Category Archives: Archaeology

Cefn Caer, Roman auxiliary fort, Pennal

Simplified reconstruction of Pennal Fort by Gwynedd Archaeological Trust: Source: Gwynedd Archaeological Trust

The nearest Roman site to Aberdovey is the fort at Pennal, called Cefn-Caer (which translates roughly as ridge/hillside of the fort), 10.5km (6.7miles) away from Aberdovey. Although there is a rock-cut track that stretches from Penhelig to Picnic Island along the estuary that is known locally as the Roman Road, this actually dates to 1827. Cefn Caer at Pennal, however, is the real thing: a Roman fort 600 yards from Pennal down a small B-road. It formed part of a network of forts and roads that were key to the Roman plans to subjugate Wales. When I first started looking into Cefn Caer for this post, it was simply because the site is part of this area’s history and I wanted to include it as a small representatives of Roman activity in Wales. The word “small” is worth noting here, because I was expecting Cefn Caer to be no more than a very ephemeral way station for travellers (mansio) or a tiny watch-post. In fact, it is a fairly substantial affair, as demonstrated by the above simplified reconstruction by Gwynedd Archaeological Trust (GAT). The GAT work at the site reveals an auxiliary fort with all the features associated with a permanent installation, which had an important strategic role.

Pre-Roman Wales. Source: Wikipedia

The Roman Empire first made its presence felt on British shores first under no less a personage than the Emperor Julius Caesar, albeit only briefly in 55BC and 54BC. Under the Emperor Claudius matters were taken far more seriously in AD 43 and there was to be no retreat, and after the invasion most of Britain was incorporated in the Roman Empire for for the best part of 400 years. The period of the Roman occupation of Britain is known as the Romano-British period (AD 43 to 410).

Iron Age Britain immediately prior to the invasion was divided into six main tribal areas, recorded in Roman documents, which were organized in social hierarchies that were based on lineage, status and military aptitude “cemented by the distribution of favours and hospitality; consequently equipment for eating looms large in the archaeological record” (Davies and Lynch 2000). Parade gear, with a particular focus on horses and chariots, is also dominant in the archaeological record. Subsistence practices depended very much upon geography, but combined herding of domesticated animals (cattle, sheep, goat and pigs) with the cultivation, where possible, of emmer wheat and barley. Hillforts are generally thought of as synonymous with the Iron Age, as places where political power was centred, but in mid-west Wales, where there are very few hillforts, suggesting that political power was more fragmented, and consisted of scattered farmsteads. Although the Tal y Llyn hoard (covered on an earlier post) found at Cader Idris is very rich, it is entirely possible that it was hidden by someone travelling through the area, rather than a local resident. Although in some areas life went on without disruption for some time, in the areas where the invaders first settled, they introduced substantial change very quickly.

The Emperor Claudius, Naples Archaeological Museum.

When Aulus Plautius, the chosen commander of the Emperor Claudius, led an invasion force to Britain and landed in the southeast, he found the prosperous and sophisticated Catuvellauni tribe dominant, their territory extending from Essex to Surrey under the leadership of Caratacus and his brother Togodummus. Caratacus and Togodummus were defeated when confronted with the 40,000 men in four legions and supporting auxiliary forces. Caratacus abandoned his family and fled to the Silures tribe in southeast Wales to rethink his strategy. Caratacus realized that the partially low-lying territory of the Silures was vulnerable and created an alliance with the Ordovices, which had highland areas in its territory, to organize resistanc,. The Ordovices were the main tribe occupying most of Gwynedd and Ceredigion, and “by creating a multi-tribe resistance he [Caratacus] offered the most effective bulwark against the Roman invasion to date” (de la Bédoyère 2003).

Cefn Caer, showing farm buildings with traces of the Roman fort in the field to its right. Source: RCAHMW (on the Coflein website) Catalogue Number C872327, File Reference : AP_2009_1671. By Toby Driver

It was not until AD 47 that the Romans felt the need to turn their attention to the tribal areas of what is now known as Wales. Wales had many benefits from a Roman point of view, including rich mineral resources, fertile valleys and a long coastline. It would also have been a good source of manpower via voluntary enlistment or conscription. Perhaps most important, strategically, without peace in rebellious Wales, all Roman-controlled land to its east was potentially under threat. The first period of military hostilities between Rome and Wales lasted between AD 47 and AD 60, with multiple campaigns against the Marches and Welsh communities, starting in the southeast. A significant event was the Battle of Caer Caradoc in AD 50, where Caratacus led armies composed of the Ordovices and Silures against the Roman military. In spite of the strategic advantage of Caratacus and his armies, holding the high ground, the Roman forces under the governor Publius Ostorius Scapula had weaponry, body armour and military experience that outclassed Silurian and Ordovician resources. Caratacus was defeated and ultimately taken into custody and carried to Rome where the Roman senate were sufficiently impressed by his speech that he earned a pardon from Claudius and lived out his life in Rome. In his book “Defying Rome,” de la Bédoyère comments that Caratacus “failed to appreciate that he was on the whole a dinosaur. While he maintained his resistance he found the only place he could do so was amongst people who had no idea what Rome amounted to.” The Romans did not have it all their own way, but although the Silures went on to defeat a Roman legion in AD 52, it was only a matter of time before Wales was brought under Roman control. There was a brief respite when the Boudiccan rebellion in East Anglia in AD 60 required the redeployment of troops. Full-scale invasion was temporarily abandoned and a strategy of containment was practised in Wales, with all of the only permanently occupied military bases lying along the border.

Wroxeter (Viroconium Cornoviorum). Source: Wikpedia. Photograph by Alastair Rae

In AD 73 under the Emperor Vespasian, Sextus Julius Frontinus was appointed Governor of Britain (AD 73-77) , and it is during his tenure that Wales was fully conquered. Three legionary fortresses were established as campaign bases, at Caerleon (Isca Silurum), Wroxeter (Viroconium Cornoviorum) and Chester (Deva Victrix), and temporary camps were set up within Wales itself, setting the scene for “a network of garrison posts, incorporating fortlets and watchtowers, eventually linked by an all-weather road system” (Arnold and Davies) which were used to maintain control over the rural and often highland zones.

Information about Iron Age and Romano-British exploitation of the western areas of west of mid Wales is particularly sparse, but it would be surprising if such rich natural resources as the Dyfi and particularly Dysynni valleys were not employed for cattle herding and some cultivation, with the surrounding highlands excellent for sheep herding. It is by no means clear if the Ordovices occupied the whole area, as the boundaries of tribal areas are not known, and it is thought that other smaller and less dominant communities also occupied parts of Wales, but it seems clear that whatever happened to the Ordovices would have had an impact on other small communities in the area. After their defeat under the leadership of Caratacus in AD 50, the Ordovician tribe again rebelled in AD 77-78 and was put down uncompromisingly by the British governor, Gnaeus Julius Agricola. Agricola went on to establish forts at Caernarfon, Caersws, Pen Llystyn (Bryncir), Tomen y Mur (Trawsfynydd), Caer Gai (Penllyn) and Cefn Caer (Pennal), most of them in river valleys or estuaries. Other sites in the mid Wales area established in this period were the fortlets at Erglodd in Ceredigion and Brithdir in Merionnydd.

Military installations c.AD 70-80. Source: Arnold and Davies 2000, p.16

The Roman architectural infrastructure in Wales took the same form as it did elsewhere, a hierarchy of military installations. The most important in strategic, organizational and to an extent administrative terms were the legionary fortresses at Chester (Deva), Wroxeter (Viroconium Cornoviorum) and Caerleon (Isca Silurum). These were, however, in a minority, and the main control over Wales was exercised by a large number of auxiliary forts dotted at strategic positions throughout Wales, often on rivers and estuaries, supplemented at intervals by small fortlets and watch towers. Legionary and auxiliary forts each refer to the type of garrison stationed there. Legions were the elite army of the Roman Empire, composed of c.5000 men, divided into ten cohorts. They served for twenty-five years and were rewarded on retirement with a choice of land or a payment. Auxiliaries were composed of non-Roman citizens, men who entered the army from throughout the Roman empire sometimes sometimes as volunteers but sometimes extracted from their homes by force. They were granted Roman citizenship once they retired. They were far more numerous than the legionary forces and were essential to the Roman occupation of Britain. Mid Wales in the Romano-British period remains poorly understood, which means that wherever a Roman site or a contemporary Iron Age is identified in the area, it is potentially of considerable importance for understanding what was happening in mid Wales at this time. The Cefn Caer fort was an auxiliary fort, the westernmost of Roman structures in Meirionnydd, established in the AD 70s.

Cefn Caer geophysical survey results. Source: Hopewell 2001

There are few visible features of Cefn Caer on the ground. The ramparts to the southwest and northwest can be made out, but elsewhere they are low banks that cannot always be seen. Before it was torn down and rebuilt in 1769 the church in the village of Pennal was reported to include a large number of Roman brick in its walls, and remaining obstructions to cultivation were probably moved in the distant past, and the land continues to be used by the local farm for cultivation. The farm buildings, including a sub-Medieval farmhouse (which can be visited), sit within the west corner of the fort and the northern corner of the fort is crossed by a small B-road Although the 1967 History of Merioneth provided dimensions derived from previous surveys of the fort, detailed knowledge of the scale and structure of the fort comes from more recent analysis of aerial photographs, the use of geophysical survey and field excavations, the latter only sampling certain parts of the site. The history of the archaeological work can be summarized as follows. The site was first noted by Robert Vaughan in his Survey of Merioneth in the mid 17th Century, and in a late 17th Century letter by the rector of Dolgellau, Maurice Jones. Amongst the 17th Century finds were a silver coin inscribed with the name of the Emperor Domition. Subsequent visits to the site reported ditches, coins, bricks, a hard paved road, pottery and a tile relating to II Augustian Legion. The main sources of information are the initial detailed report by Professor R. C. Bosanquet in 1921, which was further studied and commented upon in 1957 by H.C. Irvine in BBCS Volume XVII part 2, and these were the best sources of information on the subject prior to the work by Gwynedd Archaeological Trust (GAT). GAT used conventional survey, geophysical survey, and excavated some sample trenches to investigate further (Hopewell 2001, 2003).

Cefn Caer was a small auxiliary fort (castellum) with traces of a ditch still visible at the northwest, outside the rectangular bank that encloses the fort. It was built in AD 70s. It is more than 1.68ha (5 acres) in area, measuring 140m x 120m (c.550ft x 425ft) northeast to southwest with rounded corners. An earlier site of c.2.4ha appears to have predated it, which may have been the temporary fort established before the construction of the permanent site. The fort was located at the west end of a ridge or spur that rises 15m (50ft) above the floodplain north of the river Dyfi, c.10km (c. 6 miles) from the mouth of the estuary. This offered it the dual benefits of having something of a view over the surrounding area, and in particular the river crossing. It was only 100m (328ft) northeast of the marshy Dyfi floodplain and 1.6km (half a mile) from the river itself, where “tongues of the land extend opposite each other to both banks of the river” (History of Merioneth) providing an ideal place for fording the river, and where coastal vessels could unload. Roman forts were built to a fairly standardized template, meaning that they could be built rapidly without resorting to labour beyond the personnel they had to hand, and Cefn Caer does not deviate from this basic form. For comprehensive details see Hopewell 2003 (available to download – link also at the end of this post) but here are some of the key features that Hopewell describes, with numbers in the text referring to the site plan, copied here.

Resuilts of the GAT geophsyical survey at Cefn Caer. Source: Hopewell 2003, Gwynedd Archaeological Trust

Cefn Caer was arranged around two main axes that crossed the fort at right-angles to each other, one on a northeast to southwest axis, the other crossing it on a northwest to southeast axis, and the whole fort was surrounded by defensive ditches. At its centre, on a natural rise, were the fort’s stone-founded headquarters, the principia (principal buildings – no.5 on the above plan) measuring 25m x 28m. Several other buildings also appear to have had stone foundations. The entrance to the principia is on the south-west side, and “leads into a courtyard with a portico on four sides bounded by a cross hall at the rear. At the rear of the building stand a set of five rooms comprising a central shrine room (sacellum) with offices to either side” (Hopewell 2003). There are two buildings either side of the via principia. GAT interprets the building to the north-west (10) as the praetorium (commander’s house). In the retentura (rear part of the fort) one block of centuriae (military barracks) (12 on the above plan) can clearly be seen. The officer’s quarters stand towards the corner of the fort. Part of the space in the praetentura (the front part of the fort) appears to be taken up by two ranges of centuriae. Part of the big building complex (14) may be a stable block with the stalls. Within the fort are a number of roads, which are standard for an auxiliary fort, as follows:

The via principalis (6 on the above plan), running from north-west to the south-east across the centre of the fort.
A short length of the via praetoria (7) runs at right angles to the via principalis under the farmyard
The via decumana (8) runs from the rear of the principia to the north-eastern gate
The via sagularis (9) runs around the inside of the ramparts

Beyond the main limits of the fort a vicus developed to the northeast and northwest. A vicus is a small settlement associated with an auxiliary fort, a community of traders and their families, who supplied good to the garrisons within, but its inhabitants were rarely local, and were just as much outsiders as those within the fort. Marriage was forbidden to Roman soldiers, but there is little doubt that less formal arrangements existed, and that families of soldiers also resided within the vicus. The presence of a vicus next to the fort is indicative of its permanence and relative longevity. Below the southwestern annex there was a small circular building that was probably a small temple, shrine or tomb. A large rectangular building (33 on the above plan) measuring 34 x 22m may be a mansio (travellers’ way station). A mid 19th Century visit by the Cambrian Archaeological Association mentions the remains of a hypocaust (sub-floor heating, sometimes associated with bath complexes), and this appears to have been located in an annex to the northwest of the fort (22) where there is plenty of Roman tile on the surface.

Cefn Caer site plan. Source: History of Merioneth, page 239, figure 102.

The fort has four entrances, one in the centre of each side, and there have been some efforts to determine where the roads that terminated here linked to locally. A small B-road cuts across the north corner of the site, shown in the plan from History of Merioneth to the left, and the History of Merioneth suggests that the sudden kink in the road indicates that for a short span it follows the Roman road that emerged from the site. Evidence of the same Roman road a little further on appears to run along a nearby ridge. There was also an earlier indication that portions of a road led from the southwest gate led down to the river. The History of Merioneth suggests that this may have led to a quay at Llyn y Bwtri. The southeast gate would have faced the river crossing. Cefn Caer appears to be linked to a number of national routes as follows.

Via the fortlet at Brithdir towards Tomen y mur (to the northeast of Llyn Trawsfynydd. Tomen y Mur is considered to have been the most important Gwynedd fort due to its strategic position, its size and its complex layout, with an amphitheatre, bath house, vicus, mansio and related structures, including a possible aqueduct. Although the roads connecting it are not completely mapped, it is clear that it was an important link between mid (and south) Wales with the important sites of Caernarfon and Canovium (Caerbun) to the north, which were in turn connected to the regional capital at Chester.
Via the fortlet at Brithdir northeast towards the important fort of Chester), via smaller forts at Caer Gai and Llanfor.
Cefn Caer probably linked to another route, this time west to another ciwitas captial at Wroxeter via the fortlet at Pen y Crogbren and the forts at Caersws and Forden Gaer.
It was also clearly connected with sites to the south of the river Dyfi, in the first instance the fortlet at Erglodd and, in turn, the forts at Pen Llwyn and Cae Gaer. These were on routes to the important southern Welsh fort Caerleon.

These are all shown on the map of Roman Wales above and although the road network cannot currently be completed, the map indicates how Pennal was linked to other sites in the area, providing an important intersection at the river Dyfi between north and south parts of west Wales.

Brithdir fortlet from the air. Source: RCAHMW colour oblique photograph of Brithdir Roman fortlet. Taken by Toby Driver on 11/12/2007. Published on the Coflein website.

Another Meirionnydd fort at Brithdir, 3 miles east of Dolgellau, was found in the early 1960s and is clearly connected by a contemporary road to Cefn Caer at Pennal. It measured c.184x184ft (54m sq), so was much smaller than the Cefn Caer fort. It has not been excavated and there are no extant remains, but it shows up very clearly in aerial photographs like the one at left, and in the early 1990s geophysical survey was carried out at the fortlet. When a new housing estate was under construction nearby in the 1970s the opportunity was taken to excavate, and the results of these combined sources show a complex history at and around the site. At least two and possibly three, ditches surrounded the fort, and there are indications that a bathhouse and workshops were present. Brithdir was considered to have been built to guard an important intersection of a number of routes.

The fortlet at Erglodd in Ceredigion. Source: Gwynedd Archaeological Trust

Looking to the south of Cefn Caer, the nearest site on the other side of the river was the fortlet at Erglodd, to which it was presumably connected by a road to the Dyfi ford. You can read more about the results of the geophysical survey in the Gwynedd Archaeological Report on the subject (Hopewell 2007).

Unlike the other parts of England and Wales, there is no evidence for towns developing or villas being built in Mid Wales. Arnold and Davies say that this “may be a silent commentary not just upon native resistance but upon the inability of the agrarian base to produce the necessary surplus. Together with geographical constraints, this inhibited political co-operation and fostered continuation of highly segmented societies.”

In the period AD 78-83, again in AD 98-119 and then again in AD 125-6 troops were required in the north of Britain (eventually resulting in Hadrian’s Wall and Antonine Wall) and overseas, when some troops were again withdrawn from Wales. Some forts were abandoned whilst others, like Tomen-y-Mur at Trawsfynydd, were resized and operated with less manpower. By AD 140 very few auxiliary forts were occupied in Wales and it is probable that Cefn Caer was abandoned either at this stage, or during the 3rd Century, when most of Wales was abandoned.

A lot of unanswered questions may be tackled in the future. Gwynedd Archaeological Trust’s Roman Fort Environs Project funded by Cadw is researching the environs of a number of forts using fluxgate gradiometer survey, which should help to develop an understanding not only of the forts but of their ancillary structures, roads and supporting settlements. Gwynedd Archaeological Trust has so far carried out surveys at Canovium (Caerhun), Caer Gai (Llanuwchllyn), Caer Llugwy (Capel Curig), Cefn Caer (Pennal) and Pen Llystyn (Bryncir). These findings will be published in the future. At the same time, a number of GAT and independent projects are looking for the remains of Roman roads in areas where the linkages are known only from small sections, in order to fill the gaps in knowledge about the roads between forts and the routes they followed. Research by Hugh Toller, for example, is thought to have uncovered a number of previously unknown sections of the RRX96 road between Pennal and Brithdir.

Main sources:
Arnold, C.J. and Davies, J.L. 2002. Roman and Early Medieval Wales. Sutton Publishing
de la Bedoyere, G. 2003. Defying Rome. The Rebels of Roman Britain. Tempus
Bosanquet, R.C. 1921. Cefn Caer – Roman fort in An Inventory of the Ancient Monuments in Wales and Monmouthshire VI. County of Merioneth RCAHM
Bowen, E.G. and Gresham, C.A. 1967. History of Merioneth. Volume 1: From the earliest times to the Age of the Native Princes. The Merioneth Historical and Record Society
Davies, J. 2007 (third edition). A History of Wales. Penguin
Davies, J. and Lynch, F. 2000. The Late Bronze and Iron Age. In (eds.) Lynch, F., Aldhouse-Green, S. and Davies, J.L. Prehistoric Wales. Sutton Publishing
Gwyn, D and Davidson, A. 2007. Ports and Harbours of Gwynedd: Aberdyfi. A Threat Related Assessment. GAT Project No. 1824. Report No. 671.1. April,2007. Gwynedd Archaeological Trust
Hopewell, D. 2001. Roman Fort Environs G1632, Report 416. Gwynedd Archaeological Trust 2001. http://www.walesher1974.org/her/groups/GAT/media/GAT_Reports/GATreport_416_compressed.pdf
Hopewell, D. 2003. Roman Fort Environs 2002/2003, G1632, Report number 479. Gwynedd Archaeological Reports. http://www.walesher1974.org/her/groups/GAT/media/GAT_Reports/GATreport_479_compressed.pdf
Hopewll, D. 2007. Roman Fort Environs. Geophysical Survey at Trawscoed Roman Fort and Erglodd Fortlet. G1827(2). Report number: 667. Gwynedd Archaeological Trust. http://www.dyfedarchaeology.org.uk/projects/07romanergloddgeophys.pdf
Irvine, H.C. 1957. Bulletin of the Board of Celtic Studies Volume XVII part 2, (p.124-131)

Coflein entry on Cefn Caer:
https://coflein.gov.uk/en/site/300159/details/cefn-caer-roman-fortpennal-roman-fort
Coflein entry on Brithdir fortlet:
https://coflein.gov.uk/en/site/95480/details/brithdir-roman-site

Castell-y-Bere (1221-1295) in the Dysynni Valley

3 Replies

Ordnance Survey map showing Abergynolwyn, shaded red at bottom right and Castell y Bere in the red square (OS Explorer OL23 Cadair Idris and Llyn Tegid)

Castell-y-Bere is at Grid Reference SH6676908547, overlooking the Dysynni valley near the village of Llanfihangel-y-Pennant. It is maintained by Cadw (Cadw number ME023 ). It is a splendid place to visit. Its remains are substantial, accessed via a short and easy walk, offering spectacularly scenic views over the Dysynni valley that it protected, and is far enough off the beaten track to be wonderfully peaceful. There are various routes to Castell-y-Bere, but if you are not fond of single track roads, the easiest, and almost certainly the quickest, is to go along the B4405 from Bryncrug to Abergynolwyn, turn left in the middle of the village and follow the brown signs to Castell-y-Bere for about 15 minutes. For those that don’t know the roads, they are very good quality with plenty of passing places, and the hedges are kept cut right back, but you do have to resign yourself to the fact that you are almost certainly have to do some backing to passing places before you get to your destination, particularly during the summer when the castle has a lot of visitors. It is very well worth it, however.

There’s a parking area, and an information sign before you pass through a kissing gate and head along the path. The walk takes you through trees. The stone-cut path is well defined but quite uneven. Although it qualifies as an easy walk and there are no particularly steep bits, there are some fairly sharp drops to the side of the path, so you do have to be sure of your footing. This is even more the case with the castle itself. There are a number of flights of stairs within the castle, some of which terminate at the edge of a steep drop with no barriers. If you walk around using a bit of common sense (particularly if you have children in tow) it is perfect, and so much better than the usual ugly tubular metal barriers that disfigure most heritage sites today.

Castell y Bere aerial photograph with my annotations showing key components of the castle (Source of photograph: Coflein website)

Approaching its original entrance, the castle offers a gloomy and imposing welcome to the building that requires a climb up wooden steps, emulating the original sense of entering into an intimidating stony eyrie,dominated by walls and gate towers, with pits beneath the wooden drawbridges so that when the two drawbridges were raised and each portcullis was dropped there were formidable barriers to entry. The castle itself provides uninterrupted views over the entire landscape surrounding it, which was strategically invaluable in the 13th century when it was built. I was expecting a far more dilapidated structure, but what survives is sufficient to make the reconstruction shown on one of the signs traceable on the ground with very little effort, although it helps to have the aerial photograph to refer to. I have added labels to my photograph of the reconstruction and the Coflein aerial photograph of the castle as it is today, so that my photographs can be related to the original layout of the castle.

The castle was built in 1221 by Llywelyn ab Iorwerth (also known as Llywelyn Fawr, or the Great, c.1173-1240). Llywelyn, prince of Gwynedd, was a remarkable character, a landmark personality in Welsh history whose reign is characterized by military action to extend his power and attempts at diplomacy to retain it. It was one of several that he built, including the important castles at Dolwyddelan in southwest Conwy and Dolbadarn at the foot of Snowdon’s Llanberis Pass.

Cattle grazing at the foot of Castell-y-Bere in the Dysynni valley.

The land that Llywelyn chose for his castle was owned by Llywelyn’s illegitimate eldest son Gruffud ab Llywelyn and was taken from him by Llywelyn for the construction of the castle. The glacial Dysynni valley is wide and flat-based, providing unusually wide tracts of fertile pasture. Cattle was one of the principal sources of wealth for the Welsh princes in Gwynedd in the 13th Century, and by controlling the pastures surrounding Castell-y-Bere, Llywelyn was able to protect his herds and provide year-round pasture. Cattle are still herded in the valley, and there were plenty of Welsh black cattle in the fields below the castle when I visited.

The river Dysynni at the foot of Castell-y-Bere

The castle had political as well as economic value. From Llywelyn’s point of view, establishing a realm over the entire area of Merionnydd was part of a much more ambitious plan to extend his control over substantial of Wales that were not yet dominated by invaders from England. Castle building was a relatively new tradition for the Welsh who established undefended courts called llysoedd, which would not have stood up to much in the way of determined attack. At Dolwyddelan Castle, for example, the remains of the earlier llys survive. The Norman advances into Wales from the 11th Century put a different complexion on Welsh strategic thinking. The Norman lords who established their territory in the southeast of Wales, along what is now known as the Welsh Marches, demonstrated how vulnerable the Welsh were to potential hostilities from the east. Timber and earthwork motte and bailey castles were the first defensive structures, but stone castles soon followed.

A photograph of the Cadw sign showing a reconstruction of Castell-y-Bere by Chris Smith. I have added annotations to identify key features of the castle.

View from the middle tower towards the north tower

Llywelyn’s castle was built on a rock outcrop and incorporates much of the bedrock into its construction. As clearly shown in the aerial photograph from the Coflein website above, it was a contour fort, following the line of the rock. The castle’s current substantial form reflects various additions to Llywelyn’s original structure. Its original walls were not as substantial as Edward I’s later additions, and the surviving walls of the original structure demonstrate that this was a much less durable structure than those built by the English. English castles consisted mainly of straight walls connected by either square or round towers. In Wales contour forts were common, and apisidal D-shaped towers were characteristic. Castell-y-Bere has two D-shaped towers, one at each end of the castle, together with a round tower the middle rectangular tower. K. Steele of the RCAHMW describes how the southernmost of these D-shaped towers differs from typical design “being isolated from the main castle structure, overlooked by the rectangular keep, and accessible from the ground floor, thus rendering it defensively weak.” The castle was constructed of the ubiquitous local stone. When the castle was excavated in 1851 some high quality carved stonework was discovered, suggesting that Castell-y-Bere was one of the elaborately decorated of Llywelyn ab Iowerth’s castles.

The following section looks at the history of Gwynedd up until Castell-y-Bere was abandoned in 1295, for which the following family tree might be of assistance:

Llywelyn ab Iowerth family tree for the period during which Castell-y-Bere was occupied

Llywelyn the Great on his deathbed, with his sons Gruffydd and Dafydd in attendance. By Matthew Paris, in or before 1259. Source: Wikipedia

Castell-y-Bere remained in Llywelyn’s possession during his lifetime. Between 1218 and 1240, when Llywelyn ab Iowerth died, peaceful relations were maintained between Llywelyn and Henry III, but the situation deteriorated after his death. Llywelyn ab Iowerth died in April 1240 of natural causes, leaving two sons, his illegitimate eldest son Gruffud and his legitimate younger son Dafydd by his wife Joan. Llywelyn had disinherited Gruffud in 1220 to ensure that Dafydd ab Llywelyn would succeed him, an arrangement that was rubber-stamped by the Pope, thanks to the intercedence of Henry III. When Dafydd ab Llywelyn inherited his father’s seat, Henry re-organized. Dafydd’s disinherited half brother Gruffud was handed over to Henry for imprisonment in the Tower of London to prevent any attempt to oust Dafydd and destabilize Gwynedd, and Dafydd’s own rights were undermined. Gruffud died at the Tower in an escape attempt in 1244. Dafydd died of natural causes without an heir in 1246.

Stairs leading up to the rectangular middle tower

The power vacuum allowed Henry III to enter Gwynedd and establish Crown control over the most powerful of the strongholds in Wales, now under the leadership of Owain and Llywelyn ab Gruffudd, two of Gruffud’s sons. A third brother, Dafydd, was also a beneficiary. They inherited a Gwynedd under siege, and peace was purchased with the provision of knights and foot soldiers. Wales remained subjugated until the three brothers came into conflict with each other, Llywelyn ab Gruffudd emerging triumphant and proceeding to take over large tracts of Wales. From 1258 until 1262, whilst Henry was busy with a rebellion of the barons under Simon de Montfort, he consolidated his new territory, securing its borders. However, in 1262 he was on the march again, claiming new territories in the far south. He formed an allegiance with Simon de Montfort in 1265, formalized in the Treaty of Pipton, and although Simon de Montfort was defeated and killed in battle only weeks later, Henry III chose to honour the Pipton agreement in the Treaty of Montgomeryshire in 1267. The principality of Wales was formed, with Llywelyn ab Gruffudd officially recognized as Prince of Wales, with the right to homage of all the Welsh lords, for which privilege he paid 25,000 marks and became a vasal of the king.

Entrance into the building providing access to the north tower.

Llywelyn ab Gruffudd had made a lot of enemies, particularly in the Marches. In 1271 he attacked Caerphilly castle and extended his realm even further. Davies says that his authority “extended from Anglesey to Machen, from the outskirts of Chester to the outskirts of Kidwelly. He was lord of about three quarters of the surface area of Wales and of as somewhat lower proportion of its inhabitants. He had perhaps two hundred thousand subjects.” However, the powerful Marcher houses of Clare, Bohun and Mortimer came into direct conflict with Llywelyn, and in 1274 both his brother Dafydd and his chief vassal abandoned him, going to England. Henry III had died in 1272, but his heir Edward I was away on the Crusade and did not return to claim the crown until August 1274.

One of the rectangular structures in the courtyard

The relationship between Llywelyn and Edward I was strained from the very beginning, caused partly by Llywelyn’s marriage to Elinor, the daughter of Simon de Montfort and by Llywelyn’s refusal to travel to the English court to pay homage to the king. Edward retaliated by abducting Elinor and in 1276 Llywelyn was labelled a rebel. Permission was given to the Marcher Lords to reclaim territories that they had lost and the king himself prepared for war against the prince and took an army of 800 knights and 15,000 foot soldiers into Gwynedd. Llywelyn, cut off from food supplies in Anglesey, submitted in November 1277. The Treaty of Aberconwy in 1277 swept away Llywelyn principality in all but name. Much of eastern Wales was lost to Norman control and castles were established to maintain control in key areas of Gwynedd, giving Edward nearly complete control by 1280.

Oak bucket bound with hazel, with hazel pegs, found in the well. Source: National Museum of Wales. 53.123/4.

More uprisings followed, in particular the war of 1282-3 that spread after an attack by Llywelyn’s brother Dafydd ab Gruffudd on Hawarden and Rhuddlan Castles. Llywelyn ab Gruffud had little choice but to participate but all these attempts were ultimately futile. Llywelyn ab Gruffud was killed in battle on 11th December in 1282 and Dafydd assumed the title Prince of Wales but by early 1283, Edward I’s vast English army had the Welsh heartland hemmed in. Dafydd based himself at Dolwyddelan Castle in southwest Conwy whilst the English took Bangor, Caer-yn-Arfon and Harlech, building vast castles as they went. Castell-y-Bere was the last of the Welsh strongholds to withstand Edward’s armies, falling in April 1283. Dafydd was captured in June 1283. He was tortured and put to a grizzly death in Shrewsbury in October 1283, whilst Edward’s programme of castle building continued uninterrupted.

The rubble interior of the walls, in a section probably reinforced by Edward I.

Castell y Bere survived the 1283 battle and under Edward I a number of improvements were made. It received additional fortifications, in particular thick walls linking the south and middle towers. The large rectangular keep overlies a rock-cut ditch suggesting that it had the adjoining D-shaped tower are additions to the original castle may be from this time. Edward wanted to establish an English borough and a charter was granted, extending from Abermaw to the Dyfi, but the site never prospered. In 1284 the Statute of Wales, or the Statute of Rhuddlan, was initiated. The three counties of Anglesey, Caernarfon and Merioneth were created and placed under the management of English sheriffs, effectively splitting Gwynedd into manageable administrative chunks and ending the dreams of Llywelyn ab Iorwerth. A last ditch Welsh uprising during 1294-5 ended Castell-y-Bere. Madog ab Llywelyn attempted to take the castle from the English. He failed, but the castle was very badly damaged in the process and was abandoned. The 1850 excavations found extensive charcoal, suggesting that it may have been burned.

View along the castle towards the pastures in the Dysynni valley

The 1850 clearance of the site produced some other interesting discoveries. One of the excavators W.W.E Wynne describes opening the excavations in an article in Archaeologia Cambrensis “in the year 1850, we commenced our excavations, not with the expectation of discovering any object of superior interest, but for the purpose of tracing as accurately as possible the circuit of the walls and making a plan of the building.” It was during these excavations that the ornamental stonework and other masonry fragments were found. Other items discovered were pieces of chain-mail, corroded arrowheads, part of a crossbow, several knives, one retaining a wooden handle, part of a bone comb and large amounts of pottery, mainly glazed in green or olive. Animal bones bearing signs of butchery included roe deer and boar.

Plate from Wynne’s 1861 report of the 1850 excavations.

Views from Castell-y-Bere over the pastures that are used today for grazing cattle and sheep

References:

Stonework from Castell-y-Bere, held at Criccieth Castle Museum. Source: Hchc2009 under CC-BY-SA-3.0 licence

Avent, R. 2010. Dolwyddelan Castle, Dolbadarn Castle, Castell y Bere. Cadw

Davies, J. 2007. A History of Wales. Penguin

Glamorgan-Gwent Archaeological Trust http://www.gatehouse-gazetteer.info/Welshsites/510.html

Jenkins, G.H. 2007. A Concise History of Wales. Cambridge University Press

Steele, K. 2008. Castell-y-Bere. RCAHMW, 4 November 2008 http://www.coflein.gov.uk/en/site/93719/details/castell-y-bere.

Wynne, W.W.E., 1861. Castell y Bere, Merionethshire. Archaeologia Cambrensis Vol. 16 p. 105-10 https://archive.org/stream/archaeologiacam07moorgoog#page/n121/mode/1up

A walk along the “Roman Road” to Picnic Island and beyond

2 Replies

If you are looking for a short walk with some lovely views over the river estuary and the hills beyond, this is a nice one. If you want to go all the way to Picnic Island, a walk of just 30 minutes or so, you will need to be prepared to pick your way along some jagged rocks, but if you only want to go as far as the beach beneath the footbridge over the railway line, it’s a simple walk along well worn paths. You can also turn it into a much longer 6 mile walk by crossing the footbridge and going up the hill and circling back into Aberdovey. I’ve added a PDF at the end of this post. Patches can be a bit muddy after rainfall and on the rock this can be slippery, so suitable footwear is recommended. Before you set out, check the tides. You will need to avoid high tide, because part of the walk is cut off by water. If you do find yourself returning along the path to meet with an unpassable section you will need to cross the bridge over the railway and return along the road, but this is a busy road with no footpath so is much best avoided.

The walk is all about beautiful views over the estuaries and to the hills beyond. It starts in the Memorial Park at Penhelig. Either go under the railway bridge or cut off that rather dangerous corner on the road by taking the private road in front of the houses known as Penhelig Terrace. The Memorial Park is on the other side. It is a lovely little park with great views over Aberdovey’s sea front. It contains a memorial and a plaque in English and Welsh to mark the achievements of the 3 Troop 10, a group of German nationals who worked on behalf of the Allies during the Second World War, and who were stationed in Aberdovey for their training (which I have described on an earlier post). There is also a little shelter, slightly unkempt at the time of writing, to the memory of Mr Richard Roberts “in recognition of his munificent gift for improvements at Aberdovey 1930.”

At the far end of the park let yourself through the gate and onto the Roman road. The so-called Roman road is neither Roman nor, in modern terms, a road. It is a path carved out of the local mudstone, a remarkable feat that even the Romans, accomplished civil engineers, would have found a difficult task without the help of explosives. And why would they have gone to the trouble? The nearest Roman military structure was Cefn Caer at Pennal, 11km to the east along the Dyfi. There is no obvious benefit for a permanent stone-built pathway to Aberdovey, even if there was any sign elsewhere between Aberdovey and Pennal of a long-term Roman presence, which there isn’t. My guess was that it was built in the 1860s, part of the works for the building of the railway, but again that fails to address the question of the purpose of such a track, given that there was a perfectly good coast road at that time. In his description of the 6-mile walk, below, the author David Roberts, an Aberdovey resident, states that the track was built in 1808 for horse and carriage, but he doesn’t say where these were headed and why such a road would be required. Even Hugh M. Lewis, who was born in 1910, and grew up and lived in Aberdovey was unable to shed any more light on the subject.

Whenever it was built and whatever it was used for, it is invaluable today as a ready-made footpath for walkers. The path has two small bridges that cross little natural outlets for fresh water that pours down the hill into the estuary. In these places the fresh water-loving gut weed grows, a livid, bright green that contrasts dramatically with the black stone and the brown seaweeds. The estuary is incredibly peaceful on a sunny day as the tide retreats, the waters flat and sparkling, making a pleasant sound lapping gently at the rocks as they travel at a considerable pace to the west. The hills beyond, in Cerdigion, fresh and green, are the perfect backdrop.

The ubiquitous mudstone, into which the path is carved, slopes gently down into the waters, and is covered with some of the seaweeds that I described on my strandline walk, fascinatingly three types forming three bands as they approach the water, with those most equipped to survive out of the water for longest at the top, and the least drought tolerant at the bottom. Sea oak is at the top, bladder wrack in the middle and egg wrack at the base. some of it remaining submerged even at low tide. Unlike my strandline walk, where all the seaweeds had been detached from their rock bases, it was possible to inspect the seaweeds in situ, so I could see the holdfast with which they attach themselves to rocks, a surprisingly tiny little mass of highly tenacious material.

On the rock face above the level of the path, two plants in particular make the best of the most implausible nooks and crannies to grow: red valerian and sea thrift. Both are drought resistant, saline tolerant, prefer sandy and low-fertility soils and need full sun, so are frequently found in south-facing coastal areas. When the sea thrift goes over, which most of them have by late September, the fallen petals leave attractive skeletal globes. Watch out for sea thrift and red valerian in cracks in vertical planes of the rock to the left as you walk towards Picnic Island. Where the rock splits it reveals trapped minerals that are often beautifully coloured providing a perfect canvas for the flowers.

At all levels above the waterline are a variety of lichens dominated by yellow scales (shown to the right), which is prolific, followed by black shields and map lichen. Lichens are not single organisms but are symbiotic, depending for their survival on “photobionts” (algae and/or cyanobacteria), which provide them with the carbon that they need. The photobionts use the process of photosynthesis to manufacture their own food source, whereas the fungal component of the lichens need an external food source. This ecological strategy has obvious benefits for the fungus, which is essentially parasitic on the photobionts, but it is thought that the photobionts might benefit too, due to the provision by the fungus of a stable environment in which they can develop. There’s a lot more information on the British Lichen Society website.

The Roman road runs out and is replaced by a well maintained footpath that runs parallel to the railway line, taking a route several feet above the estuary, again with wonderful views across the estuary. At low tide the sands in the middle of the estuary are revealed, a shifting chiaroscuro of colours and deep shadows framed by the speeding waters of the retreating tide. There is rich vegetation along the footpath – blackberry and rose brambles, ferns, berberis, purple thistles, red valerian, buddleia, holly, wild oregano and much more. In autumn there are few flowers, mainly the last of the red valerian, but there is a profusion of red and orange berries.

A fork in the path offers a choice. The left fork leads up to Picnic Island and the footbridge over the railway into a lay-by and, 100m down the road, the continuation of the walk for those who want to pursue the 6 mile option. Picnic Island is not an island, just a small promontory cut off from the hillside by the railway, but it has wooden seating and is a pleasant green area with excellent views south over the estuary. It’s original name is Bryn Lestair (obstruction hill).

The right fork leads down steps to a small pebble beach, and the continuation of the Roman road for a short span, before it runs out again. The beach was the site of a shipbuilding business, some of its walls still surviving, but the site was largely destroyed by the railway. From the beach, facing the sea look left and you will see that the Roman road resumes. Follow this for a short distance and then it’s a matter of picking your way down the rocks to the beach, and along the foot of the retaining wall behind which the railway runs. This is invariably wet, with hillside water pouring from under the wall. There were almost no shells on the beach sections, only very fragmented muscles and barnacles. The barnacles were on loose bits of slate, so they were probably detached from rocks elsewhere and brought in on the tide. This is probably because the waters are brackish, combining freshwater pouring out of the river Dovey and salt water coming in on the tide.

On the other side of this small beach is another promontory with views over the estuary and east towards the Georgian Trefri Hall with its own island complete with crenellated folly. Before it was painted mustard yellow it was my favourite house in the area – that wonderful location, those stunning views, a private tidal beach and that super folly! In 2016 it came on the market and was featured in an article on the Wales Online website – for sale for a cool 1.7 million pounds. Rather more than I had in my piggy bank on the day.

The walk back into Aberdovey is simply a matter of retracing your footsteps.

If you are interested in the 6-mile walk that takes you up into the hill behind the estuary, here’s a PDF to download: 6 mile circular walk Picnic Island and hill. It is the BBC Weatherman Walking map and guide by local resident David Roberts (eight pages with photographs). I haven’t done it yet, but it looks splendid and it’s on my to-do list.

The Aberdovey Second World War pillbox

1 Reply

Walking towards Tywyn from Aberdovey you will come across a Second World War pillbox, an ugly concrete box with a small square hole in each side. It has subsided unevenly into a dip in the beach at the foot of the dunes, an incongruous contributor to the area’s heritage. It can be reached easily along the beach from Aberdovey. It’s a fairly short walk from the car park, a little way beyond the Trefeddian Hotel, which is visible through a dip in the sand dunes. If you prefer a short-cut there is a public footpath from a big lay-by on the A493 that takes you across the sand dunes and drops you very close to it. Not that it’s a tourist destination, but it is certainly a local landmark, and sitting in an unspoiled stretch of eternal pale yellow sands with the rich blue sea beyond, it has an emphatic presence all of its own. It is at grid reference SN59549635, at the end of the footpath known as The Crossing.

The pillbox is marked as a red box by The Crossing. Source: OS Explorer, Cadair Idris and Llyn Tegid. OL23. Ordnance Survey 2015

There are two war memorials in Aberdovey. There’s a lovely 1999 memorial to 3 Troop 10(1A) who were stationed at Aberdovey during the war for their training (see my earlier post about this) and there’s a little shrine and plaque listing the dead from form both wards inserted into the wall of St Peter’s Church.

In some ways, the pillbox is an even more substantial monument to the bitter truth of war, mute but evocative. The fact that it sits there, so out of place, so thoroughly ugly, is an appropriate shock to the system. As detritus of war, it is something that demands a response and forces an acknowledgement of the realities of the past in a way that a conventional memorial, however heartfelt, does not. Although it was a lovely day for a walk, the sands endlessly beautiful and full of light, when I arrived at the pillbox it was just as dismal as I remembered. Ugly, lop-sided, surreal, a scar on the landscape, a slap around the face. A savage, palpable war memorial.

Pillboxes were part of a network of small defences that were put in place along the coastline, at road junctions and on canals to counter threats of Nazi attack on Britain. The network consisted of a number of measures including offshore minefields; beach and manned seafront obstacles like barbed wire and landing craft obstructions, pillboxes and minefields; and cliff-top and dune defences including pillboxes and anti-glider obstructions. The pillboxes, 28,000 of them, were sometimes round or hexagonal to avoid blind spots, but there were were seven different types in total (Types 22-28), with variants. The Aberdovey one is a Type 26 prefabricated square with an embrasure in each wall and a door, now slightly subsided into a slight dip in the sand, 3ft or so deep. Some pillboxes were brick- or stone-built but many, like this one, were made of concrete that was sufficiently thick to be bullet proof. My thanks to the Pillbox Study Group for this excerpt, which explains the thinking behind pillboxes and other defence structures that were put in place in WWII:

On 25th June 1940, General Paget, Chief of Staff to the Commander-in-Chief Home Forces submitted General Ironside`s anti-invasion plan to the War Cabinet in the form of Home Forces Operation Instruction No.3.

SECTION 13 of the Instruction stated: “The general plan of defence is a combination of mobile columns and static defences by means of strong-points and stops. As static defence only provides limited protection of the most vulnerable points, it must be supplemented by the action of mobile columns. However mobile such columns may be they cannot be expected to operate immediately over the whole area in which it is possible for the enemy to attempt invasion by sea or air. It is therefore necessary to adopt measures for confining his actions until such time as mobile columns can arrive to deal with him. This will be done by means of stops and strong-points prepared for all round defence at aerodromes which are necessary to prevent the enemy obtaining air superiority, at the main centres of communications and distributed in depth over a wide area covering London and the centres of production and supply. This system of stops and strong-points will prevent the enemy from running riot and tearing the guts out of the country as had happened in France and Belgium.”

In total there were 6 pillboxes every 500m from south of the river Dyffryn Gwyn, which flows into the sea just south of Tywyn, to the entrance of the river Dovey. Prefabricated pillboxes were built of concrete panels and were then bolted into place on site. The pillboxes to the north of this one are badly damaged, perhaps in an attempt to destroy and remove them. Aberdovey had an Observer Corps, a Home Guard and a Coastguard Station during the Second World War.