19. Through Celts and Romans: Technology and Symbolism of Bronze Enameled Roosters

  • Federica Grossi, Università degli Studi di Milano


One of the most interesting groups of Roman bronze metalwork spans the second and third centuries AD. It comprises about ten enameled statuettes portraying roosters, possibly standing on bases. Their provenance is attested at sites of the Western Empire, possibly connected with Celtic art. The roosters are portrayed in attack position, with the beak open and the crest up. Their chests are decorated with triangles or lozenges of multicolored champlevé enamels and they have detachable backs. The recent find of a similar item, complete with its tail, has brought renewed attention to these objects, but they are still rarely published or are mentioned only in old scholarship. There are still open questions about them, such as how they were made and why, with such an elegant and precious technique. Furthermore, there are debatable aspects of both the metalworking and the shapes of these objects, which are probably connected with the Celts, as well as problems related to ateliers and workshops. Finally, it would be interesting to know more about their function, whether as lamps, containers, or simple decorative statuettes.


The symbology characterizing the Roman world, as much as its artistic production, often manifests a beautiful syncretism: there are many cases in which a theme, a figure, or a meaning is rearranged and presented in a different way, and it is often possible to see that their roots are planted far away. From this point of view the Celts were very important, because their myths, rites, and interpretations influenced Roman ones over a long span of time. If we add that the technique and technology used for the objects under discussion are originally Celtic, the bond between these two cultures will become clearer, as will the artistic debt that the Romans owe to the Celts. The purpose of this paper is to describe in detail a group of bronze enameled roosters—fascinating objects related to the larger group of enameled vessels—trying to understand their background, technology, age, and function.

Roosters and Mercury: Ancient Sources and Archaeological Finds

In the Greek and Roman worlds, the figure of the rooster was frequently associated with gods such as Asclepius, Minerva, and Mars, and represented different aspects of their personalities or spheres of influence, such as healing, readiness, and pugnacity.1 Nevertheless, the god most often portrayed with a rooster is Mercury, a central character in trade and communication: he was not only the divine messenger but also represented the means and ways of communication. Mercury was the connection between different levels of existence, as they were understood at the time: divine, mortal, and afterlife. In other words, he was an entity at the limits in every sense. No one was more suited to this role: he was born at dawn, between night and day; he was a thief and a gambler, but also a merchant and a gifted musician; excess and moderation in a single god. Probably for these reasons, he was also the chaperone of souls, and the rooster is one of his totemic animals: it is fierce and smart, and, above all, it too knows the borderlines, because it announces the rising sun with its voice, obliterating darkness.

Greeks and Romans were not the only ones to think this way. Caesar (Bellum Gallicum 6.17.1) asserts with no doubt the supremacy of the god in Gaul. He says: “Amongst the gods, they worship Mercury above all and he is the one with the most numerous representations.”2 Lucan (Pharsalia 1.444–46), on the other hand, presents three strong, cruel Celtic gods—Taranis, Esus, and Teutates—the latter two of which, according to their attributes, could have been Mercury;3 Taranis was recognized as Jupiter. The matter of their identification with Mars, Mercury, or simple local gods is still open and scholars are trying to determine the truth, but it’s difficult because attributes blur one into the other, dialects vary, and there is no definitive archaeological evidence.4 Nevertheless, in Gaul, Germany, and Britain, there are materials such as statues, bronzes, and inscriptions that demonstrate the peculiar devotion to this god of commerce and craftsmanship and that witness the bond between him and the rooster. Among them are a bronze statuette found in a deposit at Dax, in the south of France, portraying the god with a rooster and a goat at his feet;5 a stone altar from Horn, the Netherlands, with a relief decoration of a rooster on top of a caduceus;6 and a silver handle of a pan from Capheaton, England, with Mercury sitting under a pavilion and holding a bag, while a small rooster stands next to him.7 Finally, it’s very important to note findings of chicken bones. A clear example is a grave in a cemetery in Tartigny, Picardy: among the grave goods accompanying someone believed to be an official are about twenty vases, in one of which were found the bones of a chicken and a goose, with their wings and claws cut at the tips.8 Other cases are the sanctuary of Mercury at Uley, Gloucestershire, where scholars have found a large quantity of bones from goats and roosters;9 and the sanctuary of Mirebeau, Burgundy, with similar finds.10 Data of this kind are sparse because chicken bones are small and fragile, so we may not have a good estimate of the frequency or quantities of such deposits. Importantly, though, these finds are in closed, protected contexts, so they were meant to survive. And they prove that those animals were necessary, as a sacrifice for a god and as a link to the afterworld.

Analyzing the Artifacts

Presently we know of nine bronze enameled statuettes of roosters, all of them coming from the western part of the Roman Empire: five from England (London; Cople in Bedfordshire; Cirencester in Gloucestershire; Slyne with Hest in Lancashire; and Drayton Bassett in Staffordshire); two from the Netherlands (Ezinge and Buchten), and one from Belgium (Tongeren). The final one is said to come from Cologne and is now in Bonn (fig. 19.1). The best known and best preserved one is from Cirencester (fig. 19.2): it is a beautiful, complete specimen, the only one with its tail intact, which was discovered in 2011 during the excavation of a Roman burial site that was probably in use from the middle of the second to the fourth century AD. It was the grave of a child, two or three years old and of unknown gender, buried in a nailed wooden coffin with the bronze rooster and a ceramic eating vessel.11 These are simple grave goods, yet very significant ones, particularly if one remembers that the site of Uley, with the sanctuary of Mercury, is only a few miles west of Cirencester. Since many of these objects were found in the nineteenth century outside of archaeological excavations, there are only two others with known contexts, both of them from the Netherlands: the one from Ezinge was found in a burial site (see fig. 19.6),12 and the one from Buchten is probably from a sanctuary (fig. 19.3). This specimen is particularly interesting not only because it still retains its pedestal but also because it bears an inscription that may prove useful in hypothesizing about dating and workshops.13 It seems that there was another rooster from Cirencester that was found in 1870, but it has since been lost.14

Figure 19.1. Bronze rooster with enamel from Cologne. Rheinisches Landesmuseum, Bonn Image: Jürgen Vogel. LVR – Landesmuseum Bonn
Figure 19.2. Bronze rooster with enamel from Cirencester. Corinium Museum, Cirencester Image: © Cotswold District Council, courtesy of Corinium Museum, www.coriniummuseum.org
Figure 19.3. Bronze rooster with enamel from Buchten. Limburgs Museum, Venlo Image: Courtesy of the Limburgs Museum (Venlo, The Netherlands)

Looking at these objects, we can define common elements that lead us to discover subgroups among them. First of all, the mechanisms differ. Most seem to have been made the same way: with a pedestal, a body in a single piece, a tail that slots into the back, and a removable spine. Second, the decoration differs. One type has colorful chest patterns like a slanting checkerboard; if we also consider the shape of neck and eyes and the similarities among beaks and combs, we have three samples probably coming from the same production (Cologne, Buchten, and Cirencester). In a second group, the chest pattern shows many little triangles with rounded angles, which probably points to a different place of production; three samples (Cople, Lancashire, Tongeren) belong to this group (figs. 19.4–5). The three last specimens cannot be ascribed to either group: the one from Ezinge (fig. 19.6) is far too corroded to recognize its pattern; the one from Staffordshire (fig. 19.7) has lost its head; and the one from London has a different structure entirely. It seems that its body was made by the union of an upper and lower part and its decoration was based on moon crescents on chest and back.

Figure 19.4. Bronze rooster from Cople, Bedfordshire Image: Courtesy of the Portable Antiquities Scheme
Figure 19.5. Bronze rooster from Slyne with Hest, Lancashire Image: Courtesy of the Portable Antiquities Scheme
Figure 19.6. Bronze rooster from Ezinge. Groninger Museum, Groningen. Rooster. 1–300 AD. Bronze. Collection Groninger Museum Image: Marten de Leeuw
Figure 19.7. Bronze head of a rooster from Drayton Bassett, Staffordshire Image: Courtesy of the Portable Antiquities Scheme

A Celtic Technique?

Working enamel is difficult and the technique known as champlevé requires great skill, not only for the bronze craftsman but also for the glass maker: preparing and firing enamel is a very difficult process, especially when making such jointed statuettes. The first step was the preparation of supports: the various parts were poured separately into molds that probably already had their hollows to contain the enamel. Then, a mix of colored ground glass and water was spread into the hollows with a spatula or a brush; the excess water was dried; and the bronze pieces were fired. Temperatures had to be watched carefully in order to avoid mistakes or damage, and the artisan had to know which pigments needed more layers of glass and which needed fewer to give perfect results. Finally, the artisan waited for the object to cool down before proceeding with cleaning and assembly. At least, this is how Romans used this technology; but how much of it was a Roman development? It is pertinent therefore to discuss some issues about the origins of enamel.15

Let us first look at what Philostratus, a Greek sophist who lived in Rome during the reign of Septimius Severus, had to say on the subject. In one of his essays he wrote: “These pigments, it is said, the Barbarians living by Oceanus compound of red-hot bronze and they combine and grow hard and preserve what is painted with them” (Philostratus Imagines 1.1.28). There are two main points to raise about this statement. First, Philostratus speaks about Barbarians who execute this technique as artisans or well-trained workers. Second, he probably went to Britain with the emperor during military campaigns defending the limes, between AD 208 and 211.16 Furthermore, some recent studies have suggested the possibility that there was a significant workshop in Castleford, south of York, where excavations have turned up two large pits with molds for the making of enameled objects.17 On the continent, however, the situation remains puzzling: suggestions for possible workshops have been proposed in France and Belgium, but there is no clear evidence of their existence.18

In light of this information, it should be simpler to see common elements that associate individual roosters and that distinguish particular workshops, but there is something further to be taken into account. If it is true that the concept and the technology are Celtic, it is also clear that Romans refined them and made them their own. In analyzing the pieces, we can see that—based on contexts, styles, or comparisons—they should be ascribed to the second century AD. The specimen from Buchten can help illuminate the situation (see fig. 19.3). It is a rooster like the others, but on its pedestal there is a Latin inscription reading: “Ulpius Verinus, veteran of the Sixth Legion, consecrates this to the goddess Arcanua, discharging a vow to her, who merits it.”19 The inscription states that Ulpius Verinus was a veteran, meaning he served as a soldier for a long time, at least 20 or 25 years. Furthermore, he was a veteran of the sixth legio Victrix, which was in Novaesium (Neuss) from AD 71 to 100, then in Castra Vetera (Xanten) from AD 103 to 122, and finally in Eburacum (York). This is also relevant, because York is quite near the site of Castleford—the ancient Roman Lagentium—where the enamel molds were found. Finally, it’s important to remember that Philostratus personally witnessed enameling and that his journey in Britain with Septimius Severus took place at the very beginning of the third century AD. With these premises, it is possible to hypothesize that Ulpius Verinus bought the rooster near Hadrian’s Wall, brought it back to Germany at the end of his service, and offered it to the local goddess Arcanua. If this theory is plausible, it would mean that there was a workshop producing this kind of precious object connected to the military service. Notably, five sites—three in England and one each in Spain and France20—have yielded small bowls that, in detail, are distinctive in their decoration, demonstrating a link between enameled vessels and Hadrian’s Wall. These objects not only represent a stylized wall but also give us names that correspond to Roman forts along Hadrian’s Wall. They are currently interpreted as souvenirs from the Wall, perhaps produced in an atelier near the Wall itself.21


To conclude, it is necessary to devote some words to the function of these artifacts. Surely, given the difficulty of their production and the preciousness of their materials, they must have been very expensive, true luxury goods. Scholars have proposed a number of hypotheses about their use. The first and least persuasive one asserts that the roosters could have been lamps, with the cavity in the back filled with oil.22 This is not possible for several reasons: for example, it would have been too difficult to insert the wick along the body and to make it come out from the beak; also, the fire could have damaged the object causing the enamels to detach. Finally, it has been shown that the spine was soldered to the back and that it served to fix the tail, so the cavity was not functional. This discovery, made especially clear by the complete Cirencester specimen, leads us also to reject the idea that these roosters were boxes: they could not contain any kind of treasure, nor jewels like necklaces, earrings, or bracelets, nor spices, essences, or incense. This use having been ruled out, the only remaining possibility is that the enameled roosters were not functional at all but were instead beautiful decorative statuettes that were intended for private use as precious objects to display. This interpretation is not all that astonishing: given the link between Mercury and the afterworld, the statuette of a rooster could have had an apotropaic value: keeping an object of this kind could have meant that the owner trusted the god for the journey at the end of his (or her) life. Over time, this meaning was forgotten and modified. This allows us to introduce a new and stirring hypothesis. We know that the bird has been a symbol of passage or a channel into another dimension since the beginnings of the Christian religion, but the innovation is that it’s no longer represented by a rooster, but rather by a dove. Images of doves become more frequent over the centuries of Late Antiquity, and by the Middle Ages—between the twelfth and the thirteenth centuries—they will form a new type of liturgical furniture known as “eucharistic doves.”23 These spread especially across France and, most importantly, show similarities with the roosters under discussion: they have a spiritual or religious meaning; they are metallic birds composed of joined pieces; and, especially, they are enameled in champlevé technique (the most famous site of production was in Limoges).24 Hence it is possible that, after the role of roosters and their association with Mercury became increasingly obsolete, western Christians adopted and adapted an ancient model for their own liturgical needs. Surely, this wouldn’t be the first case.


  1. Pintus 1986, 243–46.
  2. Caesar Bellum Gallicum 6.17.1: Deum maxime Mercurium colunt. Huius sunt plurima simulacra. hunc omnium inventorem artium ferunt, hunc viarum atque itinerum ducem, hunc ad quaestus pecuniae mercaturasque habere vim maximam arbitrantur.
  3. Lucan Pharsalia 1.444–46: Et quibus inmitis placatur sanguine diro / Teutates horrensque feris altaribus Esus / et Taranis Scythicae non mitior ara Dianae.
  4. On the discussion on Esus and Teutates as Mercury and Mars, see Sjoestedt 1949, 21–23. See also Duval 1976, 27–31, 69–73; and Kruta 1997, 536–39.
  5. Santrot et al. 1996, 260–82.
  6. Hardenberg 1946, 5–42.
  7. Brailsford 1964, 41, plate 10.49.
  8. Massy 1986, 16–18.
  9. Van Andringa and Lepetz 2002, 89.
  10. Goguey 1979; Barral and Joly 2011; Joly and Barral 2008.
  11. Hilts 2013, 28–34.
  12. The exact word to describe the structure is hutkom or komhut, as Hoss et al. (2015, 166) say: it indicates a rectangular pit covered with a sloping roof. In contrast, Zadoks-Josephus Jitta, Peters, and Van Es (1967, 114) only mention a provenance from the artificial hillock of Ezinge.
  13. Hoss et al. 2015, 159–71.
  14. Hoss et al. 2015, 167.
  15. For a complete history of enamel and its development, see Henry (1933, 65–146).
  16. There are many problems regarding Philostratus and his historical accuracy, because there were four philosophers with the same name and place of birth. It is difficult to distinguish one from the other and it is even more complicated to establish a precise chronology for each one. However, only two of them are said to be the possible author of the text known as Imagines. They are Philostratus of Athens and Philostratus the Elder; both of them came to Rome to serve the emperor Septimius Severus, so it is possible that one of them followed the court to Britannia.
  17. Bayley 1995.
  18. Bequet 1900; Goudineau and Peyre 1993.
  19. The goddess Arcanua (or Arkanua) is only mentioned twice in epigraphy (AÉpigr 1983, nos. 723–24): she was probably a local goddess, whose sanctuary was found and excavated in the Netherlands. See Derks 2015a, 150–54; Toorians 2015; Derks 2015b, 173–76.
  20. Rudge Coppice, Ilam, Bath, in England; Zamora, Spain; and Amiens, France.
  21. Breeze 2012; Hunter 2016, 136–39.
  22. Faider-Feytmans 1979, 134–41.
  23. McLachlan 2005.
  24. Boehm and Taburet-Delahaye 1996.


Barral and Joly 2011
Barral, P., and M. Joly. 2011. “Le sanctuaire de Mirebeau-sur-Bèze.” In Aspects de la romanisation dans l’Est de la Gaule, ed. M. Redde, P. Barral, F. Favory, M. Joly, J.-P. Guillaumet, J.-Y. Marc, P. Nouvel, L. Nuniger, and C. Petit, 543–56. Collection Bibracte 21. Glux-en-Glenne: Centre archéologique européen.
Bayley 1995
Bayley, J. 1995. “Spoon and Vessel Moulds from Castleford, Yorkshire.” In Acta of the 12th International Congress on Ancient Bronzes, Nijmegen 1992, ed. S. T. A. M. Mols, A. M. Gerhartl-Witteveen, H. Kars, A. Koster, W. J. Th. Peters, and W. J. H. Willems, 105–11. Nederlandse Archeologische Rapporten 18. Nijmegen: Provincial Museum G. M. Kam.
Bequet 1900
Bequet, A. 1900. “La bijouterie chez les Belges sous l’Empire Romain (IIe siècle).” Annales de la Societé Archéologique de Namur 24: 237–76.
Boehm and Taburet-Delahaye 1996
Boehm, B. D., and E. Taburet-Delahaye, eds. 1996. Enamels of Limoges, 1100–1350. New York: The Metropolitan Museum of Art.
Brailsford 1964
Brailsford, J. W. 1964. Guide to the Antiquities of Roman Britain. 3rd ed. London: Trustees of the British Museum.
Breeze 2012
Breeze, D. J., ed. 2012. The First Souvenirs: Enamelled Vessels from Hadrian’s Wall. Kendal: Titus Wilton & Son.
Derks 2015a
Derks, T. 2015. “Inscripties en graffiti.” In Derks and de Fraiture 2015, 148–55.
Derks 2015b
Derks, T. 2015. Summary to Een Romeins heiligdom en een vroegmiddeleeuws grafveld bij Buchten (L). In Derks and de Fraiture 2015, 173–81.
Derks and de Fraiture 2015
Derks, T., and B. de Fraiture, eds. Een Romeins heiligdom en een vroegmiddeleeuws grafveld bij Buchten (L). Rapportage Archeologische Monumentenzorg 226. Amersfoort: Rijksdienst voor het Cultureel Erfgoed.
Duval 1976
Duval, P. M. 1976. Les dieux de la Gaule. Petite bibliothèque Payot 298. Paris: Payot.
Faider-Feytmans 1979
Faider-Feytmans, G. 1979. Les Bronzes Romains de Belgique. Mainz am Rhein: Philipp von Zabern.
Goguey 1979
Goguey, R. 1979. “Le Sanctuaire gallo-romain de Mirebeau.” Mémoires de la Commission des antiquités de la Côte-d’Or 31: 169–206.
Goudineau and Peyre 1993
Goudineau, C., and C. Peyre. 1993. Bibracte et les Eduens: A la découverte d’un people gaulois. Paris: Editions Errance.
Hardenberg 1946
Hardenberg, H. 1946. “L’Autel de Mercure Arverne à Horn.” AntCl 15.1: 5–42.
Henry 1933
Henry, F. 1933. “Émailleurs d’Occident.” Préhistoire 2.1: 65–146.
Hilts 2013
Hilts, C. 2013. “Corinium’s Dead: Excavating Cirencester’s Tetbury Road Roman Cemetery.” Current Archaeology 281: 28–34.
Hoss et al. 2015
Hoss, S., J. Kempken, and T. Lupak. 2015. “Bronzen beeldje van een haan met voetstuk.” In Derks and de Fraiture 2015, 159–71.
Hunter 2016
Hunter, F. 2016. “The Impact of Rome, c. AD 50–250.” In Celts: Art and Identity, ed. J. Farley and F. Hunter, 128–52. London: The Trustees of the British Museum and National Museums Scotland.
Joly and Barral 2008
Joly, M., and P. Barral. 2008. Archéologie en Bourgogne: La longue histoire d’un sanctuaire lingon à Mirebeau-sur-Bèze (Côte-d’Or). Archéologie en Bourgogne 12. Dijon: DRAC Bourgogne.
Kruta 1997
Kruta, V. 1997. “La Religione.” In I Celti, ed. S. Moscati, O. Hermann Frey, V. Kruta, B. Raftery, M. Szabó, E. Arslan, and D. Vitali, 533–42. Milan: Bompiani.
Massy 1986
Massy, J.-L. 1986. “La Nécropole gauloise de Tartigny (Oise): Étude des structures.” Revue archéologique de Picardie 3.4: 13–22.
McLachlan 2001
McLachlan, E. P. 2001. “Liturgical Vessels and Implements.” In The Liturgy of the Medieval Church, ed. Th. J. Heffernan and E. A. Matter, 369–429. Consortium for the Teaching of the Middle Ages. Kalamazoo: Western Michigan University.
Pintus 1986
Pintus, G. M. 1986. “Storia di un simbolo: il gallo.” Sandalion 8–9: 243–67.
Santrot et al. 1996
Santrot, J., B. Derion, X. Dupuis, M. Fincker, and P. Guibert. 1996. “Bronzes et fers de Dax, Landes: La cachette d’un antiquaire-restaurateur au IV s. après J.-C.” Gallia 53: 251–343.
Sjoestedt 1949
Sjoestedt, M.-L. 1949. Gods and Heroes of the Celts. London: Methuen & Co.
Toorians 2015
Toorians, L. 2015. “Naamkundige analyse van het theoniem Arcanua.” In Derks and de Fraiture 2015, 156–57.
Van Andringa and Lepetz 2002
Van Andringa, W., and S. Lepetz. 2002. “Le ossa animali nei santuari: per un’archeologia del sacrificio.” In Sanctuaires et sources dans l’antiquité: Les sources documentaires et leurs limites dans la description des lieux de culte, ed. O. de Cazanove and J. Scheid, 85–96. Actes de la table ronde organisée par le Collège de France, l’UMR 8585 Centre Gustave-Glotz, l’Ecole française de Rome et le Centre Jean Bérard. Collection du Centre Jean Bérard 22. Naples: Centre Jean Bérard.
Zadoks-Josephus Jitta, Peters, and Van Es 1967
Zadoks-Josephus Jitta, A. N., W. J. T. Peters, and W. A. Van Es. 1967. Roman Bronze Statuettes from the Netherlands. Vol. 1: Statuettes Found North of the Limes. Groningen: J. B. Wolters.