There are various ways in which a town can protect itself from the risk of flooding, but few can be more eccentric than those adopted by the medieval residents of Monmouth.

Founded by the Normans in the floodplain of the rivers Wye and Monnow, the town was (then as now) subject to regular flooding when the rivers burst their banks. Excavations conducted over the past nine years in Monnow Street, the main street in Monmouth, have shown that for nearly three centuries, residents piled up domestic rubbish on their floors in order to raise them above the level of the floodwater.

Old broken pots, domestic implements, animal bones and other kitchen waste were all strewn on the floors, and packed down with grass and rushes in order to provide a more or less stable floor surface. In one house, the floor was piled up with butchery and tannery waste - the heads and feet of animals - which must have reeked.

In some houses, compacted `rubbish floors' have been found to a depth of six feet. The excavations, conducted largely by amateur archaeologists from the Monmouth Archaeological Society, suggest that the rubbish was not laid down in response to particular floods but accumulated gradually over time.

In the middle of the 13th century Monmouth underwent expansion when it was taken over by the House of Lancaster. With royalty in town, the earth and timber defences were replaced with stone walls, and there was a building boom. Many of those who came to live in Monnow Street were extremely wealthy. One massive stone house has been found, for instance, whose owners could afford to use vessels of rare Italian glass. Their stone flagged roof was adorned with ornate ceramic louvres and finials (smoke-escapes and roof decorations). On another house, the louvre was made in the form of a horned devil that belched smoke through its wide open mouth.

However, the town's new wealth was no protection against flooding and Monnow Street's house floors continued to rise. Some of the street's new stone buildings were set on a layer of iron forge dross sandwiched between layers of clay - a damp course in all some 18in thick.

The expansion of the 13th century was followed by a century of disasters. By the turn of the 14th century the weather had deteriorated and the floods in Monnow Street increased dramatically. About half way up the street, an industrial workshop was evacuated, and excavations suggest that the heavy sand of the River Wye had already covered the forge floor when the equipment was removed. Deposits at this level have been dated to 1315-45.

Houses in the lower stretch of Monnow Street appear to have collapsed into the flood waters. In 1348 the Black Death arrived as the final straw. The thick, unbroken layers of silt that accumulated in the ruins of collapsed buildings suggest that much of the street was abandoned for some time, and there are documentary references to brambles growing over the burgage (or house) plots.

The street was eventually recolonised, continuing to thrive to the present day. But oddly enough, although the rivers have continued to flood, Monmouth's residents never again piled up their house floors as a defence.

The effect of floor-piling has been to give Monmouth some of the best-preserved Norman archaeological evidence anywhere in Britain. Much of it lies in undisturbed stratified layers, because the town's medieval residents rarely seem to have dug pits in their floors owing to the height of the water table.

However, this priceless evidence is now at severe risk from new construction. Although protected by PPG16 - under which archaeological deposits are preserved where possible in new development schemes - the Norman deposits are so ubiquitous and rich in Monnow Street that any new foundations will cause unacceptable loss. Archaeologists in Monmouth have been trying to persuade the local planning authority to work on a `presumption against development' in Monnow Street - as they do at the Roman town of Caerwent nearby - but it seems unlikely that our voices will be heard.

Stephen Clarke is Chairman of the Monmouth Archaeological Society

When Samuel Jackson, an official of Northern Ireland's Council for Metal Detecting, tried to sell a stolen Bronze Age gold collar to the National Museum of Ireland in Dublin in 1993, he was not aware that his activities had been closely watched by museum officials for several months.

The museum, believing the collar, or lunula, was a metal-detected find and therefore state property, offered to pay Jackson a finder's reward, but its price was too low and Jackson went to Christie's in London to try his luck there. The Irish state, however, threatened Christie's with an injunction if it proceeded with the sale, and Jackson was forced to come to terms with the museum.

Tintagel, linked with romantic tales of King Arthur, is one of the most visited ancient monuments in Britain. Yet although most visitors haul themselves up to the ruined 13th century castle, few venture to the top of the so-called Island beyond the castle (originally a peninsula, whose land-bridge has fallen away).

This is a pity, for on the Island's plateau are the remains of building foundations and other features - some explored in the 1930s, many not exposed until a major grass fire in 1983 - and these suggest there may, after all, be some basis for the legendary association of the site with an early medieval king.

Until 1983, Tintagel had taken its place in archaeological literature not as a site in some way associated with King Arthur, but as an Early Christian monastery, largely as the result of literary analysis in the 1920s and archaeological exploration in the 1930s. The remoteness of the site, the apparently spartan lifestyle implied, and the nature of the buildings seemed to point inevitably towards such an interpretation. This has since coloured much of the discussion about early monastic sites, and Tintagel became a `type site' against whose evidence other sites were judged.

However, the exposure of the very large number of building foundations in the fire, coupled with the enormous number of sherds of 5th- 6th century pottery imported from the Mediterranean world (more of the so-called A and B wares than from almost all other sites in Britain put together), made the interpretation seem exceedingly suspect. Was it really credible that large numbers of monks (as implied by the number of buildings) were based here, and were the focus of what can only be assumed to be a trade network of Mediterranean products, such as olive oil and wine, to the south-west of Britain?

These considerations prompted English Heritage to invite me to undertake some limited excavations on the site. Trial work soon established the existence of a previously-unsuspected lower artificial terrace, with traces of human occupation. Although quite fragmentary, enough remained of wall-fragments and living debris to indicate there had been a focus of settlement here in the 5th and 6th centuries, as marked by the presence of the ubiquitous Mediterranean pottery (alongside possible native wares). The buildings themselves were not impressive structures, and not immediately the sort one might associate with a real-world equivalent of King Arthur; but they could have been the huts of such a figure's retainers.

In more recent excavations, in the past two years, a number of hearths have been found below the building remains. It is not clear whether they were originally inside or outside buildings, but associated stake holes suggest the presence of some wind-breaks. Open working areas for craft activity are a well-known feature of secular sites of this period in western Britain.

It now seems clear that the settlement at Tintagel was already long-established by the 5th-6th centuries. Some of the hearths appear to be associated with an earlier aceramic phase; and a late Roman date is indicated by the earliest deposits examined, where pottery apparently of the 3rd-4th centuries was found, together with a piece of fine vessel glass. There may even be indications here of a phase of Roman-period occupation, hitherto only hinted at by the presence of unstratified or redeposited sherds of Roman pottery and a Roman milestone in the churchyard nearby.

Although the date of the sub-rectangular stone buildings previously examined in the 1930s is not yet certain, they may yet turn out to be part of the 5th-6th century settlement. Features cut into the underlying bedrock indicate the possibility of an earlier phase of structures and occupation, and there is at least one large fissure in the bedrock with 5th-6th century organic material underlying the wall of one of the buildings.

Elsewhere, wherever we looked, we found yet more imported pottery: one area of 2m by 1m produced more than 300 sherds - comparable with the total count from certain other sites of this period. The existence of the settlement, and the presence of so much pottery, leave no doubt that Tintagel was indeed central to the economic activity of western Britain, and that it was a high-status secular site, perhaps connected with the rulers of the south-western Kingdom of Dumnonia.

Chris Morris is Professor of Archaeology at the University of Glasgow