The final day of our annual dig on the site of James I's Theobalds Palace in Cedars Park was one of surprises.

The weather held up well for the most part, and we had lots of intrigued visitors to the site -- happily this year, as last, there was lots to show them!

As predicted late in the day yesterday, the walls in trench one are indeed a discrete square feature, most likely the supporting foundations for a column or pier base.

Although we saw a near identical feature last year, we hadn't expected to find the gaps between columns to be so close - only just over three metres. This may suggest that the columns once supported a very substantial structure, perhaps of more than one storey.

We did not, however, find evidence of a column in trench three, so we will have to continue looking for evidence of column bases away from the 'loggia' boundary wall perhaps next year.

Another feature that proved elusive was the ornamental canal bounding the loggia. Although last year we found a cut feature which looked like a contender for the canal, trench two, which was placed to pick up its projected line, failed to deliver the goods -- instead providing us with a highly compacted deposit of brickearth and rubble.

Worried that the canal may have turned and perhaps run under the main wall into the loggia itself, we extended trench one to see if there were signs of a culvert anywhere along the main wall, but here we also drew a blank.

This will require some considerable re-evaluation of the canal evidence -- whether the line of the canal is radically different than was thought, of if the interpretation of the cut we saw last year needs to be revised, will be a job for post excavation (see future issues of Society News for progress on this).

The head-scratching provoked by the lack-of-canal was as nothing, however, compared to that caused by an unassuming pottery sherd that appeared unexpectedly from trench one.

This rim-sherd is unmistakably a Roman black-burnished ware, and quite out of place in a Jacobean palace. It was not stratified in a Roman context, and there is no known Roman activity in Cedars Park -- there are several possible explanations for its presence, including agricultural activity (ploughing and field-manuring can spread unrelated artefacts over a huge area), but there have been a few isolated Roman pot sherds found in Cedars Park before, and this one certainly provides food for thought, not to mention something extra to keep our eyes peeled for next year!

We are as ever extremely grateful to Broxbourne Borough Council for their kind cooperation and support, and especially to the parks manager Adrian Hall, who foolishly valiantly lent a welcome hand as we struggled to remove some extremely compacted and sun-baked brickearth deposits.

We have tentative plans to expand the scope of the dig next year (watch this space) when we will return to find out more about James I's loggia.

Next week we will be digging in Forty Hall, Enfield on the site of Henry VIII's Elsyng palace -- with school visits and several students set to join us we hope the weather forecasts are inaccurate and we can manage to stay dry for most of the week.