Excavations at Dunragit: 2001 Julian Thomas With contributions by Ange Brennan,Chris Fowler,Kirstin Miller and Dave Robinson

Droughduil mound, looking south Page Index page 1 page 2 page 3 page 4 page 5 Index Introduction Some Observations Methods and objectives Workforce and logistics References

Three seasons of excavationtook place on the Neolithic ceremonial complex at Dunragit during 1999-2001. These investigations demonstrated that the principal element of the complex had been an enclosure composed of three concentric rings of timber posts (Fig. 1). The largest of these rings had a diameter of around 300 metres, while the innermost had a diameter of perhaps 120 metres. Both the outer post-ring and the middle were composed of large uprights interspersed with smaller members, forming a continuous palisade and comparable with the structure of the Late Neolithic enclosure at Meldon Bridge in Peebleshire (Speak and Burgess 1999). Both of these post-rings had been single-phase structures, and in both cases the posts had eventually rotted out and been replaced by silting and collapsed gravel packing. These two outermost rings of post-holes also produced very little in the way of material culture. By contrast, the inner ring had been made up exclusively of large, free-standing posts, most of which had rotted away, but some of which had been deliberately removed. These post-holes were considerably larger than those of the two outer rings, and it is possible that their digging had involved a conspicuous expenditure of effort. Where posts had been pulled out, a number of elaborate deposits had been placed in the crater left by the post-removal. These included the cremated remains of a woman and a sheep, sherds of Grooved Ware in a matrix of dark, burnt organic material, a deposit of Beaker pottery, and a mass of charcoal containing fragments of burnt bone, possibly representing a pyre deposit.

Fig. 1: Dunragit, showing features plotted for aerial photographs and trenches excavated 1999-2002.

The enclosure had been preceded on the site by a post-defined cursus monument, one side and the terminal of which had been identified in the 1999-2001 seasons. The post-holes of the cursus were easily distinguishable from those of the enclosure, for in every case the post had been burned in situ. The existence of earlier Neolithic cursus monuments and later Neolithic enclosures on the same site is known from a number of other sites, including Dorchester on Thames in Oxfordshire, Llandegai in Gwynnedd, and Thornborough in Yorkshire. This suggests that the relationship is not coincidental, and that the significance that a location achieved through the construction of a cursus was a significant factor in the positioning of later enclosures.

The aerial photographs taken by RCAHMS that were responsible for the identification of the Dunragit enclosure also show two possible sets of entrances, both located in the area immediately to the south of the modern railway line. As the owner of this field is presently not disposed to allow excavation the character of these entrances must remain conjectural. However, it is notable that the entrance through the outer post-ring does not respect that which is connected with the middle ring. The outer entrance opens to the south-west, while the middle ring entrance is aligned to the south. Given that both of these palisade-rings are single-phase structures, while the central post-circle has two phases of construction, it may be that the three concentric circles are not all contemporary. On the contrary, an interpretation is preferred in which the monument as a whole had two phases of construction, in each of which a timber circle was surrounded by a palisade, and in which the middle post-ring succeeded the outer, or vice-versa.

The more elaborate of these entrances, connected with the middle post-ring, is composed of two parallel lines of features which are presumably post-holes. It opens toward the south, and is aligned almost precisely on a large earthen mound some 400 metres away at Droughduil. This mound presently stands in a plantation managed by Forest Enterprise, and has recently been cleared of trees. It is located between the enclosure and the extensive dune systems of Luce Sands, and as such it stands in an area where the subsoil is much sandier than the outwash gravels on which Dunragit village stands. The mound is known as Droughduil Mote, and has been recorded as a medieval motte. However, the spatial relationship with the Dunragit enclosure is suggestive, and the site is by no means representative of local castle mounds. Mottes such as Mote Slapp at Sandhead are generally high-lying, and have a profile which is much more flattened than the Droughduil mound. Moreover, a number of large late Neolithic mounds are known, and many of these are directly associated with henges or palisaded enclosures. The best-known example is Silbury Hill in Wiltshire (Whittle 1997), but others include the Conquer Barrow in Dorset, perched on the bank of the Mount Pleasant henge and overlooking the palisade enclosure that it contains; The Marlborough mound (Wiltshire); the Hatfield Barrow, contained within the Marden henge (Wiltshire); and Duggleby Howe, which contained an important sequence of late Neolithic burials, and is enclosed within a massive causewayed ditch (Kinnes et.al. 1983).

Recently, John Barrett (1994: 31) suggested that Silbury Hill might effectively represent an elevated platform, which could raise a small number of people above the surrounding landscape, placing them in a position of extreme visibility. At the same time, such a group would have been able to observe activities taking place within the Avebury henge, over the top of the intervening Waden Hill, and in the West Kennet palisade enclosures. In the light of these arguments, it was judged appropriate to test the character of the Droughduil mound by excavation. One trench was to be opened running up the side of the mound, and another would investigate the presumed ditch surrounding the monument. A final, smaller trench would investigate a depression on the summit of the mound, which was thought possibly to represent an antiquarian shaft similar to that in the top of Silbury Hill, and which might represent a management problem for the monument. A secondary objective for the 2002 fieldwork season was identified as the excavation of a final 30 x 20 metre trench in the field immediately to the north of the railway line, completing the excavation of the inner and middle post-rings in this area. This would result in a strip of approximately 100 metres having been excavated beside the railway, and would enable the western side of the cursus to be identified and investigated. In the event, a series of test-pits failed to identify any ditch surrounding the mound, so that no ditch cutting was possible, and this secondary objective was pursued, given the availability of adequate labour.

Droughduil Mote
A visit to the mound in spring of 2002 had identified a series of objectives for the investigation of the site. The mound is markedly oval rather than circular in plan, with a long axis running north-south (Fig. 2). A full section of the monument would have involved serious logistical problems, for a large mound composed of sand would have required major shoring, and this was not considered practicable in the context of a university excavation. Instead, it was decided to excavate a trench up the side of the mound, bottoming on the original mound surface and revealing its morphology and construction (Trench G).

Fig. 2: Droughduil Mote, contour survey and layout of trenches.

Silbury Hill, of course, is composed of a series of stacked drums, revetted by walls of chalk blocks. If the Droughduil mound were Neolithic in date, it was conceivable that a similar structure might have existed. It was agreed with Historic Scotland that a limited part of this section would cut into the tail of the mound, to enable the structure of the monument to be revealed to some extent, and to allow samples to be taken for environmental analysis and OSL dating. A second trench, 3 x 3 metres extent was opened on the summit of the mound, allowing the depression there to be investigated (Trench H). The third trench was to be cut over the presumed quarry ditch surrounding the mound. In order to locate this trench to the best advantage, a transect of shovel-pits was laid out running due west from the mound at 5 metre intervals for a distance of 55 metres (Fig. 3). Although several of these were dug to a depth of around two metres, none produced any evidence for the presence of a ditch. The exercise was then repeated with transects of pits to the east and south of the mound. These too were unsuccessful, and it was eventually concluded that no ditch had ever existed.

Fig. 3: Droughduil Mote, layout of trenches and shovel-pits.