A History of Caerhun Roman Settlement

© Kanovium Project/Ken Evans

The Roman army first entered North Wales in the year 48 AD, led by the governor of Britannia and leader of the army, Ostorius Scapula.  Tacitus, a contemporary author, records that the Roman army defeated the local tribe, the Deceangli (the name surviving as the name of Roman Chester, Deva) and ‘reached the sea facing Ireland’, which must be Rhyl or Colwyn Bay.  The reason why the Roman army was in Flintshire is unrecorded, perhaps the Deceangli had become troublesome, probably because the army had started to exploit Flintshire lead and silver deposits, using many local men as mine workers.  Scapula, has to leave quickly, as the Brigantes of Yorkshire, had rebelled against Rome.  AD 60 the Romans return, this time under the command of Suetonius Paulinus, Paulinus a stern general, recently returned from fighting mountain tribes in Mauritania, had been given the job of annexing Wales, a task seemingly beyond other governors.  Anglesey, or Ynys Mon, the Mother island, was seen as the power base of a religious cult, called ‘Druidism’, usually Rome tolerated other religions, but not in this case because the Druids practiced ritual slaughter of humans, these Druids had fled from Gaul during Julius Caesars’ conquests, and upon Claudius’ conquest of Britain, had fled to Anglesey.  Tacitus records that the Druids lined up on the beach near Caernarvon and poured curses upon the terrified Roman army, who refused to ford the shallow Menai Strait, in the meantime, Paulinus sends in his Batavian auxilia, an aquatic people from Holland, who engage the warriors and priests on the foreshore, the Druids are destroyed, their sacred sites, woodland groves, are burnt, but as before the army has to leave unconsolidating their gains, this time the Iceni of East Anglia rebel, Paulinus flees to aid the Roman cities under siege, St Albans, Colchester and London.

Rome has big problems with Icenian revolt and Wales is left unconquered until the principate of Flavius Vespasianus, AD 78 the governor Julius Frontinus finally annexes South Wales, and turns his attentions to North Wales, the conquest spills over into the term of Julius Agricola, who returns to Anglesey and gives the local folk a good kicking, just for good measure, even though they declare  they are already beaten
 
 
  
 It is during Agricola’s conquest that Kanovium was probably founded, the excavator P.K. Baillie Reynolds, could not find specific evidence to prove this fact, so it could have been late in the term of Frontinus.  Baillie Reynolds thought the site had some sort of Celtic structure present when the Romans arrived, on the basis of a La Tene III fibula (Llandudno Museum) having been found nearby.  The Conwy valley must had had a large population, people had lived in the area for many thousands of years, and there are sites from the Mesolithic, bronze and iron ages, and many large hillforts that the Romans would sack later.  The tribe west of Caerhun, in the Snowdonia mountains, the Ordovices had not been easy to conquer, having destroyed a cavalry ala, a high status regiment just at the end of Frontinus’ period, so the Romans put a chain of forts to hold this mountain region, interestingly, Kanovium faces east towards the river crossing.  The fort is established in earth and timber, and covered an area of just under 5 acres, the number of barrack blocks suggest it was for 600 soldiers, the mixed foot and horse cohort ‘cohors quingenary equitata’  The aim of the fort was to watch the river crossing which had been at this point since the Neolithic period.  This trackway was formalized by the army and became the main road between Chester and Caernarvon, not only could the locals no longer cross the river, but they couldn’t use the road either.  The Conwy river, tidal up to this area allowed the Romans to bring in most supplies and men by ship, and established a docking complex, which remains today.

Once established, the fort would have taken many months to build, the auxilia, for whom it was built, would have maybe done some of the donkey work, we are unsure how much construction work the conscript troops would have took part in at this time, cutting down the large mounts of timber, usually oak, but the main construction would have been done by the Chester legion, legio II Adiutrix, who moved out of the province during 86.  The tree cover in this part of the valley would have been radically stripped, and the local farming folk would have to have supplied the 600 soldiers with food and fodder for the considerable amount of animals which would have been present, at least 250 horses for the cavalry, plus mules and oxen.  The garrison would have been engaged in tax enforcement on the local population, who not only lost their freedom, travel rights and most of their land, but now had to pay taxes, possibly at 90% if they had cost the Romans much time and loss of manpower, and this could be enforced for 150 years.  There is evidence that in North Wales, the Romans struggled to encourage the people to use their coinage, previously they had no coin economy, unlike southern England, and Roman coins are only recovered in a significant civilian context in the 3rd and 4th centuries.  Recently an archaeologist, informed me that the catchment area for the supply of food for the garrison would have extended outwards in a circle 5 miles wide, and that if there had been other forts in the vicinity then the circles will almost have touched.    Rob bases this evidence on work he has done in Northern England and the fort at Brough on Noe.  The local people obviously had been made to be aware that things had changed, possibly some of the men will have been conscripted into the army and moved out, this is basically a no win situation for the people of the Conwy valley.

The fort had a clay rampart 2 metres tall, about 20 feet thick and topped with a breast work which must have raised the wall to about 15 feet, surrounded by a double ditch, and entered by 4 gates, 2 with twin towers 2 with single, though possibly on post hole evidence all four gates at this time had twin entrances without guard rooms.  Attached to the south rampart was an annex, which contained, strangely, a Celtic roundhouse and much evidence of metal working, a vicus or civilian village would have been established, which by the mid to late 2nd century had become very large.  The timber fort was replaced about A.D.100 with another one, due to rotting of the timbers, and this period represents the greatest military presence at the site, from the Flavian to early Hadrianic periods.  Most recorded coin finds are of these dates, dropping off radically at Hadrian, even though the pottery continues.  Around A.D.150 the fort is given a rebuild in stone, which was red sandstone shipped in from Chester, however G.A.T informed me that the Grace Simpson pottery survey which claimed this date now has to be discounted, and that the stone wall is actually earlier. Eitherway the stone phase will have been constructed by the legio XX VV, who used tiles recording this title in the bath-house of the fort.  The inner fort ditch is filled in, with a very hard spread of clay, so as not to undermine the wall, which is laid against the clay rampart which was cut back 6 feet.  It seems that the inner ditch was not filled in right away as it exhibited some build up of silt, the berm must have been too narrow and the wall started to slip down into the inner ditch.  There are a few other problems here too, the fort has detached corner towers which suggest had been built before the rampart, all the gates are different, one is recessed into the wall while one projects beyond the wall, the 2 single gates appear more 3rd century, though the excavator records no structures of this date in the fort.  The question of the detached towers is easy to interpret, as the timber rampart was 13 feet smaller at the east face, it seems that the towers had been built to this specification, once the change of plan was brought forward it seems the towers were left were they were.  The inner buildings, also became stone, the high status buildings using the Chester stone, while the barracks used the local shale.  There is evidence that the fort wall and gates had a decorative cornice, and some stones had the decoration known as ‘dentils’ which are square projections, still present in the city walls of Tarragona, Spain, and Jerusalem.  The principia was quite large and ostentatious, having dwarf columns and small stone eagles as statuary, but it appears the work was not completed, and some columns never reached the site.  There is no evidence of coins towards the end of the 2nd century, occupation is slight but continuous, ending about AD 200

AD 200, the fort is demolished, the headquarters is partially demolished, some of the decorative stonework is flung into the well in its courtyard, and much other material which had to be dug out by the excavators.  Once clear of the rubbish the well apparently refilled with clean water, but only to a depth a 9 inches, which possibly suggests that the well was never completed as it was not lined with masonry and radically narrowed as depth was gained.  This destruction is down to a garrison withdrawal, the soldiers torched their own forts, some areas of the fort such as the principia were sacred and could not be left for the Celts to defile.  This is probably down to the Roman army withdrawing from Britain to fight for a usurper, Clodius Albinus, Albinus was defeated and the victor, Septimius Severus, sends the depleted garrisons back to their forts.  Caernarvon has provided an inscription which records work been undertaken to repair a damaged fort under the Severans, so possibly Caerhun also had work done, though judging by the pottery evidence it was confined to areas around the principia.  Either way, a cookhouse is inserted into the rampart under Gallienus, during AD 250, if a garrison was present it would have been smaller, barrack sizes had been reduced by a half, noted at Caernarvon and most forts on Hadrian’s wall, at this time, a trickle of coins and pottery leads on to AD 390.  There is an idea that the site became a civilian tax paying centre due the large size of the praetorium, but a small garrison may have been retained to guard the Great Orme mines from Irish pirates, a fort being established at Holyhead on Anglesey for the same reason.  During the early 5th century all official Roman input into the site would have ended, possibly the local people may have continued some form of Roman lifestyle, as mid 5th century tombstones recording Roman officials have been found at Penmachno.  Possibly squatters lived in the fort, or perhaps a local chief may have used it as his base.  By the 13th century the site became a quarry for St Marys Church, being largely constructed out of Roman stone.  The area south of the Roman dock became a medieval fish trap, and there are slight remains of post medieval farm houses on the site.  One was excavated on the south west corner of the rampart and produced pottery of a 17th century date.  Walking the site produces pottery of a date 1500- 1750, so people must have been living around the ruined buildings.  The dock was used to ship in stone for Caerhun Hall, which had a small rail track to carry the stone, which like the Romans used, was red sandstone, one block of the 18th century stone remains at the Roman dock.

Kanovium was excavated by P.K. Baillie Reynolds of Manchester University, between 1926-9 and provided the reasonable plan on page 2 of an auxiliary fort, it is, interestingly the second best plan to have come from Wales.  This was a time of great excavations, R.E.M Wheeler was excavating Segontium at Caernarvon, J.P. Hall, also from Manchester did some work on the fort at Capel Curig, which he named Caer Llugwy, and local archaeologist W Gardner did some work at Pen y Corddyn Mawr, Llanddulas, which has recently produced evidence of 5th century Roman military equipment, which is rather odd, and Pen y Gaer the hillfort overlooking the site.  Baillie Reynolds was brought in as the local council wanted to extend the graveyard into the south area of the fort, this, thankfully never happened, and the graveyard had to wait until the 1960’s and it went several hundred yards south of the site.  He did a very complete job and excavated the entire fort, except the churchyard, some of his plans are a bit lacking as he failed to record or see walls that still appear as parchmarks, particularly in the barracks.  This however could be because constraints were placed on him in both time and were he was allowed to dig.  The fort has had other excavations, in the 1790’s the antiquarian Samuel Lysons excavated the north gate, which he viewed as ‘twin sepulchers’ and found Roman pottery and human bones, now the site of the car park gate, he also excavated the bathhouse, very well preserved, still with floors of pilae.  The site saw some limelight in 1995 when two re-enactment societies commenced a ‘sponsored walk’ from the site which was covered by national network TV.  I have made this site for four reasons, I believe Kanovium is under exposed, tourists are not familiar with it,  it is interesting and well worth a visit - ask the birdwatchers, also it comes under threat from mundane routine farming work which is wrong, and that it introduced me to a remarkably interesting hobby.

The author wishes to point out that the specific periods at Kanovium are not as yet correctly understood especially the timber to stone phase(s), evidence is looked at differently by different experts, so while Baillie Reynolds believed the stone wall to be built under Trajan and Simpson thought Antoninus Pius, we can safely say it took place sometime between these Emperors

The fort happily recieved some much deserved attention during 2002, the much discussed Geophysics survey by the Gwynedd Archaeological trust can be viewed via the following link. or you can see the results   at Kanovium Geophysics.  During two weeks of July the fields had been marked out with red pegs and the area surveyed using  Geophysics machines relay information to a computer which then can display a plan of the underground features.  In a short matter of time the survey can show information that a Edwardian archaeologist could never hope to discover, that is while it might have taken them four years to plot the ground plan of the fort buildings at Kanovium, Geophysics can indicate this in a matter of weeks, and with no digging.  The survey has taken knowledge of the vici here into the 21st Century, and it was with the greatest pleasure that this author could add this information to this website.