HOME

CONTACT

MEMBERSHIP

NEWSLETTERS

DIARY DATES

LINKS

TRAINING DAYS

WHAT'S ON

LIBRARY

ACTIVITIES

Combe Martin

Fish Weirs

Hartland

History Days

Holworthy

Parracombe

Sherracombe Ford

Six Acre Farm

Archaeology Projects (external)

Community Landscapes Project

The Exmoor Iron Project

XArch Project

MEMBERS INTERESTS

EDITORIALS

The Swannery, Hartland, Devon. - Stephen Hobbs.
Field investigation, June 2007.
Swannery WCR07 National Grid Reference: SS 226 241 GB Grid
(Back to Hartland Project Page)

Scope of investigation:
To investigate if the recorded information that a Swannery existed at this site is correct or possible?

Site Location:
The Swannery, St Katherine’s Tor, Hartland, Devon. SS 226 241 GB Grid.
Access: This is private land under a Country Stewardship agreement, the Coastal Path passes through the enclosures.
Permission should be sought from Hartland Abbey Estates, Affeton Castle, East Worlington, Crediton, Devon and the tenant farmer at Stoke Barton, Stoke, Hartland, Devon.
For the purposes of this non invasive recording, permission was granted by the owner.

Figure 1 ; Location of site at Hartland, Devon (OS Web site 2007)

Figure 2 ; Location of St Catherine's Tor/Swannery site, Devon (OS Web site 2007)

Background:
St Catherine’s Tor and its associated valley is the reputed site of one of the ancient chapels of Hartland, a Swannery and Bradman’s well. The site is approx one and a half kilometres south-west from Hartland Abbey.

Archaeological background:
This area is in the immediate vicinity of Hartland Quay, the Hamlet of Stoke and Hartland Abbey. There are therefore numerous Devon HER listings, restricting these to those in the immediate vicinity of the Swannery the relevant numbers are: 7316-Standing Stone; 11616-Bradstone’s Well; 19997-Beacon at St Catherine’s Tor; 51978-Swannery; excluding shipwrecks.

Method:
Visual inspection and recording of the localised features, which will be followed by field walking of the immediate surrounding area. It is intended to take representative core samples from the valley base to establish if there is any trace of such as water born sediments present.

Introduction:
The HER has an entry to suggest that all or part of the river meadow which lies below St Catherine’s Tor was at one time a Swannery which may have been associated with the Abbot at Hartland Abbey (1164-1539). The evidence for this suggestion comes from a sequence of sources. Primarily an entry in the Dean Milles parochial survey of 1750, completed on behalf of Hartland parish by John Velly of Galsham Farm, and has an entry: "Catherine Tor - under which is a marsh in ye abbot's time a swan pool."  This entry is picked up by William Heard in his MSS Remembrances of Old Hartland  (North Devon Athenaeum /H.A.R 900) and then by Richard Pearse-Chope (Transaction of the Devonshire Association /66(1934)73). The statement is then often repeated in a variety of documents and has drifted into fact. Close inspection of the Dean Milles document would reveal that John Velly was often less than correct in the responses he gave to the standard questionnaire and allowed conjecture to be entered.

There are no known documentary archive materials extant for Hartland Abbey of the period up to the Dissolution, a small number of transitional documents exist within the present archive which are evidences of land transfer and ownership. Entries exist within Exeter Cathedral archive for a small number of communications with the Abbey; within both sources there is no evidence to suggest that the Abbot had or had use of a ‘Swan Pool’ An alternative is that any Swan Pool could have been part of the Manorial Estate belonging to the Dinham family (c1100 – 1501), but no mention is contained within their document archive (Arundel collection Cornwall Record Office AR/), explicit works by Fox & Padel [1] and Hannes Kleineke [2] on these records have no mention of the Dinhams using or producing Swan as a household item. Kleineke has detailed inventories of household consumption for capons, gannets a wide variety of salt and fresh water fish, meats and eggs but no mention of swan or indeed game bird such as duck. Therefore there is no factual reference to a Swan Pool and the association may have been established by a loose connection between Bradstone’s Well with its reputed healing powers, and a valley with a substantial earth works across its base in John Velly’s reply to Milles as being co-locative. ( See also the report on Clovelly Dykes by Carter 1927 who discusses this very point; his suggestion is that such as the pool and well may be found closer to Bursdon Moor/Summerwell than this site. Although it must be said that some of Carters ideas would not be sustainable today)

Geography/Geology:
The area under study runs from the ‘middle’ car park at Hartland Quay in the north as far as the northern end of the cliff top at Spekes Beach in the south. The area is of geological interest in that it is part of a sea dissected river valley system. The Spekes River, originally running north along the valley system, discharged into an ocean or larger river system at a point west of Smoothlands (Blagdon Valley). Erosion has divided the valley into isolated sections, most devoid of any river presence.

The study area contains three promontories each created as an eroded section of the original valley. The northern promontory, Screda Point, is dry and shows no evidence of the line of a stream. The second above Childspit Beach contains evidence of a dry river bed running parallel to the present cliff line, Figure 3. The depth of the erosion of this dry bed would indicate that it had been a stream for a considerable time and had discharged into the sea at its northern end at a now lost waterfall.

Figure 3; Dry River Bed at Childspit Beach, looking north (50cm marked rods)

Subsequent erosion has cut the stream to a position 150 metres east, where it now discharges over a substantial waterfall onto Childspit Beach. Inspection of the cliff face erosion shows only light evidence of the position of the original stream course. This promontory could also be the source where early man was able to extract the substantial stone slabs from the cliff face for use as monoliths once common in the immediate area [3] .

The third promontory in this area is formed by the sub-circular tor known as St. Catherine’s. This tor is regarded as the site of an early chapel attached to the Hartland church. Erosion has reduced the summit by some forty feet within the last century (Compare photographs of the periods, Cann Picture Archive & Hartland Digital Archive [4] ). Evidence of a building is described by Pearse-Chope in relation to a variety of artefacts recovered from the summit in the early 20^th Century and now displayed in the museum within Stoke church at Hartland.

The valley system is accessed by a track from Hartland Quay (now part of the South West Coastal Path) as far as the valley earthworks, this track was created in the late 19^th Century to provide a route for hunting parties the route is shown as a ‘footpath’ on the 1891 OS map (see photograph of hunting party in the Cann Picture Archive). A second track runs into the valley south of Big Shipless field this is shown as a track on the 1891 OS map.

Survey:
The study area contains a number of hedges and banks that although not unusual within the Abbey and Barton Estate are such that recording and classifying would be beneficial.

Method:
The method of classification and identification will be based on a system used by The North Devon Archaeological Society developed in association with the ‘Historic Environment Research Strategy for Exmoor’ (Objective 8.i), and recorded on a pro forma based on that used by the Cornwall Archaeological Unit in 1998 adapted to the local circumstances (Gillard, Unpublished Thesis, University of Exeter, 2002) - (download forms: Small or Large)

By measuring hedge width at base, width at top and height gives the dimensions of a cross section which is essentially a trapezium. The area of a trapezium is base (B) plus top (T) divided by 2 to give an average; times the average of the heights (H1, H2) (both sides of the hedge) - see Figure 4. The resulting figure gives not only the area of cross section, but, when multiplied by 1, gives the cubic volume per metre. Applying this formula to all the boundaries recorded, the resulting figures are rounded up or down to the nearest whole number, producing the basis of a simple numerical classification: Classes 1,2,3,4,5,6,7 based on volume. (v³ = ((H1+H2)/2) x ((B+T)/2) x 1 or length)

 
Figure 4; Formulae used to calculate the area of a cross-section of a hedge bank

Classification/description of hedges/banks will accord with the Rural Development Service Technical Advice Note 32, 2006 where relevant.

Hedge Recording: (simplified data output) See Figure 23 for locations

Hedge H1; (Figure 5) runs parallel to the small stream and is a substantial construction; it is formed as a dropped bank Corn Ditch in style with the stock coping facing north against the stream. The bank is stone faced on both sides with an earth/turf cap, the infill is packed earth. On the south face a build up of earth is in situ which could be part of the original construction or may be hill wash or collapse of the cap. The bank terminates on the west at the cliff top due to erosion; the eastern end has been reduced by the formation of the track-way and work on the modern reservoir to an extent whereby it is almost non-existent. It is possible that it formerly continued to join the hedge H8.

Figure 5; Hedge H1 viewed from the north

Hedge H2; (Figures 6-12)
This is the large bank that dissects the valley north/south at the base of St Catherine’s Tor and the subject of the enclosure bank forming the ‘Swan Pool’. In its present appearance it is as a stone faced wall on both sides with a stone coping on the north-west face in the form of a corn ditch. The fill is of compacted earth/clay layers (Figure 12 & 13) which is exposed at the south face where a cattle shelter was inserted into the depth of the bank (Figure 11). The top of the hedge is 1.1m high in the form of a domed earth cap now turfed (Figure 8). The western end of the bank abutted the stream but was reduced due to erosion to the extent whereby c.3m was cut back to form an access gateway. Originally a stile was in situ at this point and the stream/enclosure had evidence of a restricting feature such as a weir. The ‘new’ end has been stone face to form a circular final and thus corresponds with other such gateways on the greater Barton Estate.

Figure 6; H2 viewed from north west	Figure 7; H2 looking north east
Figure 8; H2 showing depth of the earth/turf cap	Figure 9; H2 showing the width of the bank looking south west
Figure 10; H2 north west corner of H2/3	Figure 11; H2 Inset cattle shelter on south east face
Figure 12; H2 showing the compacted earth infill layers (west)	Figure 13; H2 showing the compacted earth infill layers (east)

The north-east end of hedge H2 adjoins H3 but may have originally continued straight on into the incline of the natural hillside thus completing a form of coffer dam. On the west of H2 is a series of small earth embankments running parallel to the main hedge, in initial appearance it could be seen as a double bank and ditch system or could be the relict feature of an earlier hedge construction whereby the west face was banked to form a more traditional corn ditch appearance (Figure 14).

Figure 14; Diagram of possible bank ditch system of H2

Hedge H3;
This hedge forms the north boundary of the enclosure and runs parallel with the general hillside below Big Shipless field. It is double stone faced with compacted earth fill and earth/turf top. The unusual feature is that the hedge has a stone overhanging coping on both faces which would seem to be a contradictory feature as it would effectively exclude stock ingress in either direction. It is the only hedge with this feature so far evident on the Barton Estate (June 2007). The gradient of the adjoining hillside is such that an animal could with relative ease jump the gap onto the hedge (Figure 10).

Hedge H4;
A traditional Devon earth bank it forms a break between the St Catherine’s enclosures and the Kernstone valley at Wester Wood the valley floor is wet marshland (Figure 15).

Figure 15; H3 from the north west

Figure 16; Erosion of bank on the valley stream looking south

Hedge H5;
Forms the boundary between the St Catherine’s enclosures and Kernstone Cliff Field; it is a stone construction with light earth jointing and a turf cap and runs south until the Spekes Beach cliff face.

Hedge H6/H7;
These banks form the boundary between the base of St Catherine’s Tor and the valley. H6 forms a distinctive feature of the valley side and forms a bank and ditch system which could at one time been an access track which may have ran in a zigzag to the summit of the Tor (Figures 17-20). There is no evidence of any stone face or any coping stone. H6 effectively stops where the two enclosures meet at the stream, the bank being absorbed back into the natural slope of the Tor. H7 is the re-emerging H6 at a position west of the coffer dam (H2) and is stone faced on the north but unlike H6 has no significant track or ditch against the hillside.

Figure 17; H6 looking north	Figure 18; H6 looking north
Figure 19; H6 looking north	Figure 20 ; St Catherine's Tor from the east showing path to summit

The stream would appear to have been subject to embankment and two small earth banks are extant on the stream edge. These could have performed a retaining aspect for any water in the formation of the ‘Swan Pool’ (Figures 21-22)

Figure 21; Earth bank parallel to stream forming two enclosures within the valley

Figure 22; Earth bank against the west edge of 'Swan Pool' enclosure

Figure 23; Layout of Study Area showing hedge survey identifications

Earth Core samples:
A small number of sample cores will be taken from both enclosures to understand the composition of the sub-soil layers. It would be expected that if the area had been flooded for a period then a layer of water born sediment and sands would be evident.

There is no similar evidence in the river banks to support controlled flooding; this may be due to the enclosure of such water within the secondary embankment on the river sides (Figure 16).

Conclusion:
Although there is no confirmatory evidence to be gained from documentary sources the evidence on the ground could be suggestive of supporting some form of water control. The large bank across the valley is a formidable barrier and would be capable of retaining a large volume of water although it is doubtful if the height of water was anywhere close to the capacity at first indicated by the size of the barrier. It would seem probable that the secondary banks on the edge of the stream represent the extent of the water retained, thus a depth of no more thant 600mm at its western (deepest) end. To operate as a pool within the secondary banks it would need an access point for the water from the stream to become impounded at a point close to Wester Woods.

Figure 24; Schematic of the area (north top)

From the layout of the secondary banks (Figure 25) it would appear that water from the side stream running down from Big Shipless and Quarry Fields supplied a steady source of water, that a form of sluice may have operated on the main stream at a point below Kernstone Cliff Field therefore providing additional water if and when needed.The water was impounded to a depth of approx 600mm against the stream running to 0mm on the north thus forming the marsh as described by John Velly in his response to Milles.

To control the water level a weir would have been sited on the north-west corner of the pool adjacent to the coffer dam (H2). The evidence for this structure is within living memory, along with an inserted stone step stile in H2 at this point. There are two large dressed stones in the stream bed which evidently formed part of the weir (pers. comm. D. Cook & S. Littlejohn)

The secondary banking on the south of the stream may have acted in reverse and prevented water ingress into the second valley enclosure.

Figure 25; Layout of the Swan Pool area showing retaining banks