David Hopewell, senior archaeologist with Gwynedd Archaeological Trust (GAT), spoke to a packed audience at the History of Deganwy Group’s meeting on 17th October 2019. David has been an archaeologist for over 30 years, and is also a musician and sound artist. He is the author of several papers on subjects including Roman forts, and a recent book covering tonight’s topic – Roman roads in north west Wales.
GAT holds the historic environment records of all known archaeological sites within the area covered by the old county boundary of Gwynedd. Most sites are recorded as grid references, posing a problem for roads which are obviously more suited to linear references; a recent project across Wales has sifted through published and unpublished material, to produce a more user-friendly record.
The only relevant surviving Roman record is the Antonine Itinerary, which mentions one road in N. Wales: the road from Segontium (Caernarfon) to Deva (Chester). Modern material includes maps, books, aerial photography and Lidar, a surveying method producing images of the ground via laser measurements taken from a light aircraft. This gives a very accurate virtual model, which can be manipulated and overlaid onto other maps and plans.
Ordnance Survey (OS) produced a map of Roman roads in 2001, showing a few definite roads, several dubious ones and many gaps. Earlier sources include the standard work Roman Roads in Britain by I.D. Margary (pub. 1967), which introduced the numbering system still retained today. A less reliable source is S. O’Dwyer’s Roman Roads of Wales (pub. 1934), which fails to distinguish Roman roads from other old roads. The OS also produced linear files in the 1970s – maps were collated, proposed routes were walked, and useful notes were made.
A team from across the UK worked on the recent project – volunteers included experienced excavators, and researchers such as the late Hugh Toller were involved. The Royal Commission on the Ancient and Historical Monuments of Wales (RCAHMW) provided an aerial reconnaissance team, who checked for crop marks during droughts. Information from all available sources was input into a database, and status or confidence ratings were assigned to each road, ranging from ‘known’, ‘proposed’, ‘predicted’ to ‘discounted’. David showed the example of a map of Tomen-y-Mur, with a known road marked alongside a now discounted one.
Roman roads in N. Wales are all military roads, and were built solely to link forts. They date from the late AD 70s and were abandoned by about AD 150, so they were not in use for long. The classic version of a straight road running along the most direct route is not applicable in Snowdonia, where they instead follow the easiest route through the landscape, often featuring zigzags (and now often accompanied by gas pipelines!). They can be identified by their very consistent features: a typical section has a raised bed 5 metres across, known as the agger (Welsh: sarn), made of several layers of stone and clay with a compacted gravel surface – ideal for horses. There was a small ditch either side of this, and often small quarry pits several metres away.
In some places the distinctive agger and ditches are sometimes still visible, e.g. at Caer Gai near Bala, and Pen y Stryd near Tomen-y-Mur, where changes in the colour of grass and the presence of reeds indicate ditches, and the gradient of the agger is clear. Other stretches of Roman road can only be detected by aerial photography or by Lidar, which shows traces of road either side of later settlements and running through structures such as medieval field systems, often with regular gravel pits visible alongside. Such traces can then be checked by excavation – this will eliminate roads such as disused turnpikes, which are approximately the same width and can look very similar. Packhorse trails, such as the one near Capel Curig, are sometimes mistaken for Roman roads by the untrained eye; they are however much narrower than the requisite 5 metres – a Roman road may sometimes be wider than this, but never narrower, as the width had to be sufficient for two horsemen to pass.
The route of the road from Segontium to Deva includes the stretch from Bwlch y Ddeufaen to Canovium (Caerhun), and is not currently fully known, though further parts have been revealed by aerial photography. Bwlch y Ddeufaen, running from Abergwyngregyn to the Conwy Valley, was in use up until at least the 18th century and so has been heavily eroded; it is still visible in places, though not easily.
Another road runs north from Tomen-y-Mur to Dolwyddelan, and is very hard to see. Slabs of rock were found under peat cuttings and drainage ditches – excavation revealed a road under a significant depth of peat. Unusually, the road had been built upon these slabs, which were used as a raft for the agger surface, as there was a layer of peat beneath.
The Roman roads project is part of a much bigger project, looking at forts and the settlements around them. It has been carried out by other archaeological trusts in other counties in Wales, resulting in a good record, with new discoveries still being made. This has been drawn together in David’s book Roman Roads in North West Wales, which the appreciative and inquisitive audience were able to buy once the flow of questions had been stemmed! Our thanks to David for a very enjoyable and informative evening.
Lucinda Smith
Latest Research
Web Design North Wales by Indever