Cramond Roman Fort: 60 years of excavation and research

Roman Denarius coin
Roman Denarius coin

Report from the Cramond Roman Conference, Saturday 3rd October 2015 – First Supplement

By Ian Hawkins.

John Lawson opened the conference by outlining sources – John Wood 1794, A & V Rae 56-70, Nick Holmes 75, EAFS, Headland Archaeology (Church Hall) and, AOC – Campus 2003 and Fort 2008.

Rebecca Jones spoke about the Antonine Wall (first lecture by HES – 3 days old) Forts had been defined by John Gillam as primary and secondary some were bonded in and others free-standing – all had annexes Geoff Bailey has defined two Antonine occupations? Current theory is that Forts and fortlets were chosen in advance based on indivisibility and strategic location, but were not all built at the same time. Subsequently these were linked by the rampart all fortlets are primary Duntocher fortlet is retained in the annex cobbled over. Is Rough Castle secondary? The turf superstructure and wall are contemporary. Surveying camps. (Rather than temporary camps) may underlie Bar Hill and Croy Hill Bearsden is a key site – and may have had a potter? North African pottery styles as proposed by Vivien Swan. The original design altered as wall build developed to suit topography and strategy. Fortlets were built with regard to topography, not distance Cramond and its harbour were in a strategic position with good visuals and a practical, port.

Martin Cook spoke on A&V Rae’s early research, rescue, followed by pre-development and
targeted research. Evidence was of 2 occupations. There are 3 main publications Rae, Nick
Holmes and Church Hall. NH 75-81. No harbour element has been found. An extramural
settlement within a large annex was a manufacturing area, possibly to supply Carpow. Bath
house excavations in 1975/6 excavated 80% of the structure which could be dated to the Severan occupation. Charlie Hoy’s minor excavations of 2 roads etc. between 1978-86 found Road 1 to E of fort within annex and a well stratified sequence of timber buildings. 2nd phase Antonine was possibly German troops? The Campus excavations between 2000/3 exposed a road, gate, ditches but no cemetery, a much larger area of annex identified multi-phased road, gatehouse with a large highly formal gate – a Severan feature. The 2008 community excavation was in advance of limited conservation, with public access, 60 volunteers, and 1000 visitors. The NE corner of fort, blocks A and B, Granary and E gate were investigated. Some floor surfaces survived the Rae’s excavation These were not barrack, but incorporated workshops / Industrial blocks, a kiln and a
water tank. 3 phases of use were identified. 3 stone boxes set in the floors may have been for craft function, leather or non -ferrous metal working and wood and iron working and recycling of metal. The Granary & E gate was truncated. In conclusion Cramond was a most investigated and least understood fort.

Nick Holmes reviewed the numismatic evidence in a wider context. Wood had reported many coins of gold & silver with little wear (PSAS 1918) including an aureus of Geta “Republican” coins found at Cramond are all 18thC discoveries and therefore uncertain. Some Flavian and republican coins are all well-worn and well used indicated a prolonged period of circulation during the Antonine Trajan and Hadrian era. There was no evidence of late Antonine occupation apart from coins of 161-2 and 164-9 ( Lucilla). Coins of Commodus 183 and 190 were little worn, supporting possible military activity in South Scotland in early 180s. Export of silver bullion was stopped by Severus but there are a few post 193AD hoards eg Birnie, East Coast. A c202 coin was almost unworn. Later coin hoard’s geographical distribution was different – eastern coastal at Birnie and Portmoak. A Carpow inscription refers to Commodus – abandoned earlier. 1 coin of 209 from Cramond others 194 – 207 not worn or circulated If Severus brought coins with him to pay the army (as he is said to have done) they would be 207-208 therefore pre 206AD coins did not come with Severus. Cramond has a dearth of coins of 207-8 Similar pattern at Carpow (200 – 207) recovery indicates occupation pre 208. Kippielaw 31 minted in 206 (Severus and Caracalla) may have been from a military pay chest? Large sum of money hard to square with finds. No coins of 208. Was Cramond supply base not fort. Falkirk hoard 230 late 3rd century suggests payments in Scotland continued so possibly Cramond could have been used as a short stay base by military groups. Coins and pottery from later occupation in Cramond may be visitors, as are later stray finds in gardens.

Discussion points: IOM inscriptions more likely to be Severan. A falling off of coins of the
Severan period is attested elsewhere – even South Shields. All finds of that period tend to be residual from later periods Inveresk dates all end at the Antonine – early 160s AD. But the supply base at South Shields is NOT pre.207 AD David Breeze suggests: Cramond was an operations base in 206s for supplies then it was pulled back to South Shields – so both sites were used quite separately. 206 – Cramond and Carpow under Commodus? Later, things get worse in Scotland until Severus arrives. Cramond and Carpow have similar early Severan coin sequence and a lack of 208 coins. Finally there is no evidence of Camps at Cramond.

Fraser Hunter spoke about the Cramond lioness sculpture – a lioness devouring a captive
barbarian also a carving of snake depicting the survival of the soul. Funerary icon example of Roman art. Lion head exaggerated from earlier tradition of art – monster eating a human is a recurring idea Some areas are more lion-rich than others Britain, Germany Pannonia, Dacia – lionesses rare mostly lions – Corbridge. Britain/Germany sculptures depict wild animal’s boar/deer/human in association with lions. Dacia/Pannonia/Belgian sculptures use sheep/cow/goat = domestic animals Stone is white sandstone analogous with Inveresk, therefore a local stone but difficult to source however. The sculpture was created on the spot.

Some areas are less well finished – not unfinished and the carving may have been painted. There is a recess behind the head – for a lamp? The lioness was a private commission by a high ranking legionary or auxiliary officer, someone expecting to stay using locally sourced stone not dated but likely Severan or Antonine.

A monument for eternity and permanent conquest. As the river was much wider the statue would have been nearer the centre of the river. The lioness was too powerful to leave behind so was deposited in river until 1997.

Dawn McLaren has been investigating the Charlie Hoy assemblage of metal stone and glass collected between 1976 and 1990 from 12 sites and 2 roads which continued in use to 18th
century. The road was 2. 8m wide and the ditch and road showed 5 phases. There was a
limited but rich array of artefacts – Manning’s type 10 dome-headed hob nails, copper alloy sheet fragments, iron slag and an unfinished sandstone pillar 1m long with tool marks – a locally sourced rough-out reused within a pit , copper alloy and glass beads, bangles (4 are Kimmeridge shale from Dorset coast same as BB1 pottery. One made of cannel coal from local source was larger than average – property of a soldier?). Glass in small sherds indicating possible recycling?

Military equipment included sheet iron artefacts – plate armour as found at Newstead also mail and one stone sling shot ball. There was a noticeable lack of weaponry. Other finds were horse fittings, strap ends, open work mounts 2 copper open mounts, and denarii of Severan date from pit. An exceptional item and a find of international significance was a miniature weapon pendant from a ring pommel sword. Similar finds are concentrated in Upper Germany – to mark out a beneficiarii of Upper Germany. Only 4 others are recorded – from Silchester, Germany, Hungary, and Tunisia. Coins suggest a Severan date. The Cramond example is silver, all others copper alloy. The collection is quite staggering in size and adds unique and significant insight to picture of military Cramond.

Paul Bidwell discussed the pottery assemblage – There were only a very few 1st century sherds, a moratorium fragment from Elginhaugh c/f type 43 evidence of Flavian activity in vicinity of Cramond. Perhaps an Elginhaugh fort/early Cramond Fortlet link? 2 early Antonine groups, 1 redeposited and another from Block A were predominantly BB1 and 2, more than half from outside Scotland. Sources were SE England, Horningsea, North Kent, and Colchester. Dorset produced BB1. E England (Colchester) produced BB2. Most Cramond pottery is dated from 140 to post AD160. There are Problems with identifying later occupation. It was unlikely that there was a full time occupation between Antonine and Severan. We cannot use pottery evidence for a 2nd Antonine occupation. Samian ware mostly 2nd half of 2nd century, eg contemporary with Hadrian’s Wall. No Thames side wares on Antonine Wall but common at Carpow. BB2 also changed to plain product of Thames estuary. Evidence of change is probably after 180 based on BB2 from turrets on Hadrian’s Wall. In the Severan era there were 300 potters producing several million pots
per year in East Anglia in the Nene valley. Most mortarium are Mancetter- Hartshill late after 210. Amphora 85% by weight olive oil from Southern Spain – no wine amphora – wine could have only come in barrels. Amphora stamps are tricky to study – 19 from Carpow and 20 from Cramond dated 209 – 224 fits well with military campaign. North African style Severan – casseroles, tulip shaped bowls. Vivien Swan demonstrated they are quite close copies of N African types. Historical conclusions remain doubtful. None found at all at Inveresk, some at Barr Hill and Bearsden, several at Cramond and the few Carpow items may have come from Cramond. Military occupation in the 1st half of 3rd century in Lower Nene Valley Pottery suggests a military 250-270 cut off. 1 sherd of late 4thC pottery and 1 fragment of hair (not grass)-tempered pottery is recorded.

Adrian Maldonado spoke on the Britons in the immediately post Roman period – if you were not Christian you were not Roman and if you were not Christian enough then you were barbarian. Bede lists barbarians as Britons, Gododdin, Meatiae Picts Caledonians Anglo Saxons (Bernicia).

There are large cemeteries but not much bone survival. Sites with few graves most common, very rarely 50 – 100 or more. Greater density of long cist burials in Lothians very few S of Tweed Examples of grave goods in Lothians at Dalmeny Anglo Saxon style necklace, re-used Roman glass. Continuum of different burial rites – re-used henges ring ditches etc. String graves in W-E aligned long cists. Ancestral landscapes are a historical re-invention to consolidate a local political identity – indigenous folk recreating their own identity as at Thornybank. Roman forts were generally avoided – Little Kerse, Polmont, the Catstane 5/7th century shale cists or sandstone as at Thornybank. Roman masonry from Inveresk lightly furnished. N – S orientations, the cist itself is the grave goods using significant stones collected and gathered together. Roman sites – Hadrian’s Wall / Antonine Wall are not reused. The Catstane is carved to venerate a mythical ancestor representing the underlying prehistoric burial cist – the long cists are then laid around it. Cemeteries were meeting places, engines for creation of Political identities. And later coalesce into the tribal areas.

John Lawson talked about the excavation of the Bath house in 1976. 2 phases were noted. 9 adults buried in a latrine pit or drain with very mixed disturbance was thought to be probably plague deaths under a medieval midden. Now all have been dated to 2nd half of 6th century. As no plan survived drawings were produced from photographic evidence showing a range of bodies aligned roughly N – S. In the adults 2 broad phases were definitely sequenced by the insertion of large stones which sealed some burials Pathology revealed a whole range of blunt force trauma, sword wounds, some healed. Extreme violence but no decapitation. What happened? DNA analysis is planned in the next 2 years to see if they are related? Isotopic analysis revealed diet was mixed with no marine. Mobility – most were from local area. Two warriors showed healed earlier sword wounds. One is a primary burial – apparently born in Skye area. One from Peeblesshire and one from Strathclyde (Skye, Eigg, and Rum) Primary burial may be political ties. Facial reconstruction of body 4 and Murder victim (5) displayed clear blunt force trauma was the
cause of death, possibly by a spear butt. 2 other women had blunt force trauma, an adult female thrown in Burial 1 had healed sword injuries. Burial No 7 was similar. Why were they buried?

Tradition is single burial in cemetery. Did Cramond have special status, perhaps a crypt? There were several neonates – dates were not established for these. The Bath house was re-used in early medieval and cleared of windblown sand. A small Roman stone altar inscribed “pro car” was found 50 yds away on the foreshore. Was the Bath house re-occupied as a chapel – Urbs Iudeu (Bede)? The Granary bodies found in 2008 date to 9/10th century.

David Breeze in conclusion: The early Roman military invasion (Flavian) was a distinctly inland affair – they wanted to conquer the whole island so aimed north and west to get into the Highlands as efficiently as possible. They weren’t interested in the coast so no Flavian at Cramond. Later military activity wanted to protect the coast – Antonine. Bearsden was originally a rectangular fort but later split into a square fort with an annex – He spent a long time looking for the principia centrally inside the square fort until Geoff Bailey pointed
out it was central to the rectangular one. Cramond has similar anomalies – Two granaries in the central range, No commanding officer’s house – Soldiers were here for a purpose. Was there a gap between Antonine and Severan occupations? Cramond has two annexes- was the new Severan phase not an annex or fort?

A return was made to Cramond when Julian was Emperor – he knew Cramond existed.
Something was pulling the Romans back to Cramond.