Archaeology in the Severn Estuary 9

Substantial human activity ranged almost continuously from the Iron Age into the early modern period at Magor Pill (Gwent) but, because of coastal erosion, the surviving evidence takes the form chiefly of transposed pottery assemblages. A collection of 2,5I3 Romano-British sherds is described, together with an assemblage of I79 prehistoric sherds chiefly of the late Iron Age (including 1st cent AD). The Iron Age pottery is dominated by Calcite-tempered and Limestone-tempered Wares, and there are some sherds tempered with quartz sand. Dominating the Romano-British assemblage are vessels representing the local South Wales grey coarseware industry, but imported Southeast Dorset BBi is also important. The minor imported coarsewares are chiefly Southwest BBi, from Southwest England, and Severn Valley Ware and Grey Micaceous Ware,from the Severn Vale to the northeast. Also reaching the site were finewares from the Nene Valley, the New Forest and, especially, Oxfordshire. The mortaria are dominated by Oxfordshire products, but Caerleon legionary Ware is represented, and there are vessels possibly from other sources in the West and Southwest of England. The pottery evidence suggests that the Magor Pill settlement flourished in the later Roman period but that occupation had been continuous from the late Iron Age. In comparison with another wetland site in the area (Rumney Great Wharf), together with a range of dryland sites in southeast Wales, the status of Magor Pill seems to have been that of a well-endowed rural settlement with significant connections by coastal waters with other parts of the Severn Estuary, including the town of Caerleon with its nearby quay on the Usk. There is however, no evidence suggesting that Magor Pill was, in the Roman period, the first port of entry for goods crossing the Bristol Channel.

This paper examines the perception of human presence in the Brue Valley, Somerset, afforded by multiple radiocarbon dates, and the change in our understanding brought about by the first dendrochronologfical date for a trackway. It is argued that individual trackways were in use for only a few years, and that ji-om the Neolithic to the Iron Age there are a number of substantial gaps in the record for human presence in the wetlands. The implications for this argument are further explored in Coles 1999.

The existence of a salt extraction industry in Somerset during the Roman period has been known for over a century, though very little archaeological research has been published on the subject. Recent small-scale fieldwork has illustrated the potential of the archaeological evidence and demonstrated the range of influences which are gradually destroying this resource.

Excavations were carried out in 1997 at Hill Farm, Goldcliff. in advance of the construction of the Gwent Levels Wetlands Reserve on the Caldicot Level. A group of samples has been analysed to provide information about the past environments and the date of a system of banks and ditches. The radiocarbon dating of an organic layer has shown the banks and most of the ditches to have originated in the Middle-Late Iron Age (450-350 cal BC), contemporary with activity on the foreshore. Flooding seems to have ceased until late in the Roman period, when an area of saltmarsh became stabilised at the south west of the site; subsequently (c. 500 cal AD) a general marine transgression is reflected by flood silts. The environmental evidence suggests saltmarsh conditions throughout the flooding, with widespread deposition of low-energy clay silts and a strong marine/estuarine component.

A paper that describes the remains of wooden boats and concrete barges that lie between Gloucester-Sharpness Canal and the River Severn in Purton

Following preliminary survey work in 1997, small-scale excavations were carried out on an area of shrunkensettlement earthworks to the north of Puxton village, near Weston-super-Mare, on the North Somerset Levels. Two slightly raised platforms were sectioned which yielded evidence for occupation between the 11th and 17th centuries, including a cesspit, various drainage gullies, and the stone footings of a possible building. Several substantial ditches were also excavated that yielded organically-rich midden deposits. These contexts should provide excellent comparative material for the early medieval assemblages recovered during excavations to the south of Puxton church in I 996, and indeed for the Romano-British material retrieved from the sites at Banwell Moor and Kenn Moor.

A summary on the fieldwork and archaeology done in the Severn Estuary in 1998

A note that presents the background and preliminary results on holocene peat and intertidal statigraphy on Shapwick Heath, Somerset