1The treasury of Saint-Aubain Cathedral in Namur (Belgium) has housed, since the thirteenth century, a leather-covered wooden casket decorated with enamel medallions (see fig. 1).1 This casket is believed to have been created to protect the Reliquary-Crown of the Holy Thorns, the cathedral’s most esteemed object and a masterpiece of mediaeval goldsmiths’ craftsmanship. Both the casket and the crown date from the thirteenth century. The long-term preservation of this rare and delicate casket, its association with a significant example of goldsmiths’ work, and the rich shared history of the casket and crown, linked to the Crusades, justify their designation as ‘Treasures’ of the French Community (Moniteur Belge, 13/03/2012).
2Over the years, the casket suffered material deterioration. The leather had dried out, leading to cracks and tears in several areas; some fragments were at risk of detaching. Due to the presence of xylophagous insects in other objects stored in the same room, there was also a risk of infestation, although no visible damage was observed on the casket itself. Conservation treatment was therefore necessary, mainly to clean the object, treat the wood, and stabilise both the leather and the enamels. The museum saw this as a unique opportunity to undertake a material study of the object and to explore several unresolved questions: the dating of the wood; the material history of the casket; the authenticity of the enamels and the possibility of later re-enamelling; the dating of the lining fabric; the potential presence of an earlier textile underneath; and so on. Furthermore, the Namur casket belongs to a broader group of twelfth- and thirteenth-century caskets decorated with enamelled medallions, traditionally attributed to workshops in the Limousin region. Its study could provide new data for this typological group as a whole.
3The granting of a subsidy by the French Community in 2019 allowed the project to become a reality. The Royal Institute for Cultural Heritage in Brussels (KIK-IRPA) was the ideal institution to undertake an interdisciplinary treatment and study of the casket, thanks to its expertise in multiple fields and the availability of advanced analytical equipment within its laboratories. The treatment of the leather and metallic components – tasks not covered by KIK-IRPA – was delegated to external specialists, Zeina Genadry, and Françoise Urban. Each element of the casket was examined and treated as follows: wood, including dendrochronological analysis (Violette Demonty, Christophe Maggi); leather (Zeina Genadry); metal components (Helena Wouters, Françoise Urban); enamel (Helena Wouters, Géraldine Bussienne); and textile (Ina Vanden Berghe, Judith Goris, Griet Kockelkoren).
4After being transferred to KIK-IRPA, the casket attracted the attention of collaborators from the ArtGarden project (Art Technical Research and Preservation of Historical Mixed-Media Ensembles: Enclosed Garden), which was ongoing at that time. Made from various materials, the casket proved to be an ideal example of a historical mixed-media object. At the end of 2020, it was formally included in the ArtGarden project as a case study. This enabled the scope of investigation to be further expanded, with the aim of deepening understanding of material interactions and related degradation mechanisms. In this context, Lieve Watteeuw conducted a detailed study of the leather, while Marjolijn Debulpaep and Elke Otten examined the casket within its conservation environment to support the development of AGATO, a decision-support tool for museum professionals.
- 2 See O’Neill et al. 1996; Castronovo and Descatoire 2016.
5Due to its connection with the Reliquary-Crown of the Holy Thorns, the casket of Saint-Aubain Cathedral drew the attention of art historians as early as the nineteenth century and has since been the subject of a limited number of publications, mainly in the form of exhibition catalogue entries. Furthermore, it forms part of a broader group of caskets dispersed across treasuries and museums in Europe and the United States, which have also been the focus of numerous studies.2 In what follows, we explore what is currently known about the history of the Namur casket, its position within this wider group, and the questions raised by its enamelled medallions.
6The octagonal casket is made from assembled oak panels and covered in leather (fig. 1). It has a brass lock, and a copper handle mounted on the lid (fig. 2). The base is reinforced with two intersecting metal strips and fitted with four small, hemispherical copper alloy bosses, attached with nails (fig. 3). The exterior is decorated with twenty-five circular enamelled and gilded copper roundels – two on each side panel, and nine on the lid (fig. 4). Additionally, a quadrangular enamelled copper plaque with inwardly curved edges serves as the escutcheon (fig. 4, I). The interior is lined with red damask featuring a fleur-de-lys pattern, and a gilded trim is applied to the edge of the case (fig. 5).
[Fig. 1]
Casket for the Reliquary-Crown of the Holy Thorns, general view after treatment. H.: 12.7 cm; Ø max.: 32 cm. Namur, Cathedral Treasury and Diocesan Museum, inv. 7.
© Brussels, KIK-IRPA, X147012.
[Fig. 2]
Casket for the Reliquary-Crown of the Holy Thorns, lid after treatment. Ø max.: 32 cm.
© Brussels, KIK-IRPA, X147023.
[Fig. 3]
Casket for the Reliquary-Crown of the Holy Thorns, base before treatment.
Working photo (Emmanuelle Mercier).
[Fig. 4]
Casket for the Reliquary-Crown of the Holy Thorns, side panels after treatment.
© Brussels, KIK-IRPA, X146995, X146997, X146999, X147001, X147003, X147005, X147007, X147009.
[Fig. 5]
Casket for the Reliquary-Crown of the Holy Thorns, open view after treatment. H.: 12.7 cm; Ø max.: 32 cm. Namur, Cathedral Treasury and Diocesan Museum, inv. 7.
© Brussels, KIK-IRPA, X147013.
7The medallions, affixed with round-headed nails, are all of the same diameter (4.5 cm), except for the central medallion on the lid (fig. 2, 1), which is larger (5.3 cm) and distinguished further by its foliage decoration. Eighteen medallions feature a reserved gold border. Seven are framed by an enamel border: one with a turquoise enamel border (fig. 4, III.2), five with a blue enamel border (fig. 4, I.2, I.3, V.1, VI.1, VIII.2), and the medallion at the centre of the lid – one – with a border combining blue and white enamel (fig. 2, 1).
8Most of the medallions depict a single figure: a fantastic animal with a somewhat elongated snout, two lion-like paws, and a long, curled tail – three of these creatures also have wings. With raised paws and heads turned backwards, biting their flanks or tails, the animals contort themselves to fit the circular shape of the metal support. Six medallions show alternative motifs: a chimera enclosed within a starry circle (fig. 2, 3); two monsters leaning against each other on either side of a central gilded line (fig. 2, 9); a bird stretching its wings (fig. 2, 7); a man being bitten by a creature (fig. 4, VIII.1); a man grasping two snakes (fig. 2, 5); and, at the centre of the lid, a medallion decorated with foliage (fig. 2, 1). The scutcheon depicts a chimaera and a lion facing each other (fig. 4, I.1). The shape of the hasp also suggests a monstrous creature.
9Round-headed nails of varying sizes complete the decoration of the casket. A row of small nails runs along the sides of the casket and around each medallion. On each side panel, seven larger nails – differing slightly in size – are distributed across the available surface, including one placed between the medallions. The exception is the panel with the lock, which features only four nails. On the lid, rows of small nails outline both the medallions and the outer edge, while large nails alternate with decorative clusters of four small nails – a pattern found exclusively on the lid.
10The first study dedicated to the Reliquary-Crown and its casket appeared in a German periodical in 1864.3 Written by archaeologist, art historian, and Director of the Bonn Museum Carl Ernst aus’m Weerth (1829–1909), it was translated shortly thereafter in the
Annales de la Société archéologique de Namur.4 The accompanying plate presents an overall view of the casket along with details of the enamels and the red interior fabric (fig. 6). Aus’m Weerth dated both the crown and its casket to the early thirteenth century and attributed the enamelled medallions to the Rhineland region. He compared the Namur casket to one in the treasury of Aachen Cathedral, which he had examined during its restoration and later published,5 as well as to the Saint Louis casket in the Louvre. He regarded the Namur example as the earliest of the group, citing the absence of heraldic motifs. While aus’m Weerth did not rule out the possibility that the red damask lining was original, he considered it more likely to be a later addition.
[Fig. 6]
The casket as illustrated in the study by Carl Ernst aus’m Weerth, published in the
Annales de la Société archéologique de Namur.From: aus’m Weerth, 1864, pl. I.
11In 1883, Charles de Linas published a brief note on the casket, in which he identified the style of the engraved medallions as characteristic of Limousin production.6 He also pointed to the nail decoration – previously seen only on objects from the Limousin region, such as the Conques casket, the Saint Louis casket in the Louvre, and the Longpont casket – as further evidence supporting a Limousin origin. Later, in 1888, the casket and the crown were exhibited at the
Exposition rétrospective d’art industriel as a ‘Western work from the early thirteenth century’.7 Building on de Linas’ observations, Ernest Rupin, an art historian specialising in the Limousin region, included the Namur casket in his study of the so-called œuvre de Limoges, thus contributing to the hypothesis that the casket originated in Limousin.8
- 9 Courtoy and Schmitz 1930, p. 13-14, pl. II.
- 10 Lanotte and Blanpain 1969, p. 24-26.
- 11 Toussaint 1996, p. 42, 47-48, 61-62.
- 12 Jeanmart 2010.
12On the occasion of an exhibition organised in 1930, the casket was the subject of an entry in the catalogue by Ferdinand Courtoy.9 He identified the enamels as originating from Limousin and compared the casket and its decoration with those of the Saint Louis casket, the Aachen casket, and the Tongeren casket. André Lanotte later incorporated much of Courtoy’s text into his own catalogue entry for the 1969 exhibition.10 More recently, in the catalogue for an exhibition devoted to Limoges production held in Namur in 1996, Jacques Toussaint proposed a slightly earlier date for the casket (c. 1200), comparing it with the example held at the Metropolitan Museum of Art in New York, dated to around 1190.11 In 2010, drawing on historical data concerning the Reliquary-Crown, Jacques Jeanmart dated the casket between 1206 and before 1218.12
13Although no source explicitly confirms it, the casket is believed to have been designed to hold the Reliquary-Crown of the Holy Thorns. It has been kept in the treasury of the collegiate church of Saint-Aubain (now the cathedral) since the thirteenth century. The octagonal shape of the casket, which reflects the eight sides of the crown, is likely no coincidence and strongly supports this theory (a point that will be examined further). The Reliquary-Crown of the Holy Thorns was created to enclose two thorns thought to be from Christ’s Crown of Thorns (fig. 7). These relics were stolen from the collection of Passion relics held in the Imperial Palace of Constantinople in 1204, during the sack of the city by the Crusader armies of the Fourth Crusade.
[Fig. 7]
Reliquary-Crown of the Holy Thorns, gold and precious stones. H.: 8.5 cm; Ø: 20.7 cm. Namur, Cathedral Treasury and Diocesan Museum, inv. 4.
© Musée diocésain de Namur / Guy Focant, Vedrin.
- 13 Transcription of the text in: aus’m Weerth 1867, p. 407.
- 14 Transcription of the text in: aus’m Weerth 1867, p. 409.
- 15 See Cambier 2020. The gems of the crown have been recently analysed by Y. Bruni (University of Lièg (...)
14Archival sources provide two key chronological reference points. The first is a letter describing the sending of the thorns and other relics in 1206 by Henry, Count of Flanders and Hainaut – soon become the second Latin Emperor of Constantinople – to his brother, Philip the Noble, Count of Namur (fig. 8).13 It was Philip who donated the relics to Saint-Aubain, the collegiate church founded and patronised by the Counts of Namur. The second important document is an inventory from 1218, which explicitly mentions the crown relic.14 Based on stylistic comparisons with contemporary goldsmiths’ work and filigree decoration, a date of between circa 1210 and 1220 for the Reliquary-Crown is considered highly likely.15
[Fig. 8]
Seal of Henry, Count of Flanders and Hainaut and second Latin Emperor of Constantinople, formerly attached to the now-lost letter of 1206. Namur, Cathedral Treasury and Diocesan Museum, inv. 1800.
© Musée diocésain de Namur (photo: Hélène Cambier).
15Since it is neither a reliquary in its own right nor crafted from precious materials, the casket does not appear in early sources, unlike the crown, which is mentioned in several texts. To understand how the casket was used and where it was stored over the centuries, we must rely on information regarding the Reliquary-Crown (assuming the casket always served as its container). By the end of the Middle Ages, the casket and its valuable contents were probably kept in a cupboard in the choir of the collegiate church of Saint-Aubain. In the fifteenth century, William II, Count of Namur, donated such a cupboard for the church’s relics. Known as the
sacraire and gilded on the interior, it was placed behind the high altar and enclosed by two painted shutters, which served as an altarpiece. In 1578, when the altar was replaced, the cupboard was built into the wall of the choir, above the funerary monument of William II and his wife Jeanne d’Harcourt. At the beginning of the eighteenth century, the cupboard may still have existed, based on its description in the second volume of Les Délices du Pays de Liège, published in 1740.16
- 17 de Hauregard 1851, p. 18.
16The cupboard was opened, and the relics, including the crown, were displayed only on certain solemn occasions. According to the eighteenth-century Namur historian Galliot, the cathedral’s relics were shown to the faithful on the feast of Saint Stephen. In the nineteenth century, Canon de Hauregard noted that the crown was exhibited during the Feast of the Crown of Thorns, celebrated at Saint-Aubain Cathedral on the Friday following Ash Wednesday.17 The casket was also opened on other special events to reveal the crown. The Counts of Namur were allowed to swear oaths on the relics placed on the main altar, as Count William II did in 1391. The Reliquary-Crown was likewise displayed to notable visitors, such as Archdukes Maximilian of Austria and Mary of Burgundy, who visited Namur in 1477. Although the historical records mentioning the crown’s exhibition do not specify the casket, all available evidence suggests that the reliquary was removed from the casket for display.
17The status of the Reliquary-Crown as one of the cathedral’s most prestigious objects helps explain the good state of preservation of its casket. At some point, it became necessary to repair the lid’s hinge using a rough strip of leather nailed to the side opposite the lock. One could also question whether the plate covering the lock is a later addition or a repair – a point addressed later in this paper.
- 18 de Hauregard 1851, p. 49.
- 19 aus’m Weerth 1867, p. 390.
- 20 Sosson and Nickers 1906, p. 27.
18The casket does not appear in inventories until the nineteenth century, a few decades after the cathedral was rebuilt. According to Canon de Hauregard’s notes,18 the relics were then kept in the sacristy – probably the one now known as the bishop’s sacristy – or in the small room adjacent to it. When Carl Ernst aus’m Weerth examined the crown for his 1864 study (see above), he reported that it was lying in the casket, ‘wrapped in a piece of red damask’,19 which he noted was identical to the fabric lining the interior of the casket. This fabric still survives: it is a rectangular fragment preserved in the Diocesan Museum, among other textiles (inv. no. 1112). The red silk currently lining the casket was therefore in place before 1864. According to Sosson and Nickers, although they do not cite their sources, this red silk replaced an earlier white textile.20
19In 1880, the Reliquary-Crown was loaned to a major exhibition in Brussels celebrating the fiftieth anniversary of Belgian independence. At the request of the chapter of Saint-Aubain, the thorns were removed from their receptacles before the loan and placed in a new phylactery. It was probably at this time that the parchment strip used to authenticate the relics of the crown was pinned to the red silk inside the casket – the strip has since been removed and is now kept separately. At the start of the twentieth century, the fabrique cathédrale deposited the Reliquary-Crown and its casket, along with other ancient objects from the treasury, in the Diocesan Museum, inaugurated in 1905 in an annex to the cathedral. The crown and casket were then displayed in a secure cupboard alongside the most important objects of the treasury. They are now exhibited in a new scenography dedicated to the history of the cathedral treasury.
20The Namur casket is part of a group of similar objects distinguished by their materials and enamelled medallions, attributed to the Limoges region and dated from the twelfth to the thirteenth centuries. The fact that other caskets were used to house relics and precious objects, combined with the uniqueness of the Namur casket, which is the only known example with an octagonal shape, supports the idea that it was specifically designed to hold the Reliquary-Crown of the Holy Thorn.
- 21 Douai, Musée de la Chartreuse, inv. A.906.
21Except for the Namur casket (which has eight sides) and the Verceil museum casket (which has a gabled lid), all the other caskets are rectangular. They are fitted with one or two handles and a monster-shaped hasp, and are made of wood that is either painted, varnished, or covered in leather or metal foil. The decoration features openwork and/or enamelled copper medallions, completed with decorative round-headed nails. Many medallions depict iconography inspired by the Romanesque Bestiary, including lions, gryphons, dragons, and other chimaeras. The Namur casket is not the most ornate in the series; it lacks the ornamental fittings seen on the New York and Aachen caskets. Probably later and in a poorer state of preservation, the Douai casket21 is the only one that displays very large, round-headed nails similar to those on the Namur casket, and they are placed in similar positions.
- 22 Paris, Louvre, inv. MS 253. O’Neill 1996, cat. 123, p. 360-363 (B. Drake Boehm).
- 23 Now in the church of Saint-Sébastien, Longpont. O’Neill 1996, cat. 133, p. 376–378 (B. Drake Boehm) (...)
- 24 Castronovo and Descatoire 2016, p. 24.
- 25 O’Neill 1996, p. 155 (B. Drake-Boehm).
22This type of casket, prized by the nobility, was used as a shrine for precious goods, to such an extent that it was considered suitable for housing reliquaries and relics. Originating from the Abbey of Notre-Dame du Lys (Seine-et-Marne), the Saint Louis casket (1236?), kept in the Louvre,22 was used to hold relics of Saint Louis (fig. 9). The casket bearing the coat of arms of Richard of Cornwall (c. 1258), in the treasury of Aachen Cathedral, is still used today to contain the four Marian relics from Notre-Dame’s shrine during the septennial pilgrimage. The reliquary casket of Blessed Jean de Montmirail, from the former Cistercian Abbey of Longpont (c. 1270),23 has been used as a reliquary since the 17th century. The use of these three caskets to enshrine relics may not have been their original purpose, but other records offer a glimpse of what their purpose could have been in the 13th century. Pierre de Nemours (1208–1219), Bishop of Paris, possessed a casket of this type for storing valuable documents.24 The inventory of St Paul’s Cathedral (London, 1295) includes a ‘black coffret which is said to have belonged to Bishop Gilbert, containing many enamelled roundels (rotellas aymallatas), in which are contained many relics, sealed by the seal of the dean.’25
[Fig. 9]
Saint Louis casket, 1236 (?), h: 14 cm, l: 36.5 cm, w: 19 cm (Paris, Louvre, inv.
© 1995 GrandPalaisRmn (musée du Louvre) / Daniel Arnaudet.
MS 253).
- 26 Castronovo and Descatoire 2016.
- 27 Ibidem, p. 9.
- 28 Ibidem, p. 10.
- 29 Ibidem, p. 10.
- 30 O’Neill 1996, cat. 88, p. 282–285 (B. Drake Boehm).
- 31 Verceil, Museo Leone, inv. 4304.
23The caskets that belonged to Cardinal Guala Bicchieri (c. 1150/1160–1227) shed light on the value placed on these objects.26 The Cardinal and Papal Legate Bicchieri was one of the most influential men of his time. He surrounded himself with the elite of Western Europe, all of whom were major patrons. He maintained strong links with Pope Innocent III, who had commissioned enamels from Limoges, notably for the enclosure of St Peter’s confessio in the Vatican. Bicchieri owned an exceptional collection of goldsmiths’ works, liturgical ornaments, and illuminated books, which he bequeathed to his foundation, the Sant’Andrea Abbey in Vercelli. Following his death, the objects were listed in an inventory, providing insight into the nature of his collection. Limoges enamels (de opere lemovicensi)27 constitute a significant part of Guala Bicchieri’s collection.28 The inventory notably reports three Limousin caskets (scrineo operis lemovicensis)29 containing goblets, numerous rings, liturgical goldsmiths’ works, and reliquaries. Two of these caskets are still preserved today: the first in Turin (fig. 10),30 and the second in Vercelli.31 In addition, twenty-nine enamelled medallions, scattered across Turin, Paris, and New York, probably originate from other caskets belonging to the Cardinal. In this context, the association of the Namur casket with the renowned Golden Reliquary Crown of Holy Thorns, kept in a church treasury, is therefore not surprising.
[Fig. 10]
Cardinal Bicchieri’s chest, before 1227; h: 38 cm, l: 82 cm, w: 34.5 cm
© Courtesy of Fondazione Torino Musei (photo: Studio Fotografico Gonella 2012).
(Turin, Palazzo Madama – Museo Civico d’Arte Antica).
24At first glance, in terms of execution, style, or iconography, the Namur medallions do not show any notable differences; The medallions are generally consistent in appearance, except for the scutcheon, which displays a different technique and a noticeably lower standard of execution.32 Could this plate be a later addition or repair? It is possible, but the difference might also be due to the plaque being made separately from the medallions.
25However, there are differences between the medallions. Most are fixed to the wood with three nails, but five medallions have four nails (these medallions with four nails belong to the Biron type 2, as described in section Blue Enamel). In terms of borders, most medallions feature a gilded copper border, but seven are edged with enamelled borders. These seven enamelled-bordered medallions belong to Biron type 1 (see section Blue Enamel); they are secured with three nails and include the only medallion with turquoise blue enamel. Nonetheless, the importance of these enamelled borders should not be overstated, as some medallions with a gilded copper border (without enamel) also fall under Biron type 1. When comparing the Namur medallions with similar ones on other caskets, Namur’s are the only ones displaying such disparities.
26All this prompts the question as to how the medallions and the casket were made. Should the medallions be seen as a uniform series produced specifically for the casket? Or were the medallions mass-produced separately from their particular purpose? Were the medallions and the casket made at the same location? Or were the medallions sourced elsewhere, and the casket crafted independently, with the medallions and the casket connected to different places and times?
- 33 O’Neill 1996, cat. 7, p. 78–82 (E. Taburet-Delahaye).
- 34 O’Neill 1996, cat. 36, p. 154–155 (B. Drake Boehm).
- 35 Toussaint 1996, p. 22 (E. Taburet-Delahaye). The Jean de Montmirail casket shows 50 armorial medall (...)
27Comparisons of the Namur casket with others in the series support its dating to the first third of the 13th century, perhaps around 1220–1225. The famous casket of Abbot Boniface, in the treasury of Conques Abbey, predates the series by nearly a century (c. 1110–1130).33 The animals on the enamelled medallions, including lions and other quadrupeds, dragons with or without wings, and creatures with flowery tails (the Bestiary present on the Namur casket), are enamelled and stand out against the gilded copper background. This gilded background typifies the first copper enamels produced in the 12th century. Medallions featuring Bestiary figures, created in reserved, engraved, and gilded copper against a champlevé enamel background, are found on the casket in the Metropolitan Museum of Art in New York, dated to around 1190 (fig. 11).34 This method was developed from the 1180s onwards, both in Limoges and in the Rhenish-Mosan enamelling traditions. Compared to the Namur enamels, the New York pieces differ stylistically: they are marked by more refined design, much more precise engraving, use of multiple enamel colours, and detailed enamelled surfaces on the figures. This indicates a later date for the Namur casket. The Saint Louis casket (1236) (fig. 9) and the Aachen casket (c. 1258) provide additional chronological markers. In addition to Bestiary motifs, these feature enamelled medallions with courtly and hunting scenes, as well as numerous medallions or shields displaying coats of arms. This heraldic decoration, absent from the Namur casket, became popular between approximately 1240 and 1250.35
[Fig. 11]
Casket, c. 1190; h: 18.1 cm, l: 45.4 cm, w: 22.5 cm (New York, The Metropolitan Museum of Art, inv. 17.190.511).
28Cardinal Bicchieri’s chests offer an interesting chronological reference point, since we know that the Cardinal died in 1227 – the chests being dated around 1220–1225 (fig. 10). The Bicchieri medallions combine traditional Romanesque iconography (Bestiary) with images that were new at the time: courtly, chivalric, and hunting scenes. In comparison, the medallions on the Namur box remain within the scope of the Romanesque Bestiary; there is no courtly or chivalric motif, and they lack the refined Gothic naturalism of Bicchieri’s coffret.
- 36 Bryn Athyn, Glencairn Museum, inv. 12.EN.89. O’Neill 1996, cat. 38, p. 158–160 (B. Drake Boehm).
29From a stylistic perspective, the Namur medallions are quite distinct from other enamelled medallions. However, they can be compared to another type that adorns several caskets: the openwork and chased medallions. These are made of embossed, chased, and gilded copper, possibly bordered by enamel. Their iconography draws on the Romanesque repertoire, featuring animals from the Bestiary and battle scenes. They are found on the Bicchieri chests, as well as on the caskets in the Louvre, Aachen, and the Glencairn Museum in Bryn Athyn;36 they also constitute a significant portion of medallions scattered across various collections (fig. 12). In terms of composition and expression, these openwork medallions bear some resemblance to the Namur medallions. For example, on the Saint Louis casket, the star motif appears alongside a composition of two monstrous animals facing each other, mirrored around a central gilded band. The Turin box depicts an eagle with its wings outstretched. The iconography and layout of the lock plate in Namur could be seen as reminiscent of the openwork plates that cover the locks on the Bicchieri, Aachen, and Louvre caskets, which also feature animals in combat facing each other. These similarities can simply be explained by the use of shared models, which could have been widely disseminated across time and space.
]
[Fig. 12
Openwork medallion coming from a Cardinal Bicchieri’s casket, Limoges, before 1227, Ø: 8,9 cm (Paris, Louvre, inv. OA 9474).
© 1995 GrandPalaisRmn (musée du Louvre) / Daniel Arnaudet.
30Given the stylistic and quality differences with the other caskets, the absence of courtly or hunting scenes in Namur cannot be considered a chronological indicator. Even if the Romanesque animals may seem somewhat outdated for this period, they do not argue against a date around 1220–1225. However, the lack of a specific stylistic link between the Namur medallions and those of other caskets does not firmly confirm their attribution to Limoges. Bestiary figures, contorted and biting dragons, are fashionable and widespread motifs at the turn of the 12th and 13th centuries. They appear in the enamelled plaques of Rhenish shrines, such as the St. Albinus shrine in Cologne (c. 1186) and the St. Mauritius and Innocentius shrine in Siegburg (c. 1185). The Albinus shrine features an openwork crest with a frieze of dragons contorted into foliage, biting into fruits and plant leaves. We may also consider the ‘Jagdfries’ that adorns the main side of the famous shrine of the Three Kings (Cologne) (c. 1198–1206). This frieze, made of openwork and gilded copper or silver, displays medallions depicting hunting scenes, dragons, and other Bestiary animals. These Cologne examples are of a much higher standard than the Namur casket, making direct comparisons difficult. However, it is essential to note that the attribution of the Namur medallions is not yet certain: they could be of septentrional, Rhenish-Mosan, or even local origin.
31The casket consists of a wooden core built by assembling two distinct horizontal wooden panels for the lid and the base, along with eight vertical panels for the sides. The octagonal lid panel measures 29.1 centimetres in height, 29.4 centimetres in width, and up to 1.2 centimetres in thickness. The vertical panels are up to 10.9 centimetres tall and as wide as 12.9 centimetres. The wood is only visible in the gaps of the leather, which restricts examination possibilities. Similarly, inside the box, the wood is covered with fabric.
- 37 Fraiture 2007; Beuting 2011.
32The wood used is oak, which generally shows slow37 but irregular growth, alternating between periods of very narrow and wider growth rings. Based on the visible radial and/or transverse cuts, the panels were quarter-sawn or rift-sawn. The pith has been systematically removed from the boards. However, the bottom panel of the box has retained sapwood along its left edge. This sapwood area was identified by its lighter colour, as well as the presence of insect damage.
]
[Fig. 13
Assembly schema of the casket: identification of the boards and their growth direction, localisation of the nails and the sapwood.
Working document (Violette Demonty, Christophe Maggi).
- 38 Carried out by Catherine Fondaire (KIK-IRPA).
33The panels are joined using butt joints along a central axis of symmetry that contrasts with the growth directions of the boards (fig. 13). They are secured with hand-forged iron nails arranged in a relatively systematic manner. The eight corners of the box are joined together with one nail (sometimes two) placed laterally at the top of each panel, either to the right or the left (fig. 13). These were identified during visual inspection of the leather gaps and with the help of a magnet (fig. 14). Furthermore, X-ray images of the box,38 taken from various angles, confirm the presence of the nails (fig. 15 and 16). Through these X-rays, it is also evident that the bottom panel is attached to the eight vertical panels (at their lower part) with horizontally placed metal nails (fig. 16).
[Fig. 14]
Nail head visible in a gap in the leather at board 6. This nail secures the assembly to board 7.
© Brussels, KIK-IRPA, X147019l.
[Fig. 15]
Assembly of X-rays of the vertical panels forming the sides of the casket. Localisation of the assembly nails is marked in red.
© Brussels, KIK-IRPA, RD012346L-RD012353L.
[Fig. 16]
X-ray of the casket, view from the top. Localisation of the assembly nails is marked in red.
© Brussels, KIK-IRPA, RD012343L.
34All the assembly nails show signs of oxidation. Their length exceeds that of the numerous decorative nails with round, domed, gilded heads. They are square in cross-section with flat heads.
35The lid is attached to the frame with iron hinges held in place by large, round, flat-headed forged nails. These are clearly visible in the X-ray (fig. 15). Currently, a leather strap, secured with small modern nails, covers them (fig. 17).
[Fig. 17]
Panel 5, metal hinge covered with a leather strap attached with small modern nails.
© Brussels, KIK-IRPA, X126522.
36Finally, it should be noted that the bottom is fitted with two non-magnetic metal strips, with their ends extending slightly over the four panels of the box (the front, the back, and the two sides). These strips are secured beneath the bottom panel of the box with symmetrically placed nails, as well as to the four vertical panels using four decorative nails with large domed heads (fig. 18).
[Fig. 18]
Panel 3 (left side): The end of the strip rises onto the small panel and is fastened by a large decorative nail with a round, domed
head. © Brussels, KIK-IRPA, X126520.
- 39 Tyers and Parsons 2010.
37The restoration of the casket provided an opportunity to conduct dendrochronological research on its wooden core. Dendrochronology is a method used to date wood by analysing the growth patterns of annual tree rings, which helps determine the period during which the trees were alive. The width of these rings varies, with some being narrow (indicating slower growth) and others wide (indicating faster growth). Several factors influence tree growth and, consequently, the width of the rings.39 In regions where climate conditions remain consistent, trees of the same species tend to show similar growth responses. Climate, which stays relatively stable within a geographic area, can fluctuate from year to year in a non-cyclical manner, causing variations in the rings. Therefore, the series of ring widths in a piece of wood is specific to the period when the tree grew and the region where it was located. As a result, analysing a series of tree rings enables us to establish not only when a tree lived and was cut down but also to identify its region of origin.
38The feasibility of using dendrochronological dating on artwork is limited by several factors. Firstly, the wooden components must come from a species suitable for dendrochronology, such as oak (Quercus robur or Q. petraea), which is the material used for the casket. Additionally, the wooden panels should be analysable without causing damage. In the case of artworks, unlike buildings or archaeological remains, dendrochronologists do not take samples because of the high value of the object. Therefore, direct access to the end grain or the surface of the wood’s cross-section is necessary. This usually involves reaching the lateral edges of the panels if the wood grain runs horizontally, or the horizontal edges if the grain runs vertically. When possible, two sets of tree rings are measured on each panel, one at each end. This method helps prevent distortions in the tree’s growth, such as knots or irregular rings, which could obscure the dendrochronological signal. It addresses gaps caused by damaged wood and helps identify the most recent ring on the panel.
39The wooden core of the casket, being covered in leather, made access to the growth rings particularly difficult, and preparing measurement paths with a razor blade was impossible. Fortunately for the study, gaps in the leather in certain areas allowed access to the edges of some panels (fig. 19). However, these edges were not constantly exposed across the entire width of the panel. As a result, sampling was guided by the random nature of the deteriorations and the different points of access, which explains why not all the panels could be analysed or not in their entirety.
[Fig. 19]
Macrophotographic recording of a measurement path on the lid panel. The leather that was present under the decorative nails has disappeared, providing access to the growth rings.
Working photo (Christophe Maggi).
40Ultimately, the lower edges of six of the eight vertical panels were recorded for the dendrochronological study. For the lid panel, both horizontal edges were accessible and thus studied. The total number of measured ring series on the box is eight, covering seven panels. Only the bottom panel retained sapwood, but, unfortunately, it was inaccessible and could not be included in the current dendrochronological analysis.
41Some measurement paths were gently cleaned with a brush when necessary to make the growth rings visible. The widths of the rings were measured on digital macrophotographs with a precision of 1:100 mm (fig. 20).
[Fig. 20]
Macrophotography of the lower edge of panel 5 shows leather remnants and a metal strip covering part of the dendrochronological sequence, with deteriorated areas and irregular growth patterns (wide rings).
Working photo (Christophe Maggi).
42The dendrochronological comparisons showed that panels from the same tree were used within the box: three vertical panels (panels 4, 5, and 6) originate from the same oak. Their series were combined into a single ‘average’ series representing this individual tree (fig. 21). The other series do not match each other or this ‘average’ series. This could be explained either by their sequences being too short (due to the presence of leather preventing full sequence reading, as with panels 3, 7, and 8) or because they come from different dendrological contexts (as might be the case for the lid panel).
[Fig. 21]
Synchronisation between the dendrochronological series from the three planks originating from the same tree and drawing of their ‘average’ series in bold (drawn in natural value; positions at 3001 indicate they are relative, undated positions).
Working photo (Christophe Maggi).
- 40 Lambert 2006; Lambert 2011.
43All the dendrochronological series from the casket were visually compared and, with the aid of synchronisation calculations, with our reference datasets, aiming to determine the date and origin of the wood.40 Only one result could be obtained for the individual ‘average’ series 4–5–6 (fig. 22). This is dated with the last ring in 1193. The other series could not be individually dated. Since boards 4, 5, and 6 do not retain any sapwood, the result represents a terminus post quem (and not the exact year of cutting) for the felling of the tree from which they originated.
[Fig. 22]
Assembly schema of the casket, highlighting the three dated panels: 4, 5, and 6.
Working document (Violette Demonty, Christophe Maggi).
- 41 Lambert 2006; Durost and Lambert 2007.
44Considering the minimum amount of sapwood (4 rings41), this tree was felled after 1197 (1193+4). The time interval between felling trees in the forest and using the wood can vary significantly (from a few months to several years). It is therefore impossible to specify this for the casket. It could have been ready for use shortly after 1197 at the earliest, if we consider the earliest possible felling date and a short period between felling and use, but no terminus ante quem can be proposed.
45The dendrochronological series of the three dated panels aligns with chronologies primarily from northern and eastern France (the Seine and Somme basins) as well as the middle Meuse region. This indicates that the tree from which these panels come originates from this vast geographical area. While this potential French origin of the wood suggests a possible origin for the casket’s manufacture, these findings should be considered with caution. The results of this dendrochronological dating and the provenance information derived from it are dependent on the existing reference datasets for this period and region, which still require further development.
46The high demand for leather-covered wooden caskets in the 13th century, particularly in Limoges and other centres of production, encouraged lively collaboration among various artisans. Casket makers, leatherworkers, enamellers, locksmiths, and decorators all contributed their skills to craft these intricate and sophisticated objects. A range of leathers was used in making these caskets, with calfskin and goatskin being especially valued for their flexibility and texture. These materials were vegetable-tanned, a process that increased their durability and softness. In Western Europe, oak bark (Quercus spp.) and sumac leaves (Rhus coriaria) were the most common tannins employed. Unfortunately, no analytical identification of the historic tannin materials used for the casket has been performed. The leather and fabric fastenings of the casket have been altered multiple times over the years, as shown by nails found beneath the fabric. While the leather may be original, the possibility of complete restoration cannot be ruled out (see below, Metal elements).
- 42 Falcão and Araújo 2018.
47Vegetable tanning is a transformative process that stabilises animal skin, making it resistant to microbial decay and temperature fluctuations. This involves treating cleaned and prepared skins with vegetable materials or their extracts through infusions, macerations, or decoctions. Goat leather, notably used in the casket, is particularly valued for its distinctive grain pattern and durability. Under microscopic observation and macro photography, goat leather reveals a characteristic arrangement of small, irregular clusters of hair follicles distributed evenly across its surface. Its tight grain structure contributes to a smooth texture, and its subtle pebbling enhances its aesthetic appeal. Compared to cowhide, goat leather has finer, more densely packed follicles, resulting in a smoother and more polished appearance, making it ideal for specialised applications.42
48The selection of leather in the workshops was a meticulous process. Artisans prioritised the natural grain of the ‘hair side’ for external display, while the ‘flesh side’ was turned inward to reinforce structural integrity. The leather was softened by moistening, making it pliable for precise application to the wooden structure. It was then stretched and moulded over the wooden framework, such as the octagonal base of the casket. Patterns were cut from a single skin of leather to minimise seams, with separate coverings for the lid and base. During the drying process, which lasted 8 to 24 hours at room temperature, the leather was secured with temporary nails or tacks and pressed firmly against the wood using tools like bone folders (tools made of bone to shape and smooth the leather). This ensured a tight fit and eliminated air bubbles. Decorative round-headed nails were later used to create patterns and permanently secure the leather. On the casket’s lid, rows of small nails bordered medallions and edges, alternating with larger nails in clusters. A hinge was made from flexible leather pieces, which were later replaced during restoration with a thicker material.
49Over time, leather-covered caskets face significant challenges due to their organic nature. The adhesives initially used, likely cellulose- or collagen-based, weaken, causing the leather to detach from the wooden base. Exposure to moisture, heat, and contact with metals accelerates shrinkage, discolouration, and darkening.43 Shrinkage also results in cracking (craquelures) and delamination, where the dermis separates from the epidermis, revealing lighter-coloured flesh layers. These effects are most noticeable at vulnerable edges, corners, and hinges, where wear is greatest.
50The Reliquary-Crown casket holds several metal components, which can be divided into two separate groups: enamelled plates and decorative nails.
51The enamelled plates are made from almost pure copper and are fire-gilded. A more detailed account of their manufacturing technique will be given in the next section on enamels.
52Three distinct sizes of decorative nails were observed on the casket. However, as shown in the X-ray images and photographs (fig. 23–28), they were all produced in the same manner. The dome is formed from a small metal plate that is cut and die-stamped (using a forming, hemispherical, concave die and a convex stamping die). The shafts of the small nails are round and uniform in cross-section, as seen in the optical microscopic images in figures 29 and 30. The shaft of the larger nails has a square cross-section, created by hammering in a forge. In all cases, the shafts are attached to the brass dome-head by a lead-tin alloy solder that nearly fills the entire cavity (fig. 28).
[Fig. 23]
The multitude of decorative nails is clearly visible on the
X-ray. © Brussels, KIK-IRPA, Rd012341l.
[Fig. 24]
Detail of the
X-ray. The long, forged nails have a decreasing cross-section. Most of the small ones show the constant cross-section of wire-drawn nails. © Brussels, KIK-IRPA, Rd012344l.
[Fig. 25]
A large decorative nail on one side has lost its brass head. The wrought iron nail shaft is visible.
Working photo (Françoise Urban).
[Fig. 26]
In the
X-ray, two different materials are visible in the nail head: the ‘translucent’ material is the brass dome, while the opaque, white material is the lead/tin solder. Both materials appear on the large and small nails. At the top of the image, three different types of nails can be seen: forged, wire-drawn, and modern nails with a usual flat head. Some of the nails are beneath the fabric. © Brussels, KIK-IRPA, Rd012351l.
[Fig. 27]
Detail of the
X-ray. Viewing from the side, the construction with the iron shafts and the lead-tin solder on the head is clear. The solder sometimes runs down the shaft, giving the nail head a biconvex shape. © Brussels, KIK-IRPA, Rd012344l.
[Fig. 28]
One of the decorative nails has been removed for analysis. The lead running along the shaft is clearly visible.
Working photo (Helena Wouters).
53The analysis showed that three different metal alloys were used. The nail head is made of brass (a copper-zinc alloy); the shaft is made of steel; and a lead-tin solder was used to join the two parts and reinforce the head. Figure 29 presents the optical microscopy and SEM-EDX analysis results of the nail tip sample. This clearly demonstrates that the nail is made of steel (with a detectable amount of carbon) and has been tinned at the surface. Thermal tinning, a process in which an object is immersed in molten tin, has been used since ancient times to protect metals from corrosion and enhance their workability during soldering. Despite this protective tin layer, significant corrosion of the iron is evident, which suggests a certain age, but it cannot be precisely dated.
[Fig. 29]
Optical microscopy and back-scattered electron images of the nail tip sample. The X-ray analysis spectra indicate that the nail is made of steel (containing a detectable amount of carbon) and has been tin-plated on its surface.
- 44 Thomas and Bourgarit 2014; Thomas and Dandridge 2018.
54The results of the optical microscopy and SEM-EDX analyses of the sample removed from the decorative dome-shaped nail head are shown in figure 30. The results clearly indicate that the dome is made of brass and filled with an alloy of lead (Pb) and tin (Sn). The brass alloy comprises 80% Cu, 16% Zn, 4% Sn, and 1% Pb. It is estimated that the filling is an alloy containing 85% Pb and 15% Sn. The brass, with a zinc content of around 15%, is typical of medieval alloys.44
[Fig. 30]
Optical microscopy and back-scattered electron images of the
cup-shaped nail head sample. The X-ray analysis spectra demonstrate that the cup is made of brass and is filled with an alloy of lead (Pb) and tin (Sn).55Along with these decorative nails, X-rays revealed that other nails were used to secure the leather and fabric. Some of these are forged, others are drawn out, and some are of a more recent industrial origin. The method of attaching the leather and fabric has been altered numerous times, as shown in figure 26, where some nails were found underneath the fabric.
- 45 Viollet-Le-Duc 1854, vol. 3, p. 470–471
56This method of crafting decorative nails has a possible origin dating back to the 13th century. Evidence includes historical records, such as the door in Vézelay, which dates back to the 11th century.45 Both drawn and forged iron nails are frequently found in archaeological discoveries and are often discussed in the literature and archaeological reports.
57The X-ray of the lock mechanism is shown in figure 31. The lock cover is made of brass with traces of Ni and does not feature gilding. Similarly, the small mounting ring for the handle is also made of brass, without gilding; however, the copper-to-zinc proportions differ from those of the lock cover.
[Fig. 31]
X-ray of the locking mechanism.
© Brussels, KIK-IRPA, Rd012346l.
- 46 Biron, Dandridge and Wypyski 1996, p. 49.
- 47 Biron and Taburet-Delahaye 1998, p. 76.
58The enamels were all created using the ‘champlevé’ technique, in which enamel powder is applied to areas carved into the metal.46 The piece is then fired at temperatures between 800 °C and 900 °c. To ensure the cavities are fully filled, the application and firing process must be repeated, as the enamel particles shrink when melted. Once the recesses are completely filled, the surface is sanded and polished to a smooth finish. The polishing process was used to remove the layer of red copper oxide (cuprite) that covered the metal surfaces.47 The cuprite phase remains visible on the plates at the interface between the glass and the copper, as shown in the image captured by the digital microscope (fig. 32). After engraving and chasing were completed, all visible metal surfaces were gilded with mercury amalgam and then burnished to a smooth finish.
[Fig. 32]
Observation under the digital microscope (HIROX). The cuprite phase remains visible on the plates at the interface between the glass and the copper.
59Regarding the enamel on the casket, the recesses on the closing plate and the medallions (except for the central medallion on the lid) were not completely filled with enamel. The enamel appears slightly below the level of the metal, and there is no evidence of polishing.
[Fig. 33]
Raking light photographs of the lid and medallions III.1 and III.2, showing the incomplete filling of the champlevé enamel recesses (except for the central medallion of the lid) (after treatment).
© Brussels, KIK-IRPA, X147022l, X147021l.
60The glass powder fused onto the copper substrate, known as ‘enamel’, is usually a soft glass made from flint or sand, lead, and soda or potash. These ingredients are melted together to form a nearly transparent glass called flux or frit. Its hardness varies – those with higher lead and potash are shinier but softer. The clear frit functions as the base for creating coloured enamels, with metal oxides serving as the colouring agents.
- 48 µXRF equipment (ArtTAX, Bruker, Germany) and MA-XRF (M6 Jetstream, Bruker, Germany). Marina Van Bos (...)
61The surface analysis of the blue enamels on the casket for the Reliquary-Crown of the Holy Thorns (using µXRF and MA-XRF as shown in figures 34 and 35)48 reveals two main types of blue enamels. The first group contains antimoniate as the opacifier, while the second one contains a tin-based opacifier. Both are coloured by the addition of cobalt ore.
[Fig. 34]
Positioning of the crown-casket under the X-ray beam of the ArtTAX X-ray fluorescence analysis equipment (Bruker).
Working photo (Helena Wouters).
[Fig. 35]
Analysis of the casket using MA-XRF (M6 Jetstream, Bruker).
Working photo (Helena Wouters)
.
- 49 Biron, Dandridge and Wypyski 1996, p. 59.
- 50 Biron 2015, p. 212.
- 51 Ibidem, p. 212.
62Very tiny samples were extracted from three different areas of blue enamel, as shown in figure 36. The polished cross-sections were analysed quantitatively using SEM-EDX. The results for the three samples are presented in figure 37. Besides the different opacifiers used in the blue enamel, such as tin oxide in sample 1 and calcium antimonates in samples 2 and 3, two types of base glass compositions are also identifiable. While all three are soda-lime silicates (sodium alkali), samples 2 and 3 contain low levels of magnesium (MgO), potassium (K2O), and lead (PbO). In contrast, the composition of sample 1 is relatively rich in MgO, K2O, and PbO. These findings suggest that the sodium source in samples 2 and 3 is mineral-based, similar to that found in Roman-style glass produced throughout the Mediterranean basin during the Roman Empire.49 This corresponds to the first group (type 1) of enamels.50 The results for the blue enamel in sample 1 indicate a plant-based sodium source, derived from coastal plants such as samphire or salsola, which are rich in sodium. This type of composition falls into the second group (type 2) of enamels.51
[Fig. 36]
View of the enamelled plates on side 1 (top) and side 6 (bottom) of the casket, with indications of the areas of the three samples.
© Brussels, KIK-IRPA, X126543,
X126523.
[Fig. 37]
Mean compositions and standard deviations obtained by SEM-EDX for the three samples of blue enamels.
- 52 Ibidem, p. 216.
- 53 Ibidem, p. 216.
63Generally, following the in-depth analytical study conducted at the C2RMF laboratory in Paris, it was observed that most of the enamels, except for the red glasses, can be chronologically grouped into two categories based on their base glass compositions, corresponding to the periods before and after the turn of the 12th and 13th centuries.52 Objects enamelled with the base glass composition type 1 date from the beginning of this production, with the enamels from Conques dating from the late 11th century to the early 13th century (before 1220), while those enamelled with the base glass composition type 2 are dated approximately from the second quarter of the 13th century to the end of the production, in the mid-14th century. These two recipes overlap briefly. A change in recipe occurred with the replacement of mineral soda sources (type 1) by plant-based soda sources (type 2), and the substitution of antimoniate opacifiers (type 1) with tin oxides (type 2).53
64Based on the existing literature (base glass composition + nature of the opacifier), it can be concluded that the blue enamels corresponding to those of samples 2 and 3 (fig. 37) confirm a manufacturing date in the first half of the 13th century, with a mineral-based soda source and antimoniate as the opacifier.
65According to the base glass composition and the opacifier used, blue enamels similar to those of type 2 (such as sample 1 in fig. 37) are more likely to be dated from approximately the second quarter of the 13th century to the mid-14th century.
66All the enamelled plates were examined using MA-XRF scanning and µXRF analysis. It was found that 14 of the plates contained blue enamel of type 1, while 11 plates consisted of blue enamel of type 2. A notable observation is that within the same medallion, only one type of enamel was used, even though the shade of blue may vary. As shown in the element distribution images obtained with MA-XRF, presented in figure 38, the levels of the constituent elements influence the shading of the blue colour.
[Fig. 38]
The element distribution images of calcium (Ca), cobalt (Co), lead (Pb), and arsenic (As), obtained by MA-XRF analysis for the enamelled plate 2 on the cover of the casket.
- 54 Before 1530, cobalt ores were of the Co-Fe-Ni-(Bi)-(As) type, with iron ( 5%), nickel ( 0.6%), and, (...)
67Regarding the blue enamel used on the plate with the lock, a slightly different composition and texture were observed. While its alkali content (see fig. 37, sample 1) aligns with a plant-based source of sodium and the enamel contains a tin oxide opacifier (hence type 2 blue enamel), making it consistent with 13th- to 17th-century soda-lime silicate glasses made from plant ashes and crushed quartz pebbles or small stones, the granulometry raises some questions. The fact is that although the blue enamels from the 15th, 16th, and early 17th centuries are well-coloured with cobalt, their sources change around 1530.54 The analysis results for this plaque reveal slight variations in impurities. Still, they are not substantial enough to definitively classify the enamel. Additionally, the arsenic content is too low, and the tin content is too high, suggesting a 19th-century production date. Visual observation further supports this, as it allows for an assessment of colour uniformity and enamel particle size. Industrial enamel production from the 19th century onwards typically produced finely powdered enamels with a uniform colour, in contrast to the coarser, more heterogeneous mixtures of earlier periods. While there is a slight degree of homogeneity in the opacifier grains of the enamel on this plaque, it is not significant enough to be a determining factor. Therefore, the dating of this enamel remains speculative, suggesting it may be from after 1530 or potentially a result of partial reuse.
68The analysis results show that turquoise glass lacks cobalt and contains a significant amount of copper. It is therefore a copper blue, further opacified by calcium antimoniate. This finding aligns with the turquoise enamels of the first type (late 11th–13th century). Traces of iron and manganese were detected, along with minor quantities of tin oxide. The presence of SnO2 indicates that the copper source in this enamel originates from a scale formed by heating tin bronze in air. These results verify the production of this turquoise enamel, generally made during the 13th century.
69The XRF and MA-XRF analysis of the white enamel reveal that calcium, antimony, and lead are the most prominent elements. These results indicate that calcium antimoniate crystals were incorporated into colourless glass to create a white enamel, a technique commonly employed in the Middle Ages.
- 55 Biron 2015, p. 179.
- 56 Ibidem, p. 261
70Finally, this study prompts the question of whether the enamels on the casket were produced in Limoges or by artisans in the Meuse and Rhine valleys. The invention of the new technique of ‘champlevé enamel on copper’ experienced remarkable growth across Romanesque Europe in the 12th century.55 The first Romanesque champlevé enamels on copper appeared nearly simultaneously in the early 12th century in various regions, with Limoges serving as the main centre, along with the Meuse and Rhine valleys. Whether one of these regions influenced the others remains an unresolved mystery to this day.56 According to Biron, it is indeed difficult to draw a definitive conclusion.
- 57 Ibidem, p. 233–269.
- 58 Biron 2015, p. 238–240.
- 59 Ibidem, p. 255.
71The detailed technical and analytical comparative study carried out at the C2RMF laboratory in Paris provides clear evidence of close connections between Limoges and the Meuse and Rhine valleys, with some minor regional variations.57 Furthermore, from their appearance before 1145 until the end of their production around 1230, Mosan and Rhenish champlevé enamels on copper show no significant changes in their manufacturing techniques. Their technical execution closely resembles that of contemporary Limoges enamels but can be distinguished by specific technical features. These include, in particular, the way the borders of the plates are chased, the frequent use of cloisonné technique, the method for carving the bottoms of the champlevé cavities, and the mixing of colour powders. The Mosan and Rhenish enamels – always simple and pure – contrast strongly with the Limoges enamels due to the absence of raised metal elements such as small heads and applied figures, and the less frequent use of engraving and chasing on the reserved metal areas.58 Moreover, the copper plates from both production centres display striking similarities in copper purity and the nature, concentration, and proportion of impurities, indicating shared ore sources and common purification techniques between the two regions.59
72Birons’ research results reveal a regional distinction among Limoges, Mosan, and Rhenish productions. While all utilise the same types of natron-based glass, the overall recipes, choice of colourants, and methods of opacifying specific glasses vary. The recipes are more precisely followed for the Mosan and Rhenish enamels (with low standard deviations) than for the Limoges enamels (which exhibit high standard deviations). Additionally, the sources of cobalt and copper used to colour the blue and green enamels differ between these two production regions. Lastly, although the blue and yellow enamels share the same opacification process (addition of crystals or a crystal-rich glass), the green enamels vary. While Limoges uses in situ crystallisation, the Meuse valley employs the addition of crystals or a similarly rich glass.
73Regarding the enamels on the casket for the Reliquary-Crown of the Holy Thorns, comparing the results for blue enamels of type 1 (mineral soda and calcium antimoniate as opacifiers) with those for Limoges and Mosan blue enamels emphasises the difficulty in attributing them to either production centre. Given the low standard deviations for samples 2 and 3 (see fig. 37) and the FeO and CaO values (which are slightly closer to those of the Mosan blue enamels), a Mosan origin can be proposed. As for the blue enamels of type 2 (plant soda and tin oxide as opacifiers), after examining all components, it does not seem possible to attribute them to either of the two production centres.
74The interior of the casket is fully lined and padded with textiles, creating a soft and protective enclosure for the safekeeping of the reliquary crown (fig. 5).
75The decorative fabric is a red silk damask with a floral pattern, finished at the edges with a trim woven from gold-coloured metal threads. The padding consists of natural-coloured loose fibres, some of which protrude from beneath the damask, allowing for sampling and analysis.
76The lining consists of four sections of damask fabric, sewn together at the seams along the side panels of the coffret with invisible stitches. The side sections are not sewn to the bottom fabric but are secured with several headpins. The fabric is nailed to the wooden structure, with nails visible at the corners where the side panels meet the base and along the edges of the casket. At the edges, the trim and fabric are nailed into the wooden panels.
77The X-ray analysis did not reveal the nature of the padding or any underlying layers of fabric. However, it provided insight into how the upholstery is mounted inside the coffret, with both nails and headpins clearly visible in the X-rays (fig. 23).
78The design of the red damask showcases staggered horizontal rows of large floral branches with paired palmette flowers, one facing upwards and the other downwards, with a fleur-de-lys at the centre of each palmette. The orientation of the motifs alternates between rows, while delicate floral branches are dispersed in the background. Figures 39 and 43 offer a partial view of the motif. The pattern aligns with the detail in the previously mentioned illustration in Ernst aus’m Weerth’s 1864 study (fig. 6).
[Fig. 39]
Partial view of the damask motif.
Working photo (textiles workshop KIK-IRPA).
79The damask’s technical properties are as follows: woven in a satin 5 structure, with approximately 110 warp threads per cm and 30 weft threads per cm (fig. 40). The approximate pattern unit measures 18,5 cm in width and 60–65 cm in height.
[Fig. 40]
Microscopic image of damask weave, magnification 50×, 110 warp threads per cm and 30 weft threads per cm.
Working photo (textiles workshop KIK-IRPA).
80Based on stylistic comparisons with other known textiles, the damask can be dated to the mid-17th century.60 It also appears in a garment shown in The Redemption of Christian Slaves by Gaspard De Crayer, dated 1669 (fig. 41 and 42).61
[Fig. 41–42]
Gaspard De Crayer, Altarpiece of the Redemption of Christian Slaves, 1669 (Ghent, Saint-James’ church) (a), and detail, the vestment of the central figure shows a strong resemblance to the damask motif of the casket lining.
Working photos (Judith Goris).
81As described in section Function and Material History, the Diocesan Museum of Namur also conserves a rectangular cloth (46 × 37 cm) featuring a motif similar to that of the casket’s lining (fig. 43). Based on the study of the motif and the textile’s technical properties, it can be confirmed that both are made from the same fabric (see also the dye analysis). The cloth is hemmed on all four sides and was documented in 1864 as having been used as a cover for the crown inside the casket.
[Fig. 43]
Rectangular damask cloth preserved at the Diocesan Museum of Namur, featuring a motif similar to that of the casket’s lining.
Working photo (textiles workshop KIK-IRPA).
82The museum collection also includes a cope made from a very similar fabric in both motif and colour (fig. 44). While the overall motif is largely the same, detailed differences show that it is a different textile. It is possible the cope was made to match the design of the reliquary casket’s fabric.
[Fig. 44]
Cope in red damask preserved at the Diocesan Museum of Namur, featuring a motif similar but not identical to that of the casket’s lining.
© Musée diocésain de Namur (photo Hélène Cambier).
83The conservation of the casket-shrine also offered an ideal opportunity to examine certain materials that required small samples, such as the dyes used for the silk upholstery, the stuffing material hidden behind the silk, and the metal threads incorporated in the metal decoration band on the inner side of the casket.
- 62 Vanden Berghe et al. 2009.
- 63 Note: The allocation of the use of dyes from the Mexican cochineal scale insect (Dactylopius coccus(...)
- 64 Cardon 2007, p. 619–631.
- 65 Dye composition of the silk from the casket-shrine lining (relative ratio of dye molecules at recal (...)
84The dyes from the red silk were analysed using a chromatographic technique called high-pressure liquid chromatography (HPLC) with a photodiode array detection system (DAD), which separates different dye molecules. Thereafter, they are each identified by comparing their spectra with those from and in-house developed reference database. This was performed using the Alliance HPLC system from Waters, following a previously described protocol.62 Before this, the colourants needed to be recovered by acidic extraction from a small fibre sample, approximately 4–6 millimetres in length, of yarn. The silk damask was dyed red with dyes derived from the scale insect Mexican cochineal (Dactylopius coccus Costa).63 These dyes were used in the presence of ellagitannins, which acted simultaneously as organic weighting agents and mordants for the silk. The identification of the Mexican cochineal species provides indirect information regarding the dating of the silk, as this species has been imported from the Americas since the 16th century and has been well-known for crimson red dyeing of silk ever since.64 It strongly suggests that the later reupholstering of the casket’s interior was performed using the same fabric as the rectangular cloth piece from the Diocesan Museum of Namur (see fig. 43). An identical dye composition is observed in the red fibres of both textiles.65
- 66 1967; Von Bergen and Krauss 1945.
- 67 Petrarco and Kubic 2004; Haugan 2013.
- 68 Flax was identified based on the transversal knots present under polarised illumination and the fib (...)
85The fibres of the filling material, located between the silk damask lining and the wooden construction of the box (fig. 45), were identified using a Zeiss AxioImager M1 optical microscope under transmitted and polarising illumination. Fibre identification was carried out based on the fibre morphology in longitudinal view by comparison with the morphological features of reference fibres from an in-house developed reference fibre database and/or published and online atlases.66 For particular fibres, information on fibrillar orientation was obtained through the modified Herzog test.67 The fibres used as stuffing material inserted between the lining and the wooden box are made of flax (Linum usitatissimum L),68 and not white horsehair. This more expensive and prestigious alternative would have been more expected for such a shrine.
[Fig. 45]
Digital Hirox image of the filling material between the red silk lining and the wood.
- 69 Chlorine and sulphur detected on the metals refers to the corrosion of the copper of the brass (KIK(...)
86The metal decorative band at the border of the casket, positioned between the leather on the outer side and the red lining on the inner side, is crafted from wide metal strips and delicate circular metal threads wound around a textile core (fig. 46). Both items were examined using optical microscopy, and the metal’s elemental composition was analysed through scanning electron microscopy (SEM) coupled with energy dispersive X-ray spectroscopy (EDS), employing a Zeiss EVO LS15 system with an Oxford Instruments detector. The fine circular metal thread consists of a metal lamella wound clockwise around a yarn made of flax (Linum usitatissimum L). The winding is highly open, revealing much of the white textile core. The metal lamellae in both the circular threads and broad strips are made of brass with a similar average relative ratio of 85 copper to 15 zinc (by weight).69
[Fig. 46]
Digital Hirox images of the metal decoration band (magnification 20x) (a) and a detailed view of it showing the metal strip and the metal threads wound around the core of white flax (magnification 60x) (b).
87For the treatment, the wooden support received only minimal conservation measures. Although insect damage in the sapwood was confirmed as inactive, the piece was treated as a precaution using semi-dynamic anoxia.
- 70 For more information see De Streel et al. 2016.
88Anoxia is a method of insect eradication in an enclosed atmosphere that has been used for approximately 30 years. It is employed by conservation centres, museums, and private companies, sometimes in static form and sometimes in dynamic form. Since 2007, KIK-IRPA has been utilising dynamic anoxia with the Veloxy® system. The system does not use any chemical precursors. An industrial compressor circulates the air and directs it into the Veloxy® unit, which is fitted with membrane filters. The nitrogen in the air is mechanically filtered and then routed through a system of pipes into the chamber where the object to be treated is placed. The advantage of this system is the absence of fluctuations in relative humidity inside the chamber during treatment, thanks to the vapour-barrier properties of the multi-layer plastic film. Additionally, the device features a relative humidity regulation system within the chamber, which helps maintain a consistent relative humidity in the event of temperature fluctuations outside the chamber.70 Using this technique, the casket was placed in a hermetically sealed bag, and the climate was monitored for approximately twenty days.
89The leather covering the casket has shrunk and torn in several places, particularly along the edges of the wooden panels and near the metal components. Adjacent to the latter, blackening and drying of the leather are visible, likely caused by moisture exposure during a metal treatment with an unknown product. Leather loss is mainly along the studded framing of the vertical panels. Some fragments of the leather are detached or semi-detached at the top of these panels.
90The leather conservation treatment was very delicate, as the leather could tear easily with minimal pressure. The treatment was focused on securing the leather to the support in areas where it had detached, using cornstarch glue and K38 Japanese paper. In places where there was no longer contact between the leather and support, a fill was applied using a shapable paste made of coarsely shredded Japanese paper mixed with starch glue (fig. 47). For aesthetic reasons, coloured paper was added as a finishing layer to mask the white paper on the visible edges, making the intervention more discreet.
[Fig. 47]
Consolidation using long-fibre Japanese paper.
Working photo (Zeina Genadry).
91The latter piece of leather serving as a hinge shows signs of structural weakening but was not consolidated to maintain a minimal intervention approach. Despite the significant alteration to the leather and its fragility, it can be considered that, given the rigidity of the structure it is attached to, the leather should remain stable in the reinforced areas, provided the object is kept in good condition and handled as little as possible.
92The metallic parts of the casket are generally in excellent condition, except for some staining and the accumulation of leather care products, such as wax and lanolin. The restoration process allowed for a detailed examination of the metal decoration.
93The brass nail heads display a uniform and visually appealing patina, which was maintained due to their protective qualities and harmonious appearance with the overall condition of the leather and surrounding materials. In the lower parts of the dome-shaped nail heads, a brownish-black deposit – probably from wax or oil-based leather treatments – was observed. This layer was carefully removed using scalpels and wooden sticks, complemented by cotton swabs moistened with white spirit. Special care was taken to avoid contact with the adjacent leather (fig. 48 and 49).
–49]
[Fig. 48
Cleaning the four small nails in the centre of the picture. The large nail on the right has already been cleaned; the nail on the left remains uncleaned. The bottom area is very dark: a layer of brown wax, probably applied to the leather, obscures the base of the brass dome.
94Black residues were observed around the nails securing the enamel plaques and, locally, on the gilded surfaces. These residues were removed by mechanical means, using wooden sticks, and with white spirit where necessary (fig. 50 and 51). In some areas, aged and acidified wax or oil had caused superficial corrosion of the brass, appearing as dark green, greasy patches. These corrosive deposits, along with the residues that caused them, were also effectively removed with white spirit.
–51]
[Fig. 50
Cleaning an enamelled plate and its fixing nails.
Working photos (Françoise Urban).
95The patina on the nails was deliberately preserved, as it helps provide long-term protection for the metal. No extra coating was applied. The primary objective of this treatment was to maintain the ongoing preservation of the metallic elements, as prolonged exposure to oily substances, pollutants, and dust can accelerate deterioration, as indicated by the presence of copper oleate corrosion products.
96Beyond conservation, the cleaning enhanced the readability of the decorative elements: the outlines of the enamel frames and designs became clearer, and the dome shape of the nail heads regained its definition. Aesthetically, the overall appearance of the casket has improved, with the brass regaining a subtle lustre that harmonises with the preserved patina and surrounding materials.
97The enamel on the medallions shows varying degrees of deterioration, including localised pitting, cracks, superficial scratches, and minor material losses (fig. 52 a–c). In contrast, the enamel on the mounting plate is less affected by pitting. However, it displays more extensive losses, several cracks, and slight lifting in certain areas (fig. 53). Residues from leather care products were found both on the enamel surfaces and within the pitted areas (fig. 52a).
[Fig. 52]
Some aspects of enamel deterioration (
Digital Hirox images).
(a) Pitting (Leica M320 microscope with 10× magnification)
(b) Superficial scratches (Leica M320 microscope with 16× magnification)
(c) Cracks (Microscope Leica M320 microscope with 16× magnification)
[Fig. 53]
Loss of enamel on the mounting plate (
Digital Hirox images, Leica M320 microscope with 40× magnification).98In some areas, a noticeable red line appears at the interface between the copper substrate and the enamel, indicating an intermediate glassy layer rich in copper oxide (cuprous oxide, Cu₂O, also known as cuprite). This feature does not signal degradation but results from incomplete firing during the enamelling process. The layer remains stable under normal conditions (see fig. 54 and 32).
[Fig. 54]
Red line at the enamel-metal interface indicating a cuprite-rich layer (
Digital Hirox images, Leica M320 microscope with 40× magnification).99Treatment of the enamel surfaces involved removing residues with cotton swabs moistened with acetone, supplemented by mechanical cleaning with a scalpel when necessary (fig. 55a and b). Areas of weakened or lifting enamel were stabilised with Paraloid® B-72 (ethyl methacrylate/methyl acrylate copolymer) at concentrations of 3% and 5% in an 85:15 acetone-ethanol solution, applied with a paintbrush or capillary tube. This reversible, long-term stable resin is widely recommended for enamel stabilisation. All treatments were performed under a binocular microscope for precision, and the treated areas were thoroughly documented with detailed photography.
[Fig. 55]
(a) Cleaning under the microscope using a scalpel blade after softening residues with acetone (
Leica M320 microscope with 40× magnification)
(b) Same area after cleaning (Leica M320 microscope with 40× magnification)
- 71 Agent of deterioration 2018.
100To enable safe access for the research and conservation of the casket’s textile interior, while mitigating any stress or strain71 on the diverse materials that make up the object, a custom-made mount was designed and constructed. Special attention was given to the leather component that connects the bottom of the object to the lid, as it is in a highly fragile state. This mount securely held the casket’s lid open, eliminating physical forces during the examination and treatment of the interior. Consequently, the need for direct handling of the object was greatly minimised. The support structure also functioned as an aid for scientific imaging (fig. 56).
[Fig. 56]
Conservation mount.
Working photo
(Griet Kockelkoren).
- 72 Clavir 2002; Kockelkoren et al. 2022.
101In the proposed treatment plan, it was suggested that the interior textiles be carefully dusted using a museum vacuum cleaner and soft conservation tools. However, due to the potentially high intangible value72 of the material particles accumulated over the years on and between – for instance – the textile folds (fig. 57), a modified approach was chosen. A static brush was used to collect and preserve as much dust and other loose material particles as possible, ensuring these particles remain available for future research. During this process, only the most accessible loose material visible on the surface was removed – no attempt was made to eliminate all loose surface materials. The gathered particles included loose interior padding, leather fragments, larger leather particles, along with nails, pieces of metal trim (found at multiple locations), and others.
[Fig. 57]
Careful dust removal with static brush.
Working photo (Griet Kockelkoren).
102As noted in the material technical description, a (padded) red silk damask, made up of several fabric panels, lines the interior of the casket. A trim, made of metal thread, is applied and secured with nails to the interior of the casket and its lid. While the damask, likely dating from about the mid-17th century, is in good condition, the trim shows damage primarily around the closure and at the points of attachment between the lid and the base. These areas also reveal old restorations, likely evidence of the casket’s intensive use and the need for prior reinforcement.
103In accordance with the principle of minimal intervention and maximum preservation of historical traces, the trim was not removed for treatment. After several tests, it was determined that the fragile areas of the trim should be consolidated using ‘two-ply silk thread’, dyed with conservation dyes to match the original trim. A self-couching stitch was employed to wrap the ‘double-threaded silk thread’ around the fragile sections of the trim, providing the necessary strength and ensuring that the loose metal threads were held in place (fig. 58). This method was found to offer sufficient strength, considering the casket will only be opened on rare occasions, while also ensuring that the loose threads of the trim remain secure (in place).
[Fig. 58]
Detail during the active conservation of the trim.
Working photo (Griet Kockelkoren).
104Given the heritage value and current ‘function’ of the object, a minimal intervention approach was adopted for all textile elements. All material traces of ‘the object’s life’ were preserved as much as possible. No part of the textile was dismantled. Notable observations that contributed to material research were recorded, such as exposed fibre ends or larger open areas between the stitches of the seams, which offered insight into the materiality underneath surface. Other observations, including old fabric remnants hidden beneath a nail or under the red lining, also provided understanding of the material biography73 of the object and possibly its use(s).
- 74 Leather, silk, enamel, metals (brass, iron, gold), and wood.
105The casket holding the reliquary crown of Holy Thorns was transferred to KIK-IRPA for assessment of its condition and to carry out a detailed preliminary investigation, as preparation for conservation treatment. The object, composed of various materials previously discussed, showed deterioration caused by the different reactions of these materials to atmospheric changes, dust, light, and other hazards in its current and past conservation environments.74
- 75 ArtGarden or Art Technical Research and Preservation of Historical Mixed - Media Ensembles: ‘Enclos (...)
106On this occasion, the casket, as a notable historical mixed-media artefact, attracted the attention of the ArtGarden researchers.75 The ArtGarden project concentrates on preventive conservation of historical mixed-media objects and expresses the desire to include this captivating object as a case study.
- 76 ArtGarden research: solid phase microextraction (SPME) coupled to gas chromatography with mass spec (...)
107Alongside the research and conservation of the casket, the ArtGarden project expanded its investigation to include analysing the preservation environment and the potential harmful effects of interactions between the different materials within the object itself. The interior of the reliquary casket offers an enclosed space where gases from the constituent materials could affect other materials present.76
108Within the ArtGarden project, the preventive conservation unit developed an online tool to support decision-making for preserving historical composite artefacts. This instrument, called AGATO, provides an automated risk assessment of individual objects based on both the object itself and its environment.77 To test this tool and improve its functionality, an in-depth risk analysis of the casket in situ at the Diocesan Museum was decided.
- 78 Waller 2003; Michalski and Pedersóli 2016.
- 79 Brokerhof and Bülow 2016.
109After considering the risk-analysis methods of Robert Waller (CPRAM) and the ABC-method of Michalski and Pedersóli, the QuiskScan method was employed to conduct a risk assessment.78 QuiskScan, a quick and efficient risk assessment tool or method designed to provide a general overview and understanding of the risks that an entire collection faces, was selected and specifically applied to a single object: the casket. Its results highlight which risks require attention and can assist in prioritising mitigation actions.79
110In close collaboration with the museum curator, Hélène Cambier, the various steps of the QuiskScan were carried out to map the exhibition environment of the object from November 2020 to June 2021. At that time, the casket was displayed on a glass shelf within an acrylic glass display case, without any other objects. The display case was situated in the main exhibition hall of the museum, an unheated room with stone walls and a glass roof, constructed beside the Saint-Aubin cathedral in Namur. To gain a comprehensive overview of the exhibition conditions, all 10 agents of deterioration are considered: physical forces, incorrect temperature, incorrect relative humidity, fire, water, biological agents, light (including UV and IR radiation), theft and vandalism, dissociation, and pollution.80
111During a QuiskScan, characterising the collection is essential to the process. It involves dividing the collection into subcollections based on different types of museum objects, considering their materials and specific vulnerabilities. Next, their relative values are assessed in comparison with other subcollections. After this, all risks are examined for each subcollection. To do this, the vulnerability of each subcollection to any of the ten agents of deterioration is assessed, and the subcollection’s exposure to these agents is determined. These two parameters, together, help establish the risk level (fig. 59).
[Fig. 59]
Assessment of the risk level for the entire collection and the casket involves identifying subcollections (within the collection) or materials (within the casket), their vulnerability, and exposure to each agent of deterioration in the current exhibition setup.
112Ultimately, they compile all gathered data to pinpoint where the highest risk of loss of value is within the collection and estimate the potential loss in that scenario. The same risk assessment is applied to the various materials used in the casket (fig. 60).
[Fig. 60]
3D bar graph showing risks for the collection and casket: potential loss of value if exposed to each agent of deterioration.
113Gathering the data needed for a QuiskScan involves an important discussion with museum staff. During an initial on-site visit, merely 125 questions were asked to the curator to understand the context of potential risks related to daily collection management, the museum’s equipment, and its building. This interview was supported by thorough documentation. The exhibition room housing the casket was photographed, and its floor plan was created. The room, all fixtures, exhibition furniture, technical equipment, and the locations of the collection and sub-collection were marked. Additionally, a layer was added to map the exposure to risks associated with each agent of deterioration.
114In this example, the risk of damage from light and UV radiation is emphasised as it was the curator’s main focus, but all other risks associated with agents of deterioration were also examined and analysed separately (fig. 61).
[Fig. 61]
Light exposure mapped for the entire collection in the exhibition room of the casket after measuring the light levels (see the difference between high and medium radiation exposure, due to the glass section above).
- 81 Mostralog thermohygrometer and data logger, for external + internal measuring with external sensor, (...)
115A brief measuring campaign was organised to support and complete the data collected during the interview. A datalogger recorded both temperature and relative humidity inside and outside the display case of the casket. A second datalogger was installed to measure light and UV radiation. Punctual measurements of light and UV radiation were systematically taken during two separate visits.81
116The results of the QuiskScan finally allowed us to put former concerns into perspective and focus attention on more pressing issues. Light intensity, for instance, proved not to be a major concern. Although the glass roof allows daylight to enter the entire room, the measured light and UV values turned out to be lower than expected, even on bright days. A closer examination of the room’s orientation and the adjacent cathedral ultimately revealed that, for most of the time, the cathedral casts its shadow over the glass roof, preventing sunlight from directly reaching the exhibition room (fig. 62).
[Fig. 62]
Light level measurements testing various display scenarios.
117Continuous measurement of light and UV radiation provided a clearer understanding of the effects of a simple mitigation measure that was implemented. To shield the casket from light damage, a square piece of dark cloth was placed over the display case. The sides of the showcase remained unprotected. The data indicated that with the textile cover, light levels decreased, creating a safer environment for the casket during the period of the measuring campaign. Light levels never exceeded 150 lux and mostly stayed below 100 lux; no UV radiation was detected inside the display case. Surprisingly, the data showed that an unexpected snow cover on the glass roof was almost as effective at blocking harmful light levels.
118To prevent future deterioration of the casket and subsequent loss of value, the QuiskScan results indicated that priority should be given to addressing the high levels of relative humidity, fluctuations in humidity, and enhancing fire prevention measures. Additionally, further measures could be implemented to reduce the risk of theft and water damage (fig. 63).
[Fig. 63]
Results of the in-situ QuiskScan alongside the automated risk analysis of AGATO for the casket.
119The risk assessment guided decisions regarding the preventive conservation of a historic mixed-media artefact, such as the highly valuable casket for the Crown Reliquary of Holy Thorns. Simultaneously, the QuiskScan acted as a dependable reference to verify the functionality of the AGATO tool, which ultimately provides its users with an automated version of this risk analysis (fig. 64).82
[Fig. 64]
Homepage of AGATO.
120The study and conservation of the casket for the Reliquary-Crown of the Holy Thorns have shed light on this remarkable 13th-century masterpiece, emphasising the importance of interdisciplinary research in protecting cultural heritage.
121Dendrochronology established a terminus post quem of 1197 for the wood, aligning with the presumed 13th-century date. The enamel medallions, stylistically uniform and dated to c. 1220–1225, correspond with this period, with no sign of later re-enamelling. However, the lock plate appears to be a later addition due to its less refined style and distinct enamel composition. The textile lining, dated to the mid-17th century, replaces an earlier fabric, which was unfortunately not detected.
122Although the medallions show a stylistic and iconographic unity, specific differences such as nail counts, border types, and enamel compositions indicate subtle variations without denoting separate production or chronological stages. The combination of enamel groups (Type 1 and Type 2, according to Biron’s classification) illustrates the evolution of techniques within a single framework.
123Attribution of the enamels remains uncertain. The medallions do not display strong stylistic connections to Limoges production, and a northern region’s origin cannot be ruled out. Likewise, enamel compositions do not conclusively favour one production centre over another. The provenance of the wood, traced to the Seine and Meuse basins, provides clues but no certainty.
124The conservation work stabilised the casket’s leather, textiles, and enamels, while preventive conservation measures addressed environmental risks to ensure long-term preservation. However, the conservation treatment has, of course, not reduced the high fragility of the object, which must be handled as little as possible. The project highlighted the need to integrate technical, historical, and environmental studies to protect mixed-media artefacts.
Addendum
In 2021, a preventive archaeological excavation conducted at the Halle Al Chair in Namur* (a public building dating from 1588, originally designed to house, among other things, the meat market) uncovered a small masonry basin linked to an earlier phase of the Halle. In the fill of this basin, dated to the 13th century, a small discoidal copper alloy plaque, enamelled in blue and decorated with a gilded foliate scroll, was found.
At the time of finalising this article, the plaque has just been cleaned by Françoise Urban and examined for the first time. Regardless of the questions raised by the presence of such a remarkable object in this archaeological context, the similarity in design, dimensions, and materials with the upper plaque of the Reliquary-Crown casket is striking and prompts further reflection. The analysis of the metals and enamels used in its manufacture certainly merits comparison. To be continued!
(Marie Verbeek, Acheologist at the Wallonia Heritage Agency (AWaP))
*M. Verbeek, M.-L. Van Hove and J.-L. Antoine, Namur/Namur : la Halle al’Chair, étude archéologique au fil de la restauration, in Chronique de l’archéologie wallonne, 2022, 30, p. 257–262.
The study and restoration of the casket were made possible thanks to the generous support of the French Community (Wallonia-Brussels Federation).