Britannia 2

[TG 2303]. Posthumous publication of Donald Atkinson's pre-war excavations was hampered by lack of section-drawings, discrepancies between plans and difficulties in relating small finds to structures. The two successive fora are difficult to date without re-excavation, but the first seems not earlier than AD 150; it was destroyed at a date unknown and replaced by II, a smaller and simpler version, towards end 3rd century. The plan of Forum I was of Principia type, its rather small size possibly due to the use of timber instead of stone. Three possible reconstructions are discussed. Incomplete excavation of the baths suggested a mid- or later-2nd century date. A note on the samian is appended.

[TL 9622]. As the site was being denuded by ploughing half the theatre was examined, to reveal a timber phase followed by a masonry-and-turf structure. The timber theatre may have been of Hadrianic-Antonine date; the more substantial structure of Phase II, AD 150-200, had a cavea mound built entirely of packed turf (estimated at fifteen acres' worth) surrounded by a masonry wall with external pilasters. Access for the ?4000 spectators was by external staircases. The layout of the stage and orchestra suggests the performance of religious rites connected with the nearby temple, rather than plays. The theatre was abandoned by mid-3rd century at latest, and demolished. Coins and pottery are catalogued, including a few pre-Belgic sherds.

SM 915173. For the first time a Welsh rath, or small univallate settlement, was completely stripped and several phases of EIA-RB occupation were revealed. A bank enclosing ½ acre was quarried from a small ditch in Phase I (14C-dated 3rd or 2nd century BC). There were two defended entrances, timber structures lined the internal periphery and three circular huts lay in the central area; crucibles attest bronze-working. In Phase II the defences were improved. Waterlogged deposits yielded wooden objects including ?loom fragments, and there were more crucible fragments and a 14C date of 2nd or 1st century BC. Rectangular and circular structures were occupied at this time. Phase III, 2nd century AD, is represented by two circular buildings with good internal drainage and a long rectangular building at the SE entrance, while Phase IV is slight 3rd-4th century occupation. The site has now been levelled for cultivation.

The two inscriptions normally cited in support of VIII Augusta's participation in the Claudian invasion do not in fact attest it. The decorations won by Gavius Silvanus could not relate to his post as primus pilus of that legion; they fit better an earlier stage in his career, not given on the stone. The inscription of Coiedius Candidus would surely have mentioned the campaign if it were indeed the bellum Britannicum. B D

The survey, giving many specific examples, covers materials used in the walls, roofs, floors, architectural features and heating arrangements of both civil and military buildings (see also 72/1672). AT

Six of the seven complete or fragmentary beams were associated with bath-houses, particularly furnace-rooms, and are shown to have been supports for boilers set across the stokeholes to provide damp heat. The intense localised heating of the furnace would account for the fractures often seen. See also J H Wright, Metallurgical examination of a Roman iron beam from Catterick Bridge, Yorkshire, Bull Hist Metall Group, 6, 1972, 24-7.

Variant spellings, even in official inscriptions, can indicate British versions of (1) pronunciation, (2) declensions and gender. Lists of the substitutions - vowels, consonants and contractions - found in "British Latin" inscriptions are provided.

[TR 3260]. A hexagonal structure discovered in the fort in 1923 is a font, comparable with examples from Cologne, Boppard and Zurzach. Rectangular stone blocks found in the vicinity, but with one exception since removed, probably formed the foundations of a building, perhaps a church, which either leant against the fort walls or was free-standing. The church (probably of wood, and comparable with similarly-dated examples from the Rhine-Danube frontier) was of late 4th or early 5th century date. A T