An Archaeological Evaluation at Highfields, Dunblane, Central
An Archaeological Evaluation at Highfields, Dunblane, Central
IntroductionThe Scottish Urban Archaeological Trust Ltd was commissioned by Halliday Fraser Munro Planning on behalf of Barrat West of Scotland Ltd to carry out an archaeological evaluation on the site of a proposed development at Highfields, Dunblane (formerly Hillside Farm). The development was expected to encroach upon the known site of two Roman Marching Camps, one located inside the other (NGR NN 775 006). The evaluation was carried out over a period of three days, 8th - 10th March 1995.
Site Location and DescriptionThe site lies to the immediate south-west of Dunblane and is situated between 95 m and 110 m OD (Site Location Plan (22K)). The ground rises slightly to the east, but to the south-west falls away towards the River Teith. At the time of this evaluation the site comprised open, recently ploughed, farmland.
The Roman camps are located in the eastern half of the development area. The location of these enclosures was originally identified by aerial reconnaissance in 1945 by Dr J K S St Joseph, and later confirmed by archaeological excavation and survey in 1966 and 1967, by Anne S Robertson. The two camps were thought to be contemporary, comprising a small inner enclosure within a larger camp. Approximately 40% of these camps already lies under modern housing schemes, and of the remainder, approximately 15% (2.5 hectares) lies within the proposed development area.
There are no upstanding remains visible on the ground, and both camps were revealed as crop markings in the standing arable crop.
Impact of Development on the Archaeological ResourceBarrat West of Scotland Ltd propose to erect 108 houses on the 18 acre site. Although much of the northern and eastern parts of the camps has already been lost to development, the central, southern and western parts survive. The proposed development will affect the western section.
A preliminary study of the existing property boundaries and the known location of the camps suggested that only the western half of the inner camp lies within the development area, and importantly includes the western gateway and tutulus - an outer earthwork which protected direct access to the entrance.
Archaeological BackgroundAerial photographs, limited excavation and probing has shown that the alignment of the camps is slightly skewed west of north-south. The outer camp was thought to measure c 34 acres, with four non-centrally, but oppositely, placed entrances between 35-40 feet wide. These provided access through a defensive ditch approximately 6 feet wide, up to 3 feet deep and with a V-shaped profile. In the areas excavated in 1966 and 1967 the associated rampart, constructed of upcast from the ditch and crowned by a timber palisade, was found to have been lost to ploughing.
The inner camp was thought to have measured 650 feet N-S by 1040 feet E-W, with a defensive ditch of the same dimensions and characteristics as the outer camp. This also had four entrances ranging in size from 22-24 feet in width placed opposite to the entrances of the outer camp. The inner camp was located against the west ditch of the outer camp and the two shared the same section of defensive ditch. As this section of ditch was common to both camps, and the associated entrance gave access to both the inner and outer camp, the tutulus was probably located here to protect what was potentially the weakest entrance.
The layout, and the characteristics of the ditches and entrances, suggest that the two camps were in fact contemporary, the inner camp merely an enclosure within a larger camp. This interpretation is supported by the relationship between the ditch of the inner camp and the west ditch of the outer camp, where they do not intersect but clearly terminate short of each other.
The internal arrangements within temporary camps are known to follow a set pattern, however the temporary nature of these camps means the configuration has proved difficult to confirm archaeologically. A T-plan network of streets (the main street -via principalis , ran across the width of the camp between the two main entrances, while the via praetoria and the via quintana) divided the interior into smaller units or blocks. These housed the commanding officer and his senior officers quarters, with the garrison arranged around the periphery. The addition of an inner enclosure, as at Dunblane, provides an interesting opportunity to test whether this influenced the standard arrangement.
No structural features were found within either of the camps except a single posthole near the south entrance. However, it must be taken into account that work concentrated on the defences and the interior was largely ignored. The sub-soil varied across the site from clay to bed-rock, and no finds were recovered to indicate a date for either the construction or occupation of the camps. Late-medieval or post-medieval pottery was recovered, but probably derives from manuring of the fields.
Dating for this site therefore remains difficult but the location and size of the camps suggests a Flavian date, possibly Agricolan. Dr Robertson (1969) pointed out that other camps with smaller enclosures, which might have housed a small garrison, include Ardoch, Innerpeffray and Kirkbuddo. In contrast to Dunblane, all of these types of camp are much larger in size (63 acres) and all but Innerpeffray have their enclosures outside of the main camp.
G Maxwell (1989) suggests that during his penultimate campaign in AD 83, Agricola appears to have split his army into three separate units, perhaps more for defensive reasons than attacking. Therefore considerable interest has been attached to smaller camps measuring around 30-40 acres in size; Ardoch, Strageath, Dornock and Dunblane, all lie at suitable marching intervals between the Forth and the valley of the Earn.
The location of the Dunblane camps is also significant as they may be indicators of an as yet undiscovered section of Roman road. This has been identified north of Dunblane, and continues along to the east end of the Gask Ridge and beyond. To the south of Dunblane the course of this road is not known until it reappears in the western suburbs of Stirling. It is for these reasons - the possible Agricolan date for the camp and the missing section of road nearby - that the camp at Dunblane has been categorised as a site of regional importance.
Objectives of the Archaeological EvaluationThe primary fieldwork objectives were:
This work was required in order to establish whether further archaeological investigation would be required to mitigate the impact of development, and what form any further work, if any, would take.
MethodologyA series of eleven trenches, labelled A-H, and J-L were opened within the archaeologically sensitive area. Trenches A and B were located to establish the precise position and alignment of the western outer ditch. Trenches C and D were positioned to examine the western gateway in detail. Trench E was positioned to examine the tutulus (Location Plan of Trenches C, D & E (8K)). Trenches F, G, H and J, K and L were positioned to examine the possibility of internal archaeological features surviving within the camp, such as structural remains, roads or pits.
Under archaeological supervision the trenches were all machine excavated reducing the soil profile in spits. This continued until either archaeological deposits were identified or the upper distinct horizon of the natural subsoil was reached. The sections of the trenches were then hand cleaned and the base of the trenches hoed to inspect for possible archaeological remains. Identified archaeological remains, such as the outer ditch, were sampled by means of a series of hand cut slots to examine their depth, form, nature, and state of preservation.
A single context recording system was employed with detailed fieldnotes taken of each stratigraphic element. All features identified were planned at a scale of 1:20 and sections and profiles were drawn of each hand cut slot trench at a scale of 1:10. All features were photographed from at least two angles, with a metric scale, and general site views were also taken.
This work was carried out by a team of three archaeologists.
A site monitoring visit was made during this evaluation on Thursday 9th March, by Lorna Main, the Regional Archaeologist for Central Region.
Summary of ResultsTrenches A (2 x 30 m), B (2 x 30 m), C (2 x 28 m), and D (2 x 26 m).
The western outer ditch, the portion shared by both enclosures, was located in Trenches A - D. It was discovered that the outer ditch lay approximately 28 m to the east of the supposed line marked on the plan of the camp dating from the 1960s. As a result the internal dimensions of the camp are reduced from c34 acres to c 32 acres. This change in size does not alter the tentative interpretation of the camps as possibly being built during the Agricolan campaign.
In Trench A only shallow remains survived (Section 3). The ditch was c 1.8 m in width and only 0.32 m in depth. The profile of the ditch at this point comprised a flattened V-shape. The fill of the ditch was waterlain light grey silt.
In Trench B the ditch was c 2 m in width and c 0.6 m deep, with a V-shaped profile (Section 4).
In Trench C the ditch was also c 2 m in width and c 0.6 m deep (Section 6). It had a V-shaped profile with a slight step on the inner, eastern side. At the base of the ditch was a deposit of waterlain silt and this was sealed by a series of mixed natural and ploughsoil deposits.
In Trench D the ditch was c 1.9 m in width and c0.6
m deep. It also had a V-shaped profile.
At this point the ditch was partially cut into bedrock.
It is interesting to note that this portion of the ditch bows slightly to the west at the western gateway. This bow in the ditch matches the topography of the slope in the western portion of the camp.
The two butt-ends of the outer ditch, representing the western gateway, were identified in Trench D. The gateway was found to be 5 m in width, not 12 m as reported after investigation and survey in the 1960s.
Curiously the butt-ends were of differing forms. The northern butt-end was squared in plan with a flattened V-shaped profile . The inner, eastern, side of the ditch appeared steeper in section than the outer, western side. The ditch at this point had a maximum depth of 0.7 m. A number of primary fills deposited during silting were identified towards the base of the slot trench. One lens of charcoal was revealed close to the base of the ditch (context 46). This deposit was sampled for possible environmental study. The upper fills of the ditch at this point comprised lower ploughsoil deposits (Contexts 48 and 49).
The southern butt-end was rounded in plan with a shallow flattened
V-shaped profile. Like the northern
butt-end it too had a steeper inner side. The ditch at this point
had a maximum depth of only 0.42 m. The ditch contained a single
fill of lower ploughsoil.
Both butt-ends were cut into natural orange/brown sand and gravel subsoil and were sealed by c 0.5 m of ploughsoil.
Unfortunately there was no evidence representing the remains of the road through the gateway. It is most likely that any possible remnant of a road has been removed by plough action.
Evidence for possible post-holes representing a gate was not found. Again this may be due to truncation from ploughing.
Trench E (2 x 12, 2 x 10)
The tutulus was located 10 m to the west of the outer ditch and
curiously lay slightly off centre to the south of the gateway.
It was c 3.2 m in length and c 1.8 m wide. It was
cut into the natural orange/brown clayey sand and gravel subsoil
and was sealed below c 0.5 m of ploughsoil. It had a maximum
surviving depth of 0.34 m with a U-shaped profile. The butt-ends
were rounded. The upper fills of the ditch comprised lower ploughsoils
(contexts 42, 34, and 35) and a series of light grey, probable
waterlain silty clays.
No evidence of an upcast mound was encountered and no datable archaeological remains were recovered.
Trenches F (2 x 27 m), G (2 x 34 m), H (2 x 34 m), J (2 x 20 m), K (2 x 16 m), and L (2 x 20 m).
No archaeological deposits features or artefacts were encountered within the internal area.
The average ploughsoil depth was found to be only 0.45 m, and such a shallow depth would undoubtedly result in the destruction of subtle archaeological features during ploughing activity over a substantial period. Indeed, with the field having only recently been ploughed visual inspection revealed that the orange/brown sand and gravel subsoil had been truncated in many places as it appeared mixed with the ploughsoil on the surface.
This average ploughsoil depth was found to be slightly shallower than that found in the trenches excavated across the western outer ditch. This difference can be accounted for by hillwash, as the internal area of the camp sampled lies on higher and flatter ground. The position of the western ditch lies at the foot of the slope from this higher ground and so a slight build up of ploughsoil occurs.
ConclusionThis evaluation has proved successful in identifying the location, form and alignment of the remains of the western outer ditch of the camp. It has also proved successful in identifying the precise location and form of the western gateway. This information has shown a slight anomaly with the previous plan of this part of the camp, dating from the 1960s. The differences in the dimensions of the gateway can be explained as they were planned as projections from one excavated butt-end of the western tutulus, and similar dimensions to the eastern gateway were assumed. The anomaly may also have been compounded as the western ditch was only probed and not excavated. As the ditch bows slightly to the west it is possible that probing may have lost its accuracy towards the centre point of the western ditch.
This evaluation has also established that due to ploughing activity and a reduced soil profile it is unlikely that archaeological remains survive on the higher flatter ground.
Further WorkThe Trust does not believe that further archaeological fieldwork is required within the area of this proposed development. This is due to the strong evidence indicating that archaeological remains within the internal area of the camp will have been destroyed as a result of plough action. Also, all the recoverable archaeological information was retrieved during this evaluation regarding the western gateway.
The alignment, depth and form of the outer ditch was also established in the process of evaluating the archaeological potential of the site. Although primary deposits were found in some of the hand cut slots across the ditch the Trust considers further investigation unnecessary as it is most unlikely that precise dating evidence will be recovered from further excavation.
However, the Trust does believe that the information retrieved during this evaluation merits further communication in the form of a short published article. As the western gateway provides the entrance to both the outer and inner camps it is a significant feature and it is therefore important that the exact dimensions are amended from the previously published plan of the camp. These changes in dimensions and the off-centre tutulus may have implications for our understanding of the relationship of the two camps. Further research is required to analyse possible parallels.
AcknowledgementsGrateful thanks are due to Barrat West of Scotland for funding this evaluation and to Halliday Fraser Munro Planning for their co-operation. Thanks are also due to Lorna Main, Regional Archaeologist for Central Region, for her assistance.
Further ReadingCollingwood, R G and Richmond, I, The Archaeology of Roman Britain (London, 1976), 8-14.
Frere, S, Britannia (London, 1978), 130-131.
Maxwell, G S, The Romans in Scotland (Edinburgh, 1989), 38-67 & 71-74.
Robertson, A, The Roman Camp(s) on Hillside Farm, Dunblane, Perthshire Glasgow Archaeological Journal, Vol. 1 (1969), 35-36.
Tel: +44 (0)1738 622393
Fax: +44 (0)1738 631626
E-Mail: jamesm@suat.demon.co.uk
Fieldwork Reports
