NGRef: NY329591
OSMap: Hadrian's Wall, LR86.
Type: Wall Fort, Minor Settlement.
Roads
Wall: ESE (5¼) to Vxelodvnvm (Stanwix, Cumbria)
Stanegate: E (6) to Lvgvvalivm (Carlisle, Cumbria)
Wall: W (3½) to Concavata (Drumburgh, Cumbria)
Stanegate: W (6) to Kirkbride (Cumbria)

Aballava - The Orchard

The area around Burgh-by-Sands is dotted with Roman military encampments, which were all placed at this strategic location to guard two nearby Solway fords, frequently used by raiding parties from the northern tribes, especially the Selgovae to the north and possibly also the Novantae in the north-west. Aside from the Hadrianic fortifications there are two earlier auxiliary forts and a number of marching camps. The Burgh village church is built from stones taken from the Wall, and marks the location of the southern defences of the fort. Evidence of a small civilian settlement or vicus has also been found outside the fort's south-eastern defences.

Evidence From Classical Sources

The name of the fort appears first in the Notitia Dignitatum of the early-fifth century, wherein is listed the station Aballaba, between the entries for Petrianis (Stanwix, Cumbria) and Congavata (Drumburgh, Cumbria). The Burgh-by-Sands fort also appears in the seventh-century Ravenna Cosmography as Avalana (R&C#153), between the entries for Uxelludamo (another name for Stanwix) and Maia (Bowness on Solway, Cumbria).

The name is recorded in 1292 as Burg en le Sandes, and before that simply as Burch (c.1180). The origin of these names obviously stems from the Old English burh, meaning fortification or stronghold, its more modern name also referring to the location of the old Roman fort among the sandy dunes of the Solway estuary.

The Epigraphy of Aballava

There are eleven inscriptions on stone recorded in the R.I.B. for the Burgh-by-Sands fort, eight altarstones and three damaged tombstones. Only one of these stones can be dated, to the middle of the third century. All of these inscriptions are shown and translated on this page.

The Military Installations

Burgh-by-Sands Fort III
The Stanegate Fort

N.G.RefDimensionsArea
NY 3170 5866 603 x 370 feet
(184 x 113 m)
5.13 acres
(2.07 ha)

Situated ¾-mile (1.22 km) WSW of the Wall fort and settlement at Burgh-by-Sands, at Hill Farm, Longborough, a large, two-phase fort was first seen on aerial photographs in 1977 and confirmed by excavations conducted by G.D.B. Jones in 1984. Pottery recovered from the site during these investigations indicated an occupation sometime during the first half of the second century AD. The site has been occupied by two successive forts which very likely indicates a change in the type or size of garrison unit:

The fort is seemingly aligned with a Roman road which runs along the humpback ridge of Fingland Rigg to the auxiliary fort at Kirkbride. This road passes directly in front of the fort, and probably represents a westward extension of the Stanegate frontier system sometime during the Trajanic period; the fort, by association, may also belong to this system.

Burgh-by-Sands Fort I
Early-Hadrianic Fort

N.G.RefDimensionsArea
NY 3237 5824 475 x 450 feet
(145 x 137 m)
4.9 acres
(1.98 ha)

Aerial photographs taken by G.D.B. Jones in 1975 revealed the distinct outline of a Roman fort on top of the commanding hill about 0.6 miles (c. 1 km) SSW of the Burgh-by-Sands Wall fort. The site was later confirmed by excavations conducted in 1978-79 which suggested a construction date around the late-Trajanic / early-Hadrianic period. It is generally accepted that this fort was built as part of the initial scheme for Hadrians Wall, but was, for some reason, built ½-mile to the rear of the original Turf Wall. The fort exhibits at least three phases of construction and appears to have been preceeded by a watchtower sited upon the crest of the ridge.

The original fort was protected by a single ditch backed by a revetted rampart of beaten clay set upon a base of large riverine cobbles about 16 feet (c. 5 m) wide, which was later widened to about 24 feet (c. 7.5 m). These defenses measured 475 feet NE-SW by 450 feet transversely (145 x 137 m) and occupied an area of 4.9 acres (1.98 ha).

An annexe was later attached to the south-east side of the fort, measuring about 460 feet by 395 feet (c. 140 x 120 m), thereby adding an extra 4.1 acres (1.68 ha) of occupation space; these defenses appear rather less impressive on Aerial Photographs than those of the fort itself. The excavations conducted by Barri Jones in 1978-9 revealed the existence of timber buildings in both the fort and the annexe, but the distinct outlines of a stone-built praetorium and a granary building were recorded on APs in 1984.

Burgh-by-Sands Fort II
Late-Hadrianic Fort

N.G.RefDimensionsArea
NY 3288 5917 c. 565 x 445 feet
(c. 173 x 135 m)
c.5¾ acres
(c. 2.3 ha)

The Wall fort was evidently a late addition to the original Hadrianic plans because turret 71B which originally occupied the site (approx. NY 3289 5914) had to be demolished before the fort could be built. The fort was then built astride the line of the Turf Wall to avoid marshy ground to the south, with the Stone Wall being re-aligned to incorporate the fort's northern defenses, as was normal for Wall forts intended to house infantry garrisons.

The site was partly excavated in 1922 by the Cumberland and Westmorland Antiquarian & Archaeological Society conducted by R.G. Collingwood. It would appear from pottery recovered that this stone fort on the line of the Wall was built well into the Hadrianic period, very-likely at the same time as the barrier wall itself was here replaced in stone, continuing to be garrisoned until the mid-to-late-4th century.

The dimensions of the fort are not known with any accuracy, but the north-south dimension has been estimated at somewhere between 550 and 580 feet (168 - 178 m), due to the finding of the Roman Military Way in the vicarage garden. The eastern gateway was positively located during Collingwood's investigations in 1922 just to the east of the church (at NY 3290 5908) and the western defenses are marked by a distinct drop in the road near the crossroads. This would suggest an east-west dimension in the region of 430 to 460 feet (130 - 140 m), and an occupation area anywhere between 5½ to 6¼ acres (2.2 - 2.5 ha).

Burgh-by-Sands Watch Tower

N.G.RefDimensionsArea
NY 3244 5825 c. 65 feet
(c. 20 m)
c. 368 yards²
(c. 314 m²)

Also shown on the AP's of Fort I is a 19m wide circular cropmark lying close to it's south-east gate. This feature has been identified as a Roman watch tower similar to those employed along the Gask Ridge in Tayside. The defenses consisted of a circular, v-shaped ditch 7½ feet wide by 6 feet deep (c. 2.3 x 1.8 m), backed by a timber-fonted rampart varying in width between 13 and 15 feet (4 - 4.5 m); a single gateway lay to the south-east. These outworks surrounded a centrally-placed, square, four-post timber tower which has been dated by the finding of black-burnished ware pottery shards in one of the main postholes to about AD120, when this type of pottery fist began to appear. This timber watch-tower was not in service long before being demolished and replaced by the Burgh-by-Sands I fort on the same site atop the hill.

The Garrison Units

The first known unit at Burgh by Sands is the Ala I Tungrorum, a five-hundred strong cavalry force enlisted from amongst the Tungri tribe of Gallia Belgica (Belgium). An inscription (LS514) recovered from the fort attests the presence of this auxiliary unit at Aballava sometime during the second century.

I O M COH I NERVANA GERMANORVM MIL EQ CVI PRAEEST Q PIVS G F CLAVD ASINIANVS TRIBVNVS¹
"To Jupiter Best and Greatest, Nerva's First Cohort of Germans, one-thousand strong, part-mounted, under the command of the tribune Quintus Claudius Asinianus, devoted son of Gaius."
(RIB 2041; altarstone)
  1. The dedicators name may also be read ARRIVS G F CLAVD VRSINIANVS.

Cohors Primae Nervana Germanorum - The First cohort of Nerva's Germans undated altarstone dedicated to Iupitter Optimus Maximus (RIB 2041).

Epigraphic evidence has been recovered which places the Cuneus Frisiorum among the fort's garrison during the reign of Caracalla (Imp. AD198-217). By the reign of Philip the Arab (Imp. AD244-249) the Frisians had been moved to Derventio (Papcastle, Cumbria; RIB 882/3), where they took the title Aballavensium from their previous station. Other Frisian cunei have also been identified at Vercovicium (Housesteads, Northumberland; RIB 1594) on Hadrian's Wall and in its hinterland at Vinovia (Binchester, Durham; RIB 1036).

I O M ET NVMINIBVS AVGGG N MAVRORVM AVR VALERIANI GALLIENIQ CAELVIBIANVS TRIB COH P P N S S INSTANTE IVL RVFINO PRINCIPE
"For Jupiter Best and Greatest and the Divine Spirits of the three Augusti, the Numerus Maurorum Aurelianorum, for Valerian and the Gallieni,¹ Caeluibianus, tribune of the Cohort, Praepositus of the Company, fulfilled this undertaking in the presence of the commander-in-chief Julius Rufinus."
(RIB 2042; altarstone; dated: AD253-258)
  1. Valerian ruled the empire from October AD253 until he was captured by the Persian king Shapur in June 260; he later died in captivity. His son Gallienus rulled the empire jointly from October 253 and became sole ruler upon his father's capture, continuing to rule until he was murdered by his soldiers outside Milan in September 268. This text also obliquely refers to Valerian's younger son also named Valerian - not Gallienus as stipulated here - who was never invested with any constitutional powers, and died of natural causes in 258.

By the time of the joint reign of Valerian and Gallienus, Burgh-by-Sands was garrisoned by the Numerus Maurorum Aurelianorum, as attested on an altarstone to Jupiter Best and Greatest (RIB2042; dated: AD253-258). The Notitia Dignitatum also places this unit at Aballaba.

Praefectus numeri Maurorum Aurelianorum Aballaba
"The prefect of Aurelian's Own Company of Mauri at Aballaba"
(Notitia Dignitatum xl.47; 4th/5th C.)

The Gods of Aballava

Altarstone of the Demi-god Hercules
HERCVLI ET NVMINI AVG COH ...
"For Hercules and the Divine Spirit of the Emperor, the Cohort [...]"
(RIB 2040; altarstone)
Altar Dedicated to the God Latis
DEO LATI LVCIVS VRSEI
"To the god Latis, Lucius Ursei [dedicates this]."
(RIB 2043; altarstone)

Dedicatory inscriptions have been found near Burgh-by-Sands in close proximity to an image of a horned god, a Germanic deity named Belatucadrus, who was associated by the Romans with Mars. There are four altars to Belatucader (altarstones RIB 2038/9 and 2044/2045; vide infra), one of which is shared with the Roman war god Mars (RIB 2044 etiam infra), two dedicated to Jupiter Best and Greatest (2041; 2042, dated: AD253-258), and single altarstones to Hercules (RIB 2040 supra) and Latis (RIB 2043 etiam supra).

Altarstones Dedicated to the War God Belatucader

InscriptionTogo-TranslationRIB
DEO BELATVCA For the god Belatucader." 2038
DEO BELATOCADRO ANTR AVF POSVIT ARAM PRO SE ET SVIS "For the god Belatucader, Antr[o] Auf[idius?]¹ placed this altarstone for himself and his family." 2039
MARTI BELATVCAD SACT MATVSI "To sacred Mars Belatucader, Matusius [dedicates this]." 2044
BALATVCADROS CENSORINVS POR SALVTE ET SVORVM POS "For Belatucader, Censorinus has placed this for the well-being of his family." 2045

The Vicus or Civil Settlement

"At Burgh-by-Sands, buildings of an extramural settlement were seen [during aerial survey] to east of the fort." (St. Joseph, 1951)
Damaged Roman Tombstone of Julius Pius
... IVL PII... TINVS CIVES DACVS
"[To the shades of the departed ...] Julius Pius [...]tinus a native of Dacia."
(RIB 2046; tombstone)
 

There are three Roman tombstones recorded for Burgh-by-Sands, all of which are damaged. That of a Dacian (RIB 2046 supra), probably a soldier, and two other fragmentary texts; D M S ... "To the sacred spirits of the departed [...]" (RIB 2047), and another inscribed merely VII "seven" (RIB 2048).

Other Sites of Interest

The Marching Camps

There are a number of temporary marching camps in the area, four to the east at Grinsdale and one at Beaumont nearby. The proximity of the Beaumont camp to the Aballava fort suggests that it may have housed the work-force which built the fort itself, though this is not proven.

The Wall and Vallum

The site of MileCastle 73 has been identified overlooking the Burgh Marsh from the western flank of Watch Hill at Dykesfield, about ½ mile west of the Hadrianic fort. The vallum also ends here on the edge of the Burgh Marsh, 66 miles from its origin at Newcastle, but reappears again to the west of the Drumburgh fort over three miles away across the salt-marsh, to continue the final three miles to the Wall's western terminus at Bowness-on-Solway.

Roman Military Earthworks

AP's have revealed the line of a palisade and ditch, running along the ridge between Burgh village and the ford over the Sandwath at Sandsfield. It is possible that the watch-tower identified within the defences of Fort I may be associated with these earlier entrenchments, which evidently predate both of the forts. It is possible that these features represent part of a temporary north-western defence zone established during the second campaign season of governor Gnaeus Julius Agricola (AD79).

A Bit About Robert the Bruce

Close to the line of the Roman earthwork about a mile north of the village is the King Edward I Monument, marking the spot where the English king's forces were defeated in battle by Sir Robert Bruce near Old Sandsfield in 1307.

Aballava Today

Burgh-by-Sands
Admission Free Car Parking Nearby Pub
The site of the Hadrianic Wall fort at Burgh-by-Sands now lies mostly beneath the church-yard in the eastern part of the modern town where, unfortunately, nothing remains to be seen, its component stone-work having been robbed-out over the centuries to build the town itself.

Aballava Bibliography

See: Hadrian's Wall From the Air by G.D.B. Jones & D.J. Woolliscroft (Tempus, Stroud, 2001);
Hadrian's Wall Map and Guide by the Ordnance Survey (Southampton, 1989);
Air Reconnaissance of North Britain by J.K. St. Joseph in J.R.S. xli (1951) p.55;
The Roman Inscriptions of Britain by R.G. Collingwood and R.P. Wright (Oxford 1965).
All English translations, including any inherent mistakes, are my own. Togodumnus

Aballava Related Lynx