Close to the western admission building, a fine temple has been found - but it is not the building described by John Wallis! The new discovery is of a type described as Romano-Celtic, with a rectangular sanctum, with bench seating round three sides, lying within a larger rectangular perimeter. But this building has been constructed with stone from quarries only used by the Romans in the earliest occupation of the site, and it had been methodically demolished by the Romans themselves, probably in the middle of the second century. Afterwards they had used the site for cremation burials, of which seven have been discovered so far. The only clue to the deity worshipped in the building is an altar some eighteen inches high, but so far only the clear word DEO, 'to the god', can be read.

The search for Wallis' temple will now continue for the rest of the summer, and it is likely to lie somewhere close by, for the temples were usually grouped together.

Elsewhere on the site, archaeologist are excavating deep down below the later stone civilian buildings, to reveal the superimposed remains of the early timber forts. The stumps of posts and wattle walls, dating to the time of Hadrian, are on view, and it is hoped that more writing tablets will be found in the layers below these.

Excavation will continue six days a week (not on Saturdays) until the end of August (weather permitting), and visitors can both watch the work in progress and talk to the archaeologists.

The Romano-Celtic temple
We were searching for 'an ornate temple' known to have been pillaged around 1700, when masons were searching for building stones, but we failed to find it. Instead, near the western edge of the site, we came across this small Romano-Celtic temple, with central shrine surrounded by a perimeter wall. It had been demolished by the Romans themselves before the end of the second century, and the site was then used for cremation burials. It is the only one of its type to be found on the northern frontier.

The examination of pre-Hadrianic levels below the civilian and Severan stone buildings, to the west of the last stone fort:
Volunteers and the Vindolanda team braved some awful weather to examine the famous anaerobic levels sealed by the later stone buildings, reaching depths of 3.75 metres before subsoil was reached. We found traces of the four superimposed timber buildings relating to the large forts two to five (dating to the years AD 90-125 or so), and below them there were ditches of the very first small fort, which lay to the east of this site.

The best preserved building was that of period IV, which alone had been constructed with large oak uprights, which survived to a height of 0.4 metres above the floor level - and 1.6 metres more beneath the floor. It was almost certainly the western end of the large 'palatial' building discovered to the east in 1992/3, and probably was the praetorium or commanding officer's residence of that fort. The part examined this year was undoubtedly the servants quarters, which included kitchens and storerooms.

Archaeologists working on the floor of the kitchen, in the period IV structure identified as a praetorium. Flooring was usually clay, on which a build up of 'carpet' (bracken and heather) had been laid, but with subsidence from the underlying ditches, the Romans had added stone flags in places.

The most precious of all finds - one of the ink documents from the floor of the praetorium - note the V notches and tie-holes. Infra-red photography will reveal most of the text.

Part of a very unusual appliqué-ware dish, featuring the god Mercury. Other fragments included the god's money-bag, a tortoise and a chicken.

A centurion's helmet plume, made of local hair-moss - once a dark brown in its natural colour.