Introduction
Rhosesmor is a small village not far from Mold, which boasts an Iron Age hillfort a short walk from the village hall car park, with stunning 360° views over the surrounding hills and across the Dee estuary and beyond.

Aerial view of Moel y Gaer Rhosesmor by Chris Musson / CPAT. Source: Internet Archaeology
It was an iffy day when I visited, weather-wise, with a lot of cloud overhead and the threat of rain, but it was possible to see for many miles all around, and is very easy to understand why this could have been a strategically valuable location. The footpath around the top of the first, inner rampart is well worn and unimpeded, and a really lovely walk. The interior and the outer of the two banks, and both ditches, are covered with gorse, which was in flower. There are paths through the gorse on the plateau, but it makes it difficult to see the line of the second, outer rampart, which in the Iron Age was never a complete circuit.
When I was an undergraduate in the 1980s my parents moved from Hertfordshire to a village not far from Chester. Whenever I returned home my favourite pastime was taking the family dog and exploring the hills in northeast Wales. That led me to the Clwydian range and Moel y Gaer Llanbedr. As a newbie to Wales, with no understanding of Welsh placenames, I had no idea that there were actually three Moel y Gaer hillforts in the area and more further afield so when I returned to university (and bearing in mind that there was no Internet in those days) I spent some time assembling data about Moel y Gaer Rhosesmor, thinking it was Moel y Gaer Llanbedr before the penny dropped. Decades later I have at last got around to visiting the the Rhosesmor hillfort, and what a super site it is, with a complete panorama of diverse views. See Visiting information at the end of the post for parking details and the route from the car park to the hillfort.
The Archaeology
The best online source of information about the hillfort is Graeme Guilbert’s page about the site on the Internet Archaeology website, part of a bigger article about the excavation of Welsh hillforts. I recommend this, because as well as details about the site, many of which I have included below, the page has some really excellent photographs of the excavations during the 1970s and reconstructions of the two different types of roundhouse.

LiDAR view of Moel y Gaer. Source: Lock, Gary and Ralston, Ian. 2024. Atlas of Hillforts of Britain and Ireland. Available at https://hillforts.arch.ox.ac.uk
Moel y Gaer translates roughly as “the bare fort,” presumably referring to the somewhat barren moorland conditions characterizing the top of many hillforts, which accounts for why there are three of them with the same name in the area. It is perched on the southern end of Halkyn mountain, c.300m (984ft) above the Dee floodplain. The visible remains include two banks, both of which follow the contours of the hill. The higher, inner bank provides a complete circuit of the hillfort, whilst the second, lower bank, divided from the inner bank by a deep ditch, never provided a complete circuit, as the LiDAR image to the left makes clear. The most interesting and informative aspects of Moel y Gaer are, however, the structures and features that were excavated in the 1970s and are no longer visible, described briefly below.
Ironically the most clearly identifiable feature today is a rectangular reservoir, completely enclosed and covered in grass and a good perch for admiring the views. If you sit by one of the metal inspection batches, you can hear something whirring gently below, but otherwise it is just a grassy platform with some metal-topped lids.
It is because of the reservoir that the site was excavated by Graeme Guilbert and his team in the 1970s, with the bulk of the work carried out between 1972 and 1974. Although the development of a modern structure slapped on a prehistoric site was undoubtedly controversial, it did at least give the opportunity for archaeologists to investigate 0.6 hectares (1.4 acres) of the 2.7 hectares (6.6 acres), and the results were excellent. It is rare that it is possible to cut a 15ft trench through a rampart, due to time and cost constraints, but this was done at Moel y Gaer, and the information about the interior, enabled by the excavation of such a large area, is now referenced in all discussions of hillfort interiors.

Sketch of the possible Neolithic structure and well as the worked flint scatter, both in red, shown in relation to the later Iron Age structures. Source: Britnell 1991, p.58
The archaeology of the hilltop begins in the earlier prehistoric period. The earliest of these is the Neolithic, represented by worked flint and chert, including scrapers, blades and arrowheads and associated waste products. Pottery fragments were found in shallow pits. Most importantly, a possible rectangular structure, a very rare find, consisting of a short length of fence and a series of postholes, suggests a Neolithic settlement.
Two Early Bronze Age round barrows have been identified within the later ramparts. One is on the highest part of the hill near the reservoir; the other was apparently levelled during the Iron Age occupation of the hill, and when excavated was found to include a burial chamber but no skeleton (probably due to the acidic conditions), and no accompanying artefacts. Early Bronze Age barrows are common in the area, but most were left in tact. It is a puzzle why one was retained and the other levelled.
The first phase of the hillfort itself predates the banks and ditches and may date back to the Late Bronze Age, in the 8th or 7th century BC. A set of roundhouse structures and a palisade trench carved into the stone were found, although it is not clear if the roundhouses and the palisade (an elaborate fence) were contemporary or if the settlement was initially unenclosed. It had a gap of 11m, indicating an entrance, although it remains unknown if this was the only entrance into the enclosure. As Guilbert says, the size of the gap “may seem inappropriate to a purely aggressive, or even largely defensive purpose” and he suggests that the site may have been positioned as much for its visual impact as for security purposes. In 1977 some trial trenches were opened that suggest that the palisade extended around most of the hillfort perimeter. The pre-rampart palisade is not a unique feature, and there is at least one other in the region, at Eddisbury hillfort on the mid-Cheshire Sandstone Ridge.

Phase 1 of the hillfort showing the palisade slot at left and the roundhouses, many with porches. Source: Guilbert 2018, Internet Archaeology.
Also in Phase 1 there were 26 roundhouses, between c.6.5m and 11.5m in diameter. All were made of an inner and an outer ring of postholes, the inner and larger ones thought to hold roof supports and the outer one to hold much lighter wall supports. All had entrances, most of which were facing east or southeast, and some of which were provided with a square porch by the addition of two posts at their entrances. Eleven had hearths at or near the centre, suggesting that those with and without hearths were used differently. Some of the structures overlapped, suggesting that they were rebuilds. A cache of pebbles, thought to be for slingshots, were found in one of the roundhouses on the edge of the rampart. Other structures were also found, consisting of twelve small oval pits as well as a series of postholes in pairs, but these may or may not belong to Phase 1.

The footpath along inner rampart, following the line of the contour around the top of the hill, marked by the line of gorse
It is not known how long a period separated Phase 1 from Phase 2. Phase 2 of the hillfort consisted of the construction of the inner rampart and ditch. The surviving bank, which made up the rampart, was up to 6m wide and nearly 2m high. At the back it sloped towards the interior but its front was provided with a “showy” revetted stonework facing that alternated with wooden posts 0.6 – 0.9m apart. A generic image of a box rampart is shown below for reference. The rampart posts were tied into a wooden framework within the bank, and the interior was divided into compartments or cells that were filled with rubble, perhaps to provide additional structural integrity and to enable easy repairs when any one part of the rampart failed. This is referred to by Barry Cunliffe as “the Moel y Gaer style” of rampart building.

J. Forde-Johnston’s illustration of box ramparts, with my annotations showing key features. Source: Forde-Johnston 1976 p.180

Phase 2 of the hillfort showing the earliest rampart at left, the roundhouses, and the 4-posters. Source: Guilbert 2018, Internet Archaeology.
The enclosed space in Phase 2 was occupied by a mixture of twelve roundhouses, different from the Phase 1 structures, and around thirty four-posted structures that were largely set apart from the roundhouses, indicating a level of deliberate planning. The roundhouses in this phase, between 5.7 and 8.1m in diameter, only had a single circuit of stakeholes and no separate roof supports, indicating more lightweight structures, with or without roofs. As in the previous phase, entrances were mainly aligned to the east. The four-posters are known from other hillfort and lowland sites, but there are no indications at Moel y Gaer about how they were used. Interestingly, when the site was abandoned the timbers from many of the 4-posters were removed, presumably for re-use elsewhere. The roundhouses and the ramparts on the bank were simply abandoned, leading to the decay of both. The period of Phase 2 occupation dates to the 4th century BC.
Phase 3 involved the repair of the inner rampart, which was widened and heightened and topped with rubble, with new breastwork added along the top. This work was radiocarbon dated to the 4th -1st century BC. Work began on the second, outer rampart, of which intermittent sections and accompanying ditch survive. Only two stretches of the bank at the east end were built to a clearly visible height, all covered in gorse, but there are places on the walk along the upper rampart where you can clearly see the height of the second rampart, which must have been impressive in places on the other side of the inner rampart’s ditch. A palisade was built on top of it, which may have been intended to look like a more substantial wood-faced revetment from the outside. Possibly during this phase, although the relationship between the components is not clear, a set of rectangular buildings were added. These had no foundations of any type, but instead sat on sill-beams that were laid on the surface of the plateau. They were provided with stone floors and were accompanied by metalled areas interpreted as yards.
There were thought to be at least two entrances, one at the east end that remains in use,with an inturned entrance, and one at the west end that was later blocked in, a phenomenon noted at other hillforts in the region.
Moel y Gaer is interesting in relation to other hillforts on the Clwydian Range. Although the Egyptian temple on top of Moel Famau is just visible from the interior, there was no hillfort on Moel Famau. Moel Fenlli is just visible, but most of the hillforts on the Clwydian Range were focused on the Vale of Clwyd on the other side of the range. Moel y Gaer is, by contrast, focused on opposite side of the range, with exceptional views up and down the Dee and across the Wirral towards Merseyside, as well as east towards the mid-Cheshire sandstone ridge where there are other hillforts. There would have been a direct line of sight between Moel y Gaer and Burton Point, if the latter is confirmed as a promontory fort (see my post on the subject of the latter here).
Some of the unanswered questions
Even with an excavation on the scale of the rescue work carried out at Moel y Gaer in the face of the planned reservoir there are many questions that remain unanswered.
For example, the possibility of a Neolithic settlement is of considerable importance, but it is not known if the hilltop contains any more clues about this period of occupation.
In Phase 1 the 11m (36ft) gap in the palisade is something of a puzzle if the site was intended to be defensive. Speculating, it might have been intended to give easy and hurried access for livestock, perhaps in the event of a raid or for herd management purposes, but without the vital information from animal bone it cannot be determined if animals were ever present at the site.

The orientation of Iron Age roundhouses, showing that the majority are orientated between the east and southeast. Source: Oswald 1997
Similarly, in phase 2 it is suggested that roundhouses with and without hearths may have had different purposes, but these purposes cannot be assessed due to the lack of artefacts, and plant and animal remains. In both phases 1 and 2 the roundhouse entrances faced mainly east and southeast. There has been a lot of discussion about whether this sort of orientation, which was common, represented practical considerations such as allowing light and heat into the building from the early morning and excluding westerly winds; or whether more spiritual and cosmological motivations were concerned with aligning buildings to solar events.
In phase 2 it is not known what the four-poster buildings may have been used for unless data can be obtained from unexcavated parts of the hillfort. Although it has been traditionally suggested that such postholes represent buildings raised off the ground surface on posts, and that these may have been granaries, from most sites there is very little, if any evidence in the form of plant or animal remains, or artefacts, to indicate what function they may have had.
In Phase 3 it is not known why the second, outer rampart was only built in incomplete sections, although Ian Brown suggests that such sites may reflect a building strategy employing separate working gangs or family units who simply failed to complete the work by joining up the sections by the time the hillfort was abandoned. The rectangular buildings that may or may not have been contemporary with the second rampart, and both their date and function remain ambiguous.
Perhaps the biggest unanswered question, common to all of the phases, is what sort of economic practices were adopted by those who occupied the hillfort. Today there is extensive livestock herding in the area, both sheep and cattle, and this may have formed the basis of subsistence in the Late Bronze Age and Iron Age, but there may also have been other types of agricultural and horticultural activity, or trade links may have supplied the cereal and vegetable components of diet. Nor is it known to what degree the builders of the Moel y Gaer Rhosesmor hillfort were related or in communication with the hillforts on the Clwydian Range, and if they may have shared or exchanged resources, or whether they were in conflict. In theory this was the territory in the Late Iron Age of a “tribe” known by Roman writers as the Deceangli, but whether this indicates solidarity between all communities in that territory, and over what sortof period of time, is unknown. Similarly, it is not known if there were affiliated lowland settlements in the valley below. Without evidence these, and other livelihood questions remain open.
Post-hillfort history
Of the post-hillfort history the earliest is the unconfirmed tradition that one of Owain Glyndŵr’s supporters, Hywel Gwynedd, set up a camp here, but was captured and beheaded in 1406. Excavations produced no signs of them. It is, however, a matter of history that in 1814 a fire-beacon was set up here during the Napoleonic Wars to take advantage of the superb views over the estuary and river.
On your way back to the car, the Victorian church of St Paul with its fantastic roof is well worth a visit.
See the National Church Trust website for information about the history of the church, but here are the first two paragraphs:
St Paul’s was built in local limestone between 1874-6 by the architect John Hill for the villagers working in the lead mining industry, which dates back to pre-Roman times. Rhosesmor grew as a distinct lead mining village in the late 18th and early 19th centuries. St Paul’s benefited from local landowners who contributed to the formation of the church. The site was given by Hugh Lupus, first Duke of Westminster (1825-1899) with foundation stone laid by Mrs J Strutt Bankes of Soughton Hall in 1874.
By 1890 necessary repairs, including the present roof beams, were undertaken by architect John Douglas of Chester. The west end Annunciation windows, by Jones and Willis were presented by the Revd W E Torr of Eastham, on the Wirral. The availability of funds grew over the next 25 years and Mr Scott Bankes, replaced the dilapidated chairs in the nave and supplied new fittings for the chancel.
Final Comments
Visiting hillforts is all about enjoying what you can see, but it is also about using excavation reports to imagine what was once there. The quiet serenity of a hilltop site with wide views, its former ramparts smoothed by centuries of erosion, is probably completely at odds with the busy lives represented by the roundhouses and four-posters, and the ongoing maintenance after the substantial ramparts were first built.
Moel y Gaer Rhosesmor is one of the most thoroughly investigated hillforts in north Wales, and it hints at many more possibilities, via geophysical survey and excavation, for understanding both this and other hillfort sites in the region, from the Late Bronze Age to the end of the Iron Age.
With its panoramic views over the hills and the Dee, Moel y Gaer Rhosesmor is an absolutely marvellous place to visit.
Visiting

The hillfort is at top left. The parking is in the Rhosesmor community hall immediately off the B5123 (red arrow). The yellow arrows show the route to the hillfort. Source: Google Maps
The hillfort is at top left of the image. Parking is at the Rhosesmor community hall, marked on the map with a red arrow (What3Words ///litigate.messy.hems). There is a blue parking sign, but it is a bit hidden. From the car park the yellow arrows show the route.
It is a short walk along the lane towards the route up to the hillfort, with the churchyard of St Paul’s on your left.
At the junction head uphill along a private road. A very faded Public Footpath signpost points you in the right direction, but it is perfectly obvious. This is quite a short stretch, all uphill. It ends at the entrance to a house on the left and a metal gate on the right. Access to the hillfort is via the gate, which leads you onto the open hillside and its top. It is perhaps a 15 minute walk from the car park to the ramparts.
As you can see in many of the above photographs, the inner bank is fully accessible along a well worn footpath that goes around the full circuit of the hillfort, and their are walks through the gorse on the plateau itself. The lower bank and ditch are completely covered with gorse, and are not accessible unless you like to wrestle with very dense spiky gorse. The fully enclosed reservoir is covered in grass, and provides unobstructed views over the entire area.
St Paul’s Church has toilets and a small shop selling essentials as well as hot drinks. The church is very welcoming and well worth a visit for the beautifully crafted wooden ceiling over the apsidal chancel.
Sources
Books, papers and leaflets
Britnell, William J. 1991. The Neolithic. In John Manley, Stephen Grenter and Fiona Gale. The Archaeology of Clwyd. Clwyd County Council, p.55-64
Brown, Ian 2009. Beacons in the Landscape. The Hillforts of England and Wales. Windgather Press
Bruck, Joana. The Architecture of Routine Life. In Joshua Pollard (ed.), Prehistoric Britain. Blackwell
Clwyd-Powys Archaeological Trust (leaflet – no date). Moel y Gaer.
Cunliffe, Barry 2005 (4th edition). Iron Age Communities in Britain. Routledge
Garner Dan (and contributors) 2016. Hillforts of the Cheshire Ridge. Investigations undertaken by The Habitats and Hillforts Landscape Partnership Scheme 2009–2012. Archaeopress
Guilbert, Graeme 1975. Moel y Gaer: an area excavation on the defences. Antiquity 49, 109-117
Guilbert, G., 1977. The Northern Welsh Marches: some recent developments. In J. Collis (ed), The Iron Age in Britain — A review. University of Sheffield, p.41–50.
Guilbert, Graeme 2018. Historical Excavation and Survey of Hillforts in Wales: some critical issues. Internet Archaeology 48
https://intarch.ac.uk/journal/issue48/3/toc.html
Davies, Jeffrey L. and Frances Lynch 2000. The Late Bronze Age and Iron Age. Frances Lynch, Stephen Aldhouse-Green and Jerrrey L Davies (eds.) Prehistoric Wales. Sutton Publishing, p.79-138
Oswald, A 1997. A doorway on the past: practical and mystic concerns in the orientation of roundhouse doorways. In (eds.) A. Gwilt and Colin Haselgrove. Reconstructing Iron Age Societies. Oxbow Monograph 71, p.87-95
Websites
Atlas of Hillforts of Britain and Ireland
By Gary Lock and Ian Ralston 2024
https://hillforts.arch.ox.ac.uk
Butser Ancient Farm Archive
The Moel y Gaer House
https://www.butser.org.uk/page43.html
Heneb Historic Environment Record
Moel y Gaer Rhosesmor hillfort, Iron Age phase
https://archwilio.org.uk/her/chi3/report/page.php?watprn=CPAT100296



















































































































































































































































































































