Above Aerial view of modern Carmarthen in which the outline of the Roman towns remains fossilised in the street system.
Below Plan of Carmarthen showing principal natural features, historic monuments, and excavation sites.
at right angles to the streets. These buildings were of different sizes and plans, and were a mixture of houses, shops and workshops, sometimes combining all three functions under one roof. The buildings were either of close-set timbers fixed in a continuous wall-trench, or of large-sized uprights in individual postholes. These differences indicate different roof structures. The close-set uprights were probably capped by a continuous wall-plate, whereas the paired uprights might have supported pre-formed crucks dividing the building into bays. Many buildings were hybrids. Whilst the plans and to some extent the methods of building differed from native styles, which were still dominated by the round-house, the materials of which the Priory Street houses and shops were built were the same: timber, daub, and thatch.'
A failed garden city? The Priory Street 'industrial village' is paralleled at many other Romano-British towns. Typically, a mix of stone-built public monuments and small timber-built houses and workshops characterise the first 75 years or so of urban development. Then, a phase of grand town-house building would begin, continuing for another 75 years or so. Often, the artisan houses would be swept away altogether, to be replaced by spacious residences for the local gentry, with dining-rooms, porticoes, and ornamental gardens. Does Carmarthen conform to the model?
Antiquarian records and evidence from the Church Street excavations show that there were some larger houses in the 2nd century. The detailed picture available at Priory Street, however, suggests that Romanisation was late coming and modest. Heather James continues the story: 'Some of the Priory Street buildings were disused or near derelict by the late 2nd century; others seem to have been deliberately dismantled and their sites cleared. This was followed by a period of abandonment. When occupation was resumed in the early 3rd century, it was different in many ways from what had gone before. Some activities continued - there were a number of smithies and possibly some iron smelting. More large clay ovens were built. Although several of the timber buildings associated with ironworking and baking were little more than sheds, two more substantial rectangular buildings were constructed on the site, fronting the streets. But throughout the 3rd and probably well into the 4th century, the site was dominated by a single, larger, rectangular building.'
Even this building (known to the excavators as B5) was small and plain compared with many of the grand town-houses in places like Cirencester and Verulamium. It comprised two sets of twin rooms either side of a central passage, to which were later added a verandah at the front and a large extension at the back. An early timber annex and a later timber granary imply that the owner's home was also his business premises. The granary - in the burnt remains of which charred cereals were found - was a solid nine-post structure: something much more than a domestic grain-store.
There are suggestions elsewhere of higher standards - a house with a heated room excavated in 1969, and antiquarian records hinting at continued on page 539
archaeologycurrent 535 202