Cramond (probably caer Amon, (‘the fort of the Almond’) is a fishing village recently absorbed into the north-west corner of the City of Edinburgh. It lies on the east bank of the River Almond, at the river's entry into the Firth of Forth through a gorge of sandstone rock: a narrow shelf along the east bank provides quay space for a tidal harbour, whose exit is protected from north and east by Cramond Island. The parish church (from 15th century), sometime schoolhouse (18th century) and manse (18th century) overlie the principia and part of the retentura of a Roman fort. A tower (15th century) and mansion house (17th and 18th centuries) stand near. All of these incorporate Roman stones. Medieval Cramond Episcopi once overlay the praetentura; but the cottages, and Roman remains under them, have been cleared by the lords of the manor to make a park.

Bbrockley Hill in Middlesex lies one mile south of Elstree Village and is bisected by Watling Street, which now forms the boundary between the London Boroughs of Barnet and Harrow. It is generally considered to be the site of Sulloniacae which was mentioned in the Second Antonine Itinerary as lying twelve miles from Londinium and nine miles from Verulamium. Excavations from 1937 to 1974, whilst failing to confirm this, have provided evidence of a flourishing coarse-pottery industry there in the first and second centuries. The manufacture of bowls, flagons, jars (some with lids), mortaria, platters and tazze is well attested.

This paper is intended as an attempt to assess the importance of the importation of overseas foodstuffs and other organic commodities to Canterbury during the late Iron Age and Roman periods through the evidence of amphorae. The material has been collected during the programme of excavations conducted by the Canterbury Archaeological Trust at the following sites around the town (FIG. I); Rosemary Lane Car Park (The Castle Vol. i of the Archaeology of Canterbury series); 77–9 Castle Street, 3 Beer Cart Lane, 69a Stour Street (CBR I–IV in Vol. 6 forthcoming); 16 Watling Street and Marlowe Car Park sites (MI–IV in Vol. 5 forthcoming); St. Gabriel's Chapel and Linacre Gardens in the Cathedral Precincts (Vol. 4 forthcoming); the Mint Yard and ‘Aula Nova’ (Norman Staircase) also in the Precincts (Vol. 3 forthcoming); St. Radigund's Street (Church Lane, Vol. 2). Although excavations are still in progress it was felt wise to attempt an analysis of all the material recovered only up to the excavation of the Marlowe IV site in 1980, the 234 kg of amphorae sherds unearthed up to that date being regarded as a sufficiently large sample to be statistically reliable, in general terms, towards reconstructing the pattern of amphora importation into the town. Further excavations might yield further types and help to refine the quantified figures, although they should be of little overall consequence.

The Roman fort at Mollins lies at about 75 m above sea level, on the south bank of the Luggie Water, a major tributary of the River Kelvin (FIG. I). The site (National Grid reference NS 714 719) lies on the side of a small ridge, and faces northwards over gently undulating ground. The substratum is glacial till and, although the site is moderately well drained, under natural conditions the broad flood plain of the Luggie Water, which borders onto the northern edge of the fort, would have been formerly very poorly drained.

The archaeological site comprises a Roman fort which was probably built and later deliberately demolished during the Agricolan campaigns around a.d. 80 to 90.

In 1990 the Glamorgan-Gwent Archaeological Trust was approached by Gwent Health Authority with respect to the archaeological potential of the areas of three developments proposed in the grounds of St Cadoc's Hospital on Lodge Hill, Caerleon. These were a College of Nursing and Midwifery at ST 3344 9095, a sheltered housing development to be built by the Abbeyfield (Gwent) Extra Care Society at ST 3280 9097, and a housing estate on an area centred at ST 3283 9090. All three lay within the area of the Lodge Hill Roman cemetery, close to where a cremation and two tombstones (RIB 358 and 392) had been found in 1909 and a sculptured head of Attis in 1972 (see FIGS 1-2). The sheltered housing scheme had already received outline permission. At the same time a watching-brief was being carried out on the widening of Pil-bach Road and the insertion of a new sewer adjacent to the Abbeyfield site.

In late summer, sometime between cal a.d. 340–405, a hoard of tightly packed, stacked copper-alloy vessels was deposited in the Vale of Pewsey, Wiltshire. The corrosion of the vessels allowed for the preservation of delicate plant macrofossils and pollen. Analysis of this material has provided insights into the date, season and context of this act of structured deposition. A second hoard of similar vessels was deposited in the fourth or fifth century only a few miles away at Wilcot. The hoards and their deposition relate to Romano-British lifeways, at a time when the region was on the cusp of a dramatic period of change. The distribution of late Roman coins and belt fittings offers further insights into the social and economic character of Wiltshire at their times of deposition.

At present 200 procuratorial stamped tiles are known from Roman Britain, the vast majority from the City of London. A small number of other stamps are known from Westminster, Barking in Essex, Brockley Hill in Hertfordshire, and possibly Saunderton in Buckinghamshire. Eight procuratorial stamped mortaria have also been found in London. All the procuratorial stamped tiles found up to 1986, with two exceptions found too late for inclusion, are listed in RIB II. The considerable number of stamped tiles found in London after this date, almost all of them unpublished, are included in an appendix to this report.

It is an unfortunate fact that historians of the Later Roman Empire are forced to derive much of their chronological material from chronicles and consular fasti. I say unfortunate, for chronicles are not nearly as accurate as modern historians would like them to be. This is because they were an exceptionally difficult type of history to compile, prone to error at every turn, from the gathering of sources to the copying of the final text, and because most chroniclers were essentially ‘amateurs’, having little experience in doing research, thinking or writing historically, and usually separated by considerable time and space from the events they were describing. These facts constantly rankle with modern historians, and some who have had to rely heavily on chronicles rather than attempt methodological or historiographical explanations of why they believe particular entries in chronicles are incorrectly dated, have often taken the easy way out and tried to solve their problems by emending them away: they alter the texts themselves, offering as excuses the assumptions (usually only implicit) that the chroniclers were skilful historians with access to reliable sources, that they were able to present their information accurately and carefully, and that the texts as we have them now were corrupted by ignorant and careless medieval scribes.

The extensive Romano-British coarse-pottery manufacturing centre in the Farnham area of west Surrey and north-eastern Hampshire has now been known for over half a century and several isolated kilns have been investigated. Until recently however the dating of the pottery has lacked precision owing to the scarcity of stratified and dated groups near to the kilns. As a result of this and the lack of a synthesis of the kiln evidence excavators have found the pottery of little value as dating evidence. Since 1969 a series of excavations on the Romano-British site at Neatham (Vindomi), about 8 km from the main kiln group, have produced large quantities of the pottery, much of it from a series of stratified groups. A study of this material has resulted in improved dating. This paper presents a dated typology of the more widely-distributed types and re-examines the dating of the excavated kilns.

From 1824–41 thousands of Roman coins were recovered from the Thames during the rebuilding of London Bridge and during subsequent dredging operations to deepen the river channel and remove submerged remains of the previous bridge. The coins covered a wide date range, from the reign of Augustus to that of Honorius. It is generally accepted that the great majority were probably votive offerings which had been tossed into the river from a Roman bridge on much the same alignment as its medieval successor.

The analysis of organic residues from pottery sherds using Gas-Chromatography with mass-spectroscopy (GC-MS) has revealed information about the variety of foods eaten and domestic routine at Silchester between the second and fourth–sixth centuries A.D. Two results are discussed in detail: those of a second-century Gauloise-type amphora and a fourth-century SE Dorset black-burnished ware (BB1) cooking pot, which reveal the use of pine pitch on the inner surface of the amphora and the use of animal fats (ruminant adipose fats) and leafy vegetables in cooking at the Roman town of Silchester, Hants.

This paper presents the results of a series of geophysical surveys of Roman forts and their environs carried out by the Gwynedd Archaeological Trust. An almost complete plan of the early wooden fort at Llanfor has been produced. Its densely-packed interior demonstrates heavy garrisoning, probably during very early Flavian campaigning. The small square auxiliary fort of Caer Llugwy has been shown to be a contraction of a larger rectangular fort. Surveys at Pennal, Caer Gai, and Canovium revealed a wide range of extramural development including vici in the form of ribbon development.

The significance of the well-known central roundel of the mosaic from Hinton St Mary, Dorset, UK, which carries a young male head with a chi-rho monogram behind it and pomegranates on each side, has been much discussed. This paper marshals evidence which suggests that the head is not a representation of Christ, but of one of the emperors of the House of Constantine, perhaps Constantine himself. Some of the implications for the nature of fourth-century imperial culture are discussed.

Disarticulated human remains were recovered from a first-century fort ditch at Vindolanda on the north-west frontier of the Roman Empire. Ancient DNA analysis revealed the skeleton to be that of a male individual and forensic taphonomic analysis suggested a primary deposition of the body in a waterlogged environment with no obvious evidence of formal burial. Occurrences of disarticulated human remains outside a cemetery context are often overlooked in Roman bioarchaeology. This discovery adds to the growing body of literature regarding alternative funerary practice in the Empire, highlighting that the concept of burial and disposal of the dead is more complex than ancient historical sources suggest. Details of the DNA analysis are provided in the Supplementary Material available at https://doi.org/10.1017/S0068113X1900014X.

Since 1991, recording of metal-detector and archaeological finds in east Kent has shown that Iron Age coins are far more numerous than previously thought. Their quantity and distribution suggests widespread use and acceptance. Continental links, suggesting trade activity, are also evident. A number of sites producing large numbers of coins and other evidence of Iron Age (and Roman) occupation have been identified, with chronological and functional differences being apparent. Away from these sites, significant numbers of coins have been recorded from across much of the surrounding area, enabling comparisons to be made between individual sites and the background distribution, leading to a greater understanding of Iron Age coinage and settlement in east Kent.

Pendants are frequent finds on sites of the Roman period. ‘Daily life was deeply penetrated, perhaps dominated, by supernatural considerations … in consequence talismans are common’. They were used for a variety of purposes; not only individuals, but houses, walls and even towns could be protected by them. This short paper concentrates on one particular type of pendant, manufactured from the crown or burr of red deer antler which is found predominantly on sites in the North-western provinces of the Empire. Whilst a small number of these pendants have received limited attention, discussion has centred primarily on the presence of a phallus as a decorative motif (Types 4-7 below) and the lesser known, but more common, undecorated forms (Types 1-3 below) have been relatively ignored.

Almost 16 years have elapsed since the publication, in 1959, of Dr. Philip Corder's paper on the structure of Romano-British pottery kilns. During this time much new evidence has come to light, especially in the upper Nene Valley, and the series of experimental firings carried out at Wattisfield, Boston, Leeds, Barton-on-Humbers and Chichester has greatly enhanced our knowledge of the techniques employed. The present paper is primarily concerned with pottery kilns of the pre-Flavian period in the Nene Valley and seeks to bring together, under one heading, the new and largely unpublished material from this region.

The chronology of the Saxon conquest of Britain is notoriously obscure. Among its most controversial aspects is the date of the adventus Saxonum, the appearance of the Saxons as an independent force in Britain. Only two sets of literary sources throw light on the question. A single contemporary, the anonymous Gallic Chronicler of 452, briefly noted under the year XVIII of Theodosius II (equated by Mommsen to A.D. 441–2) that ‘The British provinces, which up to this time had suffered various disasters and misfortunes, were reduced to Saxon rule’. Bede, writing in the eighth century, and basing his calculations on Gildas’ sketchy sixth-century account, supplied two somewhat later dates for the coming of the English to Britain: the first, A.D. 446–7, and the second ‘in the reign of Marcian and Valentinian’, that is, between 449 and 455. Bede's dates were generally accepted in the first part of this century, and the testimony of the Chronicler of 452, which records the end, and not the beginning, of a Saxon conquest, was dismissed as unreliable. In the last few years, there has been a tendency among scholars to consider Bede's dates mistaken, and to base the chronology of the adventus on the entry in the Chronicle of 452. Bede has not lacked defenders, however. Recently, Dr Molly Miller suggested not only that the date of 441-2 for the adventus was a figment of Mommsen's edition of the Gallic Chronicle, but also that the Chronicle itself could not be considered as a fifth-century source because it had been edited and interpolated at a later date. Miller's article raises serious questions about the value of the Chronicle of 452. Can it be used as a contemporary account of fifth-century events? Or must it be treated as a compilation contaminated by Carolingian editors?