Skip to content
Skip to navigation menu


Roman fort and settlement at Caergwanaf near Miskin

Geophysical survey and excavation were undertaken on the site of the recently discovered Roman fort and settlement at Caergwanaf near Miskin during 2004. A resistivity survey concentrated on areas to the west and north of the 1.6 ha Roman fort identified in 2003, locating a series of large enclosures on its western side (with evidence for occupation inside Enclosure A) but little sign of activity to the north. Therefore, the Caergwanaf settlement appears to have extended over an area of some 450m from east to west, and approximately 300m from north to south. An intensive resistivity survey within the second to third century iron-working settlement to the south and east of the site confirmed the position of the fort's southern defences. Furthermore, this survey suggests that the southern ditch of Enclosure B followed the line of the fort defences also on the south, raising the possibility that the enigmatic Enclosure B may be a fortlet.

Levelling operations by the farmer early in 2004 removed the topsoil from an area approximately 15m by 50m along the suspected line of the western fort defences, and the opportunity was taken to excavate part of this exposed surface before further damage occurred. The double-ditch system of the western defences was found, together with the location of the western entrance. Although the archaeological deposits had been heavily truncated, the inner V-shaped ditch survived up to 3.5m wide and 1.5m deep, while the outer ditch was at least 1.5m wide and 0.6m deep. The two ditches were interrupted by a gap of approximately 3m that formed the western entrance into the fort. On either side of this gateway the outer ditches became shallower and turned inwards to join the terminals of the inner ditches, thus forming a funnel-shaped entrance that narrowed from 17.5m externally to 11m internally. This corresponds closely with the shape and dimensions of the northern entrance as indicated by the earlier geophysical survey. This arrangement of ditches is sometimes known as a "parrots beak", and is characteristic of many Flavian forts in Scotland. In Wales, a similar system was found at Pen Llystyn and one may occur at Llandeilo (G. Hughes, pers. comm. 2005).

No trace of the road into the gateway survived, but a linear ditch to the south of the entrance and linked to the outer fort ditch by a shallow overflow channel, probably represents the bottom of the southern roadside ditch. A second shallow ditch ran out of the fort on the northern side of the entrance, although on a slightly different alignment from the roadside ditch to the south and passing more directly down slope. This feature may have been connected to a possible roadside drain inside the fort that was identified by previous geophysical work.

A considerable thickness of sediment had filled the fort ditches before they were deliberately backfilled. Although the pottery assemblage recovered during the excavations was limited, the evidence indicates that the ditches were being filled within the period AD70/75 to 80/85. The backfill of several of the ditches contained slag, revealing that iron smelting at Caergwanaf began before the abandonment of the fort.

A deposit of laid stones (possibly the remains of a trackway or path), containing quantities of abraded iron smelting debris and second to third-century pottery, overlay the inner fort ditches. There was little further structural evidence, however, for occupation during this period on this part of the site (in contrast to the excavations undertaken in 2002 which recorded second and third-century material in the ditch of Enclosure C on the hilltop to the southeast and from the iron-working platforms to the east).

After this hiatus, occupation recommenced in the western part of the site during the late third or early fourth centuries, consisting of a ditch running parallel to the outside of the earlier fort defences together with an associated gully, a posthole and several amorphous hollows eroded in the subsoil. These features all yielded a very large quantity of hobnails, ranging from individual examples through to complete soles.

The 2004 excavations indicate a short life-span for the fort at Caergwanaf, which was built and occupied during the early Flavian military consolidation of South Wales. Iron smelting seems to have started on the site before the abandonment of the fort, although the 2002 evidence showed that the major phase of iron-making occurred mainly to the south and east of the demolished fort during the second to early third centuries. Later Roman occupation, on the other hand, appears not to be significant in areas investigated in 2002, perhaps indicating a shift in focus to the west of the site in the late third and fourth centuries.

The 2004 excavations were directed by Peter Guest and Dr Tim Young. The work was funded by the University of Wales (through the Board of Celtic Studies) and the National Museums & Galleries of Wales.