The Results of Caerhun ‘Environs Survey’ by Gwynedd Archaeological Trust 2002

© Gwynedd Archaeological Trust.  This image must not be subject to reproduction, transmission, display, lending or storage in a retrieval system without prior permission of copyright holder.  Surveyors David Hopewell and John Burman.

During July 2002 a Geophysics Survey of the vici, or civil settlements, was undertaken at Caerhun Roman site, it was done by the local archaeological unit Gwynedd Archaeological Trust (this will be abbreviated to G.A.T) who kindly allowed me to visit during the survey and have also supplied me with the results and have allowed them to be included in this website.  The Geophysics Survey is conducted using machines called ‘Geoscan Fluxgate Gradiometers’, these handheld units which have two sensors which are periodically activated while the operator walks in a controlled grid pattern over the field, measure variations in magnetic field (flux density).  The results are then transferred to computer, and a printout is then produced, almost a map of the underground features.

G.A.T have now kindly allowed me to publish the survey result, and even to discuss the result of the survey findings.  The features are quite well defined so please enjoy viewing what the Survey produced.
 
 
Regular readers of Kanovium Project will be aware that back in July 2002 following an approach to G.A.T I was invited to Caerhun fort by David Hopewell during the several weeks that the Geophysics survey was being undertaken.  David Hopewell, along with John Burman were responsible for the survey, and I was pleased to be able to discuss some issues regarding both the current Geophysics and matters concerning both Caerhun and Roman forts of North Wales in general.  I was particularly pleased that our forts are actually receiving some attention.  Firstly some back ground to the Gwynedd Archaeological Trust Environs Survey of the Vici at Welsh Roman forts.  G.A.T outline the proposed work on their website, including some interpretations of results.
 
 
http://www.heneb.co.uk/GAT_site/Projects_99-00/roman_forts/roman_forts.htm

 The ‘Environs Survey’ funded by C.A.D.W, will mean all the Welsh auxiliary forts are to receive a ground survey of the ‘environs’ outside of the ramparts of the actual military establishment.  Formally work on a Roman fort would only extend to the area enclosed by the defensive rampart and ditch(s),  and in particular buildings such as the headquarters (principia) and  commanders house, (praetorium) often evidence has been ‘blurred’ by earlier excavation in the 19th and early 20th century, therefore the ‘vici’ areas are not only unexcavated, due to the fact the old antiquaries, either had been unaware of their existence or were not interested, so therefore are also an unknown commodity.  This was the state of play in Roman Wales, but excavation at Hadrian’s Wall forts such as Housesteads and Vindolanda have shown that they can be quite complex and extensive, though most had declined by the late 3rd century.  Vici have long been known at Segontium, Tomen y Mur, Caer Gai and Pen Llystyn fort sites.

The vicus at Caerhun fort was first noticed as having some significance during the droughts of the mid 1970’s, and was photographed by J.K.F. St Joseph,  subsequent to the results obtained by Dr Joseph’s air photography, the entire area of the vicus which exhibited remains revealed by the burning of the sun was also declared ‘scheduled’.  Scheduled means that the area of archaeological interest is protected by the British Government, therefore farming, building work, and metal detecting is prohibited without permission from the Secretary of State ; a landowner may conduct activities already engaged in, such as re-cutting existing drains, but cannot conduct new works, deeper, or more extensive, than have formally existed on the site. 

The vicus at Kanovium fort, in common with Bryn y Gefeiliau, and forts further south such as Pennal, and Caer Gai have revealed due to G.A.T.s surveys, that these ‘vici’ have much information to offer us in the early 21st century, especially regarding the extra mural activities of the Roman soldier, or civilians in close attendance, be they possibly the taverns and tradesmens stalls, or evidence of widespread industrial activities.  So it was of the greatest interest to me when I discovered of this proposed new work, an e-mail approach to G.A.T resulted in the very kind offer that I could view the results at the site, over the ten days proposed Geophysics survey at Kanovium.

G.A.T’s aims can be listed here, and are quoted from their website -

‘A survey of Roman fort environs is being undertaken throughout  Wales by the Welsh Archaeological Trusts, funded by Cadw. The survey aims to  improve knowledge about settlement in the areas immediately surrounding Roman  forts and to help to ensure the preservation of any archaeological remains. The  study of Roman forts in Wales has tended to concentrate on the military  buildings but evidence exists, both in Wales and throughout the Roman Empire,  for the presence of civilian settlements (vici) associated with these  forts.

The project aims to assess the form and extent of the military vici at selected sites in Gwynedd. It is intended that this information  will help to ensure the continued survival of features in the environs of Roman  forts, as well as adding to our knowledge of the layout and function of the vicus buildings and perhaps reveal evidence for continued use of the  sites after the abandonment of the forts by the Romans. Recent results from  Llanfor (near Bala) have demonstrated the value of geophysical survey in the  detection of details of Roman forts and their associated vici. A similar  approach will be adopted as the primary fieldwork technique during the project.  Two or three sites have been selected as part of a pilot project to identify the  quantity and quality of the information that can be derived from this  methodology.

A programme of documentary research was carried out, drawing  together available information including plans, references, excavation reports  and aerial photographs of all relevant sites in Gwynedd. All of the sites  identified in the desktop study were then visited, and assessment surveys have  been carried out at the forts of Canovium and Caer Gai with encouraging results.  The results of these initial surveys will allow the final phase of the  fieldwork, consisting of larger areas of targeted geophysical survey, to be  carried out around selected areas of the forts.’

(Text courtesy of G.A.T)

The final result of the Environs Survey is unlikely to spur on any actual archaeological excavation.  G.A.T. informed me that C.A.D.W. do not have a policy of random excavation unless the site is under threat, either from natural or man made activity, which obviously Kanovium Roman site is not in the large scale sense.

David Hopewell informed me the survey would move to the recently discovered Roman military site at Llanfor, at the north end of Bala Lake.  Results for which have been very exciting indeed. They have certainly covered some ground, the fort vicus areas at Caer Gai, Bala, the Capel Curig fort, the fort at Pennal, and Caerhun, and Llanfor, which also included the fort area, are now surely understood to a greater level, and this can only be of excellent news to both the student and author of studies of the Roman army in North Wales.  During late 2003 the results of the survey at Caerhun were added to the G.A.T website, and I quote

http://www.heneb.co.uk/GAT_site/Projects_01-02/Pr_Apr-Sep02.htm#02

Canovium
The survey revealed dense activity extending for 270m alongside the road to the north of the fort. The results appear to show multiple phases of buildings and hearths along with an additional range of buildings to the north-east of the fort. A substantial area was also surveyed to the south of the fort but little Roman activity could be recognised here.  Text courtesy of David Hopewell.

So that is the official opinion of the survey at Caerhun.  Possibly the location was not perfect for Geophysics, the north survey area is rather ‘busy’, possibly the features revealed are post Roman, natural, or modern farming activities in the form of field drains and water pipes.  Also old field boundaries are noticed, as is the modern drive to the holiday cottage.  That is not to say Roman features are not prominent, and in our opinion this is what can be seen in the northern survey area.

A linear feature can be seen to commence shortly from site the north gate of the fort, this corresponds to the Roman north road and holiday cottage track, but only for a short distance before the modern track curves away to the N-E and is a prominent feature.  A broader (and fainter) linear feature can be seen to cross the field which can be seen to be making for roughly the same point as the modern track, to a small brook, this is believed to be the Roman north road.  Its width poses problems, either it was constructed by the Romans this wide to form a market place which was a theory of St Joseph after he viewed the vicus in drought from the air, or simply the gravel road ‘agger’ or embankment has been shifted and blurred by centuries of plough damage.  This feature was noticed by P.K Baillie Reynolds south of the fort when a Roman road was clearly seen excavated to be spread by such activity.  The north road is lined with buildings, many have a black anomaly at their entrance which is conferred to be hearths.  The structures would have been strip buildings, a long narrow building with only the short end fronting onto a Roman highway, space was at a premium and the front would likely be a shop, while the living quarters had been at the rear.  It is interesting conjecture that these hearths allowed prospective customers to actually view the tradesman at work, I have been informed pottery kilns usually occupied rooms deeper into these buildings, and not at the front rooms.  David Hopewell informed me some of the hearths have a diameter of nine feet.  They could equally be pottery kilns, brewing vats, or even pits for ‘fulling’ though the black colour indicates metal working being the most likely, and are known as ‘ferrous spikes’.  As David Hopewell says the occupation of the area is dense, it is difficult to interpret, but the buildings seem to extend on either side of the road.  Several buildings to the east of the Roman road can be seen to be larger and may represent ‘mansiones’ or taverns.  Many black anomalies are seen, some are aligned, in particular a group of seven or eight immediately east of the north gate (a feature also noted by G.A.T during their survey at Bryn y Gefeiliau, east gate, which they noted may have been unconfirmed and untypical cremation burials) but they may simply represent hearths of timber buildings which are now not detected in a geophysics survey.  Possibly that then is the extent of the Roman features.  A conjectural Roman feature is the track which seems to break away from the north road and modern track to descend N-E to the area of the stone boat house, shown on the survey as a rough oblong within a white area unsurveyed.  The origin of the dock close by is controversial, and while the track is shown as a public footpath on the 1880’s Ordnance Survey  map it could date from any period.  If however the dock is proven to be Roman work then this is possibly the dock road.  Incidentally the modern holiday cottage track only originates from the 1970’s, previous to this the line of the of Roman road was followed by walkers, farm traffic and tenants of the cottage.

Natural features are probably also in abundance, for one, a curving anomaly seen top left, can be vouched for by Kanovium Project to be culveting of a spring during  2001.  The curving scar of this work can still be viewed on the ground, Roman occupation seemed to have been uncovered in the form of clean blue clay which contained Roman pot sherds at a depth of one metre. 

Several curves and semi circles can be viewed which may simply represent soil wash or mineral deposits.  An area outside the north road, west, could be an enclosure, date is impossible to guess but it is sited in an area that an annex would have been constructed.  A south annex was detected in the 1926-9 excavations, so it is possible.  A linear feature close by, running N-S for most of the northern survey area could possibly be a modern drain, though it seems in close association to the N-W fort angle.  A large iron water main constructed in the 20th century, this runs close to the N-W and N-E ramparts, and impeded geophysics in this area, hence the white unsurveyed strip.

Finally a linear feature obliquely crossing the north area, which corresponds slightly to modern field boundaries may also be the trackbed for a small railtrack used for the moving of stone from the old dock during the construction of Caerhun Hall.

The survey moved to the south of the fort earthwork, David Hopewell records that less Roman activity was detected.  This is unusual as during the 1926-9 excavation it was noted that Roman activity was fairly frenetic in this area (Site F) that there being as many as seven Roman levels, comprising of an annex (an additional defended area of the fort) with a branch road adjacent to the southern Roman causeway, areas of cobble bases, all overlain to the S-W by a post medieval farmstead and its associated yard and outbuildings.  David thought possibly this area has seen intensive ploughing subsequent to the excavation, so what exactly can we see now?

Roman activity can be seen at the south gate, the fort ditch is clear and where it should be, and closely matches up to Baillie Reynolds’ plan.  The southern road can be seen, blurred, but clearly curving S-W to avoid the annex ditch and rampart.  The annex is mostly unrecognizable, though a section of the ditch is possibly viewed to the right of an unsurveyed area, a modern electricity pole, and possibly the roundhouse which contained numerous forges is also detected in the centre above.  A double linear feature can be seen bottom, this possibly is the twin ditches of the south road, but it is difficult to be precise.  All these features however may represent modern fence lines, as one can be seen in the Excavation Report forming the easterly boundary to the land available for excavation at Site F.

The post medieval farmstead earthworks can be clearly viewed at the S-W angle, a shed actually being butted into the rampart, and a large empty area south of the house, crossed with rigg and furrow plough marks is likely the field of the dwelling.  Seen west of the centre of this field there appears to be three sides of an oblique building, the slight nature of which, possibly conveys a Roman date.  To be seen throughout the south area are the possibly black forge anomalies, ferrous spikes, there being several clusters.  Little else remains to be seen south of the fort.

© Kanovium Project

David Hopewell and John Burman of Gwynedd Archaeological Trust discuss the Caerhun Geophysics close to the brook north of the fort site.
 

 
It was very interesting for me to view the Geophysics results on site, and equally, to recieve the result and be allowed to publish it on this website.  For this I would like to thank David Hopewell, John Burman and G.A.T for being very helpful from the outset. 

I learned my history from the many, many, books on Roman Britain, and sure I know a great deal about the period, but to talk to these archaeologists for an hour or so really told me I knew very little, I feel however that by giving access to this information,  to people like myself, G.A.T are allowing the average person a much greater perception of what archaeology is all about.  Roman military archaeology in particular can sometimes only exist as soil stains or buried ditches, here we see Kanovium’s vici to a level of detail not seen for many centuries.
It has been a pleasure to try to understand the results of the Caerhun fort Environs Survey, and not an easy task, hopefully G.A.T will come to some more concrete understanding.

Geophysics Image Courtesy and © Gwynedd Archaeological Trust