A Roman altar found beside Hadrian’s Wall once carried bright red lettering, purple and blue decoration, and a surface that looked like gold. Today much of that colour is difficult to see. A study published in February 2026 has examined the surviving traces to explore how the altar might have looked when it stood inside a temple.
The research, by Louisa Campbell, Margaret Smith and Sarah Dugmore of the University of Glasgow, concerns an altar from Carrawburgh in Northumberland. Although the findspot is south of Scotland, it offers a useful comparison with the altars discovered at Inveresk, near Musselburgh. It also reminds us that the appearance of a Roman stone in a museum can be very different from its appearance in use.
An altar in a temple to Mithras
Carrawburgh was a Roman fort beside Hadrian’s Wall. In 1950, excavators investigated a nearby temple dedicated to Mithras and found three stone altars. These date to around the third century AD and are now held by the Great North Museum: Hancock in Newcastle.
An altar was a focus for religious dedication and offerings. Its inscription could name the god, the person paying for it and the reason for the gift. The altar examined in this study was dedicated by Marcus Simplicius Simplex, a prefect, or military commander. His inscription records the fulfilment of a vow to Mithras.
Mithras was worshipped by groups of male initiates, including soldiers and officers. Their meeting places are called mithraea; a single temple is a mithraeum. These were enclosed spaces used for ceremonies and communal meals. Much of what we know comes from their buildings, inscriptions and carvings, rather than a surviving account explaining all their rituals.
For the wider background, see our introduction to gods and beliefs in Roman Scotland and our guide to Roman altars.
How can researchers find paint that has almost disappeared?
The study combines close-up photography with several forms of scientific analysis. Portable X-ray fluorescence helps identify chemical elements on the stone. Infrared spectroscopy and X-ray diffraction help investigate the compounds present in tiny samples. Used together, these methods can distinguish some ancient painting materials from the sandstone and from substances introduced during conservation.
The technical term for decoration in several colours is polychromy. A pigment is the material that gives paint its colour. The researchers found that the altar’s painters mixed pigments to produce particular shades, rather than simply applying one mineral for each colour.
The evidence is uneven. Some painted areas retain useful traces, while cleaning, repairs and deterioration have made others much harder to interpret. Detecting the ingredients of paint does not always establish its exact original shade.
Red letters, purple paint and a golden finish
One of the clearest findings concerns the letters of the dedication. These were coloured with cinnabar, also known as vermilion, which produces a strong red. The researchers also identified this pigment on the carved figure’s cloak.
Behind the inscription lay a purple mixture containing blue and red pigments. The frame included further purple and blue decoration, with yellow details. Egyptian blue and ultramarine were among the blue materials identified. On a surviving sunbeam, the evidence indicates yellow ochre mixed with orpiment, another yellow pigment.
The most unusual finding is a golden layer over the purple inscription panel. The authors interpret this as pyrite, often called fool’s gold, combined with beeswax. It would have given the surface a gold-like appearance without requiring gold leaf.
The sequence raises a question that the study cannot settle: was the golden layer part of the original decorative scheme, or was it added later? The researchers favour its application at the same period as the purple. Their further suggestion that it concealed a socially sensitive use of purple is an interpretation, rather than something the pigment tests can demonstrate.
An altar designed for lamplight
The carved figure has a radiating crown with openings through the stone. A recess at the back could hold a lamp, allowing light to shine through around the head. In a darkened temple, that light would have worked alongside the painted surface.
The proposed golden finish adds another possibility: it could have reflected the movement of nearby flames. The altar was therefore more than words cut into stone. Its colour, carved openings and position within the building helped shape what worshippers saw.
The paper includes the researchers’ own colour reconstruction in Figure 20. View the original study and its illustrations. The reconstruction is credited to Louisa Campbell; it visualises the authors’ interpretation of the surviving evidence.
Who does the carved figure represent?
Mithras is often shown killing a bull, but that is not the scene on this altar. Here the imagery is associated with the sun. Sol, the sun god, was closely connected with Mithraic worship.
The authors propose that the figure represents the Heliodromus, or Sun-Runner: an initiate holding the sixth of seven recognised Mithraic grades and embodying Sol during ritual. They also suggest that the original arrangement of the three Carrawburgh altars reflected the religious grades of their dedicators.
This is a new reading of the carvings, inscriptions and temple layout. It should be distinguished from the identification of pigments: scientific evidence for red paint does not, by itself, establish a person’s position within a religious group.
The connection with Inveresk
At Inveresk, near Musselburgh, archaeologists also found altars connected with Mithraic worship. The altar to Sol has a face surrounded by pierced rays, providing a particularly useful comparison with the use of light at Carrawburgh. Our National Museums Scotland video page introduces the Inveresk finds.
The distinction between the two sites matters. The pigment analysis discussed here was carried out on the Carrawburgh altar, not the Inveresk altar. Applying the Carrawburgh palette to an image of the Inveresk stone can help us imagine a painted monument, but the resulting image is a conjectural illustration. It is not evidence that Inveresk used the same colours in the same places.
What the study adds
The research gives us a more detailed picture of the skill and materials involved in decorating a Roman altar. It also shows why surviving stone colour can be misleading. An apparently plain inscription may once have stood out in red against a carefully painted and reflective background.
For Scotland’s Roman monuments, the value lies in both the comparison and the method. Each stone needs its own investigation, but Carrawburgh shows the kinds of evidence that can survive—and how much can be lost if a surface is cleaned before those traces are recorded.
References and copyright
Campbell, L., Smith, M. and Dugmore, S. (2026), “Multi-Technique Analysis and Digital Reconstruction of Polychromy on a Mithraic Altar from Carrawburgh Roman Fort near Hadrian’s Wall”, Colorants 5(1), article 6. Read the open-access paper. This is the principal source for the research findings summarised above.
For the original Carrawburgh excavation: Richmond, I. A. and Gillam, J. P. (1951), The Temple of Mithras at Carrawburgh, Archaeologia Aeliana, fourth series, 29, pp. 1–92. For Inveresk: Hunter, F., Henig, M., Sauer, E. and Gooder, J. (2016), “Mithras in Scotland: A Mithraeum at Inveresk (East Lothian)”, Britannia 47, pp. 119–168. Both are referenced in the 2026 study.
This post is an original plain-language summary of the research. The 2026 paper is © the authors and published under a Creative Commons Attribution 4.0 licence. Its illustrations retain their individual credits; several photographs and other images are reproduced there by permission of third parties. We link to those illustrations rather than reproducing them here. Figure 20 is credited © Louisa Campbell.
