Archaeological recording at The Leasowes, Halesowen, Dudley, West Midlands:

centred on NGR: S0 898984

by Christopher K Currie BA (Hons), MPhil, MIFM, MIFA

CKC Archaeology

Report to Dudley Metropolitan Borough Council

July 1998

Contents

Summary statement

  1. Introduction
  2. Historical background
  3. Strategy
  4. The results of the measured surveys
  5. The results of fieldwork along the streams
  6. The results of the excavations
  7. Discussion
  8. Conclusions
  9. Recommendations
  10. Specialist reports
  11. Archive
  12. Acknowledgements
  13. References

Summary statement

The Leasowes is a Grade I landscape listed in the English Heritage Register of Parks and Gardens of Special Historic Interest in England. It was created by the poet, William Shenstone between 1743 and 1763, and is considered to have had considerable influence on contemporary landscape design. It is presently a public park and golf course. It has recently been awarded a grant from the National Lottery Fund towards the restoration of the 18th-century landscape.

Dudley Metropolitan Borough Council requested C K Currie of CKC Archaeology to undertake archaeological investigations in advance of this restoration. The fieldwork was carried out between April 27th and June 19th 1998.

Surveys included making measured plans of two derelict pond dams, plus the earthworks associated with a grotto-like feature known as the Chalybeate Spring. Excavations revealed the possible sparse remains of a feature called the Dripping Cascade. The partial remains of a brick base marked the site of an obelisk, but nothing could be found of his two roothouses. The latter were believed to be ephemeral wooden structures, possibly made from roots. The remains of a cascade from the Heart-shaped Pond were well preserved at the western end where they had been buried following a possible diversion of the stream. This was found to be a stone lined channel that was possibly repaired at a later date, and given a brick floor and edges within the older stone channel.

Substantial remains of a Gothic folly known as the Priory Cottage were found next to a golf course tee. This originally appears to have been a stone structure, with a long Gothic facade facing the house, and a small cottage behind. It is thought to have been built c. 1757. Numerous fragments of painted ceramic heraldic shields were found in the rubble confirming Shenstone's report that he had decorated the cottage with Gothic shields. Between 1859 and 1887 the cottage appears to have been enlarged by the addition of a brick extension on the SW side of the building, thought to be a kitchen. The interior of the cottage was tiled with ceramic floor tiles, and a large kitchen range placed in the extension. Internally some older walls may have been taken down and replaced in brick. It is thought that a cellar was dug at this time.

The results of these excavations and surveys have added considerably to our knowledge of Shenstone's landscape and the methods he used to construct it. There were a number of surprises, not the least that his work was more complex than previously considered. His cascades and other water features seem to have been more elaborate than more recent commentators have allowed. There is tentative evidence to suggest their ornamentation went beyond the construction of the features themselves. The regular occurrence of compacted pinkish surfaces near his features suggests the use of coloured materials laid on to parts of the surrounding earth for ornamental effect. That these would become rapidly overgrown without maintenance suggests that he may have originally conceived a long-term management scheme of upkeep. His water features may have survived in recognisable form until at least 1859-85, when they seem to have been repaired in brick. Their final dereliction seems to have occurred within 50 years of this rebuilding, and much evidence seems to have been removed during a municipal clear-up this century.

The evidence suggests that the durability and extent of Shenstone's work has been under-estimated. If this can be shown to be the case for his water-features, those most prone to erosion and decay, it bodes well for other structures he made that have not yet been examined to the east of the house.

Archaeological recording at The Leasowes, Halesowen, Dudley, West Midlands (centred on NGR: S0 898984)

This report has been written based on the format suggested by the Institute of Field Archaeologists' Standard and guidance for archaeological evaluations (Birmingham, 1994). The ordering of information follows the guidelines given in this document, although alterations may have been made to fit in with the particular requirements of the work.

1.0 Introduction

(Figs. 1-2)

The Leasowes is a Grade I landscape listed in the English Heritage Register of Parks and Gardens of Special Historic Interest in England. It was created by the poet, William Shenstone between 1743 and 1763, and is considered to have had considerable influence on contemporary landscape design. It is presently a public park and golf course. It has recently been awarded a grant from the National Lottery Fund towards the restoration of the 18th-century landscape.

Dudley Metropolitan Borough Council requested C K Currie of CKC Archaeology to undertake archaeological investigations in advance of this restoration. The work was carried out between April 27th and June 19th 1998.

2.0 Historical background

The Leasowes is a large public park within the urban conurbation of West Midlands. It is surrounded entirely by housing and industry. The open areas of the park are managed as a golf course. There are also facilities for children (a playground), and angling on the Priory Pool. There are a number of streams that empty into the Priory Pool. The stream valleys are wooded for the most part. There are a number of smaller ponds along the line of these streams, but these are purely ornamental at this stage, and are not used for any formal activities.

The park was laid out as a designed landscape in the English Landscape style by the poet, William Shenstone, between 1743 and 1763. Prior to this the land had been used as farmland, and little has been recorded of its pre-18th-century archaeology or history. Shenstone ornamented the natural landscape by building artificial ponds, cascades and follies, including a mock medieval 'priory' ruin. He also improved the views and vistas obtained from the site by judicious planting.

Under Shenstone, the site gained much popular renown, and was considered to have had much influence on later 18th-century landscape design. This continued after Shenstone's death in 1763, but the encroachment of urbanisation on its views and vistas in the 19th century led to its slow decline. Despite the removal of nearly all of Shenstone's ornamental features, the overall landscape has survived well into the late 20th century.

The site has suffered from the usual ravages common to public parks in urban areas. There has been some vandalism, and the site suffers from litter and the unstructured access to dog-walkers. The open landscape of the park is currently managed as a golf course. This frequently conflicts with the original ethos of the landscape by introducing large earthworks (bunkers) into the park, and removing features that were part of the original design. Until recently, management of the golf course had been undertaken without consideration being given to the historic landscape.

The stream valleys are largely unaffected by the golf course. However, the features along these have been much neglected. Ponds have become silted or dewatered through breaches of the dams. Cascades have become much altered and ruinous, and everywhere has become covered in silt, years of decaying leaves, and rank undergrowth.

3.0 Strategy

(Figs 3-4)

3.1 General strategy

The archaeological strategy followed that outlined in the brief issued by the client, and the project design written in response to that brief.

Access for machinery was considered impractical on most of the sites, and so the trenches were all excavated by hand. Topsoil and rubble overburden were removed by mattock and shovel. Once significant archaeological layers were encountered, they were excavated by trowel and hand shovel. The work was supervised on site by C K Currie MIFA.

The trenches were recorded in plan and by sections at a scale of 1:20. The trenches were excavated stratigraphically, according each context with a separate number. Single-feature planning was undertaken where suitable remains were encountered. All features were recorded by monochrome and colour (slide) photography, using appropriate scales. The trenches were backfilled by the archaeologists before leaving the site at the completion of the fieldwork.

All finds from stratified layers were retained, with the exception of post-medieval brick and tile and oyster. The latter was discarded on site after having been suitably sampled. Twentieth-century pottery from rubble layers overlaying the remains of the Priory Cottage (demolished c. 1950) was not retained except for items of particular interest. Samples were taken from other unstratified layers for the archive where this was considered necessary.

3.2 Specific strategy

The brief mentioned a number of possibilities that were, on reflection, impractical. The Project Design outlined the practicalities of the proposed work, pointing out where archaeological work was unlikely to produce realistic results. Alternative work that could produce more useful results was suggested where appropriate. Alternatives were agreed with the client verbally and in writing. An amended Project Design was appended to the brief.

3.2.1 Stream valley: Beechwater to Priory Pool

i) A stream walk was carried out to locate worked stone, brick and other man-made objects of historic interest. Obviously modern materials (such as road stone) were not included in this survey. The materials were collected, and stored on site for possible future reuse. The client provided for a location to store these on site. It was considered impractical to ask the archaeologists to remove these materials to an off-site store at this stage as the brick quantities were quite high.

A typology of brick was made, according to size and fabric, with suitable samples retained for the archive.

ii) Using plans provided by the client, the archaeologists marked where in-situ materials existed along the stream valley. Certain structural remains needed more detailed recording. Following preliminary study of the remains, it was decided to change details made in this section of the project design. For example:

a) Profiles of derelict pond dams along the north stream were made

b) A detailed plan of the Middle Pond and Chalybeate Spring area was made.

c) A plan was made of the pond dam below the Beechwater.

iii) It was considered that at least one of the presently disused pond dams along the valley needed to be explored for information about its make up. The dam of the Middle Pond best suited this purpose.

3.2.2 The Dripping Cascade

The mud currently covering the conjectured site of this feature was removed to reveal the underlying remains.

3.2.3 The Obelisk and adjacent path

Excavation around the site of this feature was carried out to examine the full extent of the feature, its size, and method of construction. As the site of this feature was well known, it was also thought to be useful to examine the nature of Shenstone's paths in this vicinity.

3.2.4 Chalybeate Spring

The mud currently covering the conjectured site of this feature was removed to reveal any underlying remains.

3.2.5 The Priory Gatehouse

The brief proposed a geophysical survey in advance of excavation. However, the site was largely covered by a tarmac road, and the rest of the area was heavily disturbed by tree roots and undergrowth. Such conditions made geophysical survey impractical, and it was recommended that geophysical survey was not carried out here. These circumstances also made trenching on the site of the gatehouse unrealistic. However, despite heavy tree and undergrowth cover, a small trench was attempted on the site of the roothouse without success.

3.2.6 The Ruinate Priory

An attempt was made to recover as full a plan of this structure as the prevailing circumstances allowed. This was restricted by the encroachment of the golf course over the site, and the existence of mature trees. The author discussed the problem of geophysical survey on this site with one of the local archaeological curators, John Hemingway. It was considered that the undergrowth and roots were unlikely to make this a viable exercise, and it was proposed to carry out excavation without recourse to this survey. Furthermore, foundations can be seen presently exposed, and so exact location was not required to locate the feature.

It was proposed to carry out excavation to recover the plan only. Full excavation, entailing the removal of in situ remains, was not proposed. Where foundations were recovered, these were recorded and reburied for future preservation.

3.2.7 The (south) stream valley from the Heart-shaped Pond to the Priory Pool

A stream walk was carried out to locate worked stone, brick and other man-made objects of historic interest. The methodology was the same as that outlined for section 3.2.1. Using plans provided by the client, the archaeologists marked where in-situ materials existed along the stream valley.

3.2.8 The south stream roothouse

Trenching was carried out on the site of this structure to try to locate the ground plan. Brickwork could be seen in the stream side near the conjectural site of this building was also examined by excavating a small trench.

3.2.9 Seats

The brief mentions looking for the site of seats. Documentary work suggests that Shenstone made these from existing tree stumps, and temporary structures that were unlikely to leave any archaeological record. It was agreed with the local archaeological curator that trying to find these features by archaeological trenching or geophysical survey was unlikely to prove worthwhile. Alternative work was suggested elsewhere in exchange (see Section 3.2.1).

3.2.10 Statue of Faunas

It was agreed with the local archaeological curator that it was unlikely that the site of this feature can be located. The site appears to have been disturbed by a set of modern steps.

3.2.11 Geophysical survey

It was agreed on site with John Hemingway, one of the local archaeological curators, that none of the sites listed in the brief lent themselves to successful geophysical prospection. It was therefore decided that this method would not be used during these works.

3.3 Recording

Recording was undertaken using standardised pro-formas and other materials supplied by CKC Archaeology, based on English Heritage's Central Excavating Service, Site Recording Manual, version 7, London, 1992.

3.4 Post-excavation strategy

All retained artefacts and ecofacts were cleaned, conserved and packaged according to the requirements of the recipient museum. These were studied by suitably experienced staff, and the information gained from them was incorporated into the final report. The project archive was prepared according to the requirements of the recipient museum and national guidelines. Cataloguing of finds was undertaken on a suitable pro-forma designed for this purpose.

Suitable provision was made for the conservation of any finds so requiring special conditions of treatment and packaging according to the guidelines laid down by the Society of Museum Archaeologists, Towards an Accessible Archaeological Archive (London, 1995), pp. 23-24.

3.5 Report preparation

The report was written to guidelines laid down by English Heritage in The management of archaeological projects (London, revised edition, 1992). The ordering of information, and the information given, followed the guidelines given within this document, but were altered slightly according to the specific needs of the report. Additional guidance was taken from the Institute of Field Archaeologists' Standard and guidance for archaeological evaluations (Birmingham, 1993) where considered appropriate.

3.6 Copyright

C K Currie (trading as CKC Archaeology) retains full copyright of any commissioned reports or other project documents written by himself or his agents, under the Copyright, Designs and Patents Act of 1988 with all rights reserved; excepting that it hereby provides an exclusive licence to the client for the use of such documents.

4.0 The results of the measured surveys

4.1 The Middle Pond and Chalybeate Spring area (Figs. 3, 9, 12, 13; Large Plan of Fig. 9 in end sleeve; Plate 2)

A measured survey was carried out on the earthworks of the Middle Pond, and the area around the Chalybeate Spring downstream. This was made at 1:100. In addition a profile was surveyed over the dam bank of the Middle Pond.

The results of this survey are given in Figure 9. These show that the Middle Pond was dammed behind an earthen bank approximately 15m wide. The dam had a maximum height of 1.3m on the pond side. At its lowest point its height was about 0.95m on the pond side. However, there was a considerable drop on the far side, where the bank was just over 4m high. The dam is about 14m wide at present, although accumulated silt on the pond side may obscure an extra metre or two.

There were two breaches in the dam. One was close to the south side of the narrow valley in which the pond was situated. To the south of the breach was a stone wall set in a concrete backing. The stones are probably a limestone type similar to those used in modern gardens for rockwork. This wall was not thought to be of any great age. It was possibly put up as a repair, possibly to an earlier breach. The length of the wall across the face of the dam was 3.3m, and it was up to 1.5m high. At the point of the breach, it turned through a right angle for 1.5m. Whether this was to revet the south side of the breach, or to support a sluice is a matter for speculation. The top of this wall was 141.42m AOD.

Upstream of the dam, scarp banks were visible. These probably showed a former shoreline of the pond. On the north side, this bank was well pronounced, about 1m high, and extending eastwards almost 30m from the dam where it merged into the longer slope of the valley. On the south side, this scarp bank was much fainter, being a slight 'lynchet' like bank less than 0.4m high for the most part. South of this bank, the valley side is steep. Taking the level on the top of the dam wall (141.42m) as the maximum height of the latest pond level (in reality, the level would have to be at least 0.2m less than this), suggests that the pond extended only about 25m east of the dam. However, the maximum height of the dam bank is at least 0.5m above this, and this indicates that the original pond was probably slightly large, extending eastwards about 30m. This tends to coincide with the approximate extent of the scarp bank on the north side.

On the downstream side of the dam are a number of interesting features that are thought to be associated with Shenstone's Chalybeate Spring. The second breach through the north side of the dam leads into a ditch. This is cut up to 1.5m, the upcast probably partly forming a bank on its north side. On the south side, the ditch runs at right angles to the dam bank, cutting into its north edge.

The ditch is approximately 13m long. At its west end it joins another ditch at right angles to it. This second ditch runs along the bottom of the dam, parallel to that feature. On the south side of the first ditch, where it meets the second ditch, there are the remains of a brick revetment. This extends for about 5m around the bend where the two ditches join. A further short section could be observed protruding from the west bank of the second ditch. This was merely a few bricks side by side extending east from the bank.

The bricks average 23x11x8 cms, and are of a type common at The Leasowes. These bricks are larger than the bricks found on primary Shenstonian masonry. Although it is not impossible that they are of later Shenstone work, it is more likely they are post-Shenstonian. It is thought that they occur in post-1800 work, but this should not exclude the possibility that they may date from the period when tax on brick encouraged the manufacture of larger bricks (1784-1803).

The ditch at the bottom of the dam is cut into the ground by about 1.2m on average. It extends from the main stream for 14m northwards. It is about 2.5m wide on average. At the north end, it extends past the ditch coming from the pond, and cuts deeply into the valley side. At the far end this cutting is by about 2m. There appears to be a spring at the north end supplying water to the ditch.

At the north end, another ditch enters it from the east. This is about 10m long, and is parallel to the ditch coming from the pond. It tapers upwards towards the east. It was probably not a drainage ditch as such, but seems to have been created, at least partly, by the creation of a bank on its south side. This bank was probably made from upcast from the ditch leaving the pond. These ditches are sited within the area known as the Chalybeate Spring, and are thought to be the earthwork remains of that feature. John Hemingway (pers. comm.) has suggested that this was once part of the Shenstonian path leading past the spring. Remains of this path can be traced further to the west as a slight holloway.

There are features within the main stream below the dam wall. Below the wall is a mass of dislodged stonework. This includes modern stone types and concrete. These stones are often large and heavy. It was not attempted to move them, but it was noted that they lay on what appeared to be the remains of a brick surface lining the stream bed. This lining seems to extend from about 5m below the dam wall to about 8m below, and further remains can be seen where the stream has eroded the bank on the north side. Examination of these bricks suggested that they were mainly 23x11x8 cms types, as noted revetting the ditches mentioned above. There were also some thinner bricks included in this lining, but these may have been reused.

At the point where the main stream met the ditch running along the bottom of the dam, a drystone revetment could be found running around the bend in the bank. This revetment was about 2.5m in length.

4.2 The Pond below Beechwater (Figs. 4, 8)

There is a small dam about 50m below the Beechwater dam. A measured earthwork survey and profile were made of this feature. It was an unexceptional feature, forming no more than a small, thin pond below Beechwater. Its site can be clearly seen on maps as the stream doglegs to go around the feature. This dam has breached right against the steep south side of the valley, leaving almost the entire artificial bank on the north side. This bank is no more than 9m in length on a rough N-S alignment. It is about 10m in width, and up to 0.9m above normal ground level (1.5m above the stream bed). Unlike the Middle Pond bank, there is no pronounced drop on the downstream side.

5.0 The results of fieldwork along the streams (stream walk)

(Figs. 3-4, 13; large plans Archaeology 1 & 2)

5.1 Features recorded along the streams

5.1.1 The north stream: Beechwater to Lower Pool (see Archaeology/1)

i) Section 1-1: Beechwater to dam below Beechwater

The Beechwater dam has been much altered over the years. According to Dudley Council staff on site, the dam was breached until recently, and had been filled in with all manner of dumped materials. The present depth of water is said to be less than one metre, despite recent dredging. However, despite this, the water level is presently very high in relation to the dam, being to within 0.1m of the dam top. Most text books on pond construction and maintenance recommend a 'freeboard' (level between top of dam and water level) of at least 30cm (Bryant et al 1980, 23), with one metre being advised on larger ponds (Barrington 1983, 25).

Recent bore hole testing found that the top 0.3m of the dam is made up of hardcore and recent dumped materials. Below this the bulk of the dam is made up of a grey or olive grey sandy silt clay (Munsell Colour 2.5Y 4/1 or 5Y 4/2), where the percentage of clay is probably less than 50%. The bore hole excavators estimated that the dam dumping soils extend to about 2 to 2.5m, thereafter is undisturbed soil. A weak sandstone layer was found at a depth of 6.25m, with a harder layer being found at a depth of 7.30m (Borehole excavators, pers. comm.).

At the back of the dam, there is evidence for the remains of two modern water exits. Below these is a series of concrete steps, forming a cascade. On the south side of this cascade, a few lumps of a reddish sandstone can be found embedded in the side of the dam. These are thought to be remnants of Shenstone's original cascade. Similar sandstone lumps can be found reused around the modern spillway, and at the base of the cascade. None of these stones showed any signs of working, and seemed to have been used as found naturally.

The section between the Beechwater dam and the dam below it contained a stream with an even flow over alluvial deposits. Quantities of brick were found in the stream. The length of this section was about 50m.

ii) Sections 2-2 and 3-3: dam below Beechwater to entry of first side stream

The length of these sections is about 60m. The relict dam beginning this section is described in section 4.2. Below this and the next major bend in the stream, the flow speeds up, cascading over a series of sheets of sandstone. These would have made a fine natural, if low, cascade that may have been utilised by Shenstone. These sheets of sandstone continue around the bend, where they are less dramatic. Just before the side stream enters the main stream, there is much evidence of modern alteration. The sides of the main stream are lined with concrete slabs, and there is a modern concrete step in the stream. These steps are referred to on modern maps as 'weirs', although the term is ambiguous. The side stream is heavily coloured orange, possibly through ferrous compounds in the local soils. This stream runs over a modern concrete lining, and discolours the main stream for at least 15m downstream. Bricks collected here were stained bright orange.

iii) Section 4-4: from the side stream to the Middle Pond dam

This stretch was about 70m in length, and passed, for the most part, through the bed of a former pond. The last 30m or so was certainly within this pond, and it is possible that a slowing down of the stream, occurred through the full length. This caused water to back up, resulting in the deposition of much silt. This was evident by the boggy condition of the ground along this stretch. No man-made features were encountered along this length.

iv) Sections 5-5 & 6-6: Middle Pond dam to Lower Pond dam

These sections were amalgamated. Their total length was about 140m. The section immediately below the Middle Pond dam contained the exit for the Chalybeate Spring, and is described in detail in section 4.1.

Below where the Chalybeate Spring enters the main stream there are a series of apparently natural cascades. These are low drops over sheets of sandstone. Three obvious examples were noticed, although 18th-century visitors often claim four. It is possible the fourth example may have been removed during work on the Virgil's Grove area this century.

After about 55m below the Middle Pond dam, the stream runs under the first of three modern bridges within the area known as Virgil's Grove. The stream from hereon has been much altered in modern times. About 5m below the first stone bridge is a concrete step or 'weir'. Immediately after this a small rill runs down the valley side from beneath a large beech. Two more steps have been placed to catch this rill, and bring it into the stream over a small cascade. These steps are thought to be modern, although the rill was probably the main source for Shenstone's 'Dripping Cascade'. This is described in detail in section 6.1 below.

From this point it is possible to observe the Shenstonian paths winding around the valley sides from the site of the Obelisk (see section 6.2) to behind the Dripping Cascade. For most of their lengths, these paths have eroded from the valley sides, leaving only slight 'steps' where they once were. Nevertheless, it is still possible to trace an almost complete circuit from the obelisk round to the south side of the Lower Pool.

The stream then passes under one wooden bridge and then a further stone bridge. Although the odd piece of stone can be seen in the bank at both of these points, it is not possible to relate them to any earlier feature. Where the stream enters the Lower Pool, much silt has been deposited. This pond is in need of dredging as silting has probably reduced its original area by about half.

The dam of the Lower Pool is now crossed by a modern tarmac road. All traces of old features here seem to have gone.

5.1.2 The north stream: Lower Pool to Priory Pool

Immediately below the Lower Pool bridge, there are three modern weirs. These use concrete and flat slabs of stone to make three steps for the water to fall down. The stone includes limestone types similar to that sold in garden centres, roadstone and some reused red sandstone. The stream has eroded the bank away on the north side of these weirs, and now takes this route downwards. Immediately under the bridge are some large chunks of brick masonry from a relatively recent (earlier 20th-century?) structure. The bricks are mainly 23x11x8 cms types.

Just below the second step there is an eroded gully in the valley side, leading from the road into the stream. This is a convenient site for fly-tipping and other dumping. A large dump of brick and roadstone has been tipped down this gully into the stream. The brick types are all fairly modern (later 19th- and 20th-century types), and include black bricks stamped 'Partridge & Guest Old Hill Staffs'. Nearly all the man-made materials found in the stream below this point seem to be washed down from this dump.

Just below the third step, the stream splits into two and rushes past a small island. There is much natural sandstone on the floor of the stream at this point. The stream also makes a fairly steep fall between some equally steep valley sides, levelling out somewhat after about 40m.

Just before a modern bridge taking a modern path over the stream, there is a modern concrete step 'weir'. Below the bridge, the sides are crudely lined with modern roadstone for a few metres. From hereon the stream meanders quietly into the Priory Pool, about 90m below the bridge. The pool causes the stream to slow, and has deposited much silt on the valley floor for about 40m above the pool.

The total length of this section is about 170m.

5.1.3 The south stream (see Archaeology/2)

This stream runs from the Heart-shaped Pool down into the Priory Pool. It does not have the volume of water flowing down it present on the north stream, and might have a tendency to dry up in hot weather.

The first 80m falls quite rapidly down a steep valley. The cascade leaving the Heart-shaped Pool seems to be entirely modern. The bed of the stream appears to be gravelly by nature, but is strewn with sandstone lumps. Large chunks of sandstone are also embedded in the banks on both sides of the stream. It is uncertain if this is entirely natural, and much of the stone may have been placed there to create an effect. None of the stones appear to be worked in any way, and appear to have been collect as found, and placed along the stream. Many of the pieces within the stream have probably been dislodged from the banks.

After about 80m the stream is crossed by a stone bridge of modern construction. After this the stream flows over gravel, with larger stone relatively rare, becoming increasingly so as one travels downstream. About 10m below the bridge, on the east bank, there are a number of bricks apparent eroding out of the side. These bricks tend to be 22.5 x 10.5 x 6.5 cms types. An excavation was undertaken here to investigate this anomaly (see Section 6.11).

About 15m below this point, a second small stream enters from the east. This collects a number of other smaller streams. Nothing of interest was observed along these streams. The only evidence for man-made activity is under the wooden bridge that crosses the second stream south of the conjectured site of the Statue of Faunas. There are a few modern bricks embedded in the bank under this bridge. These were not collected.

The path leaves this bridge heading north-eastwards. As it rises up a steep slope containing some built in steps, there is a small mound, about 3m in diameter, with a moderate-sized holly bush growing out of it. The bush prevented any further exploration on this site, but it is suspected that this may be the site of the Statue of Faunas.

Returning to the main stream, it now runs slowly westward along a slight incline towards the Priory Pool. There was little sign of past human activity on this stretch until about 50m above the Priory Pool. Here the stream flows over a concrete floor for a few metres before dropping down about 0.6m where erosion has undermined the concrete. At this point was a small pile of mixed bricks, some modern, and some possible reused older types. After this there is a wooden walkway crossing the stream. Apparently there had been a boating pool here before the main pool proper. This has now silted up and is much overgrown with reeds and other water-loving weeds. The distance between the point where the side stream enters and the Pool proper is about 180m.

The total length of this section from the Heart-Shaped Pool to the west end of the former boating pool is about 310m.

5.2 The results of the brick/stone collection from the stream beds (see Appendix 2)

Brick, worked stone and other man-made artefacts were collected from the north and south streams. The collection units were the same as those indicated in Section 5.1. The main category of artefact was brick. Only occasional pieces of tile or other ceramic were found, and only one piece of worked stone.

The Bricks were put into seven different categories, dependent on approximate characteristics. This was based mainly on size, weight and external features. Detailed analysis of fabric types was not made. These categories are described below:

Category 1: These bricks were up to and including those 6cms thick. They were frogless, and apparently handmade. They weigh, on average, about 6lbs. It is generally considered that these were the earliest bricks found on the site. Such bricks are probably later 17th- or early 18th-century. They could have been used by Shenstone, but, if this is so, they are most likely to be reused, as they may have been made slightly before his time.

Category 2: These are bricks of around 8cms thick. They are frogless, and weigh around 10lbs. They may be handmade, but this is not always apparent. Such bricks are found on most Shenstonian sites (including the Priory Cottage), but they are not thought to be from the primary building phase. Notably they are not found in the remains of the Obelisk, a feature that was not known to have been repaired or rebuilt. Consequently these bricks are thought to date from 19th-century repairs. It is not impossible that some may date from the period when the imposition of a brick tax encouraged manufacturers to make thicker bricks (1784-1803). Nevertheless, it is more likely they are later than this, particularly as they have been found incorporated in work thought to be of a relatively late date (possibly as late as the early 20th century).

Category 3: These are bricks that are slightly wider on one face than another. Differences are between 0.5 and 0.75 cm. They are frogless, and not particularly well made. They often weigh 8lbs or more. They may have been made for a specialist purpose, such as coping or for arches. Their use at The Leasowes was probably opportunist, as they may have been left over from other work, and used for repairs. They are probably of a mid-late 19th-century date.

Category 4: These are a rare frogged brick, stamped 'PLA..'. They are probably from the first half of the 20th century.

Category 5: These bricks tend to be between 6 and 7cms. thick, and weigh about 7lbs or less. They are frogless and handmade. They have large haematite inclusions and a characteristic marbling in the fired clay. This latter shows poor clay mixing characteristic of pre-industrial bricks. They are probably of 18th century date, and are found in the Obelisk. They seem to be the main brick used in Shenstone's period, although differentiating them from Category 6 can sometimes be difficult.

Category 6: These are similar to Category 5, but tend to be slightly larger, and heavier. There are no other accurately distinguishing features, however, and slightly large Category 5 types are easily mistaken for Category 6. They tended to be between 7 and 7.5cms thick, and weigh around 7.5lbs. They are probably made throughout the 18th century, but one would expect rural handmade industries to continue making bricks like this well into the 19th century. It is more likely they originate from post-Shenstonian activity, although he may have used them in his later phases.

Category 7: This category lumps together all obviously modern brick types. It includes frogged bricks, and specialist modern types like engineering bricks. They have not been found in any in situ structures, but tend to come from either hardcore dumped on the Beechwater dam, or from dumping off the road near the Lower Pond bridge. They can be found swept far below these sites, however. Local children have also been found fond of building dams across the stream from building debris, and at least two of these were dismantled by the author for their bricks. There is no knowing how far bricks get transported for these casual purposes.

Table 1: Brick types recorded between Beechwater and Lower Pool (sections 1-6) by fragment numbers

Brick type Sections Total bricks of each type
1 % 2/3 % 4 % 5/6 %
1 2 7% 10 14% 7 10.1% 2 5.2% 21
2 8 26% 17 24% 10 14.5% 7 18.2% 33
3 4 13% 10 14% 8 11.6% 2 5.2% 24
4 2 7% nil   1 1.5% nil   3
5 2 7% 13 18% 22 31.9% 7 18.2% 44
6 12 40% 20 28% 19 27.6% 11 28.6% 62
7 nil   1 1% 2 2.9% 10 26% 13
Total bricks per section 30 71 69 39 

Table 2: Brick types recorded on lower north stream (Lower Pool to Priory Pool) and south stream

Brick type Lower north stream south stream
nos % of total nos % of total
1 nil - 2 9.0%
2 12 63.6% 3 13.5%
3 nil - nil -
4 nil - nil -
5 nil - 4 18.0%
6 3 15.9% 12 54.0%
7 4 21.2% 1 4.5%
Total bricks per section 19 22

Table 3: total bricks of each type

Brick type total numbers % of total nos. total weight (in kilos) % of total weight
1 23 9.5% 49.33 7.4%
2 48 19.9% 203.69 30.6%
3 24 10.0% 79.22 11.9%
4 3 1.3% 8.81 1.3%
5 48 19.9% 118.44 17.8%
6 77 32.0% 170.29 25.6%
7 18 7.5% 36.62 5.5%
Overall total 241 666.40 kilos

6.0 The results of excavations

6.1 Trench 1 (Fig. 5)

Trench 1 was the clearance of accumulated mud over area thought to be the main part of the Dripping Cascade. The area cleared was approximately 9m by 2m.

It appeared that a water course had originally been eroded into the hillside by the rill that presently emerges from under the roots of a large beech tree. This channel then filled with silt and mud. The area 'excavated' coincided with this natural channel, and followed that shape rather than that of a formal rectangular archaeological trench.

Once clear of mud and silt, the channel proved to be a fast flowing small rill, emanating from beneath the roots of the beech tree. This was not apparent before the mud was removed. In its uncleared state, the mud appeared to be constantly wet, with water dripping down the bank at the bottom of the channel. The clearance made the possible original appearance more obvious.

It would seem that the channel was once lined with irregular sandstone, taken from nature without any deliberate shaping. It seems that the majority of this had been removed at some time. Only a few pieces remained, mainly those on the far west edge of the channel that had probably become hidden in mud and vegetation at the time of the conjectured dismantling of the cascade.

A further interesting feature of the channel was quantities of very small pieces of crushed ceramic in the lower levels of the mud, just above the natural surface. These had been shattered into very small pieces, although it was not clear whether this was deliberate or during some subsequent action, such as dismantling of the cascade. There were a few large pieces that showed that these ceramics may have been mainly flowerpot fragments. Two diagnostic pieces of vessel base showed that these were plain earthenware pots with drainage holes in the side. Although side drainage holes are occasionally found in later flowerpots, particularly 20th-century 'antique' reproduction types, this feature is often characteristic of pots made before 1760 (Currie 1993).

Also found within the channel were a few large pieces of modern masonry. This included brick, roadstone and concrete. These tended to be situated in the centre of the channel, and may represent later activity on the site. This might include a very crude attempt to renovate the feature.

6.2 Trench 2 (Figs. 6-7; Plate 1)

Trench 2 was excavated on the site of Shenstone's Obelisk. This was easily found because four mortared bricks still protruded from the surface indicating the site. The trench was prevented from being a formal rectangle by the roots of a large yew tree that stood next to the remains. The approximate area of the trench was 1.65m by 1.2m.

Immediately below the mulchy topsoil (Context 01; Munsell Colour 5YR 3/1), the brick foundations of the Obelisk became apparent. If construction cuts existed, the large roots of the yew tree had utilised them to spread through thus obliterating any evidence. However, assuming a cut once existed, this had been dug into a brown clay soil (context 02; Munsell Colour 5YR 4/6).

The surviving remains of the Obelisk was three sides of four sided base, around a hollow centre. The north side had apparently been demolished, and all trace of it removed. There was very little superfluous debris from the demolition of the structure in the trench. The one exception to this was a lump of sandstone. This was slightly curved, about 4 cms thick, with a thick covering of mortar on the upper surface. It was the only piece of masonry found that was not made of brick.

The three sides of the brick foundations had not survived in a uniform shape. The south side survived in the most complete condition. This was possibly of five courses in height. Roots and the proximity of the large yew tree prevented excavation of the south face, and height had to be judged from the adjoining faces. The height was one course higher than those adjoining, as a result of the four bricks, the remnant of the conjectured fifth course, that had originally been above the ground surface prior to excavation. The bonding was probably alternating header and stretcher courses, making an English Bond pattern with Queen Closers at the corners. The length of this face was 0.95m, height 0.35m, and width 0.35m. At the back of the face, in the hollow, an incomplete foundation course had been put in, possibly to level the base on a sloping surface.

The surviving part of the east side was 0.55m long, 0.28m high, and 0.20m wide. It projected merely one stretcher brick length (0.22m) from the thickness of the south side, and had been the more thoroughly demolished of the surviving sides.

The west side survived to a length of 0.91m, a height of 0.3m and a width of 0.32m. There was an additional course along the base projecting outwards by about 0.08m. This did not extend the full length of the side, ending before the north edge was reached. It was thought to be made to level the structure on its sloping site. There was also a brick at the base inside the hollow on this side, probably so placed for the same reasons. The hollow in the centre was 0.41m wide, and was not exactly central, the west side wall being wider than the east.

A random selection of bricks in the structure was measured, and one was taken away as a sample. These are reported on in the specialist reports (section 10.1.2).

6.3 Trench 3 (Fig. 6-7)

This was a small trench 1.6m by 0.3m cut across a possible relict path about 15m north of the obelisk. It was represented by a small holloway about 1.2m wide with traces of a slight bank on the south side. On the north side, the slope of the hollow continued into the natural upwards slope of the hill.

The hollow was filled with an organic mulch about 0.2m deep and filled with a mass of roots (context 05; Munsell Colour 10YR 3/1). There was a deeper gully on the south side, extending down a further 0.1m. This was thought to be natural, caused by larger roots pushing deeper into the soil at this point. Immediately below the mulch was a compacted brown clay (context 06; Munsell Colour 5YR 4/6) that appeared to be the undisturbed surface of the original path. The compaction was caused by constant use over a long period of time.

6.4 Trench 4 (Fig. 6)

This was a small trench, 0.9m by 0.8m, excavated a few metres to the south of the southern end of the Dripping Cascade to explore a large stone protruding from the surface. Extension of the trench was prevented by the presence of major roots of a large adjacent beech tree.

The large stone appeared to have been crudely shaped, possibly as part of an arch over a niche of some kind. There were a number of other small stones around it. These were not shaped in any way, and appeared to be natural. It is possible, however, that they may have been brought to this spot by human agency. It is possible that these stones may represent the remains of a rock niche from which the Dripping Cascade spring emerged. This conjectured 'niche' now had a large mature beech growing over it, preventing any further exploration. The spring seemed to originate from very close under the stones here recorded.

6.5 Trench 5 (Fig. 6)

This L-shaped trench, with maximum dimensions of 1.85m by 1.5m, was excavated to the west of the Dripping Cascade to investigate a compacted feature found by Birmingham University Field Archaeology Unit (BUFAU) in 1996. This was thought to have been a path.

A wet silty loam mud made up the topsoil (context 10; Munsell Colour 10YR 2/1). This was about 0.15m deep. It came down onto a series of overlapping layers running along the contours of the hillside. The uppermost of these was an olive grey clay (context 12; Munsell Colour 5Y 5/2) at the north end of the trench. This partly overlay a surface of compacted crushed fired clay. This was an irregular shape, with a maximum width of 0.9m. The fired clay was mainly small fragments of crushed earthenware ceramics, probably flowerpots, and mortar (context 11; Munsell Colour 10YR 3/1). It was less than 2 cms. thick, and almost certainly the compacted surface found in 1996. This overlay a dark organic silty clay mud that also contained lesser amounts of ceramic, mortar and stone debris (context 09; Munsell Colour 10YR 2/1). This was also only a few centimetres thick, and overlay a sandy clay, containing many small pieces of sandstone (context 13; Munsell Colour 5Y 4/2). This layer was similar to undisturbed layers found elsewhere in the area, and was probably 'natural'.

The reason for this compacted debris being here is discussed in section 7.2. Its purpose, though debatable, was unlikely to have been a path.

6.6 Trench 6 (Figs. 6-7)

This trench, 2.5m by 0.7m, was a partial section taken through the dam of the Middle Pond to try to assess the nature of construction of this feature.

The first relevant layer in this dam was a layer of reddish brown clay dumped over the top of the dam (context 16; Munsell Colour 2.5YR 4/4). This was an average of 0.2m thick. It overlay a further dump of sandy loam of about the same thickness (context 17; Munsell Colour 10YR 4/3). This overlay the core of the dam proper. This was divided into two parts, with what may originally have been a vertical divide between them. The core proper was a yellow brown clayey sand (context 19; Munsell Colour 10YR 6/6). This was less than 50% clay, although its exact proportions were not measured. At the point excavated it was 0.95m deep. It was butted by a dump of dark grey sandy clay (context 18; Munsell Colour 10YR 4/1). Although the division between 18 and 19 had become uneven, probably through subsequent slumping, there was a suggestion that this division was once vertical. The slumping may have been caused by the rotting away of a dividing vertical feature, such as a wooden revetment. The sandy nature of the clay was probably acidic, making the life of any such feature relatively short.

A sondage was dug into the core well below the present ground surface of the old pond bed. This revealed a wet olive sandy clay (context 20; Munsell Colour 5Y 4/2) that was similar to the soil used in the base of the Beechwater dam. This was interpreted here as the base of the artificial dam bank.

6.7 Trench 7 (Fig. 6)

This was a small trench, 1.5m by 1.3m, excavated on the conjectured site of a roothouse near the south bank of the Lower Pool. It was originally intended to make this trench much larger, but extensive root disturbance suggested that this could be both unrewarding, and damaging to the large beech trees overlooking the trench.

The topsoil in this trench was an organic mulch made up of massed root fibres, rotting leaves and beech nut shells in a loose loamy soil (context 21; Munsell Colour 10YR 3/2). This was about 0.15 to 0.2m deep, and overlay a brown sandy loam (context 22; Munsell Colour 7.5YR 4/3). This contained the odd fragment of yellowish sandstone that was possibly naturally occurring. The soil was heavily disturbed by roots of all sizes. On observing this, the management was consulted. It was decided that continued excavation here was unlikely to be rewarding, and that the mature trees nearby could be damaged.

6.8 Trench 8 (Figs 10-11, 13; Large Plan of Fig. 10 in end sleeve; Plates 3-7)

This trench was dug on the site of the Priory Cottage. The site was an overgrown area of woodland between two golf course tees. Vegetation included a number of large mature trees, with a shrub understorey, and mainly bramble and nettle ground cover.

The trench was an irregular shape formed by locating the BUFAU trench excavated in 1996, and working outwards from the structural features located to try to establish the plan of the building as far as the constraints of the site allowed. The situation of a golf course tee and three large ash trees prevented the full plan from being recovered. Nevertheless, the area excavated within the current restraints proved to be sufficient to allow conjecture on the size and extent of the building.

The extent of the BUFAU trench was still visible following vegetation clearance. Parts of this trench had not been fully backfilled, leaving masonry remains still showing on the ground. This was probably the result of building materials being taken out of the overlying topsoil, and left out in a heap besides the trench, resulting in a significant reduction in the volume of soil. The amount of rubble within the topsoil was proved by the current excavation. Building materials were retained on the side of the trench, with samples being selected for later study. This also caused some reduction in the volume of soil left for backfilling.

Excavation showed that the soil overlying the ruins of the Priory Cottage comprised mainly building rubble with as little as 50% soil (by volume). The soil matrix was a dark grey-brown sandy loam (context 23; Munsell Colour 10YR 3/2), containing large amounts of brick, stone and mortar rubble. The depth of the soil over the buried walls varied from less than 0.1m to almost 1m, depending on the proximity of the golf course tee on the NE side of the site. This tee had been constructed by heaping rubble from the demolition of the cottage up to a height of over 1m, and then levelling the surface with a thin layer of topsoil, less than 0.1m deep. The edges of the tee sloped downwards over the cottage ruins, making the depth of burial variable. Away from the tee the tops of the walls were less than 0.1m below the surface, although in the inside of the rooms depths of burial varied from 0.1m to 0.6m.

The NE edge of the trench was similar to that of the BUFAU trench, although it was extended further to the NE. The overall size of the finished trench was approximately 9.2m SE-NW and 5.8m SW-NE, with extensions in the SW and SE corners of approximately 2.4m x 2m and 3.2m x 2.15m respectively. The complexity of the building remains was such that it is not possible here to describe every aspect of it in this report. If the reader requires further details they are referred to the relevant figures and the site archive.

The main feature excavated was a room, approximately 5.8m by 3.9m internally with a curved wall on the NW side. This room was surrounded by brick walls on the SW and NW sides (contexts 33 and 27/32 respectively). It had a mainly stone wall on the NE side (context 26) and a brick wall on a stone foundation on the SE side (contexts 44 and 48). There were three openings for doors on the NE (context 42), SE (context 51) and NW (context 31) sides. The first two of these led into further rooms within the overall structure, whilst the opening on the NW side seemed to lead out on to a brick and tile exterior surface (context 28).

Inside the room appears to have been floored with red ceramic tiles. These survive as a complete floor in the E corner of the room (context 41), extending NW towards the door in the NE wall, to about the mid point along this wall. Underlying the tiles was a surface of mortar a few centimetres thick (context 30). This, in turn, overlay the sandy loam subsoil that was general to the area (context 29).

Overlying the tiles, and butting against the SE wall of the room was the base of a rectangular brick structure, approximately 1.52m by 0.6m (context 47). The tile floor in line with the NW edge of this feature is slightly raised by about 1cms. Within the feature itself there was a T-shaped hollow, with the remains of an iron feature (context 52) on the NW face, and as a rectangular iron surround in front of the entrance to the T-shaped hollow area within the feature. This feature was thought to be a kitchen range, the iron features being part of a grate, the T-shaped 'hollow' area being the void in which the fire was lit. This interpretation seems to be confirmed by a chimney shown on a photograph of the Priory Cottage approximately over the site of this feature. It should be noted that feature 47 was not bonded into wall 44 behind it, but butted against it, with a void of about 3cms. between the two structures extending to within 0.2m of the two outer edges. These outer edges butted against wall 44 without a void between them.

Some further comments about this room are worth adding here. The SW wall (context 33) had wider foundations than the actual wall. The foundations were about 0.35m wide, the upper wall only about 0.24m. The latter measurement was the approximate width of all the walls of this room. This wall survived to a height of about 0.6m near its SE end, as did the adjoining wall (context 48). Both walls had the remains of plaster on their inner faces, which showed traces of having been painted a pinkish colour. The external face of wall 33 also had faint traces of a render on it. This was confirmed by a photograph of the cottage (see Plate 7) as having covered the entire SW external wall.

Wall 44 was built over a stone foundation (context 50). This foundation extended across the doorway near the south end of this wall, forming a stone threshold (context 51). About 1m from the E corner of the trench, wall 44 was bonded with traces of another brick wall (context 45). NE of wall 45 was a deep void filled with brick and stone rubble (context 46). This was thought to be the remains of a cellar. Wall 44 seemed to extend beyond the first room to continue as the NE wall of the adjoining room. Where the wall crossed the NE wall (context 26/43) of the first room, the other wall butted against it.

About 1m short of this butt, the stone wall 26 had stopped, to be replaced by a brick wall (context 43). This brickwork appeared to be a blocking in the original wall. It extended NW only as far as the cellar-like void (context 46), which seemed to extend under the join of walls 43 and 44. The brick upper wall (context 43) over stone foundations (context 50) might suggest this wall was originally of stone.

The original door in wall 26 was partly blocked on the NW side by a stone blocking (context 55). This foreshortened the original door by 0.3m. The original opening had been 0.8m wide, being reduced to 0.5m.

The NW wall also had some unusual features. Outside the postulated door (context 31) were two iron bolts (context 56) embedded vertically into the ground on the edge of the exterior brick and tile surface (context 28). This was interpreted as the remains of a foot scraper, a further indication that the surface to the NW was external to the building. The bricks making up the curving wall contained mostly types that were 8cms thick. This was different to the adjoining SW wall (context 33), which were generally 7cms thick on average. This presents a slight problem of dating. In the discussion section of this report, it is argued that the room described so far was an extension added to the original cottage between 1859 and 1887. This is based on map and pictorial evidence that seems to be reliable. It is therefore a puzzle to find that the NW wall uses different bricks to those found in the other three walls of this room. Was the NW wall a later alteration?

Outside the exterior door of this room there appeared to be a brick and tile surface (context 28). The NW edge of this was located where it had been only suspected in the BUFAU excavation. The surface was also found not to extend much beyond the NE jamb of the exterior door on the NE side. On the SW side the surface seemed to follow the side of the building in an irregular manner eventually formed a linear surface (contexts 36 & 39). This was made up of various materials, deteriorating from tiles to a mixture of tiles and brick, and finally to a mixture of brick and stone. It seemed to peter out about a meter beyond the line of wall 33, where it became a crude gravel surface (context 40). This was thought to be the remains of a poorly made garden path. The edge of the surface was brought to a neat halt against wall 33, where a short section of neatly laid stone was found crossing the area between the wall 33 and linear surface 35/39.

Along the exterior edge of wall 33 was a layer of soil contaminated with charcoal (context 58). There were faint traces of a linear line of stones extending parallel with wall 33 at about 0.8m distance. This was found again at the SE end of wall 33, where the stone edging seemed to be more convincing (context 54). This was interpreted as the edging of a possible plant bed alongside the cottage wall.

On the NW side of the postulated extension were other features making up further parts of the Priory Cottage. Door 42 led into a room that was covered in a ceramic tile floor (context 25). The flooring was complete from the NE door jamb in a NW direction, apart from in the W corner where there are a few stone tiles on the floor. NE of the door jamb the floor was absent, being replaced by a rubble fill (context 46) to the conjectured cellar. Although the trench was only extended less than 1m N of wall 26, it would seem that this formed another room of the cottage. Indications are that it was part of the original structure that had once had an external SW wall of stone. The partly blocked door (context 42) had been the external door. All this seems to be confirmed on Jenkins' drawing of c. 1770 (see Figure 14).

In the SW corner of this second room, a brick wall (context 24) butts against the stone wall 26. There was a large stone block on this corner. It acted as a continuation of wall 26, but was external to the cottage. It is possible it was part of an external corner buttress. Wall 24 may not have been part of the original cottage as it does not seem to be bonded with wall 26. Likewise the tile flooring (context 25) runs into the flooring (context 41) of the adjoining room in such a way that it seems that both floors were laid simultaneously.

The door (context 51) in the SE wall of the conjectured extension led into an area that appeared to have been the 'ruined' priory part of the building. There was no trace of floor here, the overlying rubble coming down onto a dirt surface. Wall 48 was thicker and more substantial than the rest of the SE wall (context 44) of the extension. This had brick faces, with a brick and stone rubble-filled cavity. The overall width was 0.47m, compared with 0.24m for wall 44.

It is possible this additional thickness was the result of it butting on to a large external stone buttress (context 53) that stood against the S corner of the extension. This buttress is shown on the Jenkins drawing of c. 1770, and appears part of the original structure. It is not completely clear how it relates to wall 33. Although this latter wall seems to butt against the buttress, there may have been an element of rebuilding behind the buttress when the extension was built.

Tied into the buttress is a large semi-circular stone wall (context 49). This survives to a height of about 0.4m, although it was less than this the closer one got to the golf tee. The sloping sides of the tee have been built out over this wall. Amongst the rubble here were quantities of glass slag and other materials that were clearly brought in from outside the site. These were mixed with demolition materials from the cottage itself. The remains of the stone wall were not high enough to reveal any traces of windows or other features within it. It seems to have had an external diameter of about 3.2m. From the evidence of pictures, this formed a room, long in comparison to its width, that may have extended along the entire SE side of the Priory buildings. This room seems to have contained the main external facade that gave the whole structure its 'medieval priory' appearance.

6.9 Trench 9 (Fig. 12)

This was a small test trench of T-shape dimensions excavated to try to recover the continuation of a wall (context 24) located in the north corner of trench 8.

The trench contained some rubble, but this was not in the same quantities as found in trench 8. There was no indication of walls or that the Priory Cottage extended this far. Two pipes were located heading towards the building area. The first was a modern ceramic pipe, as used in sewers. It seemed to be diving towards the cottage. This might have prevented its use as a sewer. An alternative explanation might be that it was to drain the adjoining golf tee.

A second pipe was of iron. It made two right-angled turns, before terminating in a deliberate stop joint. It is possible this was a piped water source mentioned in a 1920s lease for the cottage. Its location suggests that the NE edge of the cottage was south of this feature.

6.10 Trench 10 (Fig. 12)

This trench was excavated to try to locate the site of a structure known as Stamford's Roothouse. It was originally 6m by 1m, but was extended into an irregular shape at its N end to try to find evidence for the roothouse. A shallow topsoil comprising a sandy loam matrix with much root and leaf detritus mixed in (context 66; Munsell Colour 10YR 3/1) was seldom more than 0.1m deep. It overlay a sandy clay loam (context 67; Munsell Colour 10YR 4/6) that appeared relatively undisturbed except for roots.

Two small features were recorded cut into this soil near the N end of the trench. Initially these were thought to be small post holes. Cut 68 proved to be very shallow, less than 0.05m deep. It was probably caused by root disturbance. Cut 70 was a little deeper, up to 0.1m deep, and about 0.16m in diameter. It may have been a shallow post-hole, but this could not be said for certain. It is equally likely to have been created by the extensive root disturbance in the area.

6.11 Trench 11 (Fig. 12; Plate 8)

This was a small trench cut to examine brick rubble seen in the side of the bank of the small stream near the conjectured site of Stamford's Roothouse. Initially it was intended only to clean the section, as the site was dangerous, being on the side of a steep bank, with steep access down a wet slippery valley side. Discoveries led to the trench being expanded from a cleaned section to a small trench 1.7m by 1m.

The original cleaned section showed that there was a stone feature apparently lining a linear channel extending diagonally into the bank of the present stream. The local topography suggested that it might follow what seemed to be an earlier stream bottom that had subsequently been filled by erosion or human agency. Further excavation proved this to be probably correct.

The channel cut into the surrounding clay was about 1.5m wide. It was filled with brick rubble in a clay loam matrix (context 75; Munsell Colour 10YR 3/2). Once removed, the clear outline of a brick lined channel was revealed. This comprised a mortared brick floor with brick sides. The channel thus made was 0.5m wide and about 0.4m deep. Beyond the brickwork was a further lining of unmortared stone set into a clay surround (context 74). It was possible that this was an original lining, the brick channel being put in later. It was not possible to assess if this had been a long time after the stone channel had been made or almost immediately. The channel appeared to extend for some way into the present bank, leaving the possibility of a structure over 10m long surviving.

6.12 Trench 12

This was not a proper archaeological trench, but an attempt to clean mud out of the Chalybeate Spring to see if any In situ stonework survived. Nothing was found in addition to that already described in section 4.1. Some additional brick debris was exposed, but this related to loose bricks originating from the brick structures visible on the surface prior to the work being undertaken. A short section of linear stonework crossing the ditch was thought to have been the work of children creating a crossing (stepping stones) in recent times.

6.13 Trench 13 (Fig. 12)

Like trench 12, this was not a trench as such, but a clearance of mud from the stream side opposite the Chalybeate Spring. It was done to investigate a pinkish compacted material that appeared to be emerging from under the mud and leaf detritus.

The overlying mud was up to 0.3m deep in places (context 78; Munsell Colour 10YR 3/1). It proved to overlie a compacted surface of pinkish ceramic-like shale (context 79; Munsell Colour 2.5YR 5/4). This was similar to that found adjoining the Dripping Cascade (see section 6.5).

An area 6m long with a width varying from 0.5m to 1.5m was exposed. The surface was remarkably level, differing by less than 4cms in height, but was very thin, being less than 2cms thick in places. Two approximately linear surfaces of greenish sandstone (context 80; Munsell Colour 5Y 4/2) were observed within it. These were at approximate right-angles to the stream, and between 0.2 and 0.35m wide.

7.0 Discussion

7.1 The Middle Pond area: excavation and surveys

The dam of the Middle Pond was sited at the head of an area where the ground drops away rather more quickly than previously. This would suggest that the stream may have formed a natural cascade here before Shenstone began work. He was probably able to utilise this drop, increasing its height by building a dam, and making his arched cascade here. Unfortunately nothing of that structure now remains. It may have been cheaply built in the first place, and being on a site where water passes with some turbulence, would have been quickly eroded. However, one might have expected some remnants of his structure in the stream bed if the site had been left untouched from Shenstone's time. This does not seem to have been the case. Although subsequent work on the estate by later owners was seldom recorded, the fame of the landscape seems to have made the temptation to restore or alter irresistible.

Below the dam on the conjectured site of the cascade was evidence for at least two phases of interference. Such works seem to have cleared all of Shenstone's work away, often replacing it with structures that barely resembled the original features. The earliest of these features was the remains of a brick floor to the stream. Although reusing possible Shenstonian bricks, the latest bricks here post-date his work. The bricks were an 8 cms thick type that are found commonly all over the site at Shenstonian features that were repaired, including the Priory Cottage. That they were not found in the Obelisk suggests they post-date c. 1800. The Obelisk had collapse by the middle of the 19th century, if not some time before. It was never apparently repaired, although it appears to have been seen in a ruinous condition. It is possible that whoever put in this brick floor came to the site after the Obelisk had gone past the point of repair or one might expect it to have been repaired. This suggests that the '8cms brick phase' probably occurred in the mid to late 19th century.

The wall now facing part of the dam, with its concrete constituents, was probably later. Whereas the person responsible for the brick floor may have made some attempt to patch up or rebuild Shenstone's Cascade after a fashion, those responsible for the wall may have no such pretensions. This work was probably an early municipal attempt to tidy up the area, and control the stream. It is possible that the people responsible had little idea of conservation, and would have removed any old stonework they found as a hindrance to a municipal idea of tidiness. It was probably on this occasion that any remnants of the earlier cascade may have been taken away, and possibly broken up for hardcore. There is virtually no trace of deliberate reuse of materials in water features that have been put in during the 20th century. Revetting work in the ditches that made up the Chalybeate Spring also seems to date from these two phases, and are unlikely to be original works.

What therefore remains of Shenstone's work? It would appear that only the earthworks themselves survive, and these have probably been altered. Excavation on the dam suggests that the original core was built behind a wooden framework, creating a box-shaped core (context 19), against which the final slope was created by simple dumping (context 18). The core has been clearly compacted, although the clayey sand used was nowhere near as pure in clay content as most historic dams seen by this author. It is possible that this may not have been as great a problem as it first appears, as modern experience has shown that dams can be successfully built with a clay content as low as 20% provided the soil is carefully laid and compacted, and the base carefully keyed in (Bryant et al 1980, 22). It is suspected that Shenstone's work in this respect was not of the highest quality, but, then again, it was not as bad as some examples seen by the author.

The creation of a possible box core behind timber work is in keeping with many 18th-century treatises on dam building. The single dump of soil used does suggest that the dam was not a medieval structure as suspected by earlier writers. Medieval ponds tend to ram soils down in successive layers of less than a foot thick (Currie 1986, 1990a). Although some relatively late ponds have been found using this technique, the dam construction methods used after c. 1740 on ornamental ponds tend to be single dumps rather than the more thorough layering method. The mode of construction of the original dam therefore suggests a Shenstonian, rather than earlier date. It is suspected that this was also the case on the Beechwater dam and on the small dam below Beechwater. There is no evidence that any of the ponds on the north stream are medieval in origin, although it is not impossible that they may have been built slightly before Shenstone's time.

The dam would seem to have been repaired. The clay layer (context 16) dumped over the dam was probably an attempt to patch a leak. This is a crude method to use for this purpose, and is unlikely to have been done by someone familiar with 18th-century treatises on the subject. As Shenstone appears to have knowledge of these techniques, this repair was unlikely to be his work. In consequence, it is probably the work of the 19th or 20th century, and may be related to one of the two phases identified above.

The earthworks around the Chalybeate Spring suggest some possibilities about Shenstone's structure here. The masonry remains currently there today seem to be later. It is understood that Shenstone talks of bringing water from upstream to his Chalybeate Spring to supplement the spring on the spot. It had been thought that this might involve piping water from elsewhere. The evidence from the earthworks suggests this may not be the case. What may have been meant here was the taking of water from the pond itself by way of the side ditch that leads into the Chalybeate Spring ditch. If the feature belongs to Shenstone's time, there is no other clear explanation why this substantial ditch is otherwise present. The alternative is that it was cut later to drain the pond and cascade area for the repairs conjectured for the later 19th century. However, it could not be used to drain the pond very successfully, and its positioning suggests it may have been associated with supplementing the water of the spring.

The position it enters the Chalybeate Spring ditch is also indicative of the nature of Shenstone's ornamentation here. On first impression, the deep 'niche' made from the earthworks at the north end of the Chalybeate ditch suggests the main rockwork ornamentation (if any existed) was here. Although this may still be the case, an important part of the ornamentation must have been south of the point where the ditch from the pond meets the Chalybeate ditch. Otherwise, there would seem little point in channelling the extra water here if the ornamental rockwork was all in the 'niche' to the north. This discovery is perhaps unexpected, and may even suggest the main rockwork was south of the meeting point, or that the structures around the spring were far more elaborate and complication than has previously been given credit.

It is possible that one approached the spring from the NW. A possible holloway still survives on the ground to suggest this (John Hemingway pers comm.). However Dodsley states that when you arriveat the spring, it is on your left. Approached from the NW side it first appears on your right. From this it might be suggested that the conjectured route needs to be reconsidered. During excavation on the banks a level compacted 'platform' like area was discovered (trench 13). It was similar to other areas found elsewhere on the site, such as the Dripping Cascade. The compacted nature of the area by the Chalybeate Spring, plus its very level surface, suggests it was either a viewing platform, or a piece of ornamental surface to act as a backdrop to a view of the cascade from the opposite bank. Although it is possible it was put in by a later owner, the association of the similar surface by the Dripping Cascade with possible pre-1760 pottery suggests a Shenstonian date.

7.2 Virgil's Grove and surrounding areas

The excavation of the Obelisk shows that it had a hollow base, with some economy being expended on its construction. This may have been simple expediency, and may not necessarily be used to argue that Shenstone built all his features on the cheap. It is unlikely that the small amount of bricks saved by this construction method made any significant difference to his finances. The hidden design of the Obelisk was probably out of his hands, being the choice of the mason. It is far more likely that the masons worked on a fixed fee principle, and were therefore the ones making the decisions to try to save on materials. The author has noted on other sites that ornamental features were often built as cheaply as the mason could get away with. Any features, such as foundations and bases, that were out of view of the owner, were often done as cheaply as possible. The foundations of the 18th-century garden walls at Castle Bromwich Hall, West Midlands, were scarcely two bricks deep on a sand base, and the below-ground foundation of a statue base found there was not even mortared together (Currie & Locock 1993).

There is little else that can be said about the Obelisk. The curved stone fragment found with mortar on its upper surface suggests that the hollow was covered by a curved stone, with the main body of the Obelisk mortared onto this. Again, this is a common practice with statues, where they are mortared onto a brick base (pers. ob.).

The excavation across one of the paths near the Obelisk seems to confirm Shenstone's statement that his paths were not gravelled. It would seem they were simple dirt or grass affairs. Although the path excavated may not be the main circuit path of his landscape, it would seem to be of a contemporary date. There is no reason for thinking the circuit path was any different. Where it survives in good condition, it can be seen to be a slight holloway, similar to the path examined. One would assume that this type of wear was unlikely on a path made of a compacted gravel surface. These tend to be cambered to allow free drainage, being higher in the middle for that purpose (Currie 1996).

Excavations undertaken in the area of the Dripping Cascade seemed to reveal sparse traces of that feature. The rill can definitely be traced but the stonework left around it is fragmentary. As at the Middle Pond this seems to have been deliberately removed later. There is also evidence for some earlier 20th century activity here that may have also been cleared away more recently.

One of the most interesting features found here was the area of crushed ceramic. In some respects this seems to have been a made surface, but whether deliberately or incidentally is not known. There are a number of hypothesis that can be put forward to explain them. The BUFAU evaluation suggested they were a path, but this does not seem to be the case. The positions of the Shenstonian paths in this area are known, and this is not one of them. Furthermore, the shape and texture of the surface are not consistent with any known contemporary path types. It is more likely that the surface has been created incidentally, either by workmen making the cascade, or demolishing it. Alternatively, it is a decorative coloured banding deliberately laid to ornament the cascade.

In considering the first option, the deposition of compacted mortar around built structures is commonly found on archaeological sites. This is created by the workmen trampling their own debris into the ground as they work. In the case of the second option, crushed materials have been found on garden sites by archaeology. At Tredegar House, Gwent, crushed coal and oyster was used to make decorative banding of black and white to help pick out the pattern of a late 17th/early 18th-century formal garden (Currie and Locock 1997). Although such decoration is more usual in formal garden types, there is no reason why Shenstone should not have used it. The informal landscape at The Leasowes was a fairly early example of its kind, and borrowing ideas from earlier garden designs would not be impossible.

There are objections to both theories. It is understandable that workmen should trample mortar and brick debris into areas where they are working, but this material seems to be predominantly crushed earthenware flowerpots. Could it be that the plants ornamenting the cascade were brought there in flowerpots that were then broken on the spot and trampled in? Alternatively, perhaps the cascade was originally ornamented by plants in pots. It was common practice in the 18th century to leave plants in pots, often so they could be brought indoors in the winter if they were delicate. In the case of the Dripping Cascade, as it became derelict the pots gradually became dislodged and were broken. When the cascade was dismantled, the workmen trampled the broken pots into the ground even further. However, the tiny fragments into which the pots are broken seem to have been difficult to achieve by chance action, suggesting a deliberate act.

The objection to the decorative banding theory is that this would soon become obscured by plant growth. The same could be said of places like Tredegar House, where it is known to have been used. When people create gardens, they do so with good intentions regarding maintenance. Shenstone may have intended to keep the banding clear of weeds. To suggest he was too poor for such maintenance can not be supported. His claim to poverty was a relative thing. No genuinely 'poor' man could have afforded an estate of the size he owned even in the 18th century. Even with his relative 'poverty' Shenstone still fell within the top 10-15% most wealthy people in the UK at the time. He certainly could have afforded the maintenance if he had wanted. More realistically, he probably became bored with the minor details of upkeep. It was probably in this manner that parts of his design may have become neglected. So it is hardly surprising that no visitors mention this conjectured decoration. It had likely become obscured by plants quite quickly.

It could be considered that the crushed materials were to aid drainage. If this was the case, they might be expected to cover a much wider area than that observed. Furthermore drainage features would not be deliberately compacted as this surface had been. Even chalk, one of the most porous of materials, can be compacted to form a water-tight surface, as is known from 18th-century treatises of pond-making (Currie 1990b). This might therefore cause water to remain on top of the crushed surface. In the case of the Dripping Cascade, this may even have assisted the 'dripping' nature of the feature by forcing water to remain on the surface, dripping over the banks of the main stream.

The area around the surface contained fragments of flowerpot bases, with side drainage holes. These are generally dated to before 1760, thereby suggesting a possible Shenstonian date for the deposition of the surface.

Finally one has to consider the failed attempt to find the roothouse by the Lower Pool. This was no surprise as this feature was probably of the most flimsy construction, possibly being made, literally, of roots. In an area heavily disturbed by roots, finding such remains was going to be very difficult, possibly even impossible. It is possible that the correct site was not explored. Contemporary descriptions of where the roothouse and the nearby Priory Gate were sited are extremely vague. There is no consensus as to their position, and one can only expect to recover their remains through possible chance discovery in the future. That is, if they can still be found at all. In the case of the Priory Gatehouse, it is most likely that the site was under the present tarmac road. In which case, it is possible that all traces may have been removed. This can only be determined if, at some future date, the road is dug up in the vicinity to allow a watching brief to be carried out (see Section 9.4).

7.3 The stream walk & brick collection

As stated in section 7.1, it seems that municipal 'tidying' this century has removed all original traces of Shenstone's 'hard' features along the streams. Elsewhere, such as below the Heart-Shaped Pool, the rockwork of the cascade was so similar to natural rock that it is difficult to tell it apart from Shenstone's artifices. It is, however, along this stretch that the greatest likelihood of Shenstonian survival occurs. It is possible that much of the natural-looking sandstone littering the stream here were placed there by Shenstone. The natural sandstone does not outcrop here much, and so much of the stone in the bank may have been placed there artificially. Those pieces in the stream have merely become dislodged with time, and have come to contribute to the 'untidy' appearance of the stream.

Bricks were uncommon along this stretch. Only two were identified, and both could have come from the nearby late 18th-century walled garden. Below the first bridge, there is a deposit of brick in the south bank of the stream. This is close to the conjectured site of Stamford's Roothouse. On excavation this proved to be the remains of a cascade (see Section 6.11). There was nothing else of interest on this stream.

The north stream was uniformly disappointing. Although features were found alongside the side attributable to Shenstone like the Obelisk, the Dripping Cascade, and the earthworks of the Chalybeate Spring and the series of ponds he built, the stream itself seems to have been thoroughly cleared on Shenstonian masonry.

An analysis of the bricks is interesting for rather unexpected reasons. It is notable that the highest concentration of modern materials is directly associated with the position of the more importance of Shenstone's features. One would have expect earlier bricks to be concentrated here, but this is not the case. Instead the earlier bricks tend to be found away from areas where he concentrated his activities. This seems, on the surface, to be the exact opposite of expectations.

However, the explanation seems to confirm other evidence for deliberate clearance of debris around his more important features. The evidence of the brick collection seems to suggest that on the north stream, 20th-century municipal 'tidying' has resulted in nearly all the Shenstonian debris near his more important features being removed. They have then been replaced by modern structures that made little attempt to imitate the originals. The people making them may have had no knowledge of Shenstone, and were creating features that simply served municipal utilitarianism. In their turn, these features have also become ruinous, thereby depositing higher concentrations of modern material around old sites than would otherwise be expected. The occurrence of higher concentrations of earlier material away from his major features can be explained by the fact that these areas were isolated from municipal activity. They were thus not cleared with the rigour attending the more important areas.

The debris in the north stream below the Lower Pool seems to have largely derived from recent off-road dumping adjacent to the bridge. This is a convenient place for illegal tipping, and much of the material in the stream below this point has nothing to do with Shenstone, or repairs to his features. The cascade that fell from the Lower Pool towards the Priory Pool seems to have been a naturalistic feature similar to that below the Heart-Shaped Pool. The difference here being that there is some evidence for natural sandstone outcropping along the stream course, at least as far as the point where the modern footpath crosses the stream. Shenstone probably brought natural unworked stone in to embellish this cascade, but if he did, it is not possible to identify his work. No worked stone appears to have been used along this stretch. There is also an absence of Shenstonian bricks. This is so marked that one might think that municipal clearance is not the culprit here. Instead, it seems unlikely that he used brick in this stretch.

7.4 The Priory Cottage

The results of the excavations here exceeded expectations. The walls of the cottage and its adjoining 'ruined' folly were found to survive to over 0.7m high in places. Evidence was found for doors, fireplaces and floors, although the remains were not high enough to locate the position of windows.

The largest area excavated, delimited by walls 33, 26, 27/32 and 44, seemed to be an extension added to the original building. There was no definitive dating evidence to prove this by way of artefacts, but there was other evidence that led to this conclusion. Early maps of the site suggest that the original building was a T-shape, with a small single room cottage attached on the NW side. The long side of the building contained the main facade of the priory ruin on its SE side. The plan, as far as was recovered, suggested a more rectangular plan caused by adding a large room on the SW corner of the building. This later plan is shown on the first large scale Ordnance Survey map (dated c. 1887; see Figure 13), but the earlier plan is shown on the tithe map of 1842, and a plan of 1859 (see Gallagher 1996, 216, Figure 8). This suggests that the room on the SW side of the site was an addition added between 1859 and 1887. Furthermore, a drawing of c. 1770 shows a door in the SW side of the cottage (see Figure 14). There was no sign of a door in this position on the recovered plan, although a door was found in what was thought to be the original SW wall of the cottage. This door was in the exact position that coincided with the c. 1770 drawing.

Although the map and pictorial evidence seem conclusive, there are a number of puzzling features to consider before wholeheartedly accepting the idea that the extension was added between 1859 and 1887. The SW wall of the conjectured extension (context 33) was made of bricks that were 7cms thick on average. These were close in size to the brick types used in Shenstone's Obelisk (trench 2), although perhaps tending towards slightly larger types (by 2-3mm on average). The bricks used in the NW wall (contexts 27/32) were types 8cms thick. These were thought to have been used in the later 19th century. Initially this suggested the SW wall may have been of earlier date, with the curving wall being of the later 19th century. This might indicate that the SW room had been altered. That is that it had been built in the 18th century, but had been altered to a curving apsidal end after 1850. Initial examination of photographs of the cottage just prior to demolition showed an ornate window that can also be seen in the c. 1770 drawing. Did this suggest the wall of 1770 was the same as context 33?

The map evidence, plus that of the door shown in the c. 1770 drawing, suggests another interpretation. As a number of plans show a T-shaped building before 1859, there is reason to think they must relate an element of truth. If the conjectured walls of this structure are examined, they all seem to have originally been made of stone. Wall 26 is stone, with a brick portion added over the top of a cellar. Wall 44 can be seen to be courses of brick overlying a foundation of stone. Only walls 27/32 and 33 can be shown to be brick throughout. From this it seems that Shenstone's original building, including the cottage tacked on to the back, was made of stone. Local tradition, as well as some contemporary commentators, have reported that Shenstone built the priory from stone from Halesowen Abbey. Two types of stone (red and greenish-grey sandstone) were found built into the priory ruins; both types can be found at Halesowen Abbey. All those parts of brick seem to be later.

Explanations for the different brick types in walls 27/32 and 33 can be made to agree with the idea that the original building was of stone. Firstly, the SW wall could have been made of reused 18th-century bricks. Secondly, the bricks are not definitely Shenstonian, being on average 2-3mm thicker, and may therefore be later, either reused or of mid-19th-century date. Thirdly, the SW wall may have been built before wall 27/32, with the latter being altered later in the 19th century, or even early in the 20th century. In the case of the window, it is of an elaborate type that is likely to be reused, and reset in wall 33 in approximately the same place as it had once been in wall 26. On the present evidence, it seems the SW room was added between 1859 and c. 1887, possibly using some earlier materials, and that the original building can be identified by its mainly stone walls.

The archaeological evidence indicated that the early pictures of the priory were largely accurate. The Jenkins drawing of c. 1770 shows square buttresses on the corners of the SW side of the ruinous part of the building. This was confirmed by the archaeology. Although the pictures do not give the perspective to confirm the apsidal end of this part of the structure for certain, they do allow that this was possible. Later 20th-century photographs of the building clearly show the curved end to the Gothic facade.

The priory ruin was shown to have been part of a narrow structure, with a long side facing obliquely the mansion house. As the present mansion is not the original, it may have faced Shenstone's house more squarely. The narrowness of the space behind the facade indicates that costs were kept to a minimum, by making a maximum impact on the 'show' sides. The existence of a door from the cottage into the priory ruin argues that the ground floor of the priory may have been roofed to provide the cottage tenant with extra space. Only the upper floor was made to look ruinous to create the effect. It is notable that the windows on the SW side seem to have intact tracery or mullions, suggesting the space behind was used. As Shenstone's folly was only meant to be seen from a distance, there is no reason why the ground floor of the 'ruin' could not have been made to be used. As Shenstone himself said, his ruin was made in such a way that he could gain a rent from putting a tenant in it. The more living space he could offer that tenant, the more money he may have been able to charge.

The later 19th-century extension contained painted plaster walls, and a tiled floor. The brick kitchen range against the SE wall of the extension was built to butt onto the wall. It also overlay the tiles. This suggests it was put in after the room had been completed. Whether this was immediately after or some years later is not known. It was not part of the original room as it is not bonded into the back wall. Likewise, the door leading from the SW room into the conjectured original cottage was altered. The NW stone jamb can be seen to sit on top of the clay floor tiles. Although of stone, this seems to have been reused to narrow the door. This is probably because the original door had been the main entrance, but was now a subsidiary internal door. The SW extension appears to have been used as a kitchen.

The new main entrance seems to have been the door in wall 27/32. This can be suggested by the brick and tile paving outside it. It, in turn, was reached by a path coming from the south. On the SW side of the door were the remains of an iron feature thought to be a foot scraper.

The internal floor tiles continued from the conjectured kitchen into the original cottage in what seemed to be a continuous pattern. If the SW room was added after 1859, the tile flooring there and in the earlier room to the NE must have also been laid after this date. The door between the two was narrowed after the floor had been laid. The cellar in the east corner caused the removal of part of the original stone wall (context 26), and its replacement in brick. It is uncertain if this was done at the time the SW extension was added, or as an afterthought.

However, the narrowing of the door after laying the floor, the failure to tie the fireplace into the new NE wall and the late bricks used in the NW wall (context 27/32) collectively seem to suggest that the extension may have been altered after its original construction. Further, the creation of a cellar under part of the original ruin, plus the conversion of its internal wall from stone to brick, suggests that by the later 19th century, at least, the ground floor of the 'ruin' was being used by the cottage tenant. The discovery of the new brick internal wall may have alerted visitors to the forged nature of the building if they had access. From this evidence, it might be suggested that they did not at this time. It is possible the same situation existed in Shenstone's time.

Other evidence found in the rubble included building materials. At least one large piece of window arch was found, confirming the Early English Gothic architecture of the building. It is not known if these windows were taken from Halesowen Abbey, but their style can be found in the ruined refectory today. The stone is also a similar red sandstone. Fragments of crested ridge tiles were also found. These were not of medieval date, having a machine-like finish. It is possible they were added to the revamped cottage after 1859. A 20th-century photograph of the cottage shows what seems to be these ridge tiles on the roof of the single storey extension (see Plate 7).

This photograph also seems to show a light-coloured render on the SW wall of this extension. The archaeological excavation revealed traces of a plaster-like render on the outer wall of this building, further confirming that the extension had been added later. Jenkins' drawing of c. 1770 seems to show the cottage wall in unrendered stone, at least one storey high, with an additional attic storey indicated by dormer windows over. There is no sign of ridge tiles on the cottage roof in this drawing. Other differences between the c. 1770 drawing and the 20th-century photograph include the high chimney stack shown coming from the extension. The photograph also shows a chimney emerging from the ruin. This is not shown in 1770, nor are the additional half-light window bars in the Early English windows in the apsidal end of the ruin. These latter additions show that, even if the ruin had not been occupied in the 18th century, it seems to have been converted for habitation after 1859.

The most revealing description of the building comes from the playwright, Thomas Hull, c. 1759. He says the priory is:

'...a pleasing Deception. It appears a Ruin of a Chapel built some hundred Years ago, whereas it has not yet been built two Years - It is reared in the Gothic Taste, & notwithstanding the outward Face, has some comfortable Rooms in the Heart of it, particularly a Gothic parlor, with antique Paper, finished in a most happy Taste. This building is not merely ornamental, for the habitable part is let to an old Man & an old Woman for the yearly Rent of 4£.' (Quoted in Reily 1979, 207-08)

Shenstone himself confirms the habitable nature of his folly in a letter to Richard Graves, when he says he has made c. 1757, 'the ruins of a Priory, which, however, make a tenant's house, that pays me a tolerable poundage' (Williams 1939, letter W50.ccxvi).

The most interesting building materials found were fragments of armourial shields. These were apparently made of a light brown (beige) ceramic. The designs seemed to be mainly of a simple type in relief. Vertical bars and flowers were the main features identified. Shenstone is recorded as having decorated one of the rooms of the priory with invented coats of arms of his friends. The bulk of these finds seemed to come from the rubble within the original cottage, suggesting that it was here than Shenstone had placed these features. They were found in sufficient quantities to suggest they had been fairly numerous.

In another letter to Graves, Shenstone describes the 'Gothic shields':

'Did you see my Priory? - a tenant's house, one room whereof is to have Gothic shields round the cornice - I am in some doubt whether to make it an House of Lords or House of Commons; if the former, my private friends will have shields round my Gothic bed-chamber' (Williams 1939, letter W50.ccxxix).

Finally it might be mentioned that there was some tenuous evidence that the cottage was ornamented by plant beds around the outside. There was some evidence for a plant bed, edged in stone running along the side of the SW wall up to the conjecture garden path. There was further lesser evidence for the remnants of an edging to a border to the NE of the paved area outside the main door. The garden path, which seems to have been made of whatever materials were available, seems to head in the direction of a small detached structure shown on the c. 1887 OS 25" map. This may have been an outdoor toilet.

There was little else that can be said of the priory building from the archaeological evidence alone. Everything found seems to suggest that historic plans and drawings of the building are largely accurate. More precise dimensions for possible reconstruction were obtained (these can be extrapolated from Figure 10), however, as well as this important confirmatory information. The recovery of building materials also enables us to be more accurate about the materials used at different times, and on different parts of the structure. Taken together, the evidence should allow a far more accurate reconstruction to be attempted than on many similar restorations currently being undertaken.

7.5 Stamford's Roothouse and Cascade

The trench excavated for Stamford's Roothouse (trench 10) proved to be a disappointment. There was no evidence for the roothouse, and little evidence for any nearby structure. If the roothouse was nearby, and built of durable materials, such as brick or stone, one might expect to find debris nearby. It is likely that the roothouse was made of timber, and all traces of it have been destroyed by the extensive root disturbance in the area. The possible post-hole found in the trench may have been part of the structure, but it is thought unlikely.

An unexpected discovery of a stone and brick lined water channel in the side of the present stream has implications for the location of the roothouse. The site chosen for excavation of the roothouse was based on the previous belief that the cascade coming from the heart-shaped pond ended near the present bridge. If the stone lined channel was part of the cascade, it would seem to have finished about 20-30m further south than previously thought. Consequently, the position of the roothouse needs to be reconsidered as also being slightly downstream of the end of the cascade.

That the structural remains found by the stream were part of the cascade seems a reasonable conclusion. However, the remains found are probably not the end of the cascade. The way the stream bends here suggests that the present stream has cut through the original structure, washing part of it away. Although it is suspected that the original end was probably within 10-15m of the surviving remains, this is not known for certain.

Why have these remains survived here? The topography of the valley suggests that the stream has been diverted from just below the present bridge. This can be seen by the way the steeper slope of the valley has been 'isolated' behind a flatter area. It might be suggested that something had blocked the stream in the past. It is not known if this was human agency, neglect, or a fallen tree. The blockage caused the stream to seek out a new course, thus adding a new bend to the stream. The former channel became infilled, again by unknown agencies. There is some suggestion in the excavated remains that at least part of the backfilling was deliberate. This is shown by the deposit of brick rubble in the channel.

Considering that part of this lined cascade appears to have been eroded through by the new stream course, there is relatively little debris in the stream below it. Earlier the results of the stream walk had suggested a cleaning up operation, possibly under early municipal ownership. It is possibly at this date that the already partly filled channel was backfilled with debris taken out of the stream.

The structure itself may have been of two phases. The stone lined channel may have been the original built by Shenstone, thereby demonstrating that his cascades were more than a simple dump of natural looking stones along a semi-natural cascade, but a deliberately constructed artifice. Nowhere in early descriptions does it mention that Shenstone's cascades include brick. Although the bricks originally looked like 18th-century types (based on thickness), work on the Priory Cottage has shown that these bricks were used to construct what is thought to be later 19th-century parts of that structure. The remains of brickwork in the other cascade below Middle Pond have been shown to be post-Shenstonian. A 19th-century date has been proposed from the latest brick types. These bricks are used in the extension to the Priory Cottage, work dated to between 1859 and 1887. On the similarity of brick types, it might be suggested that the enlargement of the priory, plus the repairs to the cascade, were all carried out at roughly the same time.

On the basis of this, it might be suggested that the cascade in question was repaired in the 1859-85 period. Shenstone's stone-lined original had fallen into disrepair by this date. The blocking of the original line of the cascade, and the diversion of the stream must, therefore, have taken place after this. The infilling of the derelict channel was probably the early work of the municipal authorities in the first half of the 20th century.

8.0 Conclusions

Surveys included making measured plans of two derelict pond dams, plus the earthworks associated with a grotto-like feature known as the Chalybeate Spring. Excavations revealed the possible sparse remains of a feature called the Dripping Cascade. The partial remains of a brick base marked the site of an obelisk, but nothing could be found of his two roothouses. The latter were believed to be ephemeral wooden structures, possibly made from roots. The remains of a cascade from the Heart-shaped Pond were well preserved at the western end where they had been buried following a possible diversion of the stream. This was found to be a stone lined channel that was possibly repaired at a later date, and given a brick floor and edges within the older stone channel.

Substantial remains of a Gothic folly known as the Priory Cottage were found next to a golf course tee. This originally appears to have been a stone structure, with a long Gothic facade facing the house, and a small cottage behind. It is thought to have been built c. 1757. Numerous fragments of painted ceramic heraldic shields were found in the rubble confirming Shenstone's report that he had decorated the cottage with Gothic shields. Between 1859 and 1887 the cottage appears to have been enlarged by the addition of a brick extension on the SW side of the building, thought to be a kitchen. The interior of the cottage was tiled with ceramic floor tiles, and a large kitchen range placed in the extension. Internally some older walls may have been taken down and replaced in brick. It is thought that a cellar was dug at this time.

The results of these excavations and surveys have added considerably to our knowledge of Shenstone's landscape and the methods he used to construct it. There were a number of surprises, not the least that his work was more complex than previously considered. His cascades and other water features seem to have been more elaborate than more recent commentators have allowed. There is tentative evidence to suggest their ornamentation went beyond the construction of the features themselves. The regular occurrence of compacted pinkish surfaces near his features suggests the use of coloured materials laid on to parts of the surrounding earth for ornamental effect. That these would become rapidly overgrown without maintenance suggests that he may have originally conceived a long-term management scheme of upkeep. His water features may have survived in recognisable form until at least 1859-85, when they seem to have been repaired in brick. Their final dereliction seems to have occurred within 50 years of this rebuilding, and much evidence seems to have been removed during a possible municipal clear-up this century.

The evidence suggests that the durability and extent of Shenstone's works have been under-estimated. If this can be shown to be the case for his water-features, those most prone to erosion and decay, it bodes well for other structures he made that have not yet been examined to the east of the house.

9.0 Recommendations

The recommendations given here relate only to a statement of where further archaeological work would be useful, or where restoration might have a detrimental effect on the site's archaeology. They may contain views that are personal to the author.

9.1 Given that the remains of Shenstone's follies were often more extensive than previously thought, the most obvious recommendation is that where restorations of structures are definitely intended, as full an excavation as the circumstances of the site allow should be attempted beforehand. This should preferably be before any work is undertaken and should assist in the drawing up plans for reconstruction. This way the archaeology can progress without being hampered by any pre-conceived notions of the results.

9.2 This recommendation should be applied to the Priory Cottage (see Plates 3-7). It should be noted here that the site of this building stands on a particularly contentious piece of land. Part of the remains extends under a golf course tee. Another part of the structure is under three very large ash trees that will need permission to remove before further works can commence. The management should therefore consider very carefully if reconstruction on the exact spot of the former building is required. It should also consider if a full reconstruction of the entire cottage is necessary. It is possible that a reconstruction of the Gothic priory facade on a less contentious site adjoining is a solution that needs to be seriously considered.

However, if such a step is planned, it is advised that the reconstruction does not take place over another part of the ruins. A site outside the area of the present ruins would be preferred, so that the remains can continue to survive in situ for the study of future generations. A spot slightly to the SE of the remains, alongside the golf tee, might be considered for this.

If reconstruction on the exact site of the former priory is intended, it should be noted that this will inevitably result in damage to the existing ruins. This should not be undertaken without giving serious consideration to the alternative. The preservation of important archaeological remains should not be compromised for the sake of restoration when a reasonable alternative is available. That this alternative would be of benefit to the golf club, who have used this site for nearly one hundred years, and are themselves tenants with a claim of historical credence, is an additional reason for urging its consideration.

As the priory was meant to be viewed from a distance, it is unlike that its displacement by 10m or so would be of any great detriment to the restoration. In particular, the present ruins are in such good condition that they possibly preserve considerable information about Shenstone and his methods. It is preferable, considering the importance of his influence on landscape design, that this is preserved in situ. The result of reconstruction on the present ruins will be the need to undertake a full excavation of the site to recover the maximum amount of information.

Jenkins' drawing of c. 1770 (see Fig. 14) can be taken as an accurate basis for restoration of the building of that date.

9.3 The Obelisk site (see Plate 1) may present a different case. Here the existing remains seem to be robust enough to be reused as part of a reconstructed feature if required. There is little additional archaeological information that can be obtained from further disturbance of this site. However, it should be stressed that restoration should try not to remove any existing remains if they prove to be inconvenient. If, for engineering reasons, they can not be incorporated without damage or seriously compromising their site and position, it would be preferred if the Obelisk is reconstructed slightly to one side, leaving the original base preserved in situ.

9.4 If the road near the Lower Pool is ever be taken up or repaired, the opportunity should be taken to conduct a watching brief to see if remains of the Priory Gatehouse can be located.

9.5 Regarding the restoration of the water features and the cascades, on present evidence, restoration may not cause archaeological remains of great importance to be destroyed. Those remains that will be disturbed could be compensated for within the present planning guidelines. When restoration here goes ahead watching briefs may be sufficient to record the archaeology.

There are two possible exceptions to this.

9.6 The first is that the in situ remains of the cascade on the south stream (see Plate 8) should be explored as fully as possible before detailed restoration proposals are drawn up for this area. This discovery was unexpected, and no provision had been made for it. There is possibly a substantial length of the cascade remaining in situ. This can offer important information on Shenstone's methods, and may give clues to his techniques at other cascade sites. It is recommended that an archaeological exploration is made of this feature. Until this is done, it is not known if the remains are suitable for reuse.

9.7 The areas of pinkish compacted surfaces found adjacent to at least two Shenstonian water features make up the second exception. The regular occurrence of these suggests they are an important part of his methodology in creating ornamental features. It is recommended that further information on their extent and use is obtained by archaeological excavation.

10.0 Specialist reports

10.1 Building materials

10.1.1 Stone

From the evidence available, it would seem that Shenstone used mainly local sandstones of a red and greenish-grey type in his features. This was best seen in the Priory Cottage where a number of architectural fragments were recovered. Local tradition claims that some of the stone came from nearby Halesowen Abbey. Both sandstone types can be identified in the present abbey ruins. Much of the stone was in a damaged condition, and little could be said of it. However, one large fragment of a Gothic window arch was found. This seemed to be of the type shown in the pictures of the Priory Cottage in the 18th century. This fragment, plus others of interest, were left at the Dudley Metropolitan Borough Council Depot at The Leasowes pending decision on their future.

10.1.2 Brick

A random selection of bricks from the Obelisk was measured. The dimensions taken were:

The fabric of these bricks was a lightly sandy fabric, with large lumps of haematite in some samples (up to 20mm) and occasional calcerous inclusions (less than 1mm). The colour varied from an orange-red to a purple-red, dependent on firing. The darker colours tended to be harder fired. This indicated unevenness of firing in the kiln. The fabric also exhibited extensive 'marbling', suggesting poor mixing of the clay. These factors combined to suggest that the bricks were made by a local rural industry, and not a large mass producer.

Other brick samples were taken from the Priory Cottage and other remains within the park. Fifteen whole bricks or fragments of bricks, weighing 35.04 kilograms, were retained. These can be found listed in the site archive. Numerous other bricks or fragments of bricks taken during the Stream Walk were kept at the Dudley Metropolitan Council Borough Depot at The Leasowes pending decision on their future. These are listed by size and weight in Appendix 2.

10.1.3 Roof tile

Roof tile was largely confined to the Priory Cottage site. It comprised mainly red clay roof tile, many with nibs, from the overlying rubble. These were sampled on site and discarded. Some stamped examples were found. One fragment, weighing 220 grams, was stamped 'ROS...'. The most common stamp was 'Dreadnought', presumably the name of a local producer. Two fragments containing part of this stamp ('DREAD...' and '..NOU..') weighing 150 and 100 grms respectively were retained in the archive. These seemed to be of 19th-century, or later, date.

Three fragments of ridge tiles, weighing 2.835 kilograms, were recovered. These had a smooth machine-like finish. They were of the type that can be seen in a photograph of the Priory Cottage on the roof of the later 19th-century extension.

10.1.4 Floor tile

Fragments of floor tile could be found all over the site of the Priory Cottage. Much of it was in situ on the floor of the cottage and the surface outside the entrance to the 19th-century extension. They are all thought to date from the period 1859-85. The average dimensions of the in situ tiles were 23x23x3.5cms. A further fragment of a thinner floor tile was found in the rubble overlying the ruins.

10.1.5 Ceramic shields

A number of fragments of heraldic shields were recovered from the rubble overlying the ruins of the Priory Cottage. These were in a light brown biscuit type firing similar to that used on fine earthenware, such as tin-glazed wares, in the 18th century. This fabric was more common in fine wares before 1760, becoming increasingly rare after this date. There were traces of white paint on some of the fragments, suggesting that they were painted.

None of the fragments were whole. They represented parts of shields and floral decoration, all in relief. The shields were decorated mainly by vertical bars. No distinctive patterns were found that seemed to relate to genuine heraldic designs.

Nine fragments were collected weighing 2.455 kilograms. All came from rubble layer 23.

10.1.6 Pottery

There was little pottery of significance recovered from the excavations. Some fragments of possible 18th-century wares were found by the Dripping Cascade. Larger quantities were found in the rubble overlying the Priory Cottage. This was considered unstratified. Only one piece that may have been earlier than the mid-19th century was observed (see below), and this assemblage was mainly discarded on site. Only diagnostic wares were retained. Modern 'china' types (mostly willow pattern), by far the most frequent pottery find, were not retained. No deposits suggesting pits or earlier dates were found. It was assumed that the ceramics found in the rubble had been in use during the last occupation of the cottage, and were deposited on its demolition c. 1950.

Pottery types retained included:

Unstratified from the Dripping Cascade:

Two base sherds of earthenware flowerpots, with side drainage holes 175 grms; possibly 18th century?

From context 09 (adjoining Dripping Cascade):

Two sherds of biscuit fired earthenware with yellow glaze 30 grms; 18th/19th century?

From context 23 (rubble overlying Priory Cottage):

Manganese-glaze earthenware handle (teapot?) 35 grms; 18th century?

Two sherds of black-glazed earthenware 85 grms; 18th/19th century?

One sherd 19th or early 20th-century stoneware stamped '..PET..' on the top line, with '..01 H..' below.

10.1.7 Clay Pipe

Hardly any clay pipe was found during these excavations. This was also true of the Priory Cottage, where only one piece was found in the overlying rubble. It might therefore be assumed that clay pipes had ceased to be used before the cottage was demolished. The one fragment found was from the rubble overlying the ruins of the Priory Cottage. It was a bowl, ornamented in relief and much worn, characteristic of 19th-century date.

11.0 Archive

The archive for this work will be deposited with Dudley Museum Services. Copies of the report were lodged with the client, the West Midlands Sites and Monuments Record (SMR), and the National Monuments Record, Swindon, Wiltshire.

12.0 Acknowledgements

Sincere thanks are given to all those involved with this project. In particular, Lynn Foord, the Project Manager, is thanked for providing information on the site, storage for tools and finds, and discussing the results with the author. John Hemingway monitored the project for the Archaeology Section of Dudley Metropolitan Borough Council, and gave the excavators the benefit of his personal knowledge of the site. The staff of the Council Depot at The Leasowes provided much cheerful encouragement, provided additional safety fencing around the Priory Cottage, obtained sight of an old postcard showing the Priory before it was demolished, and allowed the author to copy it.

13.0 References

(Figs. 13-14; Plate 7)

13.1 Original sources

Maps:

Tithe map & award for Lapal, 1844

Tithe map & award for Hill, 1844

OS 25" map, c. 1887 edition, sheet (see Fig. 13)

OS 25" map, 1933 edition, sheet

OS 25" map, 1955 edition, sheet

Pictures:

D Jenkins, 'A view of the Leasowes... including the Priory & Seat...', c. 1770 (see Fig. 14)

Photograph of Priory Cottage from SE, n.d (but 20th-century?), supplied by Dudley MBC (see Plate 7)

13.2 Original sources in print

E M Betts (ed), Thomas Jefferson's garden book, 1766-1824, 1944

R Dodsley (ed), The poetical works of William Shenstone, with the life of the author, and a description of The Leasowes, 1765

J Heeley, Letters on the beauties of Hagley, Envil and The Leasowes, London, 1777

R Graves, Recollections of some particulars in the life of the late William Shenstone Esq. in a series of letters from an intimate friens of his to (W S) Esq., 1788

M Williams (ed), The letters of William Shenstone, 1939

13.3 Secondary sources

W H Andrews, J L Allen, & A St John Holt, Health and safety in field archaeology (Southampton, 1991).

Central Excavating Service, Site Recording Manual, version 7, London, 1992.

D Chambers, 'Smoothing or brushing the robe of nature: William Shenstone and the Leasowes' in The planters of the English Landscape Garden, New Haven, 1993, pp. 177-84

C K Currie, 'Titchfield fishponds project 1985: interim report', Hampsh. Field Club & Archaeol. Soc. newsletter new series 6 (1986), 19-21

C K Currie, 'The excavation of an 18th-century garden pond: the West Pond, Castle Bromwich Hall, West Midlands', Post-Medieval Archaeology 24 (1990), 93-123

C K Currie, 'Southwick Priory fishponds: excavations 1987', Proceedings of the Hampsh. Field Club & Archaeol. Soc., 46 (1990a), 53-72

C K Currie, 'Fishponds as garden features c. 1550-1750', Garden History, 18.1 (1990b), 22-46

C K Currie, 'The archaeology of the flowerpot in England and Wales c. 1650-1950' Garden History, 21.2 (1993), 227-46

C K Currie, 'Excavations in the park and gardens at Leigh Park, Havant, 1992', Proceedings of the Hampshire Field Club and Archaeological Society, 51 (for 1995), 201-32

C K Currie, Project Design for archaeological investigations at The Leasowes, Halesowen, West Midlands, report to Dudley Metropolitan Borough Council, 1998

C K Currie & M Locock, 'An analysis of the archaeological techniques undertaken during the first year's excavations at Castle Bromwich Hall Gardens, 1989-90', Garden History 19.1 (1991), 77-99

C K Currie & M Locock, 'Excavations at Castle Bromwich Hall gardens 1989-91' Post-Medieval Archaeology, 27 (1993), 111-99

C K Currie & M Locock, 'The development of garden archaeology', Institute of Field Archaeologists Yearbook and Directory of members 1997, Manchester, 1997, 23-25

L Dingwell, An archaeological evaluation at Leasowes Park, Halesowen, West Midlands, 1996, report to Dudley Metropolitan Borough Council, 1996

English Heritage, The management of archaeological projects, London, 1992 (revised edition).

English Heritage Register of Parks and Gardens of Special Historic Interest in England, London, 1987

C Gallagher, The Leasowes Restoration: proposals for replanting and future management, unpublished report to Dudley Metropolitan Borough Council, 1997

C Gallagher, 'The Leasowes: a history of the landscape', Garden History, 24.2 (1996), 201-221

Institute of Field Archaeologists, Standard and guidance for archaeological field evaluations, Birmingham, 1993.

J Owen (ed.), Towards an accessible archaeological archive, Society of Museum Archaeologists, London, 1995

J Riely, 'Shenstone's Walks: the Genesis of The Leasowes' Apollo, 110 (1979), 202-09

Appendix 1: list of recorded contexts

Context no. Trench no.
1 T/2 loamy layer (Munsell Colour 5YR 3/1)
2 T/2 clay layer (Munsell Colour 5YR 4/6)
3 T/2 mortared brick structure
4 T/2 loamy layer (Munsell Colour 5YR 3/1)
5 T/3 loamy layer (Munsell Colour 10YR 3/1)
6 T/3 clay layer (Munsell Colour 5YR 4/6)
7 T/3 cut?
8 T/3 loam fill of 07 (Munsell Colour 10YR 3/1)
9 T/5 silty clay layer (Munsell Colour 10YR 2/1)
10 T/5 silty loam layer (Munsell Colour 10YR 2/1)
11 T/5 clay layer (Munsell Colour 2.5YR 5/4)
12 T/5 clay layer (Munsell Colour 5Y 5/2)
13 T/5 sandy clay layer (Munsell Colour 5Y 4/2)
14 T/6 silty clay layer (Munsell Colour 7.5YR 3/0)
15 T/6 clay loam layer (Munsell Colour 7.5YR 3/2)
16 T/6 clay layer (Munsell Colour 2.5YR 4/4)
17 T/6 sandy loam layer (Munsell Colour 10YR 4/3)
18 T/6 sandy clay layer (Munsell Colour 10YR 4/1)
19 T/6 clayey sand layer (Munsell Colour 10YR 6/6)
20 T/6 sandy clay layer (Munsell Colour 5Y 4/2)
21 T/7 loam layer (Munsell Colour 10YR 3/2)
22 T/7 sandy loam (Munsell Colour 7.5YR 4/3)
23 T/8 sandy loam with abundant rubble (Munsell Colour 10YR 3/2)
24 T/8 mortared brick wall
25 T/8 ceramic floor tiles
26 T/8 sandstone wall
27 T/8 mortared brick wall
28 T/8 irregular ceramic floor tiles with some brick infilling
29 T/8 sandy loam layer with mortar flecks (Munsell Colour 10YR 5/4)
30 T/8 mortar bedding to tile floor (Munsell Colour 10YR 6/4)
31 T/8 brick threshold (door)
32 T/8 mortared brick wall
33 T/8 mortared brick wall
34 T/8 sandy loam layer (Munsell Colour 10YR 3/4)
35 T/8 brick surface (part of path?)
36 T/8 sandy loam layer (Munsell Colour 10YR 3/2)
37 T/8 brick surface (part of 28/35?)
38 T/8 stone surface or edging (part of 28/35?)
39 T/8 stone surface (part of 28/35?)
40 T/8 gravel layer, possibly path (Munsell Colour 10YR 3/2)
41 T/8 ceramic tile floor
42 T/8 threshold of door
43 T/8 mortared brick wall
44 T/8 mortared brick wall
45 T/8 mortared brick wall, part of cellar
46 T/8 rubble fill of cellar (Munsell Colour 10YR 6/4)
47 T/8 mortared brick fireplace
48 T/8 mortared brick wall
49 T/8 semi-circular sandstone wall
50 T/8 sandstone footings to wall 44
51 T/8 door with sandstone threshold
52 T/8 remains of iron fire-grate
53 T/8 stone buttress to wall 49
54 T/8 possible stone edging (to plant bed)
55 T/8 partial (stone) blocking of door 42
56 T/8 two iron bolts, possible remains of foot scraper
57 T/8 iron door hinge
58 T/8 sandy loam fill of plant bed (54) (Munsell Colour 10YR 3/2)
59 T/8 ceramic (circular) tile drain
60 T/9 iron (water?) pipe
61 T/9 modern ceramic pipe
62 T/9 sandy loam layer (Munsell Colour 10YR 3/3)
63 T/9 clay loam layer (Munsell Colour 10YR 5/4)
64 T/9 linear cut containing ceramic pipe 61
65 T/9 S-shaped cut containing iron pipe 60
66 T/10 sandy loam layer (Munsell Colour 10YR 3/1)
67 T/10 sandy clay loam layer (Munsell Colour 10YR 4/6)
68 T/10 cut (probable root disturbance?)
69 T/10 fill of cut 68
70 T/10 cut (post-hole or root disturbance?)
71 T/10 fill of cut 70
72 T/11 clay loam layer (Munsell Colour 10YR 3/2)
73 T/11 clay layer (Munsell Colour 10YR 4/6)
74 T/11 stone lining to water channel?
75 T/11 brick rubble in clay loam matrix (Munsell Colour 10YR 3/2)
76 T/11 linear cut
77 T/11 brick edging and floor of water channel?
78 T/13 sandy sily layer (Munsell Colour 10YR 3/1)
79 T/13 compacted pinkish surface in clay (Munsell Colour 2.5Y 5/4)
80 T/13 greenish sandstone in sandy clay layer (Munsell Colour 5Y 4/2)

Appendix 2: raw data from brick collection along streams

Key: bf brick fragment, bw whole brick, brick categories 1-7 (see section 5.2 for description)

North Stream: Beechwater to Lower Pool

Section 1

Brick type dimensions in cms. weight in kilos category other features
bw 22.5x7.5x11 3.97 3  
bf 11.5x6.5 1.82 6  
bf 11x8 1.36 2  
bf 7.5x? 1.13 6  
bf 11x7.5 0.91 2  
bf 11x8 1.36 2  
bf 11x7.5 1.82 6  
bf 11x6.75 1.59 4 frog, stamp 'PLA..'
bf 10x6.5 1.82 6  
bf 10x6.25 1.36 6  
bf 10x7.5 1.25 6  
bf 10x6 1.82 6  
bf 10x6 1.82 6  
bf 11x8.5 2.84 2  
bf 11x8 4.54 2  
bf 11x6.75 4.09 6  
bf 10.25x6.5 2.72 6  
bw 23.25x11.5x8 4.54 2  
bw 22.5x11x7 3.86 3  
bw 21x10x6.25 2.72 1  
bw 20.5x10x6 2.72 1  
bw 24x11x8 4.77 2  
bw 22.25x11x7 4.09 4 shallow frog
bw 22.5x11x8 4.43 2  
bw 21x10x7 3.18 5  
bw 23x11x7 4.31 3  
bw 21.5x10x7 3.18 5  
bw 10.75x6.5 2.6 6  

Sections 2/3

bf 11x7.75 1.36 2  
bf 11x6 2.04 6  
bf 11x6.5 1.82 6  
bf 9.5x5.5 1.25 6  
bf 10x6 2.04 6  
bf 10.5x7 1.36 6  
bf 10.5x7 1.82 6  
bf 10.5x6 1.25 6  
bf 10.5x6 2.04 6  
bf 11x8 2.84 2  
bf 11x8 3.18 2  
bf 11x7 2.72 6  
bf 9.5x6.5 1.36 6  
bf 10.5x6.5 2.26 6  
bw 23x10.5x8.5 3.75 2  
bf 11x7.5 2.95 6  
bw 22x10x6.25 2.84 1  
bw 22x10.5x6.75 3.18 5  
bw 23x10.5x7.25 3.63 6  
bw 22.5x11x6.5 4.09 6  
bw 21.5x10x6.5 2.37 7 drainage brick
bw 23x11x7 3.97 3  
bw 22x10.5x6 3.52 5  
bw 23x11x5.5 3.29 5  
bw 22.5x11x8 4.31 2  
bw 22x10x6.5 3.18 5  
bw 22x10x6 2.84 1  
bw 23x11x6.5 3.18 5  
bw 23x11x8.5 4.54 2  
bw 23x11x8 4.09 2  
bw 23x11x7.5 3.86 3  
bw 23x11x8 4.54 2  
bw 23x10x8 3.97 2  
bw 23x11x7 3.86 3  
bw 22x10x6.5 3.41 5  
bf 11x6 2.04 6  
bw 23x11x6.5 3.18 5  
bw 22x10x6 2.72 1  
bw 23x11x7 3.18 5  
bw 23x11x8 4.54 2  
bw 22.5x10.5x7 3.86 3  
bw 23x11x7.5 3.97 3  
bw 24x11x8 3.86 2  
bw 22x11x7.5 3.97 3  
bw 23x10.5x8 4.54 2  
bw 22x10.5x6 3.06 1  
bf 10.5x6 2.95 1  
bw 23x11x7.5 4.54 2  
bw 23x11x8 4.54 2  
bw 23x10x7 3.41 5  
bw 23x11x8 4.77 2  
bw 22x9.5x6.5 3.52 5  
bw 22.5x11x7 3.86 3  
bw 22.5x10.5x6.5 3.18 5  
bw 23x10.5x6.5 4.09 6  
bw 22.5x10x6.5 3.41 5  
bw 22x9.5x6 2.95 1  
bw 22.5x11x7.5 4.43 2  
bw 21x9.75x6 2.72 1  
bw 22.5x11x7.5 4.09 3  
bw 23x11x7.25 3.97 6  
bw 23x11x8 4.54 2  
bw 22.5x11x7.5 3.86 3  
bw 22.5x10x6 3.18 5  
bw 22x10.5x6 3.18 1  
bf 10.75x7.5 2.72 6  
bf 10.5x7 2.26 6  
bf 10.75x7 2.49 3  
bf 10x7 2.37 6  
bf 11.5x6 1.7 1  
bf 11x6.25 1.59 1  

Section 4

bf 10x6.5 1.36 5  
bf 10x6.5 1.82 5  
bf 9.5x7 1.59 6  
bf 10x6.5 1.36 5  
bf 10x6.5 0.91 5  
bf 11x7 0.91 3  
bf 10.5x6.25 1.82 5  
bf 11x6 1.36 6  
bf 10.5x7.5 1.59 3  
bf 10.5x7 1.82 3  
bf 11x6.5 0.79 6  
bf 10.5x6.5 1.36 5  
bf 10.5x6 1.25 1  
bf 10x6.5 1.25 5  
bf 11x6 1.25 1  
bf 9x6 0.68 1  
bf 10.5x6 0.91 1  
bf 11x6.25 1.48 6  
bf 11x6.5 1.36 6  
bf 10.5x6.5 1.48 5  
bf 11x6.25 0.59 6  
bw 23.5x11x8 4.09 2  
bw 22.5x11x6.75 3.18 5  
bw 21x10x7.25 3.29 3  
bw 22.5x11x8 4.31 2  
bw 23x10.5x8 4.54 2  
bw 23x10.75x8 3.86 2  
bw 22x10.5x6.5 2.95 5  
bw 23x11x8 4.54 2  
bw 22.5x10.5x6.5 3.18 5  
bw 23x11x7 3.52 6  
bw 23x11x8 4.54 2  
bw 22x10.5x6.5 2.72 5  
bf 10x6 2.26 1  
bw 22x10x6.5 2.6 5  
bw 22.5x11x7 4.09 3  
bw 23x11x8 4.54 2  
bw 22x11x6 2.72 5  
bw 22x10x7 3.13 6  
bf - 2.84 7 perforated brick
bw 23x11x8 4.54 2  
bw 22x10.5x6.5 2.95 5  
bw 21.5x10x6.75 3.41 5  
bw 23x11x7 3.86 3  
bw 22x11x6.75 3.13 5  
bw 22x11x6.75 3.13 5  
bf 11x6 2.26 6  
bf 10.5x9.5 3.13 7 slight frog
bf 11x6.25 2.04 6  
bf 11x7 2.26 6  
bf 11x6.5 2.04 6  
bf 10x6.5 1.59 5  
bf 11x8 2.95 2  
bf 10x9 2.49 2  
bf - 0.68 7 drainage brick
bf 11x6.5 2.26 6  
bf 9x7.5 1.36 3  
bf 10x7.5 1.59 3  
bf 11x6 1.93 6  
bf 10x6 1.25 1  
bf 11x6.5 1.82 6  
bf 10x6 1.36 1  
bf 11x6.5 1.48 6  
bf 11x6.5 1.7 6  
bf 11x6.5 1.36 6  
bf 10.25x6.75 1.36 5  
bf 10.5x6.75 1.36 5  
bf 10.25x6.25 1.36 5  
bf 11x6.75 1.02 6  

Section 5/6

bw 22x10.5x7.75 3.52 3  
bw 22x10.5x7 2.38 7 frog; stamped LBC PHORPRES
bw 23x11x7 3.4 6  
bf 10x7 2.15 7 stamped PHORPRES
bw 23x11x8 4.65 2  
bw 22.5x11x7 3.52 3  
bw 23x11x7.75 4.31 6  
bw 22x10.5x6 3.06 5  
bw 23x11x8 4.54 2  
bw 2.5x11x7.5 4.31 6  
bw 22x10x6 3.18 1  
bw 22.5x10.5x7 3.63 6  
bw 22.5x11x8 4.65 2  
bw 22x10.5x6.5 3.29 5  
bf 11x7.5 2.83 6  
bw 22x10.5x6.5 3.4 6  
bf 10.5x7.25 1.13 7 frog; stamped ..RES
bf 11x6 1.81 1  
bf 11x8 2.61 2  
bf 12x7 2.49 6  
bf 11x8.25 1.93 2  
bf 10.5x6.25 0.79 5  
bf 11x6 0.68 6  
bf 10.5x8 0.91 2  
bf 10x7.5 1.81 6  
bf 10x6.5 1.13 7 3 hole perforated brick
bf 10.5x7 0.91 7 frog; stamped LB.. PHO..
bf 10.5x7 1.36 6  
bf 10x7 1.36 7 3 hole perforated
bf 10x6.5 1.25 7 ditto
bf 10.25x6.5 1.7 5  
bf 10x6.5 1.13 5  
bf 11x8 1.47 2  
bf 10.5x7 1.24 7 frog; stamped ..C ..PRES
bf 10.75x7 1.24 7 frog; stamped L.. PH..
bf 10x6.5 1.36 7 3 hole perforated brick
bf 10x6.5 1.13 5  
bf 10.5x6.5 1.13 5  
bf 11x7.5 1.59 6  

Section 7: North Stream below Lower Pool

bw 23.5x12.5x5 3.18 7 stamp-Partridge & Guest Old Hill
   Staffs   
bw 26x13x5 3.97 7 black brick with shallow frog
bw 25x12x5 3.97 7 machine made black brick
bf 10.5x7.5 2.04 6  
bw 23x11.5x7.5 4.42 2  
bw 24x11.5x8 4.88 2  
bw 23.5x11.5x7.5 4.76 2  
bw 23.5x11.5x7.5 4.54 2  
bf 11.5x7.5 3.4 2  
bf 10.5x7.5 1.93 7 stamped NE[T?]...
bf 11.5x8 0.69 2  
bf 11x7.5 1.81 2  
bf 11.5x7.5 2.15 2  
bf 10.75x7 1.93 2  
bf 11.5x7.5 0.79 2  
bf 10.5x7.75 3.52 2  
bw 22.5x10x8 4.08 2  
bw 23x11.25x7.5 4.42 2  
bw 23x11x7 4.42 6  

Section 8: South Stream

bw 23x10.75x7 3.4 6  
bw 22x11x8.5 4.65 7 shallow frog
bw 23.5x11x7 3.97 6  
bw 22.5x10.5x8 4.19 2  
bw 23.25x11x7 3.63 6  
bw 22.25x10x6.5 3.4 5  
bf - 2.61 - semi-circular brick
bf 10x6.5 1.13 6  
bf 10.5x6 1.47 6  
bw 22.5x10x6.5 3.52 6  
bw 22.5x10x6.5 3.29 6  
bw 22x10x6.5 3.18 5  
bw 22x11x7.5 3.86 6  
bw 22x11x6.75 3.29 5  
bf 11x6 0.79 6  
bf 10x6 2.15 1  
bw 23x10.5x7 3.52 6  
bf 10x6 1.25 1  
bf 10.5x7 2.04 6  
bw 23.5x11x6 3.4 6  
bw 23x11x8 4.2 2  
bf 11x6 1.25 6  
bf 11x7.5 1.25 7 yellow brick
bf 10x8 1.59 2  

Appendix 3: glossary of archaeological terms

Archaeology: the study of man's past by means of the material relics he has left behind him. By material relics, this means both materials buried within the soil (artefacts and remains of structures), and those surviving above the surface such as buildings, structures (e.g. stone circles) and earthworks (e.g. hillforts, old field boundaries etc.). Even the study of old tree or shrub alignments, where they have been artificially planted in the past, can give vital information on past activity.

Artefacts: any object made by man that finds itself discarded (usually as a broken object) or lost in the soil. The most common finds are usually pottery sherds, or waste flint flakes from prehistoric stone tool making. Metal finds are generally rare except in specialist areas such as the site of an old forge. The absence of finds from the activity of metal detectorists is not usually given much credibility by professional archaeologists as a means of defining if archaeology is present

Baulk: an area of unexcavated soil on an archaeological site. It usually refers to the sides of the archaeological trench.

Context: a number given to a unit of archaeological recording. This can include a layer, a cut, a fill of a cut, a surface or a structure.

Cut: usually used to mean an excavation made in the past. The 'hole' or cut existed in time as a void, before later being backfilled with soil. Archaeologists give a context number to the empty hole, as well as the backfilled feature (called the 'fill').

Earthwork: bank of earth, hollow, or other earthen feature created by human activity.

English Bond: brick bonding pattern of alternate course of headers and stretchers.

Evaluation: a limited programme of intrusive fieldwork (mainly test-trenching) which determines the presence or absence of archaeological features, structures, deposits, artefacts or ecofacts within a specified land unit or area. If they are present, this will define their character, extent, and relative quality, and allow an assessment of their worth in local, regional and national terms.

Munsell colour: an objective method of defining soil colour using a specially designed colour chart for soils. The reading defines hue (an objective description of colour; eg YR means yellow-red), value (darkness or lightness of the colour) and chroma (the greyness or purity of the colour). For example 10YR 3/2 is a dark grey-brown.

Natural [layer]: in archaeological reports, this is a layer that has been formed by natural process, usually underlying man-made disturbance.

Period: time periods within British chronology are usually defined as Prehistoric (comprising the Palaeolithic, Mesolithic, Neolithic, Bronze Age, Iron Age), Roman, Saxon, Medieval and Post-medieval. Although exact definitions are often challenged, the general date ranges are as given below.

Pottery sherds: small pieces of broken baked clay vessels that find their way into ancient soils. These can be common in all periods from the Neolithic onwards. They often find their way into the soil by being dumped on the settlement rubbish tip, when broken, and subsequently taken out and scattered in fields with farmyard manure.

Site: usually defined as an area where human activity has taken place in the past. It does not require the remains of buildings to be present. A scatter of prehistoric flint-working debris can be defined as a 'site', with or without evidence for permanent or temporary habitation.

Sondage: an arbitrary hole dug during archaeological excavation. Often dug after the main excavation is complete to quickly test for information that may be required to clarify points of the main excavation.

Project Design: a written statement on the project's objectives, methods, timetable and resources set out in sufficient detail to be quantifiable, implemented and monitored.

Settlement: usually defined as a site where human habitation in the form of permanent or temporary buildings or shelters in wood, stone, brick or any other building material has existed in the past.

Stratigraphy: sequence of man-made soils overlying undisturbed soils; the lowest layers generally represent the oldest periods of man's past, with successive layers reaching forwards to the present. It is within these soils that archaeological information is obtained.

Queen Closers: method of ending a brick course with half a brick (to fill in at the corner).