclosure on the east side. In the meantime our post-excavation programme will start to conserve, analyse and research the material already retrieved from the site.

The excavations so far could not have happened without the team of hardworking and enthusiastic volunteers who joined us from the local area. Thank you so much to everyone who came and helped!

The King’s Seat Hillfort Archaeology Project is a partnership between Perth and Kinross Heritage Trust and the Dunkeld and Birnam Historical Society. Excavations were delivered by AOC Archaeology Group and PKHT. The project is funded by SSE, The Gannochy Trust and Heritage Lottery Fund. Permission to excavate on the site was granted through Scheduled Monument Consent from Historic Environment Scotland and from the landowner Mr Arnold Shnegg, Dunkeld House Hotel.