Font Size

100% 150% 200%

Background Colour

Default Contrast
Close Reset


Roman Fort (Roman)

Site Name Birrens

Classification Roman Fort (Roman)

Alternative Name(s) Blatobulgium

Canmore ID 67099

Site Number NY27NW 4

NGR NY 21900 75180

Datum OSGB36 - NGR


Ordnance Survey licence number 100057073. All rights reserved.
Canmore Disclaimer. © Copyright and database right 2022.

Toggle Aerial | View on large map

Digital Images

Administrative Areas

  • Council Dumfries And Galloway
  • Parish Middlebie
  • Former Region Dumfries And Galloway
  • Former District Annandale And Eskdale
  • Former County Dumfries-shire

Archaeology Notes

NY27NW 4.00 21900 75180

(NY 2190 7519) Blatobvlgivm Roman Fort (R) (Remains of)

(NY 2180 7513) Roman Fort (R) (Site of)

OS 6" map (1957)

NY27NW 4.01 NY c. 219 751 Birrens Tower-house

NY27NW 4.02 NY c. 219 751 Celtic Stone Heads

NY27NW 4.03 NY 218 753 Field Survey Area

Not to be confused with Birrens Roman temporary camp (NY 2245 7502), for which see NY27NW 5.

For inscribed stone (presumably from Birrens) found in a house at Middlebie (NY c. 21 76), see NY27NW 19. For stone head from 'Middleby' (presumably Birrens), see NY27NW 38.

A Roman complex of which the only extant feature is a fort, 'Blatobulgium' (from the Antonine Itinerary) (OS Roman Britain map 1956), or 'Birrens' (from the local place name). It is best known by the latter name, and is scheduled under it.

The fort was excavated in 1895 and again in 1936-7 and 1962-7, and has been examined from the air. Miss Robertson's findings on the investigations up to 1967 establish the following elements and phases:

1. A probably Agricolan enclosure to the W of the fort, with its E side overlain by the W side of the fort. An air discovery.

2. A late 1st. or early 2nd century (? Flavian) fort, coincidental with the extant fort, except that its N rampart is some distance short of that of the latter.

3. The extant fort. An Antonine work of two phases, the second a rebuilding in AD 158. No occupation beyond the 2nd century has so far been established. (But see NY27NW 10, where it is said that a coin of Maxentius (306- 312 AD) was found here in 1934).

4. A triple ditched enclosure on the W side of the fort and probably an annexe to it, mapped by Roy, noted in the Soc Antiq Scot report (D Christison 1896) and re-identified from the air.

5. Crop-marks to the NW of the fort, discovered from the air and identified by St. Joseph as (1) a courtyard building, probably a mansio, and (2) the ditches of either a temporary camp or fort annexe, also many adjacent marks of ditches, pits, and hollows whose significance is not immediately apparent.

Finds here include a sculpture of a horned head, depicting a Celtic deity, now in the National Museum of Antiquities of Scotland [NMAS}, and another female head, possibly depicting a local deity probably comes from here (A Ross 1967). Also, there is a stone slab, inscribed CISTVMVCI LO(CO) MABOMI (Gift) of Cistumucus from the Place of Maponus, though it is not clear whether this mention of Maponus merely gives the origin of the dedicator, or implicitly includes the deity as recipient.

A S Robertson 1964; A S Robertson 1965; A S Robertson 1966; A S Robertson 1967; R P Wright 1968

Only Antonine pottery found in excavated well.

A S Robertson 1969

Birrens annexe (NY 217 752). The following items were recovered during field walking over a period: coarse wares (various) 5143 frags; Samian 1539 frags; grey ware 539 frags; amphorae 517 frags; white ware 121 frags; mortarium 107 frags; lead 55 frags; coins (including 1 silver) 3; bronze 3 pieces; glass beads 3; also, iron (including nails).

S M Mason 1987.

Cast bronze cheekpiece, toggle or 'slider' (Macgregor no. 42) measuring 9 by 1.6cm; in poor condition.

Cast bronze platform terret with enamelled decoration (Macgregor no. 65), measuring about 6.4 by 4.8cm and in poor condition.

Cast bronze platform terret (Macgregor no. 66); a pair to Macgregor no. 65, and in similarly poor condition.

These objects were found (with a martingale of apparently Roman type) in 1936 during the excavation of a burnt deposit (site VII, level II) which probably represents a native uprising prior to 155AD. They are held in Dumfries Museum.

M Macgregor 1976.

A length of wall and doorway was discovered eroding from the bank down towards the Mein Water in the SW part of the annexe. The wall is c 3m W-E and stands for the most part 0.75m high, of long, thin, roughly dressed blocks of local sandstone, showing some signs of having been mortared. The doorway has a large (c 0.7m) stone slab as a threshold with two other stone slabs forming the other surviving two sides.

It is possible that this wall can be connected with one of the two sub-rectangular buildings shown on Roy's plan of the fort and annexe area (Military Antiquities of the Romans in Britain, plate XXIV); if this is the case then the remains discovered should be the N face of one of these two structures, the other three sides having been eroded by the Mein Water.

Sponsor: Manchester University Art History & Archaeology Department.

N Lockett 1996.

Listed as Roman forts (NY 2190 7518), temporary camp (NY 216 753) and mansio (NY 2166 7528). Native finds include pottery, bronze cheekpiece, toggle and two terrets.

RCAHMS 1997.

NY 2198 7530 In July 1999, the Dumfries Team of the Environment Task Force, under the supervision of the Solway Heritage archaeologist, installed a stile and waymarker to improve public access to Birrens Roman fort. The work necessitated the excavation of five post-holes, 30cm square and 40cm deep, located close to the field gate at the NE edge of the field in which the fort is situated. No archaeological deposits or artefacts were noted.

A report has been lodged with the NMRS.

Sponsors: Dumfries and Galloway Council, Dumfries and Galloway European Partnership, Landfill Tax.

R Toolis 1999.


Publication Account (1986)

Birrens was the fIrst Roman fort in Scotland to be extensively excavated (in 1895), and, excavated again in 1936-7 and 1962-7, it has proved itself the most informative of all Roman sites in south-west Scotland. It is the only major fort in the region where there are unmistakeable, if slight, visible, remains, and it is the only one known to us by its original name, Blatobulgium. Its inclusion in the Antonine Itinerary (a Roman road map), where it is named, reflected its function as an outpost fort of Hadrian's Wall, 14 Roman miles on from Netherby (Castra Exploratorum) in Cumbria. The road through Birrens was the main western route into Scotland.

What one sees today is most of the central platform of the fort eroded at the south end by the Mein Water. The rampart is visible as a low mound, but most eyecatching are the corrugations of the six outer ditches at the north end traversed by a central causeway. Visible from the air, but not on the ground, are three temporary camps to the south and east of the fort, a large annexe to the west and, to the north, a large building interpreted as an inn.

Excavation and research have given us an almost complete picture of the building sequence and layout. Laid on the site of a late 1st century fortlet, the Hadrianic fort of the 120s covered an area of 1.65 ha, had a turf rampart, timber buildings, and a large western annexe. In about AD 142, it was rebuilt and enlarged to 2.1 ha in order to accommodate a 1,000- strong garrison of the 1st Cohort Nervana Germanorum, a mixed unit of cavalry and infantry of the auxiliary army. This fort was characterised by the use of stone, which formed the bases of turf ramparts and provided dressed foundations for stone and perhaps stone-and-timber buildings. The barrack blocks were ranged on each side of a central group of administrative units, and consisted of narrow, closelyspaced pairs of buildings instead of the more usual wider ranges. This fort was destroyed and then rebuilt in AD 158 along similar lines. The new garrison consisted of the 2nd Cohort of Tungrians, likewise milliaria equitata. From c 163 for about 20 years it played an isolated role as an outpost of Hadrian's Wall, but it was finally abandoned by about AD 184, presumably as part of a reorganisation of troops in frontier posts.

Finds from the various excavations are in RMS, the Hunterian Museum, University of Glasgow, and Nithsdale District Museum, Dumfries. Over twenty inscribed stones, most now in Edinburgh, have been recovered from the site. These include a dedicatory tablet erected by the Tungrians in honour of their Emperor in AD 158. Three of the surviving altars refer to the two known milliary cohorts, whilst other inscribed fragments testify to the presence of the Sixth legion (Victorious).

Information from ‘Exploring Scotland’s Heritage: Dumfries and Galloway’, (1986).

Publication Account (17 December 2011)

There is a small cluster of camps at Birrens and Broadlea, close to the Roman fort at Birrens, on fairly level ground in a landscape dominated by Burnswark hill to the northwest. A further two camps lie about 1.5km to the northwest at Middlebie Hill.

One camp (I, not shown on illus 86), a ‘Flavian enclosure’, was recorded under the fort at Birrens during excavations in the 1960s (Robertson 1975: 73–5, fig 19). It was apparently filled in before the construction of the Hadrianic fort on the site, and finds from the ditch silting included glass and pottery, all dating to the late 1st century ad (Robertson 1975: 75). The ditch was at least 71m by 21m and measured 2.1m in width and 0.9m in depth (Robertson 1975: 73).

Three other small camps and a possible fortlet were recorded south and east of the fort (across the Mein Water,also see Broadlea), along with a possible fourth camp to the north-west (camp III ), although this may be one of the annexes to the forts (St Joseph 1951: 57–8).

Camp II was discovered in 1948 as a cropmark by St Joseph from the air (1951a: 57–8) lying to the ESE of the fort across the Mein Water. It measures 341m from ENE to WSW by 245m transversely, enclosing 8.3ha (20.6 acres). A titulus is visible close to the centre of the WSW side but none is recorded from the other sides. A linear cropmark with a rounded corner within its interior could represent a further camp, but the evidence is inconclusive.

Camp III consists of the cropmark of a rounded corner and stretches of two adjacent sides just to the north-west of the fort, and was recorded in the 1940s by St Joseph. At least 85m of the south-west side and 25m of the northeast are known, and the site is probably either the remains of a temporary camp or of an annexe to the fort (St Joseph 1951a: 57).

R H Jones 2011

Geophysical Survey (2 July 2012 - 7 July 2012)

NY 21900 75180 The 3.5ha magnetic survey of the field to the immediate W of the extant Roman fort undertaken 2–7 July 2012 recorded the detail of what is probably a substantial part of an earlier phase fort. To its W are a triple ditched enclosure and a probable annexe. The results confirm and amply extend the aerial photographic evidence. Electrical survey to the N of the extant fort in the field by Birrens Lodge indicated some high resistance anomalies, but these are unlikely to be Roman in origin.

The survey, which arose as a result of the former Solway Hinterland Archaeological Remote Sensing Project’s interest in Roman sites in Dumfriesshire, was a collaboration between Glasgow University and staff and volunteers of the Discovering Dumfries and Galloway’s Past project.

Archive: RCAHMS

Funder: Mouswald Trust and Dumfries and Galloway Natural History and Antiquarian Society

Richard Jones, Glasgow University


Geophysical Survey (12 May 2013 - 14 May 2013)

NY 21900 75180 The survey carried out in 2012 to the W of the extant fort was continued, 12–14 May 2013, northwards across a 1.8ha area. The major NW–SE road running through the earlier fort discovered in 2012 (and clearly visible in aerial photographs) extends for c80m stopping abruptly where it meets a large but amorphous-shaped feature (also visible on the aerial photograph). The only other likely Roman feature in this field is a segment of ditch belonging to Birrens Camp III. The remaining numerous anomalies, most of them linear, are probably related to post-medieval agricultural activity. The survey was carried out collaboratively between University of Glasgow and volunteers of the Discovering Dumfries and Galloway’s Past project.

Funder: Mouswald Trust and Dumfries and Galloway Natural History and Antiquarian Society

Richard Jones, University of Glasgow, 2013

(Source: DES)

Geophysical Survey (2015 - 2018)

NY 1860 7870 Burnswark; NY 11329 82085 Ladyward; NY 21900 75180 Birrens; NY 21124 75815 Middlebie; NY 1960 7665 Scalewood

RGK commenced a small survey in 2015 in order to investigate relations between indigenous peoples and Romans on the sites of Burnswark and Ladyward from a landscape archaeological perspective. In 2017, RGK continued work in both places and added two further sites. Birrens and Middlebie were therefore included in the research programme for the area, centring on investigations along the line of the Roman road. A third survey in 2018 continued the successful work at Birrens and included nearby Scalewood Castlefield enclosure as a further possible Iron Age site.

In 2015 and 2017 6.6ha were surveyed at Burnswark. In addition to earthworks in the South Camp, accessible areas in the North Camp were examined as well as an area adjacent to the E of the South Camp.

Coherent anthropogenic anomalies inside the camps that could be referred to as contemporary could not be detected with this method in most of the surveyed areas. Both campsites appear to be void of related permanent architecture that can be traced by magnetometry. Only in the area of the ‘Fortlet' in the NE area and in the W area of the South Camp several anthropogenic features could be detected. Moreover, there are indications that an enclosure, exceeding the limits of the South Camp and of trapezoidal shape is also present in the area of the ‘Fortlet'.

In Ladyward the area of the Roman fort, detected in aerial photography in 1990, was surveyed. Next to a further clarification of the inner structures and defences of the fort, linear structures are visible at the base of the ridge.

At Birrens Roman fort, areas N, W and NW of the fort were surveyed. In the W, the annexe of the fort was surveyed with 0.25m sensor distance. The inner structure of the annexe can be identified. NW of the fort a parcelled area was found, dispersed with 23 strongly magnetic anomalies that might indicate hearths or furnaces.

In 2018 the work at Birrens was continued with a quad drawn 14 sensor magnetometry device 0.25m sensor distance, average measuring every 0.15 m. The field W of the fort was completed and new areas S of Mein River were surveyed in the fields of Broadlee Farm. Magnetometry data of 20.8ha in the wider area of Blatobulgium is now available.

West of the township of Middlebie a field was surveyed in 2017, where detectorists had found several Roman metal objects. The line of a Roman road was assumed, and the hope was to trace its course by magnetometry survey. There were no certain indication of structures of roadworks. However, in the centre of the field, clusters of pit-like anomalies refer to anthropogenic activities that would require further investigation.

In the area of Scalewood prehistoric ‘Castlefield enclosure’ situated on the southern flank of Clint Hill 600m NNW of Castlefield cottage was surveyed in 2018. The magnetometry – though badly blurred by very strong magnetic anomalies – at least partly revealed a double-ditch with several interior and outer structures of anthropogenic origin in the SE part of the surveyed area. A circular ditch of 6.5m diameter is visible in the N. The N part of the enclosure is not visible in the magnetometry. The boundary between the two fields containing the enclosure showed a brim, the surface of the S field was up to 0.5m lower than the N area. The abrasion of topsoil in the S field might support the invisibility of the enclosure here.

All gradiometer works of 2015 and 2017 were conducted with a GPS-supported 5-sensor fluxgate GM650B gradiometer device. The survey in 2018 was conducted with the similar technical configuration of 14 sensors mounted on a quad drawn 3m wide rig. Complementary augering was conducted in 2015 Burnswark and Ladyward.

The works are carried out in collaboration with Trimontium Trust Melrose, Archaeology Service - Environmental Planning Dumfries and Galloway Council and Historic Environment Scotland.

Archive: NRHE and Communal Archaeology Dumfries and Galloway

Funder: RGK and Trimontium Trust

Ruth Beusing, John Reid, Axel Posluschny, Nina Dworschak – Römisch-Germanische Kommission, Frankfurt (RGK) of the German Archaeological Institut (Deutsches Archäologisches Institut

(Source: DES Vol 19)


MyCanmore Image Contributions

Contribute an Image

MyCanmore Text Contributions