Bar Hill Fort Roman Fortlet

Antonine Auxiliary Fort (AD 138–161)

Bar Hill an Antonine Fort  – Aside from the outline of its defensive earthworks, the only visible remains here are the foundations of the principia in the fort’s interior and a linear bath-building sandwiched between the defensive ramparts and the Antonine curtain-wall, the two being detached.

[Antonine] Bar Hill (Fig. 8) is a squarish fort of 3 acres, lying close to the Antonine Wall. It has a single ditch towards the Wall, double on the other three sides. The berm measured from 6 to 8 feet ; the rampart was of turf and was built on a 12-foot stone foundation. The four gates were about 12 to 14 feet wide, single, and built of timber. The two most exposed to attack were protected by tituli (cf. Brough-by-Bainbridge, Fig. 7). At the corners were traces of towers or artillery platforms in the thickness of the rampart. Inside were found the usual headquarters and granary and another building in stone, and a stone bath-house laid close against the north rampart. There were no certain traces of other stone buildings, and the barracks were of wood. The underlying Flavian fort has been described above (G. Macdonald and Park, The Roman Forts on the Bar Hill, Glasgow, 1906).” (Collingwood, pp.46/47)

The Bar Hill fort measures 375 ft. east-west by 369 ft. north-south (c.114 x 112 m) between the ramparts, giving an occupation area of almost 3¼ acres (c.1.3 ha). The fort is detached from the rampart wall itself, lying about 20 – 30 yards to the rear of it. The only other Wall fort to be built separate from the Wall was at Old Kilpatrick. A consequence of this was that the fort was defended nu a single ditch on its north side, facing the Wall, and double-ditches elsewhere. There are four gateways in the ramparts, those on the north and south placed centrally in their sides while those on the east and west are displaced slightly to the north; the gap in the ramparts to the west is not matched by a corresponding causeway across the ditches. There is a short titulum outside the ditch causeway on the south and a longer earthwork protecting the entrance gap on the east. The Military Way passed to the north, between the fort rampart and the Antonine rampart wall.

The fort was excavated in 1902-5 and 1978-82, during which a number of animal bones were uncovered, including those of Ox, Sheep, Pig, Red Deer, Boar and Fox; the latter animal very likely being hunted and killed for sport and as a means of pest control, certainly not for food. Other finds of note include a number of leather shoes, of men, women and children. Among the artefacts recovered from the interior of the fort was a wooden wheel with an iron rim, the hub of the wheel made of elm, the spokes of willow. It is likely that this wheel, and fragments of another three found in rubbish pits at Newstead, were taken as booty or tribute from the lowland Caledonian tribes. Wheel ruts worn away by cart-wheels such as these were noted in one of the fort gateways during excavations.

Possible Agricolan? Fortlet

Our last Flavian site shall be Bar Hill. Here, on the line of the Antonine Wall, the ditches of an earlier fort were found while digging was going forward in a second-century castellum ; and as it was clear that these earlier ditches had silted up naturally and become overgrown with brushwood before the later fort was built, it was inferred that they belonged to one of the castella which Agricola built on a line between the Forth and Clyde.¹ The fact that no objects of any kind (except one shoe) were discovered, was regarded less as invalidating this conclusion than as showing that the occupation was a very short one. The ditches were from 8 to 11 feet wide and on average about 4 feet deep. Their ground-plan was somewhat complicated ; but it appeared certain that the fort had occupied a rectangular space measuring 180 by 145 feet, with an annexe on the west. Some traces of an earthen rampart were found, but no stone foundation ; if the rampart was 10 feet wide, the internal dimensions would be 160 by 125 feet, or nearly half an acre. Nothing is known of the internal buildings, but a gap about 15 feet wide in the ditch indicates the position of a single gateway.” (Collingwood, p.35)

  1. Stated in Tacitus’ Agricola chap.23; see Mollins.

The Roman occupation of Bar Hill was possibly first established during the campaigns of governor Agricola sometime around 80AD. A small fort of less than ½ acre was found beneath the Antonine Wall fort. It was enclosed by a defensive system consisting of a narrow rampart perhaps 10 feet in width and two (or more) ditches between 8 to 11 feet wide and around four feet deep. There was a single gateway through these defences, near which a sandal from the late-1st C. was found in the ditch infill. The occupation period of this fortlet appears to have been very short, perhaps only a single winter, abandoned when the scene of action had moved north to the Tay. Having said all this, the fortlet lying beneath the Antonine fort at Bar Hill and another at Croy Hill, are now both thought not to be Agricolan, possibly not even constructed by the Roman military, perhaps even by the native inhabitants prior to the coming of Rome.

The Garrison Units at Bar Hill Fort

RIB 2171 - Building Inscription of the Second and Twentieth Legions

Detachments of the Second Legion Augusta and of the Twentieth Legion Valeria Victrix built this.


No commentary. Addenda from RIB+add. (1995): These detachments may have also inscribed a column-shaft at Bar Hill: see note to RIB 2312.

RIB 2173 - Distance Slab of the Twentieth Legion

For the Emperor Caesar Titus Aelius Hadrianus Antoninus Augustus Pius, father of his country, a detachment of the Twentieth Legion Valeria Victrix built (this work) for a distance of three miles.


The provenance is uncertain. Sibbald assigns it to Cadder Manor, probably because Camden cites it with the Cawder House inscription; others assign it to Balmuildy. On the basis of the distance which it records Sir George Macdonald places it somewhere east of Auchendavy.

RIB 2170 - Honorific Building Inscription of the First Cohort of Baetasians

For the Emperor Caesar Titus Aelius Hadrianus Antoninus Augustus Pius, father of his country, the First Cohort of Baetasians, styled Roman citizens for valour and loyalty, (set this up).

T AE[  ...  ...]TONINO
AV[...  ]OH
I B[...  ] R OB

For the recognition of this inscription as an original dedication of the first period on the Antonine Wall see Birley loc. cit.

RIB 2169 - Altar dedicated by First Cohort of Baetasians

The First Cohort of Baetasians, Roman citizens, (set this up).


As Macdonald (loc. cit.) observes, this must have come from the chapel of the standards.

RIB 2166 - Altar dedicated to Mars Camulus

To the god Mars Camulus the soldiers of the First Cohort of Hamians …

[...]ILITES COH [...]

No commentary. Addenda from RIB+add. (1995): Keppie (Scotland) notes the unusual band of chevrons on the capital which also occurs on RIB 1778, an altar dedicated by a prefect of the cohort at Carvoran in the period A.D. 136/8.

RIB 2167 - Altar dedicated to Silvanus, with plinth

To the god Silvanus Caristanius Justianus, prefect of the First Cohort of Hamians, gladly, willingly and deservedly fulfilled his vow.

[...]EO SILV[...]
[...]OH I HAM[...]

For the family of the Caristanii at Antioch in Pisidia see Cheesman, JRS 3 (1913) 253, Birley, Roman army (1953), 163, 169. For coh. I Hamiorum cf. RIB 1778, 1792, 1810 (Carvoran), 2172 (Bar Hill).

The First Hamian Cohort were the only regiment of auxiliary archers in the entire Roman army of Britain, recruited from a tribe native to Syria. They are recorded at other forts in Britain at Carvoran (vide RIB 1792) on the Stanegate also at Housesteads on Hadrian’s Wall, both sites in Northumberland.

RIB 2172 - Funerary inscription for Gaius Julius Marcellinus

To the spirits of the departed (and) of Gaius Julius Marcellinus, prefect of the First Cohort of Hamians.


Presumably been taken from the cemetery outside Bar Hill fort.

Of the nine inscribed stones recorded in the R.I.B. for Bar Hill, five are altarstones. These contain dedicatory inscriptions to the popular Roman gods Apollo (RIB 2165), Mars (RIB 2166) and Silvanus (RIB 2167), also one dedicated to an unknown god (RIB 2168; not shown) and another altarstone containing only the name of a garrison unit (RIB 2169).

RIB 2165 - Altar dedicated to Apollo

To the god Apollo ..

[  ..]C [...]
[.] S [6]

Wrongly stated to have been found at Falkirk and bought at Birrenswark (Cat. (1849), followed by Huebner), or ascribed to Croy (by J. Macdonald).For a modern inscription on the capital of the back see RIB 2332*.

Pottery marked with the stamps of eight Antonine potters has been unearthed at the Bar Hill fort: Avitus Form 31, Beliniccus Form 31, Cinnamus Form 37 (3), Divicatus Form 33, Felicio Form 27, Geminius Form 33, Malluro Form 31 and Peculiaris Form 31.

The Numismatic Evidence

Forty Roman coins have been recovered from the Bar Hill fort, the most yielded from any fortification along the Wall. These range from a denarius of Mark Antony found with 12 other coins in the well of the principia courtyard, to a bronze coin of Gordian III, also including 11 of Hadrian, 10 of Trajan (8 of which were found in the well), and 9 coins which cannot be classified.

RIB 2312 - Inscription

For the Emperor Caesar Titus Aelius Hadrianus Antoninus Augustus Pius, father of his country, the detachments ..


See RIB 2171 for a dedication by vexillationes leg. II et leg. XX at Bar Hill; see RIB 2313 for a milestone set up by a military unit.CIL vii 1110a is recorded only by James, Earl of Morton, who died in 1786. He reads (Morton, James, Earl of., Soc. Ant. Scotland Morton MS. no. 6, not now, 1953, available) imp . caes . t. aelio | adria … antino | avg . pro p . p | vexellinvs. As some items in this MS. are not derived from autopsy, e.g. RIB 974, 2172, and as Horsley records this inscription as being in Baron Clerk’s possession at Penicuik, it seems likely that Morton’s description of its being in the garden-wall of the manor at Kilsyth is only derivative, and that this is a doublet of CIL vii 1109, of which ll. 3-5 match Morton’s ll. 2-4. Huebner, on the other hand, CIL vii 1110a, regarded this text as a distance-slab, and Macdonald (RWS (1934) 406 n. 2) concurred.Antoninus Pius, A.D. 138-61. Addenda from RIB+add. (1995): But Keppie, loc. cit., points out that a milestone of this period would normally bear details of the emperor’s regnal year and consulships. Despite RIB 2313 it is exceptional for a military unit to be named on a milestone.

Classical References to Bar Hill Fort

The Ravenna Cosmography lists BEGESSE as one of the places in a line at Forth-Clyde “neck”. The first is VELUNIA (Carriden), the next is like Mumrills. The sixth and seventh are MEDIO (Balmuildy) & NEMETON (Old Kilpatrick). This left BEGESSE amongst three other forts to be assigned to locations. These have been assigned to locations based on size and early occupation which meant that Bar Hill would be Begesse. For more see article on: Nemthur.

Welsh Gaelic Old English Other
NA G:beag, OI: becc (little)
OI: barr (top)
OI: ess, esso(waterfall)
OI: geis (taboo)
began (to bow, bend, turn)
Beges-ig (Beg’s Island)
Berg (hill)
+ æsc (ash tree), æsce, æxe (ashes), axe (axe)
Latin: beccus (beak)

Bar Hill appears to derives from either Welsh Bar or Gaelic Barr meaning hill-top, so it literally means “Hill Hill”. So with an ending like Early Irish geis (taboo) we could consider “Taboo hill”. In Old English “esse” is the ending in a number of words like Ælmesse (alms), hægtesse (witch) & Lindesse, which is a shortening of Lindes-ig (Lind’s Island). Therefore one etymology would be “Beg’s Island”. There are raised areas like Nether Inch and Inch Less, that were effectively islands along the Kelvin valley. Likewise Old English began (bend) could refer to a river bend. Irish “beag-ess” seems possible with a translation of “little waterfall”, but the ridge on which the Antonine wall runs, does not have significant streams flowing down it. However (like many places) there are some small streamlets on bar hill. The easiest to match to the topology of Bar Hill is Old English berg (hill). If so it could be Berg – Aesc or “Ash hill”

References for Bar Hill Fort

  • The Archaeology of Roman Britain by R.G. Collingwood (1930);
  • The Roman Wall in Scotland by Sir George MacDonald (Oxford, 2nd Ed. 1934) pp.271-285;
  • The Roman Inscriptions of Britain by R.G. Collingwood and R.P. Wright (Oxford 1965);
  • The Roman Military Diet by R.W. Davies in Britannia ii (1971) pp.122-142;
  • The Roman Occupations of Scotland by B.R. Hartley in Britannia iii (1972) pp.1-55;
  • Chronology of the Ancient World by E.J. Bickerman (Thames & Hudson, London, 1980);
  • Air Reconnaissance in Roman Britain 1977-1984 by G.S. Maxwell & D.R. Wilson in Britannia xviii (1987) pp.28/9;
  • Chronicle of the Roman Emperors by Chris Scarre (Thames & Hudson, London, 1995);
  • A Survey of the Coin Finds from the Antonine Wall by Richard Abdy in Britannia xxxiii (2002) pp.189-217. 

Map References for Begesse

NGRef: NS 70746 75926 OSMap: LR64

Roman Roads near Begesse

Antonine Wall: W (2) to Avchendavy (Strathclyde) Antonine Wall: E (2) to Croy Hill (Strathclyde)

Sites near Bar Hill Fort Roman Fortlet