South Oxfordshire Archaeological Group

maintaining the best standards of independent archaeology

Ascott Park project

Ascott Park is close to the village of Stadhampton, about 7 miles south-east of Oxford on the B480 Watlington road. It has a mystery at its heart concerning a lost manor house. The great Oxfordshire family of the Dormers acquired the manor of Ascott in 1518 and lived in the old medieval/Tudor manor house, of which Ascott Park Cottage appears to be a small remnant. After the Restoration, William Dormer – who earned the appellation ‘the Splendid’ by using silver to trim his horses and carriage – decided to build a grand new manor house, complete with formal avenues and landscaped gardens in the latest fashion. The house was close to completion when it was accidentally burnt down, in October 1662. It was never rebuilt and the Dormers and their successors went on living in the old manor house. The remains of the new house were pulled down, although we have no record of when. Extensive research and fieldwork undertaken in recent years, by and on behalf of Oxfordshire Buildings Trust (OBT), to try to confirm where it stood, seems only to have compounded the mystery. So, what is the story so far?

John Sykes records that it has long been believed that the site of the house was marked by a large, rectangular hollow (? cellar) on the centreline of the main avenue and fronting a wide earth bank (? terrace) that overlooks the formal gardens to the south [see aerial photo, left – The rectangular hollow and earth bank are just below the centre of the image]. Mark Bowden of English Heritage, who carried out a comprehensive earthwork survey of the Park in 2007, fully supports this accepted view.

Roger Ainslie of Abingdon Archaeological Geophysics (AAG), who carried out an extensive geophysical survey in 2007, favours the whole area north of the terrace and up as far as the main avenue, as the site of a much larger house arranged around a courtyard, perhaps open to the east. In Ainslie’s interpretation, the hollow would be a cellar under just one small part of a much larger building. Bowden disagrees with this and interprets this area as an entrance courtyard for the house which he locates at the hollow.

Brian Dix, whose Trench 7 sectioned the bank and western end of the hollow in 2009, concluded that the hollow was not the site of the 1662 house (which must therefore be elsewhere) but might represent a second attempt to build a house in the early 18th century. An excavation in 1969, led by Susanna Everett of Oxford University, also sectioned the bank and the eastern end of the hollow, but Everett concluded that there were no buildings in the hollow. So the Directors of two separate excavations 30 years apart, in different ways, both concluded that the hollow was not the site of the 1662 house.

One final possibility, put forward in 2011 in a somewhat desperate response to Brian Dix’s unsettling findings, is that the house might have been at the southern end of the main avenue, looking across a formal garden towards the terrace; in which case the hollow could be the site of a grotto/pavilion. A layout popular in the first half of the 17th century but not so fashionable in the latter half. So, much data has been gathered and there are several theories but no firm answers. How do we solve the mystery?

  • First, a thorough and critical review of all the previous work and the conclusions reached.
  • Second, a geophysical survey to merge with and extend outwards the ‘courtyard’ area of the 2007 earth resistance survey. The primary purpose of this new survey is to see whether there are any rectilinear anomalies suggesting a substantial building in the peripheral areas; but it will also enhance Ainslie’s important ‘courtyard’ area by placing it in a wider context.

This resistivity survey was conducted in 2013 – see the following links for the survey specification and geophysical survey report:

The initial critical appraisal of earlier work was refined and completed in 2015 to include reference to SOAG’s resistivity survey. However, desk-based research continued and this appraisal then became a working document for developing ideas and discussion right up to 2018 and the completion of a Project Design for a new Research Excavation. See the following links:

Ascott Park is owned by Oxfordshire County Council and an Historical Trail was opened by OCC and OBT in 2010, with signboards pointing out the ‘mystery’ surrounding the 1662 house. It is on the English Heritage Register of Historic Parks and Gardens registered Grade II, and various buildings and structures are also listed Grade II or II*.

Recent excavations

The Ascott Park project was revived as a three year SOAG excavation in 2018. Final fieldwork took place in 2021 and the project is now in its post-excavation phase. See the following links for full reports on each year’s excavations:

Short summaries of the work undertaken from 2018 to 2021 are given below.

2018

Our main effort this year was focussed on recovering the ground plan of the house and terrace (including the ‘annexe’) and examining their structural relationships.  We also hoped to begin to understand the constructional style of the basement.

Progress in the first few days was frustratingly slow as we tried and initially failed to locate Susanna Everett’s 1969 trench and were unable to find a key feature in Brian Dix’s 2009 Trench 7. But eventually beautifully constructed terrace foundations and the house south wall began to emerge from our new trenches, located in some cases by a ‘best guess’.

A quick summary of the key findings from this year:

  • The stone foundations of the terrace retaining walls appear to have survived largely undisturbed and significant brickwork of the walls remains along the northeast section and towards the southwest end. The terrace width is 10.1m (33ft or 2 rods or 2/4 of a chain) and the length appears to be 60.35m (198ft or 3 chains) giving a 6:1 ratio.
  • The terrace wall foundations are in line with the outer face of the south house wall and the house wall and east terrace foundations are of integral construction. The terrace was clearly an original feature of the design and not a later addition.
  • The house wall on the south side of the basement is 0.9m (3ft) thick and survives on its eastern side to a height of probably 1.6-2m over at least half its length. The western end appears to have been robbed out. It is of rough-hewn, coursed limestone with a rubble core, well dressed on the inside but undressed on the side facing the inside of the terrace, again confirming that the terrace is an integral feature. On current evidence the house is a compact rectangle in plan: 18.3m (60ft) wide and 14.6m (48ft) deep, i.e. a 5:4 ratio. The west, north and east walls are almost certainly robbed out but we have not yet looked for their foundations.
  • The supposed ‘annexe’ on the east side appears rather to be either a sunken yard or (perhaps more probably) a vaulted underground storage area, i.e. an extension at basement level only. The terrace retaining wall along its south side is supported on a deep foundation, the inner face of which has been robbed out exposing the rough core.
  • There is evidence from the unbroken inner face of the south house wall (the upper courses of which we exposed for just over half its length) that the basement was vaulted but we cannot yet say what form the vaulting took.
  • A post-excavation effort over two days finally pinned down the elusive Dix T7 feature, locating the probable southwest corner of the house and the southwest corner of the terrace.

2019

The 2019 excavations on the presumed site of the 17th century manor house in Ascott Park took place over 10 days between in August and September and were a great success. SOAG members were joined by diggers from other local archaeology Groups and Projects including: BARG, BAS, MAS and RCG (Risborough Countryside Group) together with volunteers from nearby villages and towns. Our sincere thanks go to them all for their hard work and enthusiasm – the success is down to them.

Our main aim in 2019 was to complete the recovery of the ground plan of the house and terrace by reopening and extending some of the old trenches and opening several new ones. As is usual with archaeology the trenches threw up some surprises but the end result was that the most of our research questions were answered and some new ones posed. There is much in the detail yet to be analysed and understood but a broad outline of some of the main findings is as follows.

The western ends of the south and northwest terrace retaining walls were fully revealed to confirm the length of the terrace, which does have a 6:1 length/width ratio. The western terrace retaining wall is robbed out but it was contiguous with the former western garden wall, as expected. There was also a T-junction at the NW corner, mirroring that found by Dix at the NE corner and confirming that there were walled enclosures on both sides of the house; a mortared floor was revealed (picture left).

This year we exposed (in part) the entire line of the house wall along the south side of the hollow which showed that this massive structure survives for about two thirds of its length. We had thought from 2018 that the western part was built at a higher level but in 2019 it became clear that the wall was built at a deep level throughout and the western third of it had simply been robbed out. The unwanted demolition material had been thrown back into the robber trench to form the ‘cut’ that Dix found in 2009 (foreground in the photo, left). This event had apparently been at a much later date than the main demolition of the building. Many thanks are due to all who helped to excavate the deep and narrow wall trench at the western end.

The width of the house (E-W) is now determined at a nominal 60+ft (18.4m) symmetrically placed about the centre point of the terrace. The depth of the house was more difficult to confirm as the north wall has been completely robbed out with little evidence of where it once was, a problem encountered by the earlier excavators. But the best evidence we now have confirms Dix’s location for it which gives us a nominal depth of the house (N-S) of 49+ft (15.0m).

The western end of the NE terrace wall (photo, left) was excavated to a greater depth to examine the partly robbed out foundations, which here form the south wall of the ‘annexe’. The area was filled with a significant quantity of broken limestone and brick and tile demolition material, the removal of which unexpectedly revealed a mortar floor. This floor is much higher up than the floor level of the basement and appears to extend beneath the terrace retaining wall. The only part of the ground plan left undetermined is that of the ‘annexe’ on the east side of the house.

A significant length of the upper level of the NE terrace retaining wall was explored from two directions but without any clear sign of an annexe east wall joining it. However, the western half proved to be partly hidden by a later and crudely built facing wall (top of picture) perhaps intended to provide some additional support. It has no secure foundation and looks decidedly amateurish compared to the rest of the masonry.

We were delighted to welcome visitors representing the Trustees of Oxfordshire Buildings Trust. Our special thanks go to OBT for their generous gift of additional funds to pay for hire of a mechanical excavator to open the trenches.

2020

Due to the restrictions imposed by the SARS-Cov2 pandemic, no fieldwork was undertaken at Ascott Park in 2020, and the plans formed were transferred to the following year.

2021

This year drew the curtain on careful and patient investigations to find the relatively scant remains of the 17th century house.  The final year excavations at Ascott Park, delayed from 2020 due to the restrictions of the SARS-Cov2 pandemic, were finally and successfully completed in May.  It had been agreed in early 2020 that there were two, possibly related, areas that needed to be addressed in the 2021 fieldwork at Ascott Park:

  1. The lack of truly convincing evidence for the line of the house north wall and,
  2. The uncertainty over the original depth of the basement cut.

Preparations for this year’s excavation included auguring in order to get a better understanding of the local geology, probing an area where we think the house north wall should be, and probing a further area where a solid piece of stone was found during backfilling in 2019 and could possibly indicate the position of house east wall.

Five ‘(very) socially distanced’ trenches were planned which, with two people to a trench, allowed a maximum of ten diggers on site.

The excavations took place over two five-day sessions in April and May. With good ground conditions and remarkably fine weather throughout, the team made rapid progress. The five trenches were divided into three essential tasks and two reserve tasks but in the event, all were opened. Two trenches were completed and backfilled early. The remaining three reached an advanced and satisfactory state by the end of the second session.

The outstanding research questions concerning the line of the house north wall and the depth and profile of the basement cut were addressed. Some of the answers we hoped for were not forthcoming, notably we failed to locate the north wall masonry that was so clearly visible to Dix in 2009 and we did not find any evidence of the supporting structures that must once have existed in the basement. But to offset that the line of the north wall, first proposed by Dix, is now sensibly confirmed and we have a clearer understanding of the basement cut profile; the latter enables us to make an educated guess concerning the post-demolition processes that have removed so much of the building archaeology. In addition, a remarkable find of a coin – a Charles II farthing dating to the 1670s – from a sealed context within the terrace fill, confirms that the terrace was open at the time of the fire in 1662, just as we originally proposed.

The field excavation work at Ascott Park is now over and we have moved on to the post-excavation phases of the project.

Some further useful links: