Dundon hillfort. On the NW the bank has disappeared and there is only a scarp. Entrance was in the middle of the E side but has been much altered and destroyed by quarrying.

Several flint flakes, a core and scrapers, also a few pieces of BA pottery were found on the surface within the camp in 1916.

The interior rises gently towards the S. Defences consist of a slight bank running along the edge of the plateau with a pronounced terrace about 6m down the slope, perhaps a silted up ditch with counterscarp. Entrance gaps are present in two places on the E side, the N one being altered by quarrying and probably not original. The approach on this side is very steep and it seems probable that the entrance was on the W side.

Old quarry in S part opened briefly in 1970s to provide building stone.

There are also miscellaneous earthworks at ST48253230, 50m to the NW, which may be outworks.

A slight univallate hillfort occupying the top of a hill. Defences vary from a scarp to a 2m high bank with outer terrace, but on average consist of a bank 0.4 – 1.0m high, atop a scarp in places. Earthworks have been affected in places by C19 hedge banks and by quarrying. Entrance uncertain – probably on E but obscured by quarrying. Outer scarps in places mainly under dense vegetation and not reachable. The slightness of the earthworks suggest the possibility of a date earlier than the Iron Age.

Dundon Beacon (PRN 53759) on the SE corner of the fort is a mound atop the scarp, ditched through the rampart on its N suggesting it is of later date. It is possible that this is a Norman motte. The stretch of the hillfort immediately N of this is heightened to a 2m high bank apparently raised from a 6m wide terrace below, and this may represent a Norman reworking of defences with the intention of creating a motte and bailey castle.

The rampart consists of a single earthwork bank, varying in height between 0.5 and 2.5m, forming an irregular enclosure. Below the rampart the hillside is very steep, forming a natural defence and there is no signe of a ditch. Neither of the two abreaches of the rampart are likely to have been the original entrance which presumeably lay on the eastern side and has been destroyed by quarrying. The quarries appear on the 1886 OS map as of similar extent as today. The last recorded use was in 1925.

Scheduling revised with new national number (was Somerset 339) on 23 Dec 1996.

The excavation for a water pipeline in 1997 through the rampart where it had been severely eroded on the NW side by a track showed that the first defences had consisted of a timber laced stone rampart that was subsequently burnt. A later phase of clay rampart overlay this and there was a possble pre rampart phase of worn cobbles. Small sherds of iron age pottery were recoverd from the stone phase. The pipe trench was monitored where it might have cut an external ditch but nothing was seen next to the rampart.

The S part has recently been cleared of coniferous trees and is beginning to regrow. It is intended to establish a calcareous grass cover grazed by sheep. A number of deciduous trees have been left but one has already blown over. To the N old grassland is being maintained with occasional trees. One has blown over revealing the rubble core of the rampart.

Much as described in 1998. The ramparts remain stable under woodland and the interior is grazed by a small flock of sheep. Mowing of thistles and nettles contiues to improve the grassland.

Location

OS map reference: ST 485 322. Nearest town/village: Compton Dundon.

Data kindly supplied by the Somerset Historic Environment Record.

Record created in October 1984

© Copyright Somerset County Council 2007