University of Exeter




Hill Slope Enclosures 

in the 

Blackdown Hills: A Case Study

Report written by: Bronwen Chapman

Hill slope enclosures on the Blackdown Hills, Devon, have so far had little archaeological investigation, and thus in many cases have unclear dates attached to them. Suggestions have been made that these features date to the Iron Age, and into the Romano British Period (Reed and Manning, 2000), with others suggested to date to the Medieval period (Silvester 1980).

Two hill slope enclosures have previously been identified by the Community Landscapes Project at Bywood Farm on the Blackdown Hills, though they too lack clear dating evidence. The contents of this report outline the characteristics of the sites at Bywood Farm, as well as thirteen other hill slope enclosures around the Blackdown Hills. Historical documentation such as the 19th century Tithe maps and the Domesday book, added information to some sites, but only when seen to be closely linked. By further analysing the context and composition of a group of suspected hill slope enclosures within a selected area of the Blackdown Hills it was postulated that research into the creation of a workable typology could be undertaken. Results suggest that the size and shape of a feature may help to determine a type group, as shown by the two D-shaped enclosures at Bywood Farm; though the characteristics of the land in which a site is situated might also show two sites to be related, as at Bywood-Cadhayes and Dunkeswell. Few of the sites within the study have clear dating evidence available without invasive investigation taking place.


Archaeological features of varying types and ages have been identified on Bywood farm in the Blackdown Hills such as two 1st century AD slag mounds and possible field boundaries of medieval date. Whilst survey of a set of earthworks initially revealed one hill-slope enclosure, a second ring ditch was identified from the 1945 RAF aerial photography and the possibility of a third (still un-surveyed) landowner's observations. Though no dating evidence was available for these features geophysical survey has assisted in revealing further information regarding their probable outline. Investigations of other similar features in Devon and Cornwall (known in Cornwall as 'rounds'), have lead to suggested dates of Iron Age use, as at Parsonage Wood, Cornwall and Milber Down, Devon, with some continuing in use into the Roman period, suggested at North Hill Cleave, Bittadon, North Devon and again at Milber Down. Other examples have been dated to the medieval period, as happen with Dunkeswell across the valley.

The aim of this report is to design and build a working typology that can assist in the identification and dating for the hill slope enclosures at Bywood Farm. It was further hoped that such work would give context to other similar features in the wider landscape of the Blackdown Hills, indicating if a particular size and shape of a feature is reflected at other sites, possibly providing a period when those features may have been in use. Sources of information used in compiling the data needed for this study will come from Sites and Monuments Record (SMR) information and RAF photographs, as well as more recent Aerial photographs. Having identified the sites to be investigated, field visits wherever possible took place so that the modern landscape of field boundaries and other features can be understood. Modern maps, Tithe maps and Domesday records were used to give place name and dating evidence for any sites mentioned in the text or map, which related in some way to the feature. In some cases, it is hoped Geophysical survey can be used to increase our understanding of the features and aid interpretation.

In order to further investigate the enclosures at Bywood Farm on the Blackdown Hills, Devon; a study area around the sites has been outlined for investigation. The study of aerial photographs and SMR data of this area has lead to the identification of a series of apparent hill slope enclosures of a similar type, that it is hoped may help to add context to the Bywood sites. By making comparisons of the sites located in this way, through the general categories of: The characteristics of the enclosures, the geography of the surrounding area, and the modern landscape, a typology of different sites may be reached, hopefully suggesting a date for some of the enclosures use. 

The characteristics of the enclosures
The size and shape of the enclosures will be an important tool in characterising the differences between sites that may date to different periods, or be used for different purposes. For example, rectilinear sites are often suggested to date from the Romano-Celtic period, while it is often inferred circular hill slope enclosures are of a prehistoric type some medieval examples possibly exist. Characteristics such as the formation of banks and ditches, numbers of entrances and the orientation of the enclosures when known, may suggest information about the use of the sites, such as a settlement enclosure with boundaries showing the settlement edge; farmsteads or field boundaries with bank and ditches designed to enclose animals, or larger settlement areas involving more than one tenant, and grander boundaries, with numerous entrances.

The Modern Landscape
The modern landscape and land-use around the enclosures can hold information not only about how the site may have looked and been used in the past, but can introduce to us ideas about the best way to locate other similar sites in the landscape today, through aerial photographs or through the discovery of sites on the ground. The current pastoral farming method used on the land in question can affect the chances of the discovery of sites; if for example, earthwork sites do not show up in aerial photographs to the same extent as soil or crop marks.

The relationship of the enclosure to other features in the modern landscape such as other archaeological enclosures, farms, roads and public rights of way; may help to set the site in its wider context. A close relationship to other enclosures, and then later farms might show the possibility that the site chosen for settlement in that area has moved through time, or that an area has become more widely inhabited and exploited over time. The distance of an enclosure to modern roads and public rights of way might on occasion offer an insight into the track ways and paths in use at the time the enclosures were active, and can thus suggest links between different sites or resources.

In some instances, an enclosure might have a direct relationship with other archaeological features, which either cut, or are cut by the ditches or features of the enclosure. This information could be of great importance in suggesting dates for the different types of hill-slope enclosure, if the feature cutting or being cut by the enclosure is of a known archaeological date. It could for example prove an enclosure to be of a younger or older period than the dated feature, depending on if cut by, or cutting that feature. Information such as this is often only found if the site has been excavated, so knowledge of previous investigations is also of great importance. Other archaeological features within 500m of an enclosure might indicate a possible relationship between the sites, which may again help us with the dating of enclosures. 

The Geography of the Surrounding Area
Knowledge of the local geography of the area where the enclosure is positioned can help to offer information about the site's past use. The exact location of the site on the side of the hill, be it the top of a slope, near the valley bottom or somewhere in between, may show a general trend in the positioning of enclosures, as might the elevation of a site on a hill.

The characteristics to be assessed Size: The size variables of the features will be split into numbered groups, relating to the type of site they are most likely to represent.
1 = 0-15m Sites likely to be representative of Bronze Age barrows 2 = 15-50m Sites which may represent minor settlement or field boundaries 3 = 50-100m Larger settlement sites, for example a farm stead 4 = >100m Thought to represent later manorial settlement sites, with larger expanses of land.


Shape: The shape of the sites may underline a difference between possible rectilinear Romano-Celtic sites, and prehistoric or medieval circular farmsteads, as well as indicate any annexes attached to a site.
1 = Circular
2 = Sub- circular
3 = Rectangular
4 = Sub-rectangular
5 = Square
6 = Sub-Square
7 = Annexed

Banks and Ditches
The construction of bank and ditches at an enclosure may help to indicate use, such as a stock-proof enclosed field, or a settlement enclosure.
1 = Single Ditch
2 = Double ditched
3 = Bank
4 = Bank and ditch - valleted
5 = Double bank and ditch - Multi-valleted 
6 = Ditch with no apparent bank

The number of entrances visible at the site.
1 = One
2 = Two
3 = Three
4 = Four
5 = >Four

Orientation of the site
The orientation of the sites entrance(s)
1 = N-S
2 = E-W

The Modern Landscape
Visible today

From SMR (sites and monuments record) data, or aerial photographs
1 = Earthwork
2 = Soil mark
3 = Crop mark
4 = Standing feature

Modern Farming Practices
To give an indication of the probable preservation of a feature.
1 = Arable
2 = Pastoral
3 = Mixed
4 = Wooded

Distance to Road
The distance of a feature to the modern road, might correlate with a track way or path in use at the time of the sites use. This will be split into distance groups.
1 = 0-99m
2 = 100-499m
3 = 500-999m
4 = >1000m

Distance to public right of way
Public rights of way might depict a relic track or road used when a site was active. These will again be split into distance groups.
1 = 0-99m
2 = 100-499m
3 = 500-999m
4 = >1000m

Sites relationship to each other, in distance
The distance between known surviving enclosures, could represent a link between two settlement sites, a settlement and related filed systems, or a later site which has moved away to replace a previous site. This distance will be shown split into groups.
1 = 0-99m
2 = 100-499m
3 = 500-999m
4 = >1000m

The relationship in distance, to the nearest modern farm
Modern farms might show a link with enclosure sites, if the chosen settlement site has moved over time. This distance will be shown in numerical groups.
1 = 0-99m
2 = 100-499m
3 = 500-999m
4 = >1000m

Relationship to datable features
Field boundaries and other datable features might cut, or be cut by an enclosure, thus helping suggest a date range for the site.
1 = site cut by a feature
2 = site cuts through a feature
3 = feature (such as field boundary, path…) respects site
4 = no relationship to datable feature

Relationship to archaeological features (within 500m)
1 = prehistoric
2 = Roman
3 = early medieval
4 = medieval
5 = post medieval
6 = unknown date
7 = none

Previous excavation
To show if a site has been investigated.
1 = Excavated
2 = Not excavated

The geography of the surrounding area
The contour of the site
The positioning of a site on the valley floor, mid-slope or hill-slope, could distinguish the sites by type.
1 = Valley floor
2 = Low slope - less than 15 degrees, easy to plough 
3 = Mid slope - 15-30 degrees, possible to plough
4 = High slope - >30 degrees, difficult to cultivate
5 = Hill top

Elevation of site
The height of the terrain the site is situated on, given as a distance range.
1= 0-49m
2 = 50-99m
3 = 100-149m
4 = 150-199m
5 = 200-249m
6 = 250-299m
7 = >300m

Local Terrain
How level the terrain of the site is.
1 = Level
2 = Slight gradient
3 = Steep gradient

Short descriptions of each site

Hillend Farm
This possible enclosure in the parish of Luppitt, referred to here as Hillend Farm 1, is a small rectangular earthwork measuring roughly 90m by 30m (DSMR: Devon Sites and Monuments Record: NUM=ST10NE/126). The site sits just 230m away from a sub-square enclosure, which will be discussed later under the label Hillend Farm2. Hillend Farm 1 lies within 150m of the nearest public right of way, with a distance of 350m to the nearest metalled road, leading to the modern Hillend farm. 


The enclosure at Hillend 1 must be of an older, or contemporary date to the field boundary skirting it, as this hedge respects the extent of the feature, bending around its length. The 1840's Tithe map shows that the boundary skirted the site in 1844. The feature sits in a field named 'Long Meadow' suggesting probable meadow status for the land. The 1974 RAF aerial photograph (DSMR: Devon Sites and Monuments Record: 30-26) of the site does not clearly show an enclosure, though there is clear disturbance running along the curved field boundary. This may be the remains of an earthwork from an enclosure, or a geological or quarrying mark. In a pastoral meadow, the possibility of a marl pit causing the disturbances seems slight, however the field just to the North of Long Meadow is called Pit Land, suggesting the presence of a quarry or pit.

Hillend Farm 2
The Sub-square enclosure at Hillend 2, in the parish of Luppitt (DSMR: Devon Sites and Monuments Record: NUM=ST10NE/127). It is noticeably smaller than Hillend 1, at just over 20m by 25m. The nearest modern farm and metalled road are situated just over 100m from the enclosure, with a public right of way within metres of the site, apparently respecting its boundaries rather than continuing to travel in a straight line, which would involve travelling across the enclosure. Today the footpath appears have limited use, and no clear destination that might suggest its use in the past. The feature at Hillend 2 shows up faintly on the 1974 RAF aerial photograph (DSMR: Devon Sites and Monuments Record: 30-26), lying in the field named on the Tithe map as 'Great Mead' and adjacent to 'Green', possibly indicating marshy ground. 

Ewin's Ash 
The Circular enclosure at Ewin's Ash in the parish of Luppitt, is only visible today as a crop mark (Devon Sites and Monuments Record: DSMR: NUM=ST10NE/153), and appears to measure around 65m in diameter. It is unclear due to its current eroded state, the form of any banks and ditches that may have been present in the past. The field in which the site lies has been ploughed extensively in modern times, as its position on the flat hilltop makes it ideal arable land. Windsor farm is the closest modern farm to the site, 460m away, and the enclosure is in close proximity to a metalled road, just 60 meters away. In a similar way to the Hillend examples, Ewin's Ash is sited near to the circular enclosure at Hill View, just over 200m away. There is no sign of the feature at Ewin's Ash on the Tithe map or on the ground, in the field named 'Ewin's Ashe Field' where it lies, though on the 1974 RAF aerial photographs, a faint circular crop mark can be seen (DSMR: Devon Sites and Monuments Record: 30-40).


Hill View
The Enclosure at Hill View located in the parish of Luppitt (DSMR: Devon Sites and Monuments Record: NUM=ST10NE/154), can faintly be made out on aerial photographs as a circular crop mark 40m in diameter, positioned just over 200m from the Ewin's Ash enclosure and 200m from Windsor Farm. The nearest metalled road to the site is less than 50m away, and leads directly to the farm. As with the very similar site at Ewin's Ash, the Hill View site can be identified on the 1974 RAF aerial photograph (DSMR: Devon Sites and Monuments Record: 30-40), but not on the Tithe map or in the field.

Cleave Farm
The roughly circular enclosure at Cleave Farm (DSMR: Devon Sites and Monuments Record: NUM=ST20NW/10) located in the parish of Upottery, is visible today as a soil mark on the 1974 RAF aerial photographs (DSMR: Devon Sites and Monuments Record: 22-102), sized around 35m to 40m in diameter. The enclosure is positioned 40m from the nearest public right of way, which skirts around the modern Cleave Farm, located just under 100m from the site. The nearest metalled road is sited the opposite side of the modern farm, over 200m away. The Tithe name of the field the feature is situated in is 'Sawpit Field', with other field names around the feature include 'Lower Orchard', Old Orchard', 'Oak Close' and 'Coppice'. These names suggest a wooded area with wood-working. The feature visible in that field may be related to a workers cottage.

Middleton Barton
This enclosure noted on the Sites and Monuments Record at Middleton Barton Farm in the parish of Clayhidon (DSMR: Devon Sites and Monuments Record: NUM=ST11SE/87), is a small rectilinear feature, located in a field which appears on the 1947 RAF aerial photograph to be quite disturbed (DSMR: Devon Sites and Monuments Record: 22-77). This particular feature appears to be around 30m by 10m in size, and is located 80m from the nearest metalled road. This road leads directly to Middleton Barton farm, located 160m away.

Investigations at Middleton Barton farm, both from aerial photographs and at the site itself, suggest this small feature is likely to be one part of a larger dispersed settlement of cottages and enclosures, spread over three fields. Similar settlement sites can be seen at Hense Moor, also on the Blackdown Hills, where a series of small plots have been recorded. An apparent 50m diameter circular feature visible on aerial photographs, was investigated but proved to be the product of a series of relic field boundaries belonging to the past settlement. The field containing the feature noted in the Sites and Monuments Record (SMR) has the name 'Seven Acres' on the 1844 tithe map and has surrounding it 'Grange Eight Acres' and 'Little Croft', all suggesting the size of the fields. 'Way Close', a field next to the feature, suggests a road ran near by.

Collinshayne Copse
The feature at Collinshayne Copse in the parish of Hemyock, is visible on the 1947 RAF aerial photograph as a small circular crop mark, 30 metres in diameter (DSMR: Devon Sites and Monuments Record: 22-73). The feature has been entered on the Sites and Monuments record as an earthwork ring and possible enclosure (DSMR: Devon Sites and Monuments Record: NUM=ST11SW/159). The site is located 50m from the small settlement at Collinshayne, adjacent to the nearest metalled road, and 460m from the nearest modern farm, Crocker's farm. Collinshayne Copse is situated 800m from the site at Burrow Hill and 1km from the site at Lemon's Hill (ST 15481218). Further similar features appear on the 1947 RAF aerial photograph in the surrounding area, suggesting that even if this site was not formed due to the extraction of marl, marl pits may have been active in the close vicinity. The Tithe names for the field containing the feature, and those surrounding it suggest that the area was already wooded by 1844. The names include 'Oak Close' where the feature lies, 'Coppice' and 'Sawpit Close'.

Burrow Hill
This enclosure near Jewell's farm in the parish of Hemyock (DSMR: Devon Sites and Monuments Record: ST11SW/100/1), is clearly visible as a rectangular crop mark on the 1974 RAF aerial photograph, 50m by 40m (DSMR: Devon Sites and Monuments Record: 22-72). The Air photograph appears to show a small relic field boundary to the east of the feature, perhaps previously forming a small field or enclosure. Tithe map information shows a now missing field boundary respected the lines of the feature, running along its outside (figure 3.4). This boundary split the modern field into two, 'Great Castle', possibly representing a past archaeological feature, to the north of the feature, and 'Little Castle and Linhay' to the south, and incorporating the feature, suggesting a cattle house lying within the old 'castle' field name. A public footpath runs through the bottom of the field, 40m from the feature. The nearest metalled road is 230m away, and the closest modern farm is Newton Farm, 500m away. The site lies within 800m of the sites at Collinshayne and Lemon's Hill.

Dunkeswell-Connet's Farm
The site at Connet's Farm in the parish of Dunkeswell, is the only example of a hill slope enclosure within this study which has been partially excavated. The site can be seen on the 1947 RAF aerial photograph as a 40m-diameter, sub-circular valeted platform, positioned on a shallow slope, leading into a steep drop (DSMR: Devon Sites and Monuments Record: 22-77). The site sits adjacent to the modern road, and within 30m of the closest modern farm, Connet's Farm.

Silvester's excavations at Connet's Farm revealed medieval pottery evidence from the inside of the enclosure although features located in the interior were not investigated further due to constraints of time and money. Mixed pottery evidence lead Silvester to conclude the site was in use between 1250, and the 14th centaury. At least one sherd of pottery similar in style to Neolithic Peterbourgh wear was also excavated from the site, with others of a very similar type previously being accepted as being of a medieval date. The field name where the site lies was noted on the Tithe map as 'Castle', and to the south of the main field, 'Little Castle'. The field boundaries and roads surrounding the two 'Castle' named fields, creates an oval enclosure of over 100m in diameter, possibly representing a past manorial boundary. The flat area of castle field remains the only undeveloped land lying next to the road in the village, suggesting that the land was still held by an important landowner, not involving it as part of the settlement development. The site lies 550m from another circular enclosure at Dunkeswell-Hutshayes. 

The site at Hutshayes, again in the parish of Dunkeswell, can be seen on the 1947 RAF aerial photograph as a roughly circular feature, 85m in diameter (DSMR: Devon Sites and Monuments Record: 30-2). The nearest metalled road is 80m away from the site, leading to the closest modern farm, Hutshayes. The enclosure is 550m from the site at Connets farm, though substantially bigger in size. The feature has been cut by a field boundary, and thus must predate it. The tithe information for this feature is interesting, showing a difference in the shape of the fields containing the feature between 1884 and today. On the modern and 1947 aerial photograph, the field boundary to the south east of the enclosure is not visible, but can clearly be seen on the tithe map, curving around the edge of the circular feature. The feature sits inside the tithed field named 'Hill Croft' and adjacent to 'Great Meadow'. 

Lemon's Hil
The feature at Lemon's Hill appears on the aerial photograph to be a roughly circular enclosure, measuring 30m in diameter (DSMR: Devon Sites and Monuments Record: 22-83). The site lays on the low slope, just off the flood plain of a small stream. The closest metalled road to the feature is 350m away to the west, with a public footpath running at right angles, 250m to the south. Bolham Farm is the closest modern farm to the Lemon's Hill enclosure, which lies in close relation to the similar enclosures at Collinshayne and Burrow Hill. The area around the site is now known as Potter's Croft, possibly having been used for pottery manufacture in the post-medieval period. The enclosure might relate to this activity, though no clear link has been noted. The name of the field containing the feature is 'Potter's Croft mead', which lies to the south of the river, behind 'Long Ham' a name commonly found beside a river. 

The enclosure at Hillside in the parish of Luppitt, is a sub-square feature, 75m by 75m. The site is located 40m from the closest metalled road, and 90m from the nearest footpath, cutting through the bottom of the field where the site now lies. It is clear from the aerial photograph of the area, that the feature has in the past been cut through the middle by a hedge, splitting the field it is sited in, though this hedge has now been removed. The nearest modern farm to the feature is Hillside, 200m away to the west. The 1974 RAF aerial photograph of the site shows evidence that the feature may not be an enclosure as such, but a series of relic field boundaries left from medieval strip field systems (DSMR: Devon Sites and Monuments Record: 30-45). There is a slight possibility that an enclosure was in existence before these boundaries were laid, with the later boundaries respecting the edges of an enclosure, or that a later enclosure was put into a larger, no longer divided field in the post medieval period, however having studied the aerial photograph, this seems unlikely. 


Hole Farm
The site at Hole Farm in the parish of Clayhidon, is a multi-valeted circular enclosure, 90m in diameter. The site can clearly be seen on the March 1948 and the 1947 RAF aerial photographs as a crop mark (DSMR: Devon Sites and Monuments Record: 22-76, 22-83). The enclosure is sited near the hilltop, adjacent to a metalled road 60m away to the east. A footpath runs across the top of the field where the feature is now sited, 90m away. The nearest modern farm is Hole Farm, 220m away. The 1840's tithe map of this area shows the modern field containing the feature was during that period, divided into two separate fields. Unfortunately the map is damaged over the point where the feature would have been depicted, though it is unlikely the feature would have been visible and deemed worthy of noting at that time. The field containing the feature is named 'Great two acres', below it is 'Little two acres and to the side is 'Long four acre'. These names can tell us nothing but the size of each plot.

Bywood Farm-Style
The Bywood-Style site located in the parish of Dunkeswell, is a sub-circular enclosure 40m in diameter on a very flat piece of land, positioned on the hilltop. Having been cut by two hedges and a drainage ditch, parts of the site are almost impossible to distinguish, but where the feature is visible, it appears above ground and on the aerial photograph to have been constructed with a ditch but no bank (DSMR: Devon Sites and Monuments Record: 22-97). Geophysical survey evidence from the enclosure contradicts this, showing a likely bank and ditch formation. In one of the three fields that now house the feature, a slag mound dated to the first centaury AD (Wiecken, 2004), appears to lie against, or even possibly over the site. Unfortunately, it is in this field that the feature has almost entirely been destroyed so it is hard to be certain that the enclosure predates the 1st century AD slag mound, or if the mound is of a younger or contemporary date to the enclosure. 


It is possible that the slag deposit represents a rubbish pile, discarded over the enclosure bank, thus showing a contemporary date for the feature. The three boundaries the feature lies under belong to fields named on the tithe map as ' Lower Style close', 'Mid Style close' and 'Great close'. No clear meaning has been found for the name 'Style'. The site at Bywood-Style lies 80m west of nearest metalled road and 300m from Bywood Farm, the closest modern farm. The site is located 700m from another circular enclosure at Bywood farm, named here Cadhayes. 


Geophysical survey around the feature has partially revealed a second rectilinear bank running around the outside of the circular feature, though the extent of this second feature has not been concluded, thus no sound suggestions can yet be made about its date or function.

Bywood Farm-Cadhayes
Bywood-Cadhayes in the parish of Dunkeswell, is a valeted circular enclosure, 40m in diameter, the site was identified with help from the landowner on the ground and is not viable on the RAF or 'getmapping' aerial photographs. The feature is situated at the top of a hill slope, just above a steep incline. It appears that the entrance of this site is positioned at the south or southwest, and possible a second entrance opposite, facing north. Erosion from cattle in the field might explain away these apparent entrances. Its also appears that a series of platforms might have been located within the feature. Geophysical survey work carried out on the site has identified the feature not to be circular, as it appears to be from the remaining earthworks, but a D-shaped enclosure with side annex. The field names for the site indicate late enclosure or a re-working of the land before the 1884 tithe map. The names include 'Seven acres', 'Eight acres' and 'Ten acres', all suggesting the size of the fields. The nearest footpath to this site is 20m away, and metalled road 300m away. Bywood farm is again the closest modern farm, as at the Style site, the farm is located 400m from Cadhayes.

Circular enclosures - 15m-50m in diameter
The features of 15-50 metres in diameter, and a circular shape include Hill View, Cleave Farm, Lemon's Hill, Collinshayne Copse, Bywood-Style, Bywood-Cadhayes and Dunkeswell-Connets. A comparison of the contours and elevation of the sites where these features are situated show a lot of similarities between the sites, with only Hill View and Bywood-Style being noticeably different, both lying on the hilltop, at a height of over 200metres above sea level. 


The five other sites in the group lie between the low and mid slope of the hill, between 100 and 200 metres above sea level. This comparison may indicate a preference for the positioning of hill slope enclosures on a gently slope of a hill, rather than the windy hill-top. Due to the poor preservation of many of these sites, a comparison of entrance orientation or the construction of banks and ditches is in most cases impossible, however at both Bywood sites, as well as Dunkeswell-Connets, it has been noted that the features all have a single bank and ditch formation. Only four of the features lie within 500metres of an archaeological feature of known date, though no absolute link can be made between the enclosure and that feature at any site. Lemon's Hill lies close to possible post medieval pottery manufacture, Connet's farm was given a medieval date by the excavator Silvester, and the two Bywood sites lie in close proximity to Romano-British slag mounds (Wiecken, 2004). It should be noted that the sites at Bywood now appear not to be circular, rather D-shaped and thus may differ from the other features. However geophysical work at other sites may show a generic 'circular' group is too general as other shapes and styles may be revealed.

50m-100m in diameter enclosures
Three features have a diameter of 50-100 metres, and appear to be circular; these are Ewin's Ash, which lies within 200 metres of the smaller circular enclosure at Hill View, Dunkeswell-Hutshayes and Hole. The available evidence of Contour and Elevation of the three features suggest that the larger circular enclosures are often positioned at a height from 200 to 300 metres above sea level, and on the hilltop. With such a statistically small group however, this suggestion may not prove correct for all 50-100 metre circular enclosures on the Blackdown Hills. The feature at Hole appears to be different from all other circular features of any size, being the only enclosure to clearly be of a double bank and ditch formation. No datable features lie in the proximity of any of the three features.

Non-circular enclosures
The five non-circular enclosures in the study area are Hillend 1, Middleton Barton and Burrow Hill (Rectangular), and Hillend 2 (square). The feature at Hillside has now been discounted as an enclosure feature, as it appears now to have been an illusion of a square feature, made up a of a series of crop marks left from a strip field system. All features but Hillend 1, sized at 90m by 30m, are of a 15-50 metre size. No clear comparison can be made between these sites, based on their positioning on the hillside or the elevation above sea level, as this ranges from 150 metres at Middleton Barton, to 300 metres at Hillend 1. Once again the low numbers of features being examined here makes comparisons of this nature difficult. The number and positioning of entrances is seemingly impossible, as is the understanding of the composition of banks and ditches. The feature at Hillend 1 is skirted by a boundary of unknown date, leaving no clear dating for any of these features. 

In almost all cases, regardless of size or shape of the feature, the distance to the nearest road, public right of way or modern farm is no more than 500 metres.

The small quantity of data available in this case study may make comparisons between features unreliable for a larger scale survey, however observations from the characteristics of sites can be made from these results. A general position of low to mid hill-slope can be suggested for the 15-50 metre circular enclosures, at an elevation of less than 200 metres; while the larger 50-100 metre circular enclosures appear to lie at the hill-top. The mixed modern farming practises on the hills has left many of the features with poor above ground preservation, resulting in a shortage of information about the bank and ditch formation, entrances and orientation missing in many cases. These results can only suggest the characteristics of features, which retain visible banks and ditches, while those which are lost through poor preservation, or sites which never possessed earthwork enclosures, but perhaps a palisade, are ignored. 

Domesday Entries
The entries from the Devon Domesday Book may be helpful in suggesting if a settlement was in place in the area, or close to any of the enclosures incorporated into this study. The area around Luppitt contains the circular enclosures of Ewin's Ash and Hill View, as well as the probable strip field system located at Hillside. There are two entries in the Domesday for Luppitt, Being 'Luppitt' and 'Greenway, Luppitt'. Both holdings are quite large, with three villagers and four smallholders with a value of 40 shillings at Luppitt, and three villagers and one smallholder, worth 20 shillings at Greenway. However, the sites at Ewin's Ash and Hillview were not incorporated directly into the Domesday records, as the land where they are located remained common land until relatively recently. This suggests the circular enclosures at Ewin's Ash and Hillview may be dated to a period earlier than 1066.

The entry in Domesday that lies closest to the feature at the modern Cleave Farm is Upottery. This holding is very large, with eighteen villagers and four smallholdings. The value of the holding, including the land and stock owned by the lord and the villagers, was 100 shillings. The shear size of the holding, and its population within, suggests that the circular enclosure at Cleave Farm may be linked to the pre or post-Norman Conquest holding of Luppitt, however it is just as likely to date to an earlier period still.

Bywood farm's entry in the Domesday Book shows that is had four villagers as well as the lord, with an over all value of 10 shillings (Morris, 1985, 16-11). Today there are at least two features visible on the modern Farm of Bywood, both with a possible link to slag mounds dated to the first century AD. However, without a definite link to these Romano-British features, it is possible that Bywood-Style and Bywood-Cadhayes may be linked to the Pre-Norman period or just as likely, to the prehistoric period.

The holding at Dunkeswell, according to its Domesday entry, was held by Ralf De Pomeroy, the Norman Lord in the area. Although the overall value of the Holding is much smaller than that at Upottery, being just 50 shillings, there were 11 villagers farming the surrounding land. The Tithe field name of 'castle field' suggests the possibility that one of Ralf's main tenants may have held a large manorial residence at Dunkeswell, possibly at the site of Dunkeswell-Connets. The Circular feature at Dunkeswell-Hutshayes shows no clear connection to the Domesday entry, however it may have belonged to one of the 11 villagers within the holding, or to an unrelated period.

The possible deserted medieval settlement at Middleton Barton, lies close to the Domesday land holding of Hole Farm (Morris, 1985, 16-8), however the two may not be in anyway linked, as the holding was quite small, with only two villagers and a value of five shillings. It is possible that the settlement at Middleton Barton may have become occupied after the writing of the Domesday Book, however if the settlement at Middleton Barton was made up of a series of small plots, it is unlikely to have been taxed separately, and thus included in the book.

The two Domesday entries of Hemyock and Bolham Water are positioned around the features at Lemon's Hill, Collinshayne Copse and Burrow Hill. Both land holdings were large, Hemyock having 12 villagers and 12 smallholders and having a value of £6 by weight, and Bolham Water having eight villagers and six smallholders, worth 35 shillings. The circular enclosures at Lemon's Hill and Collinshayne Copse and the rectangular enclosure at Burrow Hill, may be related to the occupation of the Villagers and Smallholders at the time of the writing of the Domesday Book, however none are linked closely to either entry. 

It must be noted however, that not all farms in the area during the compiling of the Domesday Book, would have been recorded in the book, as they may have been incorporated into a larger holding.

Evidence from Geophysics
Geophysical survey at the two features at Bywood Farm has increased our understanding of the below ground archaeology of both sites, that were previously believed to be roughly circular enclosures, especially that at Style. Survey results from both sites suggest a D-shaped appearance, with a bank and ditch construction. An Earthwork survey carried out by Community Landscape project volunteers of the Cadhayes site did produce a slightly D-shaped appearance though the annexed enclosure which appears from the geophysical evidence to exist under the ground, is not visible at all from the surface. 

The feature at Style can be seen on the RAF aerial photographs, showing what appears to be a ditch but no bank. A faint earthwork has been noted in the field, seeming to once again suggest no bank enclosing the feature. Geophysical results show however, that a bank and ditch formation does appear to exist under the ground, in a similar way to Cadhayes. The geophysical results have also led to the questioning of the 1st century AD radiocarbon date of a slag mound lying adjacent to the feature. As the keyhole excavation was taking place into the slag mound to obtain enough charcoal for dating, a trench may have been cut into the ditch of the enclosure. It was believed prior to the geophysical survey results, that the extent of the enclosure at Style, continued well into the field containing the slag mound, thus it was not suspected that a ditch from the feature would effect the dating of the slag.

Geophysical survey at other hill slope enclosures within the case study may help produce a clearer interpretation of the sites, as it has at Bywood Farm. It would be especially helpful if future geophysical work could take place at Dunkeswell-Connet's Farm, to show if similarities with Bywood-Cadhayes are visible beneath the ground as well as above, or if the two are of a totally different style below ground.

No clear chronological dating can be assigned to the varying hill slope enclosures within the case study on the Blackdown Hills, however comparisons between sites can be made which may indicate a type group for a particular style of feature. On occasion, such as at Bywood-Style, a 1st century AD date may be linked in some way to the enclosure, and might be used to suggest a period of use for a similar feature elsewhere in the case study, such as at the similarly shaped D-shaped enclosure at Bywood-Cadhayes. The clear similarities in elevation and positioning of the probable circular sites, of 15-50 metres in diameter, could suggest a similar period of use for these features, or a continuation of tradition through time. Without clear dating evidence this remains impossible to determine. The 13th century date accepted at the site of Dunkeswell-Connets holds similar characteristics to the features at Bywood Farm, which both lie in close proximity to 1st century AD slag mounds. If there were to be a connection between the Bywood features and the mounds, it may suggest the characteristics of the sites, such as their position on the hill-slope, remain in use through time, from atleast the Romano-British period at Bywood, to the medieval period at Dunkeswell. The presence of field boundaries respecting or cutting some sites can add to the possible dating evidence, such as at Hillend 1 and Burrow Hill, however, these boundaries are themselves of unknown date.

Any further work on the hill slope enclosures of 15-100 metres in size, on the Blackdown Hills, would be aided greatly with a wider survey of geophysics on the features, in order to reveal more about the below ground characteristics of the features. This would be especially helpful at sites such as Ewin's Ash, Hill View and Dunkeswell-Hutshayes, where the only evidence available for the features at present is in the form of aerial photographs, showing faint outlines of the size and possible shape of the features. The geophysical survey already carried out at Bywood Farm has revealed that the remaining earthworks, soil marks or crop marks of a feature might not reflect exactly the true characteristics of the enclosure.




                                                           This project was supported by

                                                                     Heritage Lottery Fund         Devon County Council