CPAT logo
Cymraeg / English
Back Home
Projects index

Potential Cursus Monuments in Mid and North-east Wales

Survey and Excavation 2008-09


Cursus monuments are characterised by roughly parallel banks and ditches, with the more complete examples having terminals at either end, forming elongated enclosures, some with central linear mounds. Excavations in England have dated them to the Neolithic, between 4,000 and 2,500 BC. They take their name from the Latin for 'racecourse', and are so named because early archaeologists believed them to have been used by the Ancient Britons for racing their chariots, a theory which has now been dismissed. Their true purpose, however, remains a mystery, although it is generally believed that they served as routeways for ceremonial processions. Although many cursuses extend over considerable distances, in some cases up to 10km, most are rather shorter, typically around 2km or less in length.

cursus distribution

Right: Known and potential cursus monuments in mid and north-east Wales

To date there are only two confirmed cursus monuments in mid and north-east Wales, on enear Welshpool and the other in Radnorshire’s Walton Basin. The Welshpool Cursus is located at Sarn-y-bryn-caled to the south of Welshpool (SJ 21720487), and forms part of a ritual complex which also includes the well-known timber circle. The cursus is c. 380m in length, aligned north-west to south-east, and comprises roughly parallel ditches 8-10m apart. Trial excavations during the early 1990s provided a radiocarbon date of 3891-3889 cal BC or 3796-3662 cal BC for charcoal recovered from the base of one of the ditches. The Walton Green Cursus (SO 26465988) has well-spaced, parallel ditches and square terminals. Excavations have demonstrated that the ditches were shallow and narrow, in keeping with other cursus monuments excavated in Wales, although unfortunately no dating evidence was revealed.

The present project was undertaken with funding from Cadw to investigate five potential cursus monuments.


Cropmarks and excavated features at Collfryn

Right: Cropmarks and excavated features at Collfryn.

Cropmarks had revealed two narrow, parallel ditches 8m apart, oriented north-north-east to south-south-west, and extending for around 80m, lying in close proximity to a large ring ditch. The cropmarks are located around 5km north of Guilsfield in Powys (SJ 21891649) in the low hills to the south-east of the River Vyrnwy at an altitude of 130m.

A programme of geophysical survey failed to identify the potential cursus, although subsequent trial excavations clearly demonstrated that the parallel cropmarks were formed in response to two field drains, and the site has now been dismissed.

The geophysical survey was successful in providing more detail of the large ring ditch, which is around 37m in diameter, with a ditch 2.3m wide. Both the geophysics and the cropmarks have indicated a possible gap of perhaps 9m on the eastern side, although the results were not sufficiently detailed to be certain that this is a deliberate entrance rather than a response to ground conditions. The ring ditch is unusually large, however, although its ditch is not particularly wide, and it is possible that this may be a hengiform monument rather than a plough-levelled round barrow.

Gerwyn fechan

Vertical aerial photography in May 2006 revealed two roughly parallel ditches, between 10m and 18m apart, extending for at least 200m. The site is located in the Sesswick area, to the south of Wrexham (SJ 3688245834). Although trial excavations had been planned for 2008-9 it was not possible to gain access. However, it is hoped that the work can be undertaken in 2009-10.


Cropmark evidence from several seasons of aerial reconnaissance has revealed a more or less continuous feature defined by two ditches between 54m and 70m apart, extending for at least 800m from SO 24366050 to SO 25086087. There are also two further lengths of linear cropmarks which may extend the northern ditch for another 1.6km as far as SO 26546155.

The site lies in the Walton Basin, 1.3km north-west of Walton village, in an area rich in prehistoric archaeology. The Walton Green cursus lies 2km to the east-south-east, while the Four Stones stone circle is close to the south-western end, and the north-eastern end extends into the interior of the Hindwell palisaded enclosure.

Photo 95-17-17 © CPAT

Left: 1995 aeiral photograph showing the two parallel ditches of the potential cursus at Hindwell.

A magnetometer survey was undertaken in September 2008 in four areas at the south-western end of the cropmarks in an attempt to identify the extent of the ditches in that direction. The presence of both ditches was confirmed in the easternmost areas, but not further to the west. It is possible, however, that this negative evidence is perhaps due to insufficient magnetic variation between the subsoil and the ditch fill as even where it was possible to confirm their existence, the ditches were difficult to detect.

A single trench 66.8m long and 1.5m wide was excavated by machine across the line of the ditches to the west of the geophysical surveys. The results were somewhat mixed, however, in that it was only possible to identify the southern ditch of the potential cursus. This proved to be very substantial and its size, at 3.9m by 1.8m, would place it amongst the largest cursus ditches so far identified. By comparison, the Rudston A cursus has a maximum ditch width of 4m, while the Greater Stonehenge Cursus measures around 3.3m by 1.5m towards the western terminal, although elsewhere it is far more slight at around 2.5m by 0.45m, conforming to the generally accepted pattern of cursus ditches being more substantial closer to the terminals. The nature of the fills suggested that the initial weathering of the sides had been followed by a period of stabilisation. A series of gravely deposits against the northern edge of the ditch suggest a later period of erosion which is likely to have been derived from bank material on that side of the ditch. The upper ditch fill consisted of a deposit of clay-silt containing flecks of charcoal, the thickness of which suggests deliberate infilling. Although no dateable artefacts were recovered from the ditch a radiocarbon date of 3660-3520 cal. BC has been produced from the lower ditch, suggesting that this may be broadly contemporary with the Womaston Causewayed Enclosure.

Further excavations were undertaken in September 2010 to investigate the northern ditch which proved to be of similar dimensions. Again, no artefacts were recovered, although it is hoped that several charcoal samples will be suitable for radiocarbon dating.

Photo 2790-0080 © CPAT

Right: The large ditch revealed during excavations at Hindwell.

The excavation also identified a short length of ditch which contained a significant quantity of Roman pottery dating to the late 3rd – 4th century, as well as a fragment of a rotary quern. The ditch, which was aligned north-south, was around 1.1m wide and up to 0.6m deep, and butt-ended at its southern terminus. Immediately adjacent to the ditch on the west side was what appeared to be a posthole, 0.6m across, containing packing stones, and extending beyond the excavation to the west. Although the feature was not excavated its proximity to the Roman ditch suggests a possible association. A second Roman ditch was identified further to the south, which was V-shaped and around 0.9m wide and 0.35m deep.

If the cropmark evidence is to be believed the Hindwell cursus may extend for more than 2.4km, with neither terminal having yet been identified. As with the scale of the ditch, the length would also place the monument amongst the larger examples in Britain. Morphologically, the Hindwell cursus is unusual in that the ditches are not parallel with each other, varying between 54m and 70m apart. This irregularity is not unknown, however, as the Greater Stonehenge Cursus varies between 100 and 150m in width.

Holywell/Whitford Dyke

A linear earthwork adjacent to a hengiform monument on the former Holywell Racecourse has been suggested as a possible cursus, although it has previously been regarded as part of Offa’s Dyke, and more recently as a separate monument known as the Whitford Dyke. The earthwork can be traced intermittently for around 9km from a point south of the henge (SJ 15317466) to Trelawnyd (SJ 08347988). The proximity of Gop Cairn to the northern end may be significant.

Photo 08-C-0277 © CPAT

Left: 2008 aerial photograph showing the hengiform monument and linear earthwork on the former Holywell Racecourse.

A programme of geophysical survey was conducted by ArchaeoPhysica to investigate the area of the henge, together with the section of linear earthwork to the north-west, with the principal intention of locating an trench excavated by Fox in 1925. The survey identified the linear earthwork, although the generally low variation in the magnetic response meant that its components were less well-defined than might have been expected, and it was not possible to gauge the width of the ditches. Although the position of Fox’s trench was tentatively identified, this later proved to be an unrelated feature. The low magnetic response was also a feature of the henge, with the result that very little detail was apparent in the geophysical data. The ditch could be identified, being more easily identified on the south and south-east, where the earthworks are more pronounced, and barely visible on the western side.

The original intention was to identify Fox’s excavation, re-excavate the trench and cut back one section to provide undisturbed material from the ditches and bank which it was hoped would contain some datable material. In the event it proved extremely difficult to locate the trench and the excavation was therefore located on the basis of measurements given by Fox in his original site archive. At this point the bank was clearly visible as an upstanding earthwork around 10.5m across and 0.3m high, although neither ditch was readily apparent at ground level.

The removal of the topsoil revealed in situ bank material with the upper ditch fills on either side. It was evident that there had been considerable animal disturbance throughout the excavated area, with the bank material in particular being mixed with topsoil. Where it was less disturbed the bank material was no more up than 0.15m thick and perhaps 5.4m wide, lying directly on top of the natural subsoil, with no indication of a buried ground surface beneath.

Whitford Dyke - photo 2791-0033 © CPAT

Right: Excavated section of the Whitford Dyke viewed from the west.

Fox recorded considerable difficulties in identifying the edges of both ditches, and the recent excavations encountered similar problems due to the variable nature of the glacial till which overlies the limestone bedrock in this area. The western ditch was particularly difficult to determine, although as excavated it appeared to be around 4.2m wide and up to 0.75m deep. The eastern ditch was better defined and measured 3.7m in width and up to 0.55m deep. It was notable that there appeared to be some disturbance within both ditches, close to the northern baulk, and although no clear edge could be determined it is possible that this was related to Fox’s excavation.

The principal aim of the study was to determine the date of the monument but unfortunately the excavation produced no artefactual evidence and, although bulk soils samples were recovered from the ditches, it seems unlikely that they will contain any material suitable for radiocarbon dating. Furthermore, the investigation of other sections by augering suggests that the nature of the subsoil is such that organic preservation is likely to be poor and charred material is evidently scarce.

The question therefore remains as to whether the linear earthwork is more likely to be prehistoric in date, or belong to the earlier medieval period. On balance, the latter is perhaps more likely, with the incorporation of two prehistoric monuments along its route, and Gop Cairn possibly marking the north-western end. As a boundary feature the earthwork runs more or less along the spine of the Flintshire plateau, dividing the hinterland of the Dee Estuary from the Vale of Clwyd. This is not to say, however, that the entire monument is of the same period and the section immediately to the south-east of the hengiform site does not readily match the alignment of the earthworks to the north and south and this could, therefore, be another earlier monument which was incorporated into a much later boundary.


A pair of approximately parallel linear cropmarks were noted in a field to the south of Ceunant Farm (SJ 16201377) on an aerial photograph taken by CPAT in 1984. The site lies alongside the main A 495 road, about 1km north-east of Meifod village, and is sited on the edge of the flood plain of the River Vyrnwy at about 80m OD. The cropmarks appeared to represent two roughly parallel ditches around 14m apart, extending for approximately 85m from north-east to south-west.

84-MB-0624 © CPAT

Left: 1984 aerial photograph showing cropmarks near Meifod.

Geophysical survey confirmed the presence of the ditches, although their form suggested that they are most likely to be associated with former field boundaries, or a trackway, and the site has now been dismissed as a potential cursus.

Privacy and cookies