Font Size

100% 150% 200%

Background Colour

Default Contrast
Close Reset


Temporary Camp (Roman)

Site Name Dunning

Classification Temporary Camp (Roman)

Alternative Name(s) Kincladie Wood; Duncrub; Haughend

Canmore ID 26662

Site Number NO01NW 7

NGR NO 02508 15025

Datum OSGB36 - NGR


Ordnance Survey licence number AC0000807262. All rights reserved.
Canmore Disclaimer. © Copyright and database right 2024.

Toggle Aerial | View on large map

Digital Images

Administrative Areas

  • Council Perth And Kinross
  • Parish Dunning
  • Former Region Tayside
  • Former District Perth And Kinross
  • Former County Perthshire

Archaeology Notes

NO01NW 7 02508 15025

Roman Camp (R)

(site of)

OS 1:10,000 map, 1994.

Extends onto map sheet NO01SW.

In April 1940 O G S Crawford noted at the S end of Kincladie Wood (NO 023 152) a bank and ditch, 430' long with the top of the bank about 3' - 4' above the bottom of the ditch which is on the N side. He considered it a cross-dyke barrier across the NE approaches to Dunning, but commented that "the possibility that it might be a Roman camp cannot be excluded." He further identified it with the "trench" mentioned by Macfarlane in 1723 as being situated in a level muir NE of Dunning and "capable to contain several thousands of men." (The writer, William Hepburn, schoolmaster of the parish of Dunning, uses the word "trench" to mean entrenchment or earthwork enclosure).

Air photographs reveal very little in the way of reliable crop- or soil- markings, but sufficient survives to suggest that the remains are those of the greater part of one half of one side of a Roman temporary camp of the 'Stracathro' class, the straight element (oblique) of the entrance defences peculiar to this class being identifiable on the E edge of the wood, the road from Forteviot to Dunning having destroyed the tip of the feature, and with it the entrance gap and complementary clavicula.

The place-name "Duncrub" about 1 mile to the W raises the interesting possibility that this may be the site of the Battle of Mons Graupius (Mons Graupius = Dorsum Crup = Duncrub is an acceptable equation.) Furthermore, the local topography fits very well the description by Tacitus.

R W Feachem 1970; D R Wilson 1970.

NO 025 148. Air reconnaissance and excavation during the last three years of this 115 acre camp (of the 1st campaign (83 - 84 AD), not of Stracathro type) has revealed axial dimensions of 2.200' E-W by 2,300'. Six gates with titulum have been postulated, two each on the N and S sides and a central one on the E and W. The only extant remains, on the N side in Kincladie Wood, is a rampart 400' long, 10' wide and 2' high with a V-shaped ditch 11' wide and originally 3 1/2' deep. The E end of the rampart terminates at the titulum of the western of the two gates.

J K St Joseph 1973.

The only remains are those in Kincladie Wood.

Visited by OS (J P) 30 April 1975.

No trace of the remainder of the camp can be seen. A gas pipeline has cut through the N defences just E of the visible N gateway. Two phases of use were discerned. Excavation showed that the ditch hereabouts, which was 3m wide and 0.75m deep, had lain open long enough for a layer of 0.17m of silt to accumulate, before its width was reduced by the insertion of a clay bank to 2.4m.

L J F Keppie 1988.

NO 021 149. A trench measuring 35m by 32m was excavated in October/November 1992 over the western entrance to the Roman camp at Dunning. The work was conducted in advance of a housing development. This part of the camp perimeter appears as a cropmark, and the trench was positioned from the rectified plot of an aerial photograph.

The western entrance to the camp was identified by excavation to be 15.5m wide. It contained no evidence for any gateway features, and was guarded by a titulus c14.3m long which lay c12m W of the alignment of the perimeter ditch of the camp. The ditch of the titulus was up to 3.3m wide at the surface and 1.6m deep, with a V-shaped profile and a squared channel running along the base. There was no evidence for recutting of this feature. Eleven joing sherds of a mid-2nd-century BB2 bowl were recovered from the second fill from bottom at its southern terminus.

The perimeter ditch of the camp was up to 3.5m wide and 1.5m deep. Where excavated, it had a V-shaped profile with a squared channel at the base. To the N of the entrance, the perimeter ditch had been constructed in two discrete lengths, containing the same sequence of fills and separated by a 0.8m-wide unexcavated rib of subsoil. The surface of this rib lay c. 0.4m beneath the preserved surface of the ditch, which narrowed noticeably in width to c. 2m at this point. A narrow channel extended transversely across the surface of the rib. Recutting of the perimeter ditch was identified in the terminus sections to either side of the entrance, but nowhere else. The layered sequence of fills within the perimeter ditch and the titulus indicates gradual infilling. No trace survived of the ramparts of the camp.

An irregular tapering depression, up to 5.3m wide and containing the discontinuous remains of two superimposed pebbled surfaces, ran from adjacent to the S end of the titulus to the centre of the camp entrance. A sherd of 16th to 17th century AD pottery was recovered from the layer sealing the pebbled upper surface. These surfaces are not necessarily Roman in origin, and may be associated with the continuing use of the entrance as a natural route through the disused defences of the camp.

Sponsor: A & J Stephen (Builders) Ltd.

CFA 1993.

(Centred at NO 0229 1519). The surviving portion of the N side of this temporary camp lies within the S part of Kincladie Wood and there is no change to the existing description.

Visited by RCAHMS (JRS, IF), 24 October 1995.

NO 023 146 Investigation in September 1997, by machine-stripping, of a site due for development as housing within the SW corner of the Roman temporary camp by Dunning (NMRS NO01NW 7) revealed no features or finds of Roman origin. The area had been a garden in recent decades, and a fertile garden soil overlay natural to a depth of 0.3-0.4m below present ground level. Finds from the soil include pottery and glass sherds of recent origin, except for two tiny sherds of glazed medieval pottery, and a small number of animal bones and teeth.

Towards the SE of the area two stone features were uncovered: an irregular setting of flat stones and the base of a drystone dyke. The setting was sunk into a hollow or cut into natural, and the hard-packed earth between its individual stones contained charcoal and two tiny sherds of medieval pottery. The feature was not fully investigated due to lack of time, but may have been a setting for fires, or a stack base. The wall was not fully uncovered, but it ran WNW-ESE across the southern corner of the site, and appeared to have no obvious relationship to the modern property boundaries. No dating evidence was found associated with it. A 1.15m wide gap in the wall was presumably a gateway. Both features are likely to have been of medieval or later agricultural origin.

A sandstone slab, 360 x 305 x 70mm, was found by chance a few metres NE of the excavated area on the surface of the adjacent field. A rounded hollow 120mm across and 20mm deep has been pecked in the upper surface, probably with a stone tool. The peck marks inside the hollow have been worn down by grinding. The slab has many recent plough scratches.

Sponsor: Abertay Property Co Ltd.

N M Robertson 1998.

Scheduled as 'Dunning, Roman camp 420m NW of Haughend... the remains of a Roman temporary camp surviving as a cropmarked structure and a linear earthwork.'

Information from Historic Scotland, scheduling document dated 5 June 2009.


Project (17 June 2009 - 28 July 2009)

Watching briefs were kept between 17 June–28 July during the refurbishment of parts of the Forteviot-Dunning-Aberuthven 11kV overhead and underground power line.

J Lewis 2009

Watching Brief (17 June 2009 - 28 July 2009)

NO 02906 15229 – NO 02212 14616

Thirteen poles carrying an overhead line, together with several associated stays, were replaced in the limits of a Roman temporary camp (NO01NW 7), a scheduled ancient monument, on the E side of Dunning. No archaeological features were uncovered in any of the trenches.

J Lewis 2009

Publication Account (17 December 2011)

Macfarlane noted a ‘trench capable to contain several thousands of men’ just to the north-east of Dunning, which he presumed to relate to the battles in the 14th century (1726, in Mitchell 1906: 121). This was noted by Crawford, who, following Macfarlane’s reference, also recorded a small stretch of upstanding bank and ditch in Kincladie Wood, commenting that it could be a fragment of a Roman camp (1949: 59). Feachem and officers of the OS archaeology division re-examined the site in 1970, and suggested that the remaining stretch in the wood was the remains of the north side of a camp of Stracathro-type (Wilson 1970: 274).

Parts of four sides of the camp were subsequently recorded as cropmarks from the air, with further information recorded through excavation (St Joseph 1973: 218–19). It lies on undulating ground immediately northeast of the village of Dunning, with housing covering the south-west angle and parts of the west side. The Dunning Burn, a tributary of the River Earn, lies immediately to the west. The camp is rhomboidal in form, measuring about 705m from west to east by 660m transversely, enclosing some 47.3ha (117 acres). The surviving stretch of rampart in Kincladie Wood on the north side of the camp is about 120m in length, and about 3m in width and up to 0.6m in height; its accompanying ditch is about 3.4m in width and 0.3m in depth. An additional linear cropmark is visible in the north-east corner of the camp and may also be Roman, potentially representing a reduction in the length of the camp to some 565m, reducing the overall area to about 39ha (95 acres). This may be the same linear cropmark noted by St Joseph (Dunwell and Keppie 1995: 61n).

An entrance gap protected by a titulus has been recorded, and excavated, in the centre of the west side, and a further titulus is visible in the western portion of the south side. In addition, a titulus has also been recorded in the west part of the north side in Kincladie Wood (RCAHMS St Joseph Collection: Notebook 6; not depicted on illus 118). The camp has been subject to excavations on several occasions, with St Joseph recording, in the 1970s, that the ditch measured about 2.7m in width and 1m in depth on the west side and up to 1.5m in depth on the north and east. The north titulus ditch was up to 0.6m in depth (RCAHMS St Joseph Collection: Notebooks 5 and 6).

Further excavations just east of the north entrance recorded a U-shaped shallow ditch, some 3m wide and 0.75m deep, that had possibly been narrowed and reused (Dunwell and Keppie 1995: 53). Subsequent excavationsat the western entrance also suggested recutting at themain camp ditch termini, although not of the titulus ditch. There was a further break in the ditch c 11.75m north of the gate. The titulus ditch (some 14.5m in length) was up to 3.3m in width and 1.6m in depth. Eleven sherds from a BB2 bowl, probably dating to the Antonine period, were recovered from the fill of this ditch (Dunwell and Keppie 1995).

Dunning lies close to the place named ‘Duncrub’, which Watson first linked with Mons Graupius (1926: 56). Feachem took this reference further, proposing the area as the site of the battle, and publishing an inaccurate plan of the site, complete with Stracathro-type gateways (1970). This argument has recently been supported by Fraser, who unfortunately fails to correct the assertion that the camp had Stracathro-type gateways and also stated that it faced the Cairnie Braes to the north (2005: 72–6) ignoring the evidence for the traverse gateways and the likelihood that the camp probably faced east or west (the latter towards Duncrub). The evidence that Fraser advances is insufficient to assert that the site related to the battle with any confidence. (For further discussion of postulated sites for Mons Graupius, see Chapter 3.)

R H Jones

Field Visit (22 June 2013 - 27 June 2013)

NO 0229 1519 A topographic survey of the upstanding area of the Roman temporary camp in Dunning was undertaken, 22–27 June 2013, by students on the SERF field school. The survey focused on the Kincladie Wood section of the defences, recording the main bank and ditch of the camp, as well as the surviving remains of the titulus bank and ditch beyond the N entrance. At the W end of the rampart, the survey may suggest the beginning of the corner of the camp, in contrast to the extant plan which shows this in the field beyond. At the E end of the rampart, the survey may have identified the entrance before being cut by the road. An additional low bank to the S of the main rampart was probably associated with drainage in the boggy area of woodland.

Archive: University of Glasgow and RCAHMS (intended)

Funder: Historic Scotland and University of Glasgow

Jeremy Huggett, University of Glasgow, 2013

(Source: DES)

Excavation (21 June 2014 - 10 July 2014)

NO 02508 15025 As part of the SERF project, two trenches were excavated, 21 June – 10 July 2014, along the NE rampart of the Roman temporary camp where the bank and ditch survive as upstanding features in the community owned Kincladie Wood.

The primary trench (13 x 4m) was located across the bank and ditch as close to the B934 road as was possible. This modern road apparently utilises the original entrance to the Roman camp. A second trench (1 x 6m) was opened up to examine the W end of the titulus protecting the entrance and a potential Roman road surface.

In the main trench across the rampart two phases of ditch were identified. The primary ditch (4 x 2.5m) was V-shaped in profile with a flat bottomed (‘ankle-breaker’) feature in the bottom. The rampart that accompanied this phase survived to a height of 0.5m above the original ground surface

and a breadth of 3m. It was constructed of layers of turves separated by dumps of the clayey subsoil, presumably from the ditch. Two distinct layers of turf were identified. The rear of the rampart appears to have been furnished with a timber revetment of substantial planks set nearly vertically 0.5m into the subsoil. There was a suggestion of a timber post to

support a breastwork on the rampart, but this feature had been heavily disturbed by roots and burrows. Immediately prior to the construction of the rampart a fire had been built and later sealed by the body of the rampart. Following a period of abandonment, the ditch was more than half filled with material eroded from the rampart.

A second phase of defensive works followed the original line of the rampart and ditch. Again, the ditch had a V-shaped profile with a flat bottomed slot in its base. It was a less substantial feature (2.5 x 1.5m). No sign of a rebuilding of the rampart was noted but, even in its well preserved

condition, it is heavily eroded and disturbed by roots and burrows.

In the modern period the Roman ditch was utilised as a soakaway drain for the road, which involved a shallow re-excavation of the ditch filled with large stones. Finally the area of the ditch and beyond to the N was covered with a course paving of irregular stones and cobbles. This has been interpreted as a hard standing constructed for the use of travelling folk who utilised the wood until the mid-20th century.

Excavations at the titulus confirmed that this was a defensive bank associated with the Roman camp. Only one phase of ditch was identified, with a V-profile and flat bottomed slot. Conditions were challenging here, because the ditch was waterlogged, despite very dry weather during the excavation, and because it had been badly disturbed by a drainage ditch

associated with the modern road. No sign of the Roman road was identified.

Apart from finds associated with the modern use of the woods by travelling folk, there were no artefacts.

Archive: University of Glasgow and RCAHMS (intended)

Funder: Historic Scotland and University of Glasgow

Stephen Driscoll – University of Glasgow

(Source: DES)

Watching Brief (17 November 2014)

A watching brief was kept on 17 November 2014 during the excavation of a trench for a new pole on an overhead power line within a scheduled monument (SAM3675) containing a Roman temporary fort, identified principally by aerial photography. Nothing of archaeological significance was uncovered in the trench, only backfilled materials from the insertion of the earlier pole.

J Lewis 2014.

OASIS ID: scotiaar1-195924

Geophysical Survey (2015)

A programme of geophysical survey was undertaken in 2015 across a number of known archaeological sites in Perth and Kinross:

NO 02182 15001 (Canmore ID: 26662, SMR: 3675) Dunning – Roman temporary camp

NO 01900 15900 (Canmore ID: 26621, SMR: 9158) Leadketty – Enclosures, causewayed enclosure, ring ditches, square barrow and pits

NO 03940 17530 (Canmore ID: 26608) Forteviot – Roman temporary camp

NO 05353 16830 (Canmore ID: 26565) Forteviot – Henges, enclosures, square barrows and pits

NO 02600 15700 (Canmore ID: 84940, SMR: 8918) – Pit alignment.

Archive and report: National Record of the Historic Environment (NRHE)

Funder: Historic Scotland


Tessa Poller and Dene Wright – University of Glasgow

(Source: DES, Volume 16)


MyCanmore Image Contributions

Contribute an Image

MyCanmore Text Contributions