Britannia 25

Reports the discovery and scientific analyses of three coarseware vessels containing copper alloy coin blanks, pellets, and iron coin dies, found in the vicinity of the Roman small town Magiovinium during 1990. `The pottery' by P T Marney (3-4) speculates that the three local sand-tempered fabric vessels are late third or early-fourth century in date. `The coin blanks and pellets', detailed by M Ponting & R J Zeepvat (5-15), were subjected to chemical and physical analyses. The former involved sampling metal from the entire cross-section of the objects using a drilling technique to minimise visible damage, the specimens then being analysed by atomic absorption spectroscopy. The physical examination involved optical microscopy and metallographic examination, followed by scanning electron microscopy and EDX analysis. Two types of copy are thought to be represented in the hoard, and their metallographic composition appeared subject to a far higher level of standardisation than had previously been thought. `The dies' by M Cowell & J Lang (15-17) were subject to radiography, atomic absorption spectrometry and colorimetry, as well as visual examination, it being concluded that the artisans who made them knew how to carburise, or select carburised, material, and to quench and temper metal. The diagnosis of a coin manufacturing hoard is thought to be the most likely and the effect of the metallographic analyses has been to suggest a far higher degree of skill and control during manufacture. However, no finished coins were present, and the circumstances of deposition are unknown.

First of two parts, ultimately aimed at establishing which auxiliary units roamed RB and where their forts of garrison were at different periods, mainly on the basis of inscriptive evidence.

A consideration of expenditure, consumption and other non-productive economic activity that may be associated with villas. There is a deliberate attempt to avoid direct equations between size/complexity and wealth/status. The simple four to five room villa plan, exemplified by Park Street and found in Gaul, is likened to the basic mansiones plan. This association with the military is thought to indicate prestige and therefore represent the elite of the areas they are found in, much as the more elaborate Sussex villas, such as Fishbourne, signify their local elite.

Reports a new reconstruction of the Silchester church for the site museum, using the pes monetalis as a unit of measurement.

Isolates and discusses principles of design indicating a coherent system of planning, as observed in military buildings. Historical sources, architects'/builders' equipment, building methods, and aspects of building design including plans and layouts, are brought to bear. Examples include Caerleon, Cowbridge, Inchtuthil, and Hod Hill, with analysis based on modular systems rather than exact measurement. There is an `Appendix: Roman building plans' (163-4).

The mud cliff and high intertidal clay-peat shelf between the R Rhymney and Towick Reen was surveyed to relocate features recognised during earlier aerial and ground surveys, and identify any that had been exposed during subsequent coastal erosion. Fieldwork concentrated on features seen in the mud cliff around the headland of ditch B4, and the embayment to the north, site C. Features in the mud cliff sections were cleaned and a short length of each excavated, in order to obtain greater artefactual and palaeoenvironmental assemblages.Built ditches occur in sub-parallel groups, suggesting they were laid out in large blocks. A droveway cut into, or formed the boundary of, one block. A second block aligned with surviving drains. It is conjectured that the Wentlooge Level was laid out in the Roman period, being drained by extensive ditches and protected from inundation by the sea wall. Circa six of the forty-three ditches identified had become infilled by the time the sea defences were set back and are fully sealed by Wentlooge palaeosol. Twenty of the identified ditches are infilled with pale brown Rhymney formation sediments, indicating activity during the repositioning of the sea-walls. Natural palaeochannels are also in evidence.The material remains are then reported, beginning with Romano-British pottery, which was recovered from all features exposed in the mud cliff. Fired clay was recovered from almost all the artificial features in sites B and C. A bronze finger-ring of third-fourth century date, a fragment of iron knife blade, similar nail shaft and third unidentified iron fragment, a small chisel-cut fragment of cast lead sheet, two glass beads and a fragment of window glass were also found. The `Textile' is documented by J P Wild (196), iron-making slag from the beach area between sites B and C is noted, and a `Flint arrowhead' described by S H R Aldhouse-Green (196). Details of unworked stone are followed by `The animal bone' by S Hamilton-Dyer (197-200), and `The pollen' by M Keith-Lucas (201-2). `Plant macrofossils' by M Robinson (202-5) and `Charcoal and wood' by Rowena Gale (205-6) complete the specialist reports.In conclusion, plant and animal bone evidence combine to provide a picture of an economy including animal husbandry, the raising of horses, wetland reclamation and probable cereal cultivation. Iron-making was also evinced, though there is little indication of wealth. Earlier occupation evidence in the form of the Neolithic arrowhead, and early-mid IA palaeochannel filling is noted. The case for RB initiation of Wentlooge wetland reclamation is thus thought to be strengthened. IH

A single core taken from the Moss was analysed, indicating little activity between the Neolithic and IA, followed by massive deforestation during Roman occupation, in association with military activity rather than agriculture. The results are placed in the context of existing archaeological knowledge.

Describes two marble heads, an uncommon find type. The Shapwick example is thought to be a portrait bust, and the Sapperton, a goddess.

Reports what is known of a decorated lead tank found in Icklingham in 1725. Antiquarian accounts had previously formed the basis of record for this piece. However, the discovery of an illustration made in 1789 prompts speculation that a previously unprovenanced tank in Cambridge Museum of Archaeology and Anthropology is in fact none other than the lost lead tank of Icklingham.

Latest in a series of articles disputing interpretations of the Gallic Chronicle, this time taking issue with arguments surrounding imperial seniority and regnal years (see also 92/526).

Tribute to the eminent Romanist whose distinguished career included important works on the towns, countryside and place-names in Roman Britain, the OS map of Roman Britain, and lecturing at Keele, as well as time spent being president of the Society for the Promotion of Roman Studies and review editor of Britannia.