Subscribe now for full access and no adverts
The rediscovery of painted imagery within Bacon Hole has brought renewed attention to a cave that was already recognised for its remarkable archaeological and palaeoenvironmental record. Recent research undertaken by the First Art Team (working alongside specialists from the Universities of Southampton and Swansea, and Nanjing Normal University in China) has combined archaeological recording and an arsenal of scientific techniques – pigment analysis, through Raman spectroscopy (m-Raman), and uranium–thorium dating – to investigate whether the prehistoric ‘paintings’ first identified in 1912 are indeed the result of human activity, and, if so, whether it is possible to determine their age.
Bacon Hole lies within the parish of Pennard on the southern coastline of Gower. Today, the cave entrance looks across the Bristol Channel, but during the Late Upper Palaeolithic (around 14,800-10,000 years ago) this landscape would have appeared very different. Sea levels were considerably lower at this time, and large areas that now lie beneath the sea would have formed open plains. The vast, treeless landscape would have supported migrating animals like bison, elk, horse, mammoth, and reindeer, providing opportunities for hunter-gatherer communities moving through this changing Ice Age environment.
As for the cave itself, Bacon Hole has a large entrance, facing south-west, approximately 18m (60ft) wide and 6m (20ft) high – but thousands of years of erosion, marine activity, and freeze-thaw processes have changed its original shape. It is likely that the entrance once extended much further south, before sections of the limestone ceiling collapsed and retreated to the position visible today. The original entrance would have sat over a rock platform that once formed part of a raised beach dating to the last interglacial period, between 130,000 and 115,000 years ago. Inside, a main gallery extends around 56m (180ft) into the cliff, and there is also a smaller eastern side chamber, approximately 7.5m (25ft) in length, where the painted panel is located. This side chamber contains areas of naturally occurring red haematite, a mineral widely used by prehistoric communities throughout Europe to create pigments. Had one such community used this mineral to make their own marks on the cave wall?
Early investigations
Scientific interest in Bacon Hole began during the 19th century: in 1850, Colonel E R Wood carried out early investigations within the cave and recovered important Pleistocene animal remains. Although these excavations lacked the detailed methods of modern archaeology, they demonstrated that Bacon Hole preserved an important record of ancient environments. Further archaeological work then took place throughout the 20th century, including excavations by Rutter and Mason between 1948 and 1956, investigations involving the Natural History Museum during the 1970s, and later work in the 1980s, which revealed a complex sequence of deposits. The layers documented by these researchers attest to changing environmental conditions across a period spanning 130,000 years, revealing a number of significant climate indicators that showed, for example, evidence of an interglacial before the onset of the last Ice Age.
The cave deposits contained evidence for both warm and cold climatic episodes, too. During warmer periods, animals such as red deer would have occupied the surrounding landscape, but during colder phases associated with the Devensian glaciation, the environment would have become open and harsh, supporting species adapted to colder conditions, including reindeer and wolverines, all of which would have represented tempting prey for prehistoric hunter-gatherers. While Bacon Hole produced evidence for later human activity, though – including Iron Age pottery, Roman material, early medieval objects like Saxon- style beads, medieval pottery, and other artefacts showing that people had repeatedly visited the cave over relatively recent years – direct evidence for Upper Palaeolithic occupation inside Bacon Hole remains limited, despite this long palaeoenvironmental sequence.
The wider Gower landscape, however, contains some of the most important Ice Age archaeology in Britain. Nearby Goat’s Hole at Paviland contained the famous Upper Palaeolithic burial known as the ‘Red Lady’ of Paviland, an ochre-adorned individual who was, at the time of discovery in 1823, initially interpreted as a Roman woman, but has more recently been revealed to be a young man who died c.33,000-34,000 years ago. Meanwhile, Cathole Cave, only a few kilometres to the north-west, is home to an engraving of an animal figure, probably a stylised reindeer, which was dated to the Late Upper Palaeolithic (dated to 14,505±560 years ago; see CA 259 and 270).
Could Bacon Hole be similarly significant? The most controversial discovery associated with the site came in 1912, when Professor William Sollas and the French prehistorian Henri Breuil examined the small side chamber east of the main gallery. On the limestone wall, they identified a series of red horizontal lines, partly hidden beneath calcite deposits. They suggested that these markings represented prehistoric cave art and possibly the first known example of Palaeolithic painted imagery in Britain. At the time, this was an extraordinary claim. Breuil was one of the leading authorities on European cave art and had studied many of the famous painted caves in France and Spain. The suggestion that similar symbolic activity existed in Wales placed Bacon Hole within a much wider European tradition of Ice Age artistic expression.
In 1913, W L Morgan published an account of the discovery and produced the first-known photograph of the painted panel: an image showing a series of horizontal bands crossing the cave wall. Morgan supported the interpretation proposed by Sollas and Breuil, and recognised the potential importance of the find. However, doubts soon developed. By 1928, many researchers had rejected the idea that the markings represented human activity, and some suggested that the red lines were natural mineral-staining rather than deliberate painting. The presence of calcite covering sections of the panel made detailed examination difficult, and the discovery gradually disappeared from wider archaeological discussion.
Rediscovery and research
For almost a century, the panel remained largely forgotten – but in August 2022 it was rediscovered by Dr Barbara Oosterwijk from the University of Exeter, along with several members of the First Art Team (an international group of researchers specialising in analysing rock art from around the world) including me, while surveying the cave on behalf of the National Trust, in whose care the site lies. This development reopened an important archaeological debate. Had two of the most experienced prehistoric researchers of the early 20th century correctly identified Ice Age art back in 1912, or were the markings simply natural features? Following confirmation of the rediscovery in 2022, the National Trust and Cadw were informed, and funding from the National Trust and the Bradshaw Foundation allowed a new programme of scientific investigation to begin.
The first stage of our project involved detailed photographic recording of the side chamber containing the ‘painted’ panel. There, the team used high-resolution photography combined with DStretch image enhancement software – a technique that increases subtle colour differences and is now widely used in the study of ancient rock art. This survey confirmed the presence of the ten horizontal red lines that were originally recorded in 1912, but it identified additional areas of pigment, too, including possible finger marks, geometric forms, and pigment spreads that may have been created by blowing or projecting colour on to the wall surface.
There were a number of rather more recent markings as well, in the form of historic graffiti. One notable contributor was Jonny Bates, a fisherman from nearby Oystermouth, who created painted images within the cave during the late 19th century, but other names and dates written in graphite survive across parts of the chamber. These later additions demonstrate the continuing relationship between people and the cave, but also complicate the study of earlier imagery.
To establish whether the red lines were created by humans, we needed to understand the pigment itself. Small samples were collected and examined using advanced scientific techniques including Fourier transform infrared spectroscopy (FTIR) and micro-Raman spectroscopy. These methods allow researchers to identify the chemical composition of extremely small samples, and the results demonstrated that the red pigment was dominated by haematite, the iron oxide mineral that, as mentioned above, was commonly used throughout prehistory. Crucially, though, analysis using a scanning electron microscope (SEM) also identified traces of clay minerals. This suggests that the pigment may not simply represent natural staining but could have involved the preparation of a mixture before application. Importantly, too, the haematite was concentrated within the painted areas rather than following natural geological formations within the limestone geology. The arrangement of the lines also provided illuminating evidence, as the panel consists of ten repeated horizontal marks placed in a deliberate and structured pattern. Such regular spacing and organisation are difficult to explain through natural mineral processes.
Clues from calcite
The team also examined how the pigment related to later calcite formation that had formed over the painted imagery. In limestone caves, water movement slowly creates mineral deposits that can cover earlier surfaces. Where calcite forms above pigment, the calcite must be younger than the painting beneath it. This relationship allowed us to undertake uranium–thorium dating, a method that measures known radioactive decay within calcite and can establish when the mineral layer formed. Although it does not directly date the pigment, it provides a minimum age because the painting must already have existed before the calcite covered it.
During the first sampling campaign in 2023, several calcite samples produced relatively young dates associated with later phases of mineral growth. However, one sample produced a much older result of approximately 17,100 years ago. This result is extremely significant, but must be interpreted carefully. It does not mean that the painting was created exactly 17,100 years ago. Instead, it indicates that the pigment may already have been present before this calcite formed. Further sampling is needed to confirm this early date, but the result raises the possibility that the panel belongs to the Late Upper Palaeolithic.
A second sampling campaign in 2024 provided additional information about the history of calcite-formation within the cave. The results showed further phases of mineral growth, demonstrating the complex geological processes that have affected the panel since its creation. If the earliest dating evidence is confirmed, Bacon Hole would represent one of the most important examples of Ice Age symbolic behaviour in north-western Europe.
Around 17,100 years ago, Wales was emerging from an intensely cold period of the last glaciation. Although the ice had retreated, the environment remained challenging. At this time, average summer temperatures were around −10°C. Vegetation was limited and the landscape was largely open; the area now forming the Bristol Channel would have been dry land crossed by animals and potentially by human groups during the warmer summer months.
The caves of Gower would have been highly visible places within this landscape. Even if Bacon Hole was not used as a place to live, it may have held another form of significance for the people who entered it and marked its walls. Based on the length of the cave, there could have been two zones of activity: a public space located within the entrance area, where a family unit could have socially convened, and another within the dark, deeper recesses of the cave that might have been regarded as a private, even sacred, space. It is within the latter zone that the painted panel is located – and it should be noted that, during use, the side chamber would have been completely devoid of natural light. This organisation of cave space appears to have been common during early prehistory, with clear separation of social and ritualised spaces seen at many other sites.
Back to the future
The rediscovery of Bacon Hole also demonstrates the importance of revisiting old archaeological discoveries. Early researchers such as Sollas and Breuil worked without the scientific techniques available to us today, yet modern evidence suggests that their original interpretation deserves serious consideration.
Our recent project’s combination of visual recording, chemical analysis, and dating has transformed our understanding of the panel, with the evidence indicating that its red markings were intentionally applied using haematite pigment, and were later sealed beneath natural calcite deposits.
Today, the side chamber containing the painted panel is protected by a steel grille installed by the National Trust. This protects both the fragile archaeological evidence and the important bat populations that use the cave. Further research will continue to investigate the age and significance of the imagery, however, and it is hoped that additional sampling, recording, and analysis may reveal more about the people who created these markings and their place within the wider story of European prehistoric art.
More than a century after Sollas and Breuil first entered the chamber, Bacon Hole has returned to the centre of discussions about Britain’s earliest artists. What was once dismissed as a geological curiosity may instead represent a rare survival of human symbolic expression from the final stages of the Ice Age. The story of Bacon Hole is therefore not only about a painted wall hidden within a Welsh cave. It is about how archaeology changes as new methods allow forgotten discoveries to be seen again, and how traces left by people thousands of years ago can still transform our understanding of the prehistoric past.
Source:
George Nash is an Associate Professor at the Geosciences Centre, IPT (u.ID73FCT); ITM (Earth and Memory Institute), Portugal; a founder member of the Welsh Rock Art Organisation, and honorary Researcher at the University of Liverpool.
G H Nash, H Collado, H Gomes, S Garcês, V Lattao, P Rosina, E Marrocchino, N Eftekhari, B Oosterwijk, A W G Pike, D L Hoffmann, C D Standish, J F Hiemstra, and Q Shao (2026), ‘Rediscovered Late Upper Palaeolithic painted imagery at Bacon Hole, Gower Peninsula, South Wales’, Quaternary 9 (43): https://doi.org/10.3390/quat9030043.
All images: George Nash unless otherwise stated
