The evidence for a major, post-Boudiccan Neronian building campaign in Calleva is set out and discussed and its wider context considered. It is suggested that there was deliberate investment in the civil development of the client kingdom south of the Thames in contrast to the re-establishment of direct military control to the north and in East Anglia where revival of the towns was slow to take off.

The Romano-British settlement at Woodcock Hall, Saham Toney, Norfolk (FIG. I) has been known since the mid-19th century when a local landowner, Thomas Barton of Threxton, donated some artefacts to the Castle Museum in Norwich. The site lies mainly in the parish of Saham Toney, but remains have been found in the adjoining parishes of Threxton and Little Cressingham. The site has not been identified with any place mentioned in the British section of the Antonine Itinerary, but Camboritum, the most likely site, is now thought to refer either to Hockwold-cum-Wilton or to Lackford, although the suggested translation of the name: ‘the ford at the bend of the river’ would suit the Woodcock Hall site well. The site has never been professionally excavated, either wholly or in part.

The evidence of the character and purpose of settlements previously described as defended ‘small towns’ is reviewed in the light of knowledge accrued since the implementation of Planning Policy Guidance 16 in 1990, the same year as the publication of Burnham and Wacher's survey, The ‘Small Towns’ of Roman Britain. This review focuses on four of the more extensively excavated settlements: Alcester, Cambridge, Godmanchester and Worcester. In the absence of convincing urban attributes, it is suggested that this category of settlement should more appropriately be regarded as defended villages (vici). These cluster in and around the West Anglian plain and on Ermine Street, suggesting a strategic function to protect grain and other food supplies and their movement, potentially either to the northern frontier or south to London and, perhaps, export to the Continent.

In 367, the enemies of Roman Britain concerted their attacks in what our sole literary source, the contemporary historian Ammianus Marcellinus, calls a barbarica conspiratio. In the archaeology of Hadrian's Wall, this conventionally dates the end of Period III; Period IV begins with the subsequent restoration of the frontiers by Count Theodosius. Damage and reconstruction are archaeologically attested, but how far they were due to enemy action is open to question. This paper is limited to a reconsideration of the literary dating evidence.

The 79 rotary querns described here are part of the Silchester Collection housed in Reading Museum. These were retained during excavations at the Roman town carried out between 1890 and 1909 and published annually in Archaeologia by the Society of Antiquaries. Although eleven items are marked with what are thought to be the original markings of the excavators indicating the insula from which they came, most are unmarked. It is assumed that the unmarked querns in the store are from Silchester primarily because the Silchester Collection has always been stored separately from the rest of the archaeological collection in Reading Museum. Some of the querns from Silchester may be from the Joyce excavations in the town but material from other excavations, including those by Boon, Cotton, and Richmond, was boxed separately. The finds numbers used here are those allocated by the museum and where a finds number is referred to in the text, this should be prefixed with the museum code 1995.90. The 79 querns comprise 34 upper stones, 42 lower stones, and 3 unknowns.

The Museum of London selected four individuals for multidisciplinary scientific analyses in order to establish their ancestry, aspects of their personal appearance and health. We also reinterpreted their burial context in order to better understand how identity was constructed and expressed in this unique Roman settlement. Our study discovered the presence of people with Black and White European ancestry, some of whom had migrated from the southern Mediterranean. The most surprising result was that Harper Road woman's chromosomes were male. Overall, our experience of undertaking a multidisciplinary study served to further underline the need for these different techniques to be used in combination when investigating past identities. The mtDNA results were very broad and required the mobility isotopes to better understand their significance, while the aDNA evidence confirmed the osteological analysis. In terms of public engagement at the Museum of London, the ability to determine hair and eye colour had a significant impact.

While working on some Celtic iron swords a number of references were encountered which contrasted the swords of the Celts and their methods of fighting with those of the Romans. Polybius (200–118 B.C. or later), writing of the battles of Cannae (216 B.C.) and Telamon (225 B.C.), says that the Gauls fought with great courage ‘in spite of the fact that man for man as well as collectively they were inferior to the Romans in point of arms. The shields and swords of the latter were proved to be manifestly superior for defence and attack….’ This was because the Gaulish swords ‘could only give one downward cut with any effect but after this the edges were so turned and the blade so bent that unless they had time to straighten them out with their foot against the ground, they could not deliver a second blow … their blade has no point.’

Rivenhall lies 10 miles south-west of Colchester and a mile north of the main Roman road to London, in a sheltered valley adjacent to copious springs which feed a tributary stream of the river Blackwater (FIG. I). The subsoil geology is mainly boulder clay containing random gravel pockets; this is capped by brickearth in places. There are numerous prehistoric and Romano-British sites in the vicinity.

Historically dated pottery assemblages have been widely used in the establishment of ceramic chronologies, especially those currently used for samian ware.2 The methods employed in the construction of these chronologies have hitherto been comparatively unsophisticated and the results have rarely been subjected to critical scrutiny. Further, in most of the development and use of these chronologies it has been the presence of specific types (for instance stamps or decorated sherds) that has been considered significant rather than the overall composition of the assemblages. This is despite the fundamental assumption on which the chronologies are based, that assemblages deposited at or near the same time, within the same distribution network, will be similar in composition. This assumption has itself rarely been tested to establish the limits within which other factors determining pottery supply, use and deposition, may obscure these underlying chronological structures, and render them unreliable for the precise dating upon which Roman archaeologists rely.

The Roman town of Venta Icenorum lies adjacent to the church of Caistor St. Edmund some three miles south of Norwich, Norfolk, where the circuit of its defences still survives impressively though everything within them has been flattened by agriculture. The publication of a famous vertical air photograph (PL. I) taken by the R.A.F. in the dry summer of 1928 aroused great interest, and funds were collected for the excavation of the town. The direction was entrusted to Mr. (later Professor) Donald Atkinson. Work started in 1929 and continued every year with the exception of 1932 down to 1935. Publication, however, did not keep abreast with excavation. A fairly full account of the first season's work was indeed printed, and the pottery-kilns excavated in 1930 were fully published; finally a report appeared of the pottery and coins found in 1929–31. For the rest there are the brief annual reports in ‘Roman Britain in 1929–35’ (JRS xix–xxvi), and a short summary of results compiled by Professor C. F. C. Hawkes with Atkinson's collaborations and published in 1951.

During the course of discussions on the garrison of Inchtuthil with a pupil who is preparing the site for publication, it became clear that one of the site's most serious problems concerns the composition of the first cohort of the legion. The plan of the fortress shows the barracks of nine cohorts arranged with their centurion's quarters next to the via sagularis, each with six barracks per cohort. The position of the legion's first cohort, however, which is situated on the via principalis to the right of the principia, is occupied by a cohort of ‘double’ strength; the plan shows ten barrack-buildings (FIG. I), which are fronted by five courtyard houses such as would suit the five centurions of cohors i, and nearer the principia there is extra accommodation, apparently consisting of one and a half barrack blocks and a row of tabernae, which it is not the purpose of this paper to discuss. It is clear that, whatever function these barracks served, they differ from the main range of ten.

Focusing on late Roman bracelets, and also including other relevant material culture types, this paper brings together an examination of spatial distribution, distribution by site-type, and selected specific burial contexts to investigate provincial Roman material of non-local origin. Using this methodology, it is suggested that migrant communities can be identified at Krefeld-Gellep in the Rhineland — thus demonstrating that this type of multi-layered approach can assist in unravelling the complexity of the surviving evidence. The study also shows that a bias towards military sites/large towns is a distribution pattern typical of material originating from a different area of the Roman Empire.

The Roman military presence at Dalswinton is reassessed using a range of remote sensing techniques (geophysical survey, LiDAR and aerial photography). At Bankfoot the absence of internal buildings suggests the postulated vexillation fortress was a more temporary structure; while numerous pits/ovens were identified across the interior of the large Stracathro-type camp. The primary fort at Bankhead was provided with in-turned entrances and two small annexes attached to the north-west and south-east quadrants of the fort. A third much larger annexe extended southwards down to the river. Only pits and furnaces were recorded within the annexes, two of which were expanded in Phase 2. Various buildings, including legionary and auxiliary barracks, were identified in the expanded fort of Phase 2, whose orientation remained unchanged. A mixed garrison of legionaries and auxiliary cavalry is indicated for both periods of occupation. Finally, the fort was deliberately demolished. The Roman attribution of the three nearby enclosures at Butterhole Brae can no longer be supported.

This paper is an attempt to address some of the theoretical issues relating to the study of Roman treasure hoards from Britain. Some years ago the main problems were succinctly and helpfully outlined by Reece, but the study of treasures has remained unfashionable amongst those who have been interested in theoretical approaches, and in recent years few scholars of Roman Britain have devoted much thought to their classification and interpretation. Although many of the points raised below may appear obvious, it is clear that this subject requires more intensive debate, and an attempt to set out some basic parameters is overdue.

The Roman fort of Carlisle was first discovered by the late Dorothy Charlesworth in 1973 at Annetwell Street. Excavation by her and after 1980 by the Carlisle Archaeological Unit located the southern defences of the primary fort and identified a sequence of Roman forts extending from AD. 72/3 down to the fourth century. In 1981–2 excavation at Castle Street produced a sequence of timber phases closely matching the fort stratigraphy, with military association in the finds assemblages, but lying outside the fort defences. The early phases were arguably part of a defended annexe or within a military vicus, though proof was lacking.