Brandon Camp lies just south of Leintwardine in the central sector of the Welsh Marches. A Roman road running south from Wroxeter through the Church Stretton gap crosses the river Teme at Leintwardine and skirts the east side of Brandon Hill before traversing the Aymestrey gap and running on to Kenchester; thence branches lead to both Gloucester and Usk. The hill projects into the flood-plain to form a notable feature in this sector of the Teme Valley (PL. XI A), rising to a height of 165 m above O.D.; it has a broad almost level summit which is easily ascended from the south but is protected around the northern edge by low precipitous cliffs. The north-western end is occupied by a pear-shaped hillfort of about 8 acres (3.2 ha) with an excellent outlook west and north. An upstanding rampart 3 m high marks the south and east sides of the Camp, but along the steep scarp it is only just discernible.

The project was a rescue excavation before housing development, mainly between Hadrian's Wall and the Vallum, close to the fort of Petriana at Stanwix. Traces were found of a pre Wall field-system. A minor Roman road was found running parallel to and 80 m south of the Vallum. A substantial Roman boundary of postsettings for posts over 2 m high was found running between, and at right angles to, the Wall and Vallum. In one area south of the Vallum a series of ditches of the second century A.D. suggested meadowland associated with a vicus of the nearby fort. To assist in its conservation the fragmentary remains of the Wall were traced and their position recorded. No trace was found of a Military Way; it may have been obliterated by an historically-recorded medieval road found parallel to and immediately south of the Wall. The position of Milecastle 65 was located by geophysical survey and found by test excavation to survive at foundation-level.

In a recent paper the late Dorothy Charlesworth published a valuable survey and gazetteer of finds entitled ‘Roman Carlisle’. This was the first attempt to bring together and summarize all the available evidence for Luguvalium, Rome's most north-westerly town and civitas capital. The paper dealt with both finds recovered during the previous century, many of which lack a secure provenance in the city centre, and with the results of observations and very small-scale excavations undertaken since the Second World War. At the time the article was written no major area excavations had been completed on Roman deposits and the first, Blackfriars Street, was still in progress. Since 1977, Carlisle Archaeological Unit has completed work on Blackfriars Street and is currently excavating in The Lanes, Castle Street, and Annetwell Street. The purpose of this paper is to draw attention to discoveries, mainly in the Lanes area, in advance of the first publication which is not anticipated before the late 1980's. The paper is divided into two sections of which the first (by M. McCarthy) outlines salient structural features, whilst the second (by T. G. Padley, Finds Assistant with the Unit, and Dr M. Henig, Institute of Archaeology, Oxford) deals with finds. It should be borne in mind that further work on all aspects of the site for final publication may result in the modification of some conclusions.

Two stone sculptures from Caerwent — a disembodied human head and a seated female figure — are the focus of this article. Using icon-theory, it is proposed that the Caerwent sculptures (albeit recovered from different chronological horizons) were perhaps produced at the same time, maybe even by a single stonemason. Issues of materiality, including choice of stone and style, are seen as key to their understanding, in terms of Silurian identity and religion. Moreover, the emphasis on mouths and ears invites interpretation of these images as those of speaking and listening Oracles, conduits between earthly and spiritual worlds.

Analytical earthwork and geophysical surveys have advanced our understanding of the lozenge-shaped Roman fort at Whitley Castle (Northumberland), which is notable for the exceptional depth of its outer defences. Built at a higher altitude than any other fort in England, it was almost certainly positioned to control the production and shipment of lead and silver from the Alston ore-fields. Its curious shape, tailored to that of the natural knoll, necessitated some adjustment of a standard fort plan, but accommodated six buildings to the rear of the central range and four to the front. An extramural settlement and terraced fields have been recorded to the west and north, and a swathe of ground to the south may have provided space for a parade ground. Post-Roman activity is evident from the cultivation and settlement remains that override the defences; two bastle-like buildings and an eighteenth-century farmhouse once stood within the fort itself.

The Roman fortress at Carpow, Perthshire, was excavated in 1961–62 and 1964–79. The gates of the fortress are the key to our understanding of its occupational history. The suggestion made here is that they were initially constructed in timber, at the same time as the associated turf rampart, but were soon replaced in stone specifically to support heavy dedication slabs set above them. A new arrangement of the surviving inscribed fragments at the East Gate is proposed, and new readings adopted, which confirm their traditional Severan dating. The Supplementary Material which includes detailed descriptions of the fragments and possible restorations is available at https://doi.org/10.1017/S0068113X19000138.

This paper represents the final publication of excavations by Brian Hartley in the northern annexe of the Roman fort at Slack, West Yorkshire, carried out half a century ago, but still providing rare evidence for activity within an annexe and the nature of its defences. Hartley concluded that the annexe enclosed a civilian vicus of Trajanic–Hadrianic date, although its defensive enclosure was provided only in the Hadrianic period. This report upholds that conclusion. In the light of recent claims of later Roman occupation at Slack, Hartley's conclusion that the site was abandoned by c.a.d. 140 is examined and upheld. It is proposed that a second annexe at Slack enclosed the fort baths (and possibly a mansio) and by examining the Slack annexes in the context of other fort/annexe plans an attempt is made to distinguish between annexes enclosing fort vici and those enclosing official or military activities.

The site whose excavation is the subject of this article was discovered during routine inspection of air photographs undertaken by the Royal Commission on the Ancient and Historical Monuments of Scotland as part of the preparatory work for the Inventory of Lanarkshire. The prints in question (National Air Survey sortie no. 541/A/530, prints 3137-–8) revealed the existence of what appeared to be a small ring-ditched enclosure a little way to the east of Beattock Summit, immediately to the north-east of the Roman road; it was situated at a height of 335 m (1100 ft.) O.D. on the lower slopes of Nap Hill, some 600 m (1970 ft.) north-north-west of Upper Howecleuch farmhouse (National Grid reference: NS 999153; FIG. 1).

The location of the province of Valentia has been, and indeed still is, a subject for debate. First, in the early thirteenth century Giraldus Cambrensis placed it in Scotland, but this identification is rejected by modern scholars as there is no archaeological evidence for an official re-occupation of southern Scotland in the fourth century, and in general terms such a re-expansion at that date is inherently unlikely. Secondly, it has been argued that it was one of the two provinces into which, it is maintained, Hadrian's Wall and its hinterland were divided. Thirdly, it has been suggested that it may have been Wales, although this identification lacks academic support at present. Fourthly, it has recently been claimed that in fact it was the diocese of the Britains which was renamed Valentia in 369, and that the whole concept of a fifth province is based on a misunderstanding.

The fort at Birdoswald is situated on the line of Hadrian's Wall, some 25 km (15½ miles) east of Carlisle (FIG. 1). The site lies between the forts of Great Chesters 10 km (6⅓ miles) to the east, and Castlesteads 12 km (7⅓ miles) to the west.

The standard pattern of Gallic local magistracies, so familar to all students of the Roman west, is apparently very straightforward. The progression of the tribal decurion from quaestorship (dealing with finance) to aedileship (dealing with administration) to duumvirate (the chief magisterial office, dealing especially with justice) is repeated by modern authorities time and again more or less as an absolute fact of civitas-life. In such circumstances one would be perfectly justified in assuming that the existence of this system was supported by a strong body of indigenous evidence; however, any close examination soon reveals that the theory as postulated is not derived from information available from the Three Gauls. A careful reading of Jullian's influential account of the Gallo-Roman magistracies shows that the greater part of his examples were drawn, not from Gaul as a whole, but from the grossly atypical province of Narbonensis, with its mass of colonies. Moreover, when even Narbonese evidence was lacking, such emphasis on highly-Romanized administrative practices allowed him to introduce the results of the vast amount of work done on the coloniae and municipia of Italy and Spain. In short, as far as the civitates of the Three Gauls are concerned, the prevailing orthodoxy as to the local career-patterns supposedly followed by their ruling aristocracies is derived not from peregrine Gallic communities but from Roman and Latin settlements clustered around the western Mediterranean.

A geophysical survey of the fort at Birdoswald, together with an area 80 m to the east and 120 m to the west of the fort walls, undertaken by the authors in 1997, was published in this journal. A detailed account of the findings is set out there, and it is not proposed to go into further detail. A further survey was undertaken in 1998, which encompassed the remainder of the vicus together with the cemetery to the south-west, with the benefit of a grant from English Heritage. More buildings were recognised below the escarpment to the south of the fort during the work on site, and two areas were surveyed in spring 2000 with the aid of a grant from the Roman Research Trust. The findings of these two latest surveys are set out in this paper, and some of the earlier findings have been reassessed. The completed survey has revealed the extent and form of the settlement, and it is probable that more is known of this fort than any other on the northern frontier.