The problem of the names of the Roman forts at the west end of Hadrian's Wall and along the Cumbrian Coast has been eased somewhat by Mark Hassall's inspired elucidation of that sector per lineam valli which lies between Carvoran and Stanwix: he argues that Birdoswald was Banna, and Castlesteads Camboglanna.

Artificial illumination is one of the most taken-for-granted features of modern life but in antiquity the availability and use of lamps was much more restricted. In Roman Britain, lighting equipment was rare and the use of ceramic lamps appears to have been largely limited to military sites and large urban centres. The so-called ‘lamp factory’ in Colchester is not the only site in Roman Britain to have revealed evidence for lamp production but it is probably the most outstanding. The extensive evidence for Roman lamp production near modern West Stockwell Street in Colchester takes the form of excavated structures as well as of a large and closely dated assemblage of lamps and moulds.

This paper describes a type of pottery made in the same region as Dorset Black Burnished Ware that can be shown to be current during the late fourth and early fifth centuries. This pottery — here named South-East Dorset Orange Wiped Ware — can be used as a diagnostic artefact to identify sites and features of the very late Roman period in Dorset. It also appears to be associated with a new architectural tradition typified by the ‘sunken featured buildings’ present at the late and post-Roman site of Poundbury.

Excavated in 1992–4, the villa yielded a portion of a whetstone which, on the basis of general shape, the presence of rebated long edges and microscopic petrography in thin-section, was with little doubt made from a sandstone in the Weald Clay Formation (Lower Cretaceous) of the north-west Weald. It is representative of a widely recorded, major stone-based industry in Roman Britain, with finds known to range from the Channel coast to the northern frontier zone.

The shrine was discovered in the Spring of 1980 and the excavation was carried out in June of that year by the Field Research Group of the Society of Antiquaries of Newcastle upon Tyne. The authors are indebted to Mr and Mrs D. A. Hofmann who made the initial discovery of the Yardhope relief, to the landowner, the Ministry of Defence, Otterburn Training Area, for permission to excavate, to Tyne Tees Television for some financial assistance, to Mr A. Reed for drawing their attention to the Hollinside relief and to Mr N. Gill for permission to photograph it. A special debt is owed to Drs D. Breeze and B. Dobson for assistance in the final drafting of this article and last, but by no means least, thanks must also be recorded to Ronald Charlton and John Day for their interest and willingness to help with the excavation, despite cramped working conditions and frequent hatches of midges.

We have recently been presented with an excellent summary of what is known of granaries in civilian contexts in Roman Britain by Miss Pat Morris in her Agricultural Buildings in Roman Britain. Most prominent in her discussion, because they are most easily recognised, are granaries with raised floors. Among a number of granaries which Miss Morris discusses under the heading ‘Unclassified’ is a room in the villa at Pitney in Somerset. Professor Applebaum first suggested this was a granary, largely because of what seem from the plan to have been external buttresses. The purpose of this article is to propose a new class of granaries, including the room at Pitney, on the basis of their flooring.