Welsh Journals

Fig. 1 Location plan of cropmark features 22-36 and Sites I-IV (based on Crew 1980, fig. 3). further 8.5 m to the south-east, had been undertaken by Mr J. Connell in 1975. That work revealed evidence of occupation in the 2nd to 4th centuries A.D. (Connell 1975, 51), probably as part of the vicus. A section cut through the main Roman road, which lay in the western half of the trench (fig. 2, Section A-B), showed it to be about 8 m wide and of two phases of construction, although it was clear that the upper layers had been greatly disturbed by ploughing. The first phase consisted of two pebble layers (101, 100) flanked on the eastern side by a shallow drainage ditch (96); this was sealed by further layers of pebbles and silt representing the make-up of the second phase road. Similar silt and pebble layers were visible on the western side of the road; on both sides these layers had been cut by additional ditches parallel with the road (81, 22/67). No first phase ditch was identified on the west side of the road, but it is possible that it had been cut away here by the second phase (81). Phase 1, which saw the construction of the main road south from the fort, remains undated because of lack of finds from sealed contexts in the the small area excavated, but it could well belong to the initial occupation of the fort in the Flavian period.