Dartington Hall stands on high ground above the River Dart, to the north of Totnes, in the County of Devon (NGR SX 798 628). The Hall has been described by Pevsner as the 'most spectacular medieval mansion of Devon' (1952, 99). It was the principal seat of John Holand, half-brother to Richard II, and has had a chequered history. Additions to the present structure were made under the ownership of the Champernowne family in the 16th century and again in Georgian times (Hoskins 1954, 381-2). Further remodelling of the Elizabethan living rooms were carried out in the 19th century under Henry Champernowne (Hussey 1938, 232). The estate declined as a result of the agricultural depression of the later 19th century and the Hall was much neglected when purchased by the Elmhirsts in 1925. It was under Dorothy Elmhirst's direction that the gardens were laid out in their present form.
During the Second World War (1939-45) German bombing of Exeter was thought to have caused the destruction of much of the estate archives, then deposited in the County Record Office. This has caused a number of problems for the future management of the gardens, as it is not possible to determine to what extent the Elmhirsts had utilised existing garden features, nor what form the earlier gardens took, from archival sources. To enable a better understanding of the garden, and as an aid to future management decisions for the continuing evolution of the gardens, an archaeological evaluation, followed by excavations, was decided on.
The earlier work was reported to the Trustees in 1991 (Currie 1991). The funding for subsequent excavation was recognised as a possible problem, and it was felt that the best way to find the money was to run a training excavation in conjuncture with the Dartington Hall Centre. Although the first year, reported here, was a trial for this idea, its success will hopefully result in the continuation of the event on an annual basis, with a review of the future plans to be made at the end of each year. As a result of this possible continuance, this report should be seen as an interim statement only . It is hoped to combine the annual reports into a more definitive document at the end of the excavation programme. An outline of the proposals has been presented to the Trustees and the County Archaeological Officer in the form of a project design (Currie 1993), written to specifications for such work laid down in The management of archaeological projects (English Heritage, revised edition, 1991).
The evaluation undertaken in 1991 deduced a number of points about the gardens.
2.1 Just inside the south-west boundary of the present garden a number of mature oak trees were observed that stood on the line of a former field boundary, possibly of medieval date. This conjectured boundary extended to the line of old chestnuts that stand at the top of the uppermost terrace on the south-west side of the Tiltyard.
2.2 The Chestnuts on the top (south-west) terrace of the Tiltyard were observed to be very old trees. The age of the oldest of these was estimated to be between 300 and 400 years old.
2.3 The water reservoir to the west of the line of chestnuts had a stone culvert leading from it that had been truncated by the terraces. It was concluded that the reservoir was therefore older than the terraces.
2.4 Some of the walks to the west of the Tiltyard (Rhododendron, Camellia, Spring Walk) can be shown to follow the approximate line of garden walks on Ordnance Survey maps of 1885 and 1904 indicating that the present design re-used pre-existing pathways. The original pathways were thought to be of 19th century date.
2.5 The Bowling Green. The north and south boundaries of this garden indicates that it was once on a lower level. On the south, the revetment wall can been seen to exhibit two phases. The lower part of the wall, up to a height of over one metre, is of a grey-blue stone described by the gardens manager as local stone (Graham Gammin pers. comm.). The upper courses, up to 0.5m. thick is of a pinkish stone, thought to be of non-local origin. Furthermore, the ruined wall that was once part of the outer courtyard of the Hall shows signs of being buried to build up the ground on both its north and south sides.
2.6 A number of factors led to the conclusion that the terraces around the Tiltyard was more likely to be post-medieval garden features. A map of 1885 showed clear evidence that a formal garden existed at that date at the west end of this area below the terraces.
2.7 Excavations in the Tiltyard indicated that the levels had been altered on at least one occasion.
Subsequent historical research and excavation has confirmed most of these observations.
A quick survey of previous literature on the site prior to the excavations brought to light a number of important points. These are associated with the origins of the idea that the terraces to the south of the Hall were part of a medieval Tiltyard.
3.1 The origins of the 'Tiltyard' tradition
This idea may have originated in the minds of the Elmhirsts shortly after they purchased the Hall. An early photograph has a note attached by him which says 'Terraces probably untouched since 14th century. Mr Lynch [the head gardener] "did them up".' (Snell 1989, 26). Initially the area below the terraces were turned into an open-air theatre, with a raised stage at the east end. The conversion to the Tiltyard's present form was carried out in 1954-55. By this time Dorothy Elmhirst was certain of it having been a Tiltyard when she wrote that it would be: '.....turned back to its historic fourteenth century form as a tiltyard, flattening out the whole space to the level of the present stage with a flight of steps at the top.' (Snell 1989, 62).
The tradition that the area was a Tiltyard may not pre-date the Elmhirsts' arrival at Dartington in 1925 (although it should be noted that the idea of the bear-baiting pit to the east probably derives from hearsay passed on by the Champernownes). When Christopher Hussey wrote his Country Life article in 1938, he said of the terraces:
'It has been suggested by Mr. Elmhirst, in explanation of this remarkable arrangement - and I feel in agreement with him - that the amphitheatre was formed out of a natural dell to provide a tiltyard, and that the arcade [the ruined arches on the south side of the private lawn] is the remains of a kind of "grand stand" incorporated in the outer side of the garden court.' (Hussey 1938, 208)
The Elmhirsts had learnt by this time that a former owner of the Hall, John Holand, Duke of Exeter and half-brother to Richard II (1377-99), was reputed to be a keen jouster. From this association, the terraces came to be seen as of 14th century date, as were the ruined arches on the south side of the private lawn. This idea was taken up by Anthony Emery, an architectural historian, who went on to quote dimensions for the ideal tournament ground, 60 by 40 yards, as defined by a medieval duke of Gloucester, as corresponding with the area of the lawn at the bottom of the terraces (Emery 1958, 201-2). The fact that this area is not 60 yards by forty, but an irregular shape that only attains a width of forty yards at one point, was excused on the argument that the dimensions were not exact because of the natural form of the dell in which it was built (Emery 1958, 202).
To help prove his theories, Emery got the archaeologist, Colin Platt, to do an excavation on the private lawn to confirm the conjectured date for the building of which the ruined arches had once formed a part. The idea behind this was that Emery felt that as this building was clearly at a strange angle to the 14th century Hall, but was parallel with the terraces, the symmetry was a result of the arches and the Tiltyard being contemporary. That is he expected Platt to confirm the 14th century date to establish John Holand as the builder of the terraces (Platt 1962, 208).
Despite a heavy bias on the expected final outcome of the excavations, Platt found the arches to date from c. 1500. He concluded therefore that there was 'no positive evidence whatever for the dating of the so-called tournament ground...' and that 'any argument for an early dating of the south courtyard that depended on the alignment of the buildings with the tournament ground and on John Holand's acknowledged prowess as a jouster would be unsatisfactory' (Platt 1962, 219).
In spite of this conclusive dismissal of the old arguments, the tradition continues to this day. As late as 1989, Reginald Snell appears to disregard Platt's evidence when he says of the Tiltyard:
'.... there is very good historical evidence that it was used as such [as a tiltyard] in the early part of the fifteenth century.' (Snell 1989, 16)
The persistence of the myth in Snell's opinion is even more remarkable in that, in his book on the gardens, From the bare stem (1989) he makes it quite clear that he is aware that the 'Tiltyard' floor was both a formal garden, known as the Dutch Garden, and that the levels were much higher before the Elmhirsts' purchase of the Hall. In this book, he presents evidence that shows that the 'Tiltyard' does not appear to be known by that name until given it by the Elmhirsts on no evidence other than their personal opinion, or local hearsay.
The Elmhirsts were clearly aware that the level below the terraces was once much higher than the present floor. They were also aware that the area was known as the Dutch Garden, and had been in that form since at least the 1850s.
The earliest dated evidence is the 25" Ordnance Survey map of 1885. This shows a formal parterre at the level of the Irish yews known as the Twelve Apostles. An undated photograph in the Hall confirms this. On the reverse, written presumably by one of the Elmhirsts, are notes which date the picture to c. 1864. This is remarkably early for an outdoor photograph, but not impossible. The yews are shown much smaller, and this has led to the writer to suggest that they were probably planted 25 years before the picture (c. 1839). The notes state that the revetment wall to the Bowling Green was raised in the first years of the Elmhirsts' residence to prevent children falling over it. Another photograph, hanging in the bar of the White Hart, shows the Dutch Garden at an apparently later date (see Plate 2). This is reproduced in Snell's book, but without a caption. The much greater size of the yews indicate the later date.
An Ordnance Survey map of 1904 shows the valley floor still at the same level as the yews, but no longer shows the parterre. Although this leads one to suppose it had been abandoned by this date, this does not seem to have been the case. Snell describes the area when the Elmhirsts first arrived:
'At one point an old cider press was discovered in the undergrowth, and moved to the centre of a square lawn in the Dutch garden that the earlier Champernownes had made at the upper end of the Tiltyard. At its lower end there had evidently once been a formal rose garden, only slightly below the level of the Twelve Apostles, but this had been filled in, and was now .... little more than a large and unsightly hole.' (Snell 1989, 22)
Snell further records that the terraces were much degenerated when the Elmhirsts first arrived, and are described from a contemporary photograph as being merely 'gently undulating slopes'. They appear to have been drastically restored by Stewart Lynch after 1928 (Snell 1989, 26), and so it is difficult to be sure how close they come to their original profile on the south side. On the north side of the Tiltyard their shape seems to correspond well with those of the present.
It has long been considered that the Dartington archives were destroyed during enemy bombing of Exeter in the Second World War. However, a reading of the historical sources referred to by Emery (1958, 1970) and Platt (1962) seems to suggest that records have survived in the Exeter City Library. These include some of the Champernowne's records, miscellaneous 18th-century sketches and other pictures, a volume of proposals for alterations to the Hall in 1805 by George Sanders, plus a family history of the Champernownes written in 1954 by Miss C. E. Champernowne (Emery 1958, 184, 186, 191, 194).
There are also some valuable clues surviving in the archives still at Dartington. As well as transcripts made of some of the Champernowne's deeds, there is a fine collection of drawings, paintings, prints and early photographs available. These include Buck's print of the main courtyard of 1734, and a curious faded watercolour by Archdeacon Froude of 1801 showing the south side of the Hall. This seems to show a substantial wall, with a possible buttress, or projecting building, in the foreground that can not be related to any surviving building. The picture is badly faded, and is not clear on detail or perspective, making interpretation difficult.
A copy of the 1839 Tithe Map survives in the Dartington archive (see Plate 1). This shows a square building on the lawn in front of what are now the private apartments. During Platt's excavation, a building of similar dimensions was found near to this spot with a piece of possible 17th-century pottery under the wall. Although Platt thought that it was possible that this building was post-17th century, he preferred to think of the evidence as being residual contamination (Platt 1962, 216n).
Platt found little evidence for 14th and 15th century occupation of the area, although a possible earlier 13th century building was uncovered. He considered that the buildings that were once sited on the private lawn dated to c. 1500 (Platt 1962, 219). However, this seems to be the earliest possible date on the evidence he gives. One is left with the impression that so deeply seated is the idea that the buildings would be of a 14th century date that he was reluctant to backdate them too much, in spite of his assertion that the type of long gallery that the ruined arches seem to be part of is more likely to be of a date no earlier than the 16th century (Platt 1962, 216). There is no reason, on the evidence produced by him, that the buildings could not date later in the 16th century. Platt concluded that the buildings on the lawn, with the possible exception of the 'square' building shown on the Tithe Map, were pulled down in the later 17th century. The area may then have been levelled for a formal garden. Later, probably in the early 19th century, more levelling occurred to make a lawned area (Platt 1962, 219).
It is hoped that the work at Dartington by the Gardens Archaeology Project will be published as a single unit. Therefore, for the sake of continuity, it has been decided to run the trench numbers in a continuous sequence, beginning with the two small trenches opened in 1991. The 1993 trenches run on from these; hence the first trench of the 1993 season was called trench 3.
Four trenches in all were excavated in 1993. These were numbered trenches 3-6 inclusive.
4.1.1 Trench 3 (Figures 3-4; Plates 3-5).
This trench, 5m by 1m with extensions, was excavated in the private lawn to try to recover information about the building shown on the 1839 Tithe Map.
Immediately below the turf was a dark grey-brown clay loam (context 10), up to 0.2m thick. Like most of the soils in the gardens this contained abundant fragments of grey Devon slate, which is the parent rock for the area. There were occasional fragments of mortar within the soil. After about 0.2m, the soil became increasingly contaminated with mortar, slate, stone and other building debris (context 19). At the south end of the trench this was a brown loamy clay (context 23). Cut into this were traces of a linear plant bed (feature 25), containing a more loamy soil similar to context 10. This cut was up to 0.12m wide and parallel to the south wall of the private lawn.
Context 19 was less than 0.2m thick. When removed, it revealed stone foundations (structures 18 and 27). These were surrounded by rubble layers containing much slate and broken stone (contexts 28, 29, 30, and 31). Context 30, adjacent to the join between structure 18 and context 30, and on the west side of structure 18, contained large amounts of ceramics, including delft, Barnstable wares, and Westerwald stoneware.
The structures were made of mortared stone. Structure 18 was 0.45m thick and several courses higher than structure 27. The coursing of structure 18 was of a more uneven nature, using slabs of stone less finely dressed than in 27. Structure 27 seemed to cut across 18, truncating the latter. This was shown by the uneven finishing of structure 18 where it was butted by 27. Structure 18 took a right-angled turn to the east 2.4m south of 27.
Structure 27 was demolished to a lower level than 18. On average the difference in these levels was 0.15m. There was a feature within 27 on its north side. This was an alcove set back 0.38m from the northern edge of 27. The sides of the alcove were very even, giving the impression that it had been made with the aid of shuttering. The full length of the alcove was not found as it continued under the baulk of the trench.
The trench was not excavated further beyond the cleaning up of the foundations of 18 and 27. Excavation ceased at a depth of approximately 0.6m.
4.1.2 Discussion of trench 3. Early photographs of the private lawn show that the last few metres before the south wall may not have been completely levelled. The lawn here is shown dropping down a bank, probably no more than 1m below the average level of the area to the north. This has subsequently been levelled up, and this work is probably indicated by the build-up level, context 23. The plant bed cut into this level is parallel with the right-angled turn in structure 18. However, this is thought to be a coincidence, the bed being parallel with the south wall of the garden, not structure 18, which had almost certainly been demolished some time before the levelling layer, context 23, was laid down.
There is a reasonable sequence of photographs of the private lawn from the 19th century through to the present. None show planting in the area excavated, and it must be assumed that the plant bed was an ephemeral change made at some time after the bottom of the lawn was levelled. This bed appears to have been removed without being recorded in any known photographs. Nevertheless, the root remains were well-established enough to suggest it must have existed for at least five to eight years.
The structures themselves were not dated as the excavation did not dig down far enough to determine this. However, the quantities of ceramics recovered from the trench give a good idea of the date of demolition. Platt (1962, 219) has suggested a date in the late 17th century for the demolition of the buildings beneath the private lawn, with a further build-up to lay out the lawn c. 1805. He also suggests that the building he refers to as the 'tower', on the approximate site of the building shown on the Tithe Map, may have survived to a later date. Bearing in mind that Platt excavated a much wider area, the findings of the 1993 largely coincide with his findings, with some additional information.
Firstly, the structure 18 seems to bear no relation to the conjectured plan Platt suggested for the main buildings, which he dates to c. 1500. If 27 can be seen as part of the structure he refers to as the 'tower' (Platt 1962, 209-10), this seems to post-date 18. Furthermore, the plan of the 'tower' as shown by Platt (1962, 211) seems to show this structure may have contained at least two phases. Taken with the suggestion that a structure here may have survived until 1839, it is possible that 27 is a later building altogether than the main c. 1500 phase. However, this should not leave out the possibility that a building of c. 1500 existed on the site, but had been radically altered later.
It is therefore possible (but not certain) that structure 18 represents part of a pre-sixteenth century building. The lack of 14th and 15th century finds suggests that it is unlikely to be of this date. This leaves the fragmentary remains of 13th century structures recovered by Platt (1962, 212), with which this wall may be associated.
There were no sign of foundation cuts during the 1993 excavation, and the demolition levels were very mixed. Platt (1962, 214-5) has suggested that the lawn may have been disturbed by Saunders c. 1805 as his plans seem to show the outline of buildings beneath the lawn, suggesting that some crude excavation may have taken place at this time. Although Platt records 19th century ceramics at some depth in parts of the lawn, indicating disturbance, this was not apparent during the 1993 excavations.
A point of slight discrepancy between this author and Platt can be discerned in the latter's dating of the pottery. Both excavations recovered a decided lack of 14th and 15th century wares, indicating little activity in this period. However, Platt's dating of his later pottery assemblage, which is very similar to the 1993 assemblage, to the later 17th century can be questioned. All the main elements of these assemblages continued to be made in the eighteenth century. For the most part this was the earlier part of that century, but the tin-glazed wares (delft) contained to be made into the 1760s, as did some of the coarse earthenwares. Admittedly neither assemblage contains any of the distinctive wares common in the later eighteenth and early nineteenth centuries such as creamwares and pearlwares. Further, salt-glazed wares, which start to appear from c. 1720, are absent in the 1993 assemblage. The evidence suggests therefore that although a late 17th century date is possible, there is an equally good chance that the demolition occurred in the early years of the 18th century before 1720. It is also not impossible that it could have occurred later; that is any time before the mid-eighteenth century.
Although the 1993 trench was too small a sample to establish anything but tentative suggestions, it is possible that the buildings under the lawn were demolished in the 18th century before c. 1760. This may have been coincidental with the Georgian additions to the present structure, but a precise date for these is not known, although Emery (1970, 183) suggests a date between 1737 and 1741.
4.2.1 Trench 4 (Figure 3; Plate 6).
This trench, 6m by 1m, was excavated on the terrace level containing the Twelve Apostles to try to locate traces of the Dutch Garden known to exist there in the 19th century.
This trench was dug to a level 0.6m below the present turf-line. No trace of the earlier parterre was found. There were traces that the terrace had been formed by the dumping of soils. The pottery recovered indicated some activity in the area early in nineteenth century, but beyond this little could be said.
4.2.2 Discussion of trench 4. The lack of any evidence for a parterre is unexpected, particularly as the photographic evidence for its existence is so good. There is no doubt that the terraces were considerably recut by Stewart Lynch after c. 1928 (Snell 1989, 26), and this could account for the removal of the evidence. The evidence further suggests some build-up for the creation of the Dutch Garden, but this should not be misinterpreted as indicating a former Tiltyard existed. Some levelling of the valley floor would have been desirable in laying out the Dutch Garden, and in keeping its level above the water table to prevent water-logging. This work may post-date c. 1760, but more work is necessary to confirm this.
4.3.1 Trench 5 (Figure 4; Plate 7).
This trench, 1.8m by 0.85m, was dug to examine a stone culvert known to exist on the first terrace level.
No trace of the Dutch Garden was found in this area, nor was there any conclusive evidence for a cut for the stone culvert (structure 17, 24). This latter structure was discovered about 0.1m below the present turf. It was apparently undisturbed at this point.
The vertical wall of the structure (structure 24) consisted of rough unmortared blocks local slate, up to 0.15m thick, and at least 0.6m high. These were constructed so that the internal wall of the culvert was even, but no attempt was made to obtain any semblance of evenness in the outside wall, which had an untidy appearance. This was overlain by large slate slabs (structure 17) up to 0.6m wide and 0.1m thick. The only find from the surrounding fill was a piece of medieval coarseware, probably of 12/13th century date.
The culvert contained a modern ceramic pipe within it, but there were no traces of it being interfered with since the culvert's construction. Elsewhere along the line of the structure, where modern disturbance has taken place, the culvert has been repaired with brick. The presence of the ceramic pipe can easily be explained as having been fed along the culvert from disturbance elsewhere.
4.3.2 Discussion of trench 5. It is curious that a major drainage culvert, as this structure clearly was, should have been built on a higher level if the original Tiltyard floor had once been at the present level. Such a layout makes no practical sense, particularly as the original structure is a drystone construction. This would not have been completely watertight, and would have led to seepage downhill on to the lower level. The insertion of the ceramic pipe clearly shows that this became a problem when the Tiltyard floor was lowered after 1928.
However, the terrace on which the culvert sits seems to be made ground. The bottom of the structure seems to sit on rock strata cut out of the hillside. Earth seems to have then been built up around it, rather than it being a structure that was built into a cutting. This relates to the evidence in trench 4, which suggests that the valley floor had been built-up to create the Dutch Garden. The evidence indicates that the culvert is an original feature dating from the first serious piece of landscape alteration in the valley. This suggests that the first use of the valley was to build the culvert, and then to level the valley floor to a height only 0.1m above the top of this structure. Therefore, it seems that the level of the Dutch Garden was the first man-made level, not the present surface as thought by the Elmhirsts.
This is supported by the fact that the Rose Garden in 1925, that is the garden slightly below and to the east of the Dutch Garden, was 'only slightly below the level of the Twelve Apostles' (Snell 1989, 22). This would have had to be the case or the drainage for this half of the 'Tiltyard' would have been ineffective. Many of the subsequent problems of drainage that have occurred on the 'Tiltyard' floor are to result of lowering the ground level both below the original water table, and below the original man-made drainage system.
Unfortunately, no good dating evidence for the original construction of the culvert was found. The best that can be said at present was that it occurred after the 12/13th century, most probably in the post-medieval period.
4.4.1 Trench 6 (Figure 4; Plate 8).
This trench, 2.1m by 1m, was dug in one of the upper terraces to examine the suspected conduit leaving the well house to the south of the chestnuts.
When the topsoil was removed, it was noted that a pile of irregular stone had been used to block the entrance of a conduit. This was more substantial than the culvert on the lower terrace. The creation of the upper terrace on which the chestnuts stand had clearly cut across the original structure. It can be assumed that once the conduit had joined up with below ground culverts which can still be seen through later inspection hatches in the vicinity of the sculpture known as the 'Swans'.
The conduit was over 1m high and 0.6m wide. Where it was first exposed its side walls were made of unmortared stone (structure 14). These were surmounted by massive slabs of slate (structure 15) up to 0.1m thick and 0.9m wide. Where the culvert had been truncated a simple wooden support had been inserted (structure 16). After 2-3m inside the conduit, the sides were cut into solid rock, with occasional drystone reinforcement to prevent roof collapses. The massive roof slabs extended the full length of the tunnel, indicating the course of the conduit was buried beneath earth cover.
It was possible to crawl into the structure. The passageway was over 20m in length and was terminated by a blocking wall. This was covered in a coating of water-lain limestone commonly found in caves and similar below-ground cavities. Consequently it was not possible to determine the material used for this wall. A hole was noted in the base through which a later lead pipe had been inserted. Where this pipe left the conduit to be continued as a plastic pipe was a stop-cock. Also within the blocking wall was a further blocked hole about 0.5m above the floor of the conduit. When the conduit was first exposed a message was found enclosed in a bottle inside the tunnel. This had been left by the workmen who presumably had put in the plastic pipe. The message said:
'February 1973. Today is clear and fine. We are building terraces up straight (?) and laying hose (?) pipe to the swans. Peace to the United Nations of the World. Rene Redway. (?) Hawkins. 1973.'
4.4.2 Discussion of trench 6.
This conduit appears to have originally led from the stone structure above the terraces known as the well house. This conduit would have channelled a water supply to the valley head, and possibly to the Hall itself. It is characteristic of medieval well houses existing at many sites nationwide. The earliest seen by this author is the well house at Offwell supplying the Augustinian priory at Southwick, Hampshire, with drinking water. This probably dates to the second half of the 12th century (Currie 1990, 58). Another example has been recorded supplying the medieval town of Southampton, also in Hampshire, and dated to the 14th century (Davies 1988).
The conduit has been clearly truncated by the construction of the terrace on which the chestnuts stand. These trees are said to be up to six hundred years old (Hussey 1938, 206). This is, however, unlikely. Nevertheless they are trees of considerable antiquity, and the oldest (they are not all the same age; some of the eastern trees are later replacements) could be between 300-400 years old, although they are more likely to be at the younger end of this range. This suggests that the terraces existed in the period 1593-1693. They could possibly be older, as the chestnuts could have been planted on a pre-existing feature. It is unlikely that the terraces pre-date the stone arches at the south edge of the private lawn, as they are all constructed parallel with this feature. Nearly all authorities seem to agree on this point (Emery 1958, 202; Platt 1962, 219). As the arches can now be dated to after c. 1500, it seems that the terraces date to either the sixteenth or seventeenth century. Further excavation is needed to get a more precise dating.
Anthony Emery's extensive writings on the architectural history of Dartington (1958, 1970, 1975) make interesting reading. His detailed account of the Hall and its history (Emery 1970) is a model of its kind. However, when one reads all three accounts, and is aware of how his views developed, an air of uncertainty develops.
Firstly, he changes from the most unequivocal champion of the garden terraces forming a 'Tiltyard' in 1958 (Emery 1958, 201-2) to complete dismissal of the theory by 1975 (Emery 1975, 151-3). It is difficult to conceive of such a complete turn-about. In his latter essay he says:
'..further research and the excavation of the south court has convinced me that the terraces are a post-medieval feature and that they bear no relation to the construction of the gallery or the south court.' (Emery 1975, 151)
He then says that few tournament grounds have survived in England, and that those that do are simple affairs on flat ground 'like cricket pitches' (Emery 1975, 152). In 1958 one of his arguments for a tournament ground was the similarity between dimensions mentioned by a medieval duke of Gloucester and the 'Tiltyard' at Dartington (Emery 1958, 201-2). In 1975 he argues the complete reverse, saying that the dimensions are not similar in any way (Emery 1975, 152). One can only assume that he had taken someone else's word on this in 1958, but had measured them for himself by 1975, thus showing clearly how false his earlier statement was.
One must admire Emery his own personal chastisement. The Tiltyard theory was destroyed in 1962 by Platt's excavation results. What makes Emery's turn-about so strange is his almost complete rejection of Platt's dating evidence (Emery 1975, 148). To the last, Emery continues to argue the south court buildings are John Holand's creation (Emery 1975, 148), whilst now maintaining the 'Tiltyard' was not a tournament ground. Yet the principal argument for the Tiltyard being so was the fact it aligned with the south court buildings, that had once been thought to date to the 14th century. It was the proof that the building dated to after c. 1500 that demolished the 'Tiltyard' theory.
The author can not support Emery's dismissal of Platt's findings. His arguments against Platt (outlined in Emery 1970, 187-97) demonstrate a lack of understanding of archaeological principles and techniques. One can only assume that after his previous support, on such flimsy evidence, of the tournament ground theory that he felt he had to give way here, but that he was unwilling to concede any further ground, causing him to stick rigidly to his earlier statements on the date of the buildings. As an architectural historian he felt he could concede on garden features, but not on the subject nearest his heart, architecture. In all other respects Emery's work on Dartington is a splendid example of its kind, but one feels that his views on the south court must be open to question.
The excavation done in 1993 not only supports Platt's findings, but could possibly push the dates forward in time. There is no reason, on the 1993 evidence, why the buildings in the south court could not date to after c. 1500, although it is difficult to push this very far into the second half of the 16th century. Likewise, Platt's dating for the demolition in the late 17th century, could be pushed back as late as 1760, although a date between c. 1670 and 1720 is the most likely.
In turning back on the tournament ground theory, Emery offers some very interesting observations. He notes the similarity between the terraces at Dartington, and the early 18th century 'amphitheatre' at Downton, Wiltshire (1975, 151-2), a site known to this author, who is in full agreement that distinct similarities exist between the two sites. In fact steep terraced gardens on hillsides are known well into the 18th century. The walled terraced garden at Castle Bromwich, West Midlands, was still being worked on in the late 1740s, and it was only the death of the owner in 1747 that seems to have curtailed the work (Currie and Locock 1993, forthcoming). It would seem that imposing terraces, with little planting on them (certainly the case at Downton, if the matter is disputed at Castle Bromwich), were a characteristic of late formal gardens, some of these still being built at a time when such features would have been considered unfashionable in London socialite circles.
Emery (1975, 152) introduces another very interesting point by drawing attention to the fact that the Tiltyard is not referred to by any 18th century writers, and is not shown on the Tithe Map of 1839. From this he suggests that it is possible that the terraces may have been formed as part of 19th century Romanticism at some time between 1839 and c. 1870 when the first photographs show the terraces in existence.
Such an argument is certainly plausible. Such gardens were becoming popular in the 1840s and 1850s. This is almost certainly the approximate dating of the Twelve Apostles, and it is tentatively supported by the pottery found in trench 4 (which indicates some levelling in the first half of the 19th century).
Against this are the chestnuts. If they are contemporary with the terraces, then they would seem to date to the late 17th/early 18th century. It must be stressed that this evidence must be used with caution as the chestnuts argument is a 'chicken-and-egg' situation. Like the south court arches, it is nowhere certain that the terraces are contemporary with either feature, and they (the terraces) could have been constructed as an afterthought to either (or both). The only agreed point between all commentators is that the terraces are unlikely to pre-date the south court. A further argument against the late date is the introduction of the Tithe Map as evidence. As these maps are concerned primarily with tithe commutation, it is not uncommon for them to leave off the internal detail of gardens. This has been observed by the author on other sites, and the terraces not being shown on such a map is not good evidence for them not existing.
A combination of the results of the 1993, and a much need synthesis of the documentary evidence for the gardens, seems destroy the last semblances of the Dartington terraces being associated with a medieval tiltyard. All the recognised authorities now seem to agree that they were made as garden features in the post-medieval period. Nevertheless, the evidence as to exactly when they were built is not conclusive, and further work is needed to resolve this problem.
At the moment three possibilities seem to be available.
Further work would be necessary to determine which of these dates, if any, is correct.
Of the non-garden features found on the private lawn, it can be concluded that there is unquestionably two phases of building in the area known as the south court. Generally the 1993 excavations confirmed Platt's dating, but showed the situation to be more complex than he envisaged. The exact dating for the construction of either phases can only be conjectured at present, but it is possible that a building on the site of Platt's 'tower' was still in existence in 1839. Once again further excavation is necessary to clarify the dating.
At least one thing can be said with certainty, and that is that the history of Dartington Hall and its environs has been simplified in the past by trying to relate activity on the site to famous historic personages. The reality is often that far more mundane characters may have been equally instrumental in the historical evolution of any site, and, at Dartington, that this may be far more complex than has been hitherto suggested.
As there is a possibility that excavations might continue in future years, it has been decided to leave a full assessment of the finds until the end of the project.
At Dartington Hall:
Ordnance survey maps 1:500 scale 1885 edition and 1904 edition.
Tithe Map of Dartington, 1839 in archive room.
Watercolour of south front by Archdeacon Froude, 1801, in toilet on first floor of private wing.
Print of Hall and courtyard by Nathanial Buck, 1734, in toilet on first floor of private wing.
Photograph of Dutch Garden in Tiltyard, undated but with pencil notes on back suggesting c. 1864, in toilet on first floor of private wing.
Photograph of Dutch Garden in Tiltyard, undated but probably c. 1900, in bar of White Hart; reproduced on page 19 of Snell 1989 (see below).
C K Currie, 'Southwick Priory fishponds: excavations 1987', Proceedings of the Hampshire Field Club and Archaeological Society, 46 (1990), 53-72.
C K Currie, An Archaeological Evaluation of the gardens at Dartington Hall, Devon, unpublished report to Trustees of Dartington Hall Gardens, 1991.
C K Currie, The gardens of Dartington Hall, Devon: archaeological project design, unpublished report to Trustees of Dartington Hall Gardens, 1993.
C K Currie and M Locock, 'Research excavations at Castle Bromwich Hall, West Midlands, 1989-91', Post-Medieval Archaeology, 27 (1993, forthcoming).
P N Davies, 'Work at the Conduit Head, Southampton', Hampshire Field Club and Archaeological Society newsletter, new series no 10 (1988), 9-12.
A Emery, 'Dartington Hall, Devonshire', Archaeological Journal, 115 (1958), 184-202.
A Emery, Dartington Hall, Oxford, 1970.
A Emery, 'Dartington Hall, Devonshire', M J Swanton (ed.), Studies in Medieval Domestic Architecture, London, 1975, 134-152.
C Hussey, 'Dartington Hall, Devon part 1',Country Life, 27th August 1938; 'Dartington Hall, Devon part 2', Country Life, 3rd September 1938.
N Pevsner, The buildings of England: South Devon, Harmondsworth, 1952.
C Platt, 'Excavations at Dartington Hall, 1962', Archaeological Journal, 119, (1962), 208-224.
R Snell, From the bare stem. Making Dorothy Elmhirst's garden at Dartington Hall, Exeter, 1989.