Excavations at the site of the temple treasure of Senuna

Following the discovery by a metal detectorist in 2002 of a hoard of temple treasure1, excavations and field surveys have been conducted 2003-20052. Geophysical and fieldwalking surveys have revealed a LPRIA and RB linear settlement extending along the banks of a stream for at least 500 metres. Magnetrometry over approximately 7 hectares shows sub-rectangular RB ditched enclosures on either side of a ditched trackway. To the south lies a palimpsest of sub-circular ditched enclosures of probable LPRIA date3. The hoard was found on the edge of an isolated polygonal enclosure, about 14.0m across, lying to the east of the RB trackway.

Three seasons of excavation has sampled approximately 50% of the ground area of this polygonal enclosure, as well as deposits extending to the south and west. The enclosure was defined by the laying of a superficial chalk pebble surface, probably in the second half of the first century AD. This chalk surface seals earlier soil deposits and chalk pebble surfaces. The sequence started when any original topsoil was stripped, perhaps in the first half of the 1st cent. AD, to expose the natural clay on which was laid a flint gravel surface. An organic soil formed on this surface until the first of three subsequent chalk pebble surfaces was laid. Between each chalk pebble surface organic soil deposits formed, rich in artefacts and ecofacts, apparently as a result of ritual feasting and the deposition of votive material, including large numbers of Iron Age and a few Roman coins, many brooches and other personal dress items, animal bones and oyster shells.

To the west of the enclosure, a gravel yard surface, of more than one phase, overlay an organic soil resting on the final chalk pebble surface. The temple treasure hoard had been in a pit dug through the gravel yard, probably in the fourth century AD, into the latest chalk surface.

The west side of the enclosure was defined by a narrow, flat-bottomed, vertical-sided, gully, probably a beam slot for a fence. It had been dug through the latest chalk surface. After a gap of several metres, the terminus of a similar, but possibly earlier, gully was found on the south side of the enclosure. The gap between the terminals aligned with an area of worn and repaired chalk surface outside the enclosure on its south-western side and with the west side of a chalk rubble floored structure within the enclosure (see below).

At the centre of the enclosure, two clay-built hearths were exposed. One was based on the natural clay from which any pre-existing topsoil had been stripped. This hearth was associated with much artefactual and ecofactual material including a great quantity of very fragmentary calcined bone. There was also fragments of burnt bone pins. The hearth seems to date from the end of the Late pre-Roman Iron Age. Several Bronze Age artefacts were arranged in an arc around the north-western side of this hearth. On its east side, a small pit held a structured deposit of animal bone and pottery, including parts of an imported fine ware beaker. A few metres further east, a second hearth was associated with quantities of ovicaprid bones and a midden containing a substantial deposit of oyster shell.

The interior of the enclosure comprised a hollow c 1.0 m deep filled by an homogeneous, organic soil. Scattered through this deposit, but mostly concentrated near the chalk perimeter, was much material, including RB potsherds, Iron Age and some Roman coins, brooches, dress pins, glass shards, some pipe-clay figurine fragments, animal bones, oyster shells and tiny fragments of calcined bone.

Within the enclosed soil deposit was the chalk rubble floor of a small building, 2.10 × 1.40 m, with wall footings trenches and evidence for timber corner posts. Extending west from it was a line of collapsed Roman roof tile and chalk, flint and sandstone rubble.

Within the enclosure, four other structured deposits were revealed. Three lay in shallow scoops in the organic soil filling, close to the chalk perimeter of the enclosure. One contained the dismembered, but many still articulated, parts of several pigs. With these remains was much early Romano-British pottery, some metalwork and coins, a couple of BA implements, and a headless pipe-clay figurine depicting an equivocal image of Apollo holding a lyre.

Another deposit consisted of a socketed iron spearhead together with three IA coins, one each of bronze, silver and gold, and a rim sherd from a glass vessel.

A third deposit comprised one silver and six bronze Iron Age coins and three early Roman bronze coins, with sherds of early Romano-British pottery, animal bone, calcined bone fragments, two iron implements, a Late Bronze Age socketed spearhead, a Polden Hill bow brooch, at least 75 cut-up pieces of iron mail armour, a pipe-clay head depicting a male character from Roman comedy, together with a crude pipe-clay stand.

A fourth deposit was not in a definable scoop but spread over an area about 4.0 × 3.0 m, with some outliers. It extended from the third deposit towards the western end of the small building. All the objects were Bronze Age metalwork. There were 8 axeheads, 13 socketed spearheads, 2 spear shaft ferrules, 2 tanged chisels, 2 awls, a socketed gouge, a dagger and part of a sword blade. Most items were fragmentary on deposition and were made from the end of the Early Bronze Age through to the Late Bronze Age. They were redeposited on this site in the first to second centuries AD.

On the first day of excavation in March 2003, an inscribed, circular, silver base, almost certainly originally attached to the female figurine in the hoard, was found close by the findspot of the temple treasure, although possibly deposited separately from the hoard. The dot-punched inscription reads: D SENVNE FLAVIA CVNORIS V S L M


1 Reported in Britannia 2005
2 Fieldwork was in March 2003, September 2004 and September 2005, by the North Hertfordshire Archaeological Society in partnership with The Heritage Network Ltd, directed by G R Burleigh. The fieldwork has been financed by The Townley Group (British Museum), the NHAS, the BBC, Herts CC and CBA Mid-Anglia. Permission was given by the landowners, the late Mr M G Sheppard and Mr S Sheppard.
3 Magnetometer survey by Prof M Noel of GeoQuest Associates and the University of Durham.