The Kanovium Roman Fort and Vici Reproduction

Drawing by David Swarbrick

The New Reconstruction of The Roman Fort and Settlement at Caerhun

The reconstruction, the work of David Swarbrick shows the fort and vici of Kanovium as it would have looked around A.D. 150, and I say ‘would’ rather than ‘may have’ looked because the drawing is based on features securely located from a number of sources.  These are Paul Kenneth Baillie Reynolds 1926-9 excavations of the auxiliary fort and south vicus, air photography which revealed parch marks, this work commencing during the mid 1970s droughts, but continuing on up to the present day by R.C.A.H.M.W, and finally the certain obvious features seen on the Environs Project Geophysics study of fort vici by Gwynedd Archaeological Trust during July 2002.  The above reconstruction is also the result of liaison between David Swarbrick and The Kanovium Project.

The Geophysics Environs Project by Gwynedd Archaeological Trust of the Kanovium vici has shown us the extent of activity in the north vicus (top) and the lack of activity in the south (bottom) therefore this has allowed David Swarbrick to create a vivid impression of how the ‘castellum’ which was the name the Romans gave to these small forts, and the associated civil settlement looked 1850 years ago.  While the Chardon Trust Museum at Llandudno used to feature a reconstruction of the site, the knowledge of the vici was then not as advanced as today, therefore the up to date information from G.A.T and David’s accurate perception of the ground allows us to view Kanovium in a way nobody has been able to since the Roman army evacuated the site around the mid to late 2nd Century A.D.

The site is viewed looking north, around the reign of the emperor Antoninus Pius, A.D. 140, though the fort has looked similar to this for around 30 years, when around 110 A.D the then emperor Trajan, or at least his representative in the new province of Britannia, the Governor, gave the command that these Flavian conquest forts, these frontier posts, built initially in clay and timber around A.D. 75-80, should be made permanent by the addition of a stone curtain wall and stone inner fort buildings.  At Caerhun fort the process seems to have been done ‘piecemeal’ and rather slowly, the important inner buildings such as the principia or headquarters complex, centre, (according to Roman fort building tradition) probably received the first stone footings, but soon after the work commenced on the gates of the fort.  The gates seemed to have been constructed on a clockwise fashion, work commencing at the north gate, top, closely followed by the east gate, right, these two gates displaying typical Hadrianic features and proportions, the north gate guard towers are recessed into the rampart, while the ones of the east gate stand slightly flush of it.  The south gate, bottom, followed, maybe twin guard towered in its timber phase (A.D 78-80 to 90-100) post holes remained to prove this, possibly around A.D 120 the proposed plan to build on the same dimensions was changed, and the result is the curious narrowed roadway, single towered gateway, this is discussed elsewhere in this website, but it was curious, unusual and was not satisfactorily untangled by Baillie Reynolds.  Finally the gate, west, was completed in stone, not long after the south gate, as it is very similar, but contains no evidence for the  changing features so evident at the south gate.  It has been proposed that these single tower gates are more akin to gates in forts (and Roman towns in Britain) during the 3rd century A.D.  During the period of gate refurbishment the angle towers were also constructed, all four can be seen, ten feet square, likely of fine cheshire sandstone for the outer course,  one is now covered by the church lane, and top right by the churchyard wall and graves.  These can be seen to be detached from the fort wall, the reason for this is unknown as it is not readily seen in other Roman forts ,while butting up of walls and towers was common, these towers were not, they are totally free standing.  The reason for this could be simply because when these towers were constructed the proposed line of the new stone wall was not yet decided on, as the timber period east rampart was 13 feet further west.  They are only joined to the rampart by the packing clay bank behind it.  No interval towers had been detected by Baillie Reynolds.  Finally at the end of the refurbishing stage the curtain wall is added, early in the reign of Pius (recent thought is that it may indeed have commenced slightly earlier) the clay bank is cut back 6-8 feet and a fine wall of red sandstone is leaned against the old timber fort rampart.  A panic sets in not long after (a slight layer of ‘quick’ silt was detected in the ditch) when the fort wall could be seen to be slipping into the ditch ,the berm or interval between rampart and ditch, was too narrow to support the new wall and half the ditch is filled in with a pack of hard clay, this then obliterterated the central ditch midrib.  During this time the area of the south retentura (barrack 1, in the main, bottom right) received a packing of clay from 0-1 feet thick N-S, this levelled this quadrant of the fort, and the inner buildings stood level, while the fort rampart follows the contour of the hillock upon which it stands.

The North Vicus

 We see Kanovium in the mid 2nd century A.D, the stone phase completed, an extensive vicus can be seen north of the fort, much industry is taking place at various locations around the settlement. The vicus development features include not only strip buildings fronting on to the very wide north road, but also higher status courtyard buildings, these can be seen, one of these buildings is nearly as large as the principia, which itself is abnormally large, featuring three ranges of rooms around a central courtyard, this can be seen above the fort corner, top right.  These courtyard buildings are thought to be ‘mansio’ or official post stations, individual employees of the Roman postal service could stay overnight and receive a change of horses free of charge at these buildings, which usually featured a private bath-house, and may also have served as tavernas or inns.  The book Roman Frontier in Wales, M.G. Jarrett, Cardiff 1969 claims that at the time of writing no mansio buildings had been detected definetly in Wales, though they were suspected tentavively at Bryn y Gefeiliau and more likely at Brecon Gaer, Leintwardine, and Llandovery auxiliary forts, and also at Caerwent Roman town.  Therefore the discovery of such large buildings at Caerhun suggests that one must indeed be possibly a mansio, and this is significant.  Several buildings of this design are to be seen fronting  the north road, and these could equally be private accommodation for government officials, in charge of tax collection, this feature was noted at the Caernarvon fort, but in a late 2nd century context after a garrison withdrawal, and  inside the fort wall.  They could also be tavernas, private houses or brothels.  The high status of these buildings could imply that the settlement had actually had some status conferred on it, its occupants, possibly it had become a the ‘civitas capital’ of the Ordovices or even Deceangli ‘tribes’? If this is so then the village folk would have become known as ‘vicani’, and the settlement would have been of a higher status than other villages not attached to forts.  If this status had been granted then the settlement would have been subject to the same set of laws as the city of Rome, (John Wacher, The Towns of Roman Britain, Batsford 1983).  Above the rampart top left, N-W, can be seen what is possibly another annex area, possibly conjectural, this level patch of ground was ideal for the parking of carts, stacking of firewood, tethering of animals, or even camping ground for passing army units etc, of which there was absolutely no room inside the fort.  Enclosed by a slight bank and ditch, probably just to delineate it from the surrounding ground.  The north road is seen to be very wide, J.K.S St Joseph (air photography)  had the  idea that this was to create a market place, while the density of buildings here certainly would have been meant a flourishing economy, we will never be sure unless excavation can shed light on this.  The road is seen to run down the hillock toward a small brook, and suddenly reduces in size drastically as it is seen to run through a collection of tombs, several are quite large and must represent burials of officers from the garrison, though it was also normal to return cremated ashes to other parts of the empire to be disposed of by the soldier’s family.  Also fronting on to the north road are strip buildings, probably shops or workshops, each has possibly a large hearth in the street front rooms, these could possibly be forges and it is likely much time was spent in metal working, lead ore being readily available locally, and copper and iron ores several miles east around Abergele and Llandudno.

The South Vicus

Bottom of the reconstruction can be seen the south vicus, this area has no such occupation as the north.  An annex to the fort proper was constructed soon after founding in A.D 78-80, initially it had a slight rampart and ditch, but this was never fully completed, present then was a roughly built wattle and daub structure, in which smelting of metals took place.  A branch road can be seen entering the annex via the main south Roman road, indicated by Baillie Reynolds, utilized in the reconstruction as a cart parking area, beneath which can be seen an area of cobble foundation, again verified in the excavation, this appears to be for an annex rampart that was never constructed.  It is interesting to note if this rampart had been constructed then a gate would have been needed to gain access to the annex.  By A.D. 140 the structure may have vanished, but the annex was still used, a rough structure can be seen fronting onto the south road (conjectural), the annex was used as and when it was needed, and the south, 2nd century road (laid slightly above the former 1st century causeway) can be seen to kink slightly west to avoid the now filled in annex ditch, which can be seen to have never been fully completed.  The main fort ditch here is crossed by a wooden bridge, as it had been in the timber phase, (abutments in the form of large boulders remained south of the ditch lip, uncovered in the 1929 excavations) the single towered gate can be seen, the guard tower eight feet back from the front of the new stone fort wall.  West of the south road is an empty area, though there is a Roman structure just beyond the fort corner, this was located by Baillie Reynolds and equally in the Geophysics survey.  It must be stressed that because Geophysics has not located known features does not mean they are not still extant so this whole area may not have been blank at all.

East of the fort wall can be seen the River Conwy, from the east gate a rough gravel track takes a devious route down the steep slope to the bath-house,  The bath-house can be seen to be a large complex building which features later additional ranges, most being hot rooms.  The reconstruction clearly shows why no east road was necessary, there was nowhere for it to go other than to the river marsh, the ditch was crossed by a gravel bank, this explains the gravel path, which surprisingly was not slabbed, which would be expected as the bath-house had a long life and it would be thought a decent surfaced path would have been a high priority.  It seems likely that a path was also provided around the fort corner and ditch here, to join up at the north gate, David Swarbrick has conveyed this, and indeed it is likely from evidence west, that other such paths were provided around the fort outer perimeter.  Several main fort drains exit into the fort ditch above the bath-house (south gate, and S-E corner) therefore the ditch is shown as waterlogged, which is how Baillie Reynolds interpreted this as back in the 1920’s sections of ditch soon became flooded after excavation, west gate.  The baths are shown to be on an excavated terrace just above the flood plain of the river, it is likely that the bath-house area would have been protected by a dike, a dike is still visible today, but the construction date is unknown.  The dike leads down river, north to the area of the fort dock.  The dock is not proven to be of Roman date, but indeed in the opinion of Baillie Reynolds after sectioning it was that the earliest level (one of five) was Roman, a brushwood and clay foundation having typical Roman characteristics, seen in the section E-T at the S-W rampart of the fort, in 1929, so it is safe to add this to the reconstruction.  The dock is lined by buildings, possibly storage buildings, one is to be seen on the south quay of the dock, this is visible as footings today. From the dock to the area of vicus beyond the north gate was a link road, Baillie Reynolds could find no trace by trial trenches, but it certainly seems present on the 2002 Gwynedd Archaeological Trust Geophysics work, so again this is added.  Interestingly the 1880 Ordnance Survey map of Caerhun shows a footpath ran here down to the dock, which often happens with Roman routes, and this is what may be appearing on the Geophysics Survey.  West of the dock can be seen a line of trees, these represent the small brook which issues from springs (rises)  far left top, this brook which is still active despite several modern drainage attempts represents the boundary of the north vicus, these springs would almost certainly have been used by the Romans as a clean supply of water, quite possibly an aqueduct (typical examples are constructed in wood or stone slabs and could run under or over ground, Dorchester, Prestatyn being examples).  It was known that the praetorium had a private bath-house and latrines, and also the external baths would also have need a good strong water supply.  Drainage work in 2000 produced Samian ware and fragments of amphora, beyond this is the probable cemetery plot.  From the west gate a road issues, again illusive in 1926-9, but clearly visible on air photographs, in common with the east gate it has no bridge, the ditch being interrupted by a gravel crossing bank.  This was added at a late hour as David Swarbrick had not shown a west road until I viewed an air photograph, Baillie Reynolds concluded no road was present here, it eventually seems to deviate north west and may be avoiding the fort parade ground.

The Central Buildings of the Roman Fort
 
Baillie Reynolds’ excavations are our principal source for the evidence for the fort plan, and inner buildings proper, air shots of crop marks show additional features not on the plan, and are the typical examples seen in a classic Roman fort of the 1st and 2nd centuries A.D.  Though the gates, and, angle towers are not typical this does not alter the general perception gained.  All the fort was stripped except the church area, but the ends of the barracks and store block in this sector protruded beneath the graveyard wall, and along with Roman predictable building practices in fort design it was possible to add these to the plan.  What do we see in the fort reproduction?  The fort contained seventeen buildings, we look directly north down the via principalis road and left, west, of is the main range of buildings, three, the private residence of the unit commander, the praetorium, while roughly centre is the command centre for the fort, the headquarters building, or principia, beyond this, north, is the fort forage store, horrea, or granary, the example here is double (with a covered yard between) David Swarbrick has been able to include the buttresses present on these buildings walls.  Right, or east, of the via principalis, can be seen the ‘retentura’ or ‘buildings in front of the tents’, this being archaic Roman terminology from the construction of camps when all structures were leather tents, but the main range was known as ‘the tents’.  We see six barrack blocks here, roofed in red imbrice and tegula  tiles (many found by Baillie Reynolds) the smaller ends of these front onto the via praetoria, this road from the east gate runs right up to the entrance arch of the principia.  Fronting on to the via principalis are two slightly narrower long buildings, VIII, and X, use unknown. VIII, bottom, which in the timber phases contained a water cistern, which may have been utilized by the baths, or maybe it is a water tank for horses and is a stable, this could equally have been a smithy as metal dross was located during the excavation of this building.  Building X, top is now beneath the modern carpark, but Baillie Reynolds dug the then present surface up, and found this building was of greater build quality than its bottom counterpart, and may have been a ‘fabrica’ or fort workshop, or another  storage building.

The area of the pretentura (or ‘behind the tents) which is behind the central range are slightly more problematic, the top two are certainly barracks XIX, XVII, as they feature the ‘centurial thickening’ at the end of each block, the bottom two have been often called stables, but a porch is present on one, again this  usual in centurial quarters, so Baillie Reynolds may have failed to locate the thickened portion on building XVIII, bottom left, with porch, XVI is the next building right, or east.  What David Swarbrick has very clearly shown here is the closeness of the praetorium to stable or barrack XVI, this must surely represent space being taken up by the additional courtyards being added to this building at unknown dates, if the building had been laid out like this purposefully then it is another indication of the awful building practices undertaken by the classic period Roman army.  It is current to look through ‘rose tinted spectacles’ in regard to the ‘High Empire’ and attribute bad design to the 3rd and 4th Centuries, but here we may see this is different.  The slight gap between these two buildings was never fitted with a path (it is at most three to four feet wide) and soon became a rubbish dump, much pottery being excavated by Baillie Reynolds, some of which may never have been catalogued.  Constructed within the clay packing bank of the fort wall can be seen cookhouses bottom right, top left, much ash and oven floors being found to prove this fact, while above the south cookhouse can be seen an oven, this built securely above the ‘intervallum’ roadway of the timber fort.  It would seem the intervallum road had indeed not been constructed all around the inner circuit of the rampart in the stone phase.  Known as ‘cloak street’ or via sagularius because of the high winds being channelled by the narrow streets and buildings? (E. Shirley, Building a Legionary fortress, Tempus 2000) Reynolds claimed none existed in this south area, he indicates the north are may have been fitted with one, but his trenches in regard to this aspect seem to come as rather an after thought in the excavations and only took the form of trials.

The forage store can be seen to be standing in an area of free space, close to the north gate and possible dock road, this was to allow produce to be temporarily stacked, before being removed into the building, also to make room for large carts to maneouvre, but also it may also be to place it away from other buildings with the risk of fires, as David Swarbrick illustrates, fires are a big factor in a Roman fort, the baths had to be kept fired up all day long, every day of the year, barracks, principia, and areas in the streets have provided hearths, and we already know the extent of the industrial work being undertaken in the north vicus.  The praetorium is large and overgrown and the principia is shown to be maybe a three floored building, the highest point providing ‘chlerestory type lighting’ for the standards and Imperial insignia below in the chapel of the standards.

I am sure you will agree David Swarbrick’s reconstruction is a graphic indication of how a Roman fort looked, what we have to remember is that the majority of Kanovium was robbed down to the foundations, so to actually reproduce it to this standard is in this author’s opinion no mean feat, and I would like to take this opportunity to thank him for his care and consideration in the undertaking of this reconstruction.