Renewed excavations at the Roman city of Silchester, near Basingstoke, have shown that the grid-planned Iron Age town discovered last year was much larger than initially thought.

Silchester, known as Calleva Atrebatum in Roman times, seems to have been at its largest and most densely populated before its destruction by fire in the later 1st century AD, possibly in the rebellion of Boadicea.

“Not only is there evidence of closely packed timber buildings . . . within our excavation trench, but there is also a great number of wells, greater than in any subsequent period,” said Professor Michael Fulford, of the University of Reading. “Indeed the total number of wells which predate the destruction is greater than the total number thereafter for the rest of the life of the town within our excavation trench.

“Although we cannot yet be certain either how many there were in total, or how many were exactly contemporary, the density, with intervals of less than ten metres between wells, does seem to imply a large population, with each household looking to supply its own needs.”

Although the later Roman city was powerfully defended, it seems to have been less heavily settled than before, and some inhabitants were independent-minded enough to rebuild their houses on the old Iron Age orientation, at 45 degrees to the Roman street grid although within the borders of the Roman insula or urban “block”.

“Where modern excavations have explored the Roman defences, they have consistently found mid-1st century AD occupation beneath them. This suggests that the early town extended beyond the limits of the later Roman town and evidence of this has now been recovered outside the northeast Roman defences,” Professor Fulford said.

“More important, there is little evidence for the continuation of the new Roman street grid beyond the Roman-period defences. The early town is not only densely occupied, but it also extends beyond the limits of the surviving walls, which were first constructed about AD200 also.”

Earlier Silchester was certainly defended. A geophysical survey outside the Roman town walls by Dr John Creighton, of the University of Reading, is identifying new configurations of early earthwork defences that appear to define the town limits. Such earthworks are known around other Iron Age tribal capitals, including the precursors of Roman Camulodunum (Colchester) and Verulamium (St Albans).

“This emerging picture of a densely populated early Roman town, feeding off the wealth of the client king, Togidubnus, and money pouring in in the wake of the Roman conquest of southern Britain in the AD40s, recalls the fast-growing, densely populated modern conurbations in sub-Saharan Africa, Central America and South Asia,” Professor Fulford said.

Evidence for the initial Roman conquest of the Iron Age town may also have come to light. The Roman historian Suetonius described the future emperor Vespasian in AD43-45 as having subdued “two powerful tribes and more than 20 fortified towns” or oppida. It has long been assumed that Maiden Castle, near Dorchester, was one of these oppida, and its occupants, the Durotriges or one of its tribes, but the identities of most of the towns remain a matter for speculation.

The discovery that Silchester was already a large, planned and defended settlement in Iron Age times makes it entirely possible that it was one of them: “This reminds us of the large, mid-1st century AD timber building, also sharing the Roman orientation, found beneath the forum basilica in the 1980s. While initially interpreted as military, though subsequently regarded as the first phase of the civil forum, perhaps its blatant disregard of the existing Iron Age layout of Calleva encourages us to think once more of a military interpretation,” said Professor Fulford. “Given that it is generally accepted that Calleva was controlled by the Catuvellauni tribe. . . it is almost inconceivable that there was not a Roman fort at Calleva by about AD44.

“The fact that Dr Creighton’s geophysical survey has not identified any traces of a Roman fort in the environs of the town turns the spotlight back on the town itself. This tantalising evidence of early Roman structures in Insula IX sharing the same orientation as the contemporary courtyard building beneath the basilica is leading us towards a military interpretation, occupation by the Roman army within the native town.”

Since no fort has been identified beneath Silchester, this may have been an initially peaceful occupation. But Professor Fulford expects the 2010 season to yield comparable surprises to last year and this.