The site had been very heavily dug over and damaged by nighthawk metal detectorists in the 1970s and 1980s, and there must have been a reason for such intensive robbing. The site had long been thought to be a Roman era mausoleum or temple (hence prompting the interest of the detectorists), but little evidence was found of this in the early years of investigation, and attention focussed on a range of buildings at the north of the site suggesting possibly a farmstead. The full temple complex was only fully revealed in the final year of excavation.

A full report of the excavations and the project, including specialist finds reports, is now available as SOAG Bulletin Number 74 (2022). Interim formal reports for each year’s work are also available Bulletins number 69 (2015), 70 (2016), 71 (2017), 72 (2018) and 73 (2019).

This page provides snapshots of the excavations completed in each of the years that SOAG worked at the site.

2014

The work at the High Wood site in 2014 was in preparation for planned excavations in 2015. Clearance of saplings and undergrowth revealed the extent of damage caused by the illicit diggings of the ‘night hawks’ who had looted the site. In places their pits have uncovered Roman flint-built masonry and in one area, where the night hawks dug trenches along its north and west sides, the stump of a massive Roman structure was revealed, only approx. 4.5 m square but with walls approx. 1.5 m thick.

This appeared to be possibly the base of a mausoleum, shrine or altar. The extensive looting suggested the former presence of votive deposits and, together with traces of several other buildings, it was possible that there may have been a Roman religious site in High Wood. The walls of the square structure incorporated some fragments of broken quern stones.

Some limited geophysics in other areas of the site did not reveal any significant features. On the advice of the Portable Antiquities Scheme Finds Officer for Oxfordshire and West Berkshire, who put us in touch with a trustworthy metal detectorist, we invited colleagues from the Oxford Blue Metal Detecting Club to run their machines over the site in June. As expected, the site had been well and truly stripped by the night hawks. However, six coins were recovered, including a Late Iron Age gold quarter stater. Seemingly out of context, this coin could possibly represent an heirloom offering to whatever god may have been worshipped at High Wood in Roman times.

2015

SOAG’s investigations into this intriguing Roman site were the first serious attempt to understand the purpose of any possible structure(s). The only other investigations here were the adjacent excavation work into a ‘mound’, undertaken by the Henley Archaeological and Historical Group in the late 1970s. The results were inconclusive then, and, despite stalwart efforts of this year’s volunteers to date, excavation has not yielded any clues for the purpose of the structure(s).  Geophysical surveys undertaken in 2014 and this year provided us with very little information.

A strong masonry, flint and lime mortar foundation has been exposed in the now-combined test pits 3 and 4, which have been extended north and slightly east in an effort to determine the direction and extent of this feature. Stratification is very affected and does not follow normal patterns. But despite this, a large quantity of a wide range of pottery from the 1st to the 4th century has been recovered, including one piece of Samian base with the stamp of PINNAE (mid/late 2nd c). Other finds have been scanty with no bronze artefacts or coins (as expected) except for two tiny fragments of sheet bronze as well as small amounts of bone, some burnt; quantities of calcined flint pieces; but particularly interesting are the number of pieces of quern/millstone of various provenance which total some 16 pieces to date. Eight test pits have been dug, two of which have been combined as above, with five revealing no evidence of structures. 

A separate 3 x 4 m grid, with extensions, to the south of the interrupted east/west cross-site test pit trench, has taken up a lot of our time.

This contained nothing, but on the final clearing to the north, an agricultural digging fork and spade were recovered with, finally, five modern (within 30 years or so) plastic bags packed with Roman pottery!  On processing (washing) the contents of one bag, four pieces of small conglomerations of iron were noted and these were found to be chain mail!  (The cache was subsequently identified as having been built by an illegal metal detectorist in the 1970s and he hid the plastic bags there!).

2016

2016 has been a successful season – over 40 volunteers have clocked up over 330 days on site. This season’s discoveries – 14m of walling still continuing into the eastern and western baulks at each end of the trenches; a corner in the middle and a probable doorway threshold, internal mortar flooring; a paved level formed of tightly packed flints (possibly a yard area?); and a wall first found in 2015 which to our surprise did not meet the east-west wall – all pose wonderful questions as to both the nature of the structures and of the whole High Wood site, to be resolved in future years’ work.

In addition to the extensive damage caused by illegal metal detecting, we are becoming more and more intrigued by the possibility that the site has been excavated by extensive antiquarian work: only in the last days of the season were small areas of undisturbed stratigraphy reached. Following exploratory work in 2015, sixteen varying sized grids have been opened within three overall trenches and have provided much more information about the site. Despite the very serious damage caused by illegal metal detecting activity since the 1980s in particular and, we suspect, earlier antiquarian probings, we have reached uncontaminated stratigraphy in certain areas of the site.

Despite variable (a euphemism for cold, rain and hail) weather patterns, progress was made in opening up two grids just to the North of the test pits of last year. In both of these, which are divided into quadrants, sections of foundations have been revealed and it will soon be possible to understand some of the alignment of the structure(s), although we have taken the trenches down only to a maximum of 50 cm so far. Again, some of the masonry is quite substantial and very similar to that found last year. Very large quantities of CBM have been encountered, including some fine complete floor tiles. Three coins have been found, of which one is in excellent condition and which are awaiting identification, as well as a wide range of pottery.

2017

Test pitting and excavations in 2015 and 2016, respectively, have produced evidence of walls and floors of a substantial Romano-British structure with a threshold of occupation of the site in the Late Iron Age. The extent and function of the building is not yet known but targeted geophysical survey and test pitting should inform the excavation strategy. This year it was planned to open a trench joining the space between the walls and floors revealed in 2016 in order to gain a better idea of the layout of the building.

In addition, when the team were asked to test pit and auger along possible wall lines to help inform future seasons’ planning, they kept finding walls extending in all directions into the woods! The final act in the field was for the team to protect the exposed areas of the site for the duration of the winter season by a layer of sturdy terram. The winter season sees a lot of work by the team to process the finds, sort the paperwork, commission finds analysis, consult experts and decide on the programme for next year.

2018

The results of the 2017 test pitting exercise were as exciting as they were unexpected. The complex is clearly much larger than originally envisaged, with some 35m of wall having been revealed, and there are indications that this may extend for up to at least a further 10m to the east. Furthermore, the walls and building collapses within the pits appear to be undisturbed and there is a good chance of discovering intact stratigraphy. Our objectives for the 2018 season were therefore:

• to extend the previous main trenches eastwards to see whether the range of rooms continues,
• to examine the north-west corner and seek evidence of further rooms and construction and
• to conduct a series of test pits to further prove the main wall and identify the full extent of the site.

Test pitting along the western line of the site enabled us to find the S-W corner and excavation on the south wall appears to have confirmed its alignment – a north-south size of approx 40m. The archaeology on these boundaries is not as well preserved as on the north and western boundaries.

Following initial processing of this year’s finds, we were able to start looking at finds recently received from the Henley Archaeology and Historical Group (HAHG), who, from 1977-82, dug a mound at High Wood which they thought would be the base of a windmill (the field is called ‘Windmill Field’). Instead, they found Roman building remains, including pottery, coins, nails and CBM. So instead of a windmill base they found what seemed to be the spoil heap of a dig from many years before (trees growing on top of the mound we now know are over 150 years old). We have now inherited many bags of pottery, CBM and other artefacts. The coins were given some years ago to the Henley River and Rowing Museum.  Interestingly we have identified some shards of medieval pottery from the mound, leading us to suspect the first digging may have been to re-use the flints for local house-building, though an antiquarian dig cannot be ruled out at this stage.

2019

Our objectives for this year were to:

• try to get a full plan of the site, hopefully finding the last 2 elusive corners of the large building enclosure which appears to extend over at least 1600 square metres,
• finish investigating the range of rooms along the northern wall, including looking for evidence of blacksmithing,
• explore the building more towards the centre of the complex of which only one wall and a corner have been found so far.

We opened the season’s initial two trenches to further investigate the range of buildings which incorporates the northern wall of the complex. The north wall was apparent in both trenches and the parallel inner wall was missing where expected.  Trench 8 did include a later, apparently less expertly built wall which forms part of a building situated within the complex. A trench extension followed the line of this later, rather shoddy wall, aside from which the whole trench has been almost entirely devoid of building material, and in general has produced disappointingly few finds.  More recently, a pit to the south of the wall was discovered and an amount of pottery has been found, including a fine Moselkeramic beaker sherd and also bone, teeth, and shell.

In the other trench (Trench 9) the northern wall of the building complex exists but again the inner wall has mostly disappeared. The tumble of flints and underlying CBM is also clearly missing from a whole area, a further demonstration that the site has been cleared, possibly in antiquity.

Later on, however, post holes were found right in the centre of the gap meaning that, if these were contemporary, the gateway could only have formed a pedestrian entrance.

Other test pits were dug amongst the mature trees in order to trace the extent of the eastern and western boundary walls but disappointingly the walls seem to disappear into a spread of flint rubble and CBM so we may never discern the exact position of the southern wall.

2020

We had planned our operations to start at end of March, however the Covid pandemic intervened! When eventually the rules were relaxed, we were able to consider some work for the year. Covid19 rules and regulations limited what we could do but, with a restriction to six people allowed to meet in the open air, and maintenance of 2-metre distancing, disinfection, etc., we felt that something could be arranged.

We conducted a Risk Analysis and concluded that we could not undertake a full project in 2020 but could open one trench with six excavators – plus a remote finds processor, and the site director. The limitations implied that we should restrict our volunteer arrangements to our core team (based on previous commitment to the project). Although we were conscious that we could be accused of elitism, this approach seemed to be accepted within our volunteer community.

We decided to open one trench of 8m x 6m over the western enclosure wall to ask 3 questions:

1. We knew of the presence of remains from an unrecorded test pit in 2019. A proper record was required.
2. Was there an entrance opposite the gateway seen in the eastern wall?
3. Was there a roofed room abutting the enclosure wall and could we identify an inner wall, post pads, post holes or beam slots to support a roof?

In the event, we were able to record a run of wall; there was no entrance way; and, although there was a significant roof fall, there was no sign of support for this.

2021

We entered upon our last season of examination of this site with an understanding as follows:

• In summary, a range of rooms measuring externally 47m x 4m which formed the northern extent of a rectangular enclosure whose full extent could not be determined due to utter destruction to the south. These buildings overlay evidence of earlier use of the site. However, in 2019 we had identified a well-constructed gateway in the eastern wall and work on the western wall in 2020 had not found an opposite gateway but it was noted that in both areas there were significant roof falls. This led us to conclude that the enclosure wall supported some form of “cloister” or covered walkway.

When excavations commenced in 2016, we had believed that this was the site of a temple. This understanding was based on little evidence although there was an unsourced local rumour to this effect. It should be noted that excavations up to 2020 had revealed nothing to support this identification and we entered upon our work this season to examine an eastern and northern wall of what appeared to be a rectangular building lying at an approximate mid-point immediately to the south of the northern range of buildings.

We were surprised – to say the least – to discern the walls of two rectangular buildings which were laid out as the ambulatory and cella of a Romano-Celtic Temple.

All walls in the ambulatory and cella are approximately 60cm in width and, although many of the outer flints were knapped, the construction was somewhat irregular in contrast with the regular walls of the northern range and enclosure of what we can now identify as the temenos. It might be concluded that the former were built for supportive strength rather than aesthetic appearance; they were possibly intended to be covered by wooden flooring, whereas the latter were meant to be seen.

Evidence of robbing was provided by an undated metal tool found wedged under the base of the north-east corner of the ambulatory within the point of its west return. It was not possible to give a date to this robbing.

The presumptive western ambulatory wall was not present, with neither indentations in the natural layer nor mortar flecks evident. A sondage across the likely line also failed to produce any evidence.

The Romano-Celtic Temple at High Wood

Throughout the inner area, the stratigraphy was a uniform sandy soil containing sparse flints and a few small coins only. The brown subsoil, which underlay several parts of the walls did not continue into this area. The finds of two clay pipe fragments in this matrix, and one lying on the natural clay, supported the conclusion that this area had been subject to previous excavation and back fill with this sandy soil. These fragments of clay pipe have been dated to the second half of the 18th century by the diameter of the bores (2.4mm diameter = 6/64ths), which is thought to imply this date range. This accords approximately with the dating of the nearby spoil heap of the ‘mound’ and indicates the likely era of the earlier excavation.

All the walls in the ambulatory and cella have been subject to subsidence in various degrees (for example, northern cella wall by 3°; ambulatory eastern wall by 2‒3° south to north). This dip was aligned towards a pit which underlay the north-western corner of the cella – which has been truncated by and was found to have tumbled into a pit. The pit, which had not been excavated previously, was cut into the natural with near- vertical sides and was 84cm deep, approximately oval and measuring 4.5m x 2.7m. It was filled with wall  rubble and CBM and contained a number of interesting finds which were interpreted as ‘votive’ gifts to the temple. These included a sherd of a C2‒3 Moselkeramik indented beaker, a Palaeolithic hand axe and 12 coins of dates ranging from the 2nd to 4th century AD. These were interpreted as having lain originally under the floor of the cella – possibly some having fallen through cracks in the floor – as suggested at the Romano-Celtic Temple at Bourton Grounds, Bucks. This pit was interpreted as the surface manifestation of an underlying solution sinkhole in the underlying chalk, such as are not uncommon in the Chilterns.

There was little evidence to indicate the form of an overlying structure but it is likely that the cella was a substantial half-timbered tower infilled with plaster and floored with wooden planks, similar to that proposed at the Wood Eaton Temple. It is possible that the 18kg of painted plaster found in the mound spoil heap originated in this structure. There were a few small fragments only of roof tile in this area but the mound had been found to contain a metric tonne which probably came from here and the adjacent rooms in the northern range of buildings.

For further details, refer to the full report of the excavations and the project, including specialist finds reports, in SOAG Bulletin Number 74 (2022).