Abstract
Ptolemaic cartonnage masks were produced by layering textiles – or reused papyrus sheets – with plaster and glue. Despite the use of the same basic components, the process of manufacture could vary depending on shape, size, time and place. This article aims to clarify the production methods and the different phases of manufacture of these masks, using different imagery techniques, such as X-ray, CT and surface scanning. To provide a better understanding of their production, we examined these masks layer by layer, considering the use of a mould to shape the face from inside, the layering of textile, the application of gesso to strengthen the mask or to sculpt features, and finally the decorative layer of paints and gilding.
Introduction: Evolution of masks and techniques employed
Graeco-Roman cartonnage masks are the result of a long manufacturing tradition. 1 The earliest examples of masks found in a funerary context are from Hierakonpolis. 2 While facial features could be painted directly on bandages or modelled in linen and plaster during the Old Kingdom, it is toward the end of this period that full masks started being produced. 3 Known as helmet masks, they not only covered the face, but also the back and both sides of the head. The use of these masks, usually in cartonnage, flourished during the Middle Kingdom. Their idealised features denoted the new status of the deceased as a divine being. Funerary masks continue to appear during the New Kingdom, although the few examples known are more likely custom made. 4 The use of masks then seems to fade, often being replaced by full cartonnage cases that were particularly in use between the Twenty-Second and Twenty-Fifth Dynasties, 5 and, while examples start to gradually reappear during the Twenty-Sixth Dynasty, probably due to archaism, their production is only clearly revived from the Ptolemaic Period. The tradition then truly flourished, with masks often partnered with other cartonnage elements, such as foot cases and trappings placed on the mummified body. Masks continued to be produced during the Roman Period, although the image of the deceased diversified and appeared on wooden panels, plaster masks and shrouds. They largely lost their stereotyped features to become increasingly naturalistic. The manufacture of these masks varied greatly during the two millennia covered by their use.
The covering of the face was not only a way to physically and magically protect it and to substitute the mummified head in case of loss, but it was also meant to elevate the deceased from a human to a divine status. Specific colours – i.e., gold for the flesh, blue for the hair – are all nuances that would allow an association with the gods. The mass-production of many of these masks would certainly have prevented the facial features from appearing too naturalistic, but this would not abate their magical power – rather the contrary. Idealisation – therefore uniformity – was a solicited trait that would promote the association of the deceased with the gods. 6
Unlike wooden mummy portraits that have been the focus of research for many decades, Ptolemaic cartonnage masks have been little studied. 7 Often from unknown archaeological contexts or without reliable dating criteria, Ptolemaic cartonnage masks are usually broadly dated and their provenance can only be established with great difficulty. One of the first to address these questions was Martin A. Stadler, in his study of the Wurzburg masks. 8 The whole topic, however, still needs a comprehensive study, and a consistent comparison of the entire corpus could allow new insights on date range, provenance and workshop. 9
The present study does not primarily focus on the delicate questions of dating or provenance, but concentrates on the manufacture of the masks, another aspect which has been little discussed until now. The production of cartonnage – with a focus on full body pieces – was the subject of a few studies, using both observation 10 and experimentation, 11 which complement the approach attempted here. Along with close observation of the pieces and their comparison with other existing items, this study aims at using imaging techniques to virtually explore the layers of which the masks are composed. Through the lens of production technology, it is hoped that this research will not only bring new insights into manufacturing techniques, but also into the craftsmen who produced these masks and ultimately their owners, providing a sense of their social, economic and practical significance.
Five Ptolemaic masks in cartonnage
The British Museum’s collection comprises about 40 Egyptian masks. 12 They cover most periods, starting from the Middle Kingdom, with the majority of examples dating to the Graeco-Roman Period. This paper will concentrate on five examples dating from the Ptolemaic Period. The selection was based on several criteria, including their state of preservation, their style, and their date and provenance (where known). We will start by describing these five pieces and presenting any known associated objects, as well as any information collated on their acquisition and provenance history.
British Museum EA 50668
H. 40 cm; W. 21 cm; D. 24 cm
Dating: Early to mid-Ptolemaic Period 13
Provenance: Atfih, Egypt
Acquisition: Donated by the Egypt Exploration Fund, 1911
Published: Taylor, in Mack (ed.), Masks: The Art of Expression, 179, fig. 123.
This mask’s monochrome dark blue/black wig is entirely outlined by a white band (fig. 1). The face is uniformly gilded, with black painted eyes and eyebrows, the sclera being left white. A stylised collar is painted between the wig’s lappets, with alternating floral and dotted motifs spread over nine registers. The pigments used are both dark and light blue, red, white, and black for the outline.
14
The back and sides of the mask are painted in plain dark blue. Mask EA 50668 (© Trustees of the British Museum).
The museum register records that this mask was found in Atfih and was donated to the museum in 1911 by the Egypt Exploration Fund (soon to become the Egypt Exploration Society). Although there is no mention of a specific excavation, it was very likely found during the 1910–11 season led by John de Monins Johnson, who directed two seasons of excavation in Atfih. 15 Primarily focusing on papyri, he excavated a large number of burials and found up to 600 mummies with cartonnage material during this single season. 16 The mask does not seem to be mentioned in the distribution list of objects given to the British Museum after the excavation, 17 but is likely listed in June 1911 by H. R. Hall among the items presented to the British Museum by the EEF: ‘From Atfîh (Aphroditopolis): Stone mould for an ushabti. Gilt cartonnage head & feet cases for a mummy. Two pots with demotic inscription. Ptolemaic Period. Three alabaster vases. One scarab. XIIth Dynasty’. 18 The cartonnage head case is possibly the mask discussed here. The document unfortunately offers no information on the context in which the mask would have been found or on the state of preservation of the mummy associated with it.
The British Museum records, however, that the mask came with an ensemble of cartonnage trappings that were found covering the mummy (fig. 2). The set consists of: a pectoral in the shape of a broad collar with alternating dotted and floral motifs, topped by a winged scarab with a papyrus column in between and flanked by a falcon head on each side; a rectangular piece showing a goddess spreading her wings flanked by the four Sons of Horus; a leg piece with alternating dotted and floral motifs; a foot case showing feet in red sandals and soles on the base of the case. Cartonnage trappings EA 50668 (© Trustees of the British Museum).

The decoration (style and colours) is consistent over the five pieces, with recurring motifs such as the dotted and floral pattern. This homogeneity suggests that they could well have belonged originally to the same set.
British Museum EA 49376
H. 36 cm; W. 27 cm; D. 16 cm
Dating: Early to mid-Ptolemaic Period
Provenance: Sedment, Egypt
Acquisition: Donated by the Egypt Exploration Fund, 1910
Published: Taylor, in Mack (ed.), Masks: The Art of Expression, 180, fig. 124.
Because of its beard covering most of the cheeks, its thick eyebrows and eyelids, mask EA 49376 is particularly startling (fig. 3), but it is the treatment of the mouth that makes it unique (the teeth are usually not visible between the lips). While the face is rather densely decorated, the blue wig remains unadorned, with a broad necklace consisting of a simple dot and flower pattern appearing between the two lappets. Mask EA 49376 (© Trustees of the British Museum).
The piece was acquired by the British Museum in 1910, again from the Egypt Exploration Fund. It was found during excavations taking place in Sedment. During the winter 1909–10 season, the Graeco-Roman branch of the Egypt Exploration Fund, led by Aylward Manley Blackman and John de Monins Johnson, 19 explored the area in the search for papyri. They worked in the Ptolemaic sector of the Sedment cemetery between November and December, until they believed that the potential of the site for ‘papyrus mummies’ was exhausted. 20 The archaeologists did not produce any detailed report, and we do not know more precisely in which context the mask was found and whether the individual to whom the mask belonged was buried with any other artefacts.
The painted teeth on the mask are such an unusual detail that we could not help wondering whether the feature was truly original or whether it could be a modern addition, or even part of a repair. A photograph in the archives of the Egypt Exploration Society attests, however, the presence of this feature soon after its discovery (fig. 4).
21
Although we cannot completely exclude the possibility that the mask had been retouched between the time of its discovery and when the photograph was taken, this would seem unlikely in the context of these excavations. Mask EA 49376 soon after its discovery (courtesy of the Egypt Exploration Society; AB.NEG.11.608).
British Museum EA 51146
H. 38 cm; W. 22 cm; D. 29 cm
Dating: Late Ptolemaic Period – early Roman Period
Provenance: Tomb E 422, Abydos, Egypt
Acquisition: Donated by the Egypt Exploration Fund, 1912
Published: T. E. Peet, The Cemeteries of Abydos. Part II: 1911–1912 (MEEF 34; London, 1914), 92, pl. XXXVIII, 3 and pl. V, 5; D. Antoine and M. Vandenbeusch, Egyptian Mummies: Exploring Ancient Lives (Sydney, 2016), 178–9.
In perfect condition, mask EA 51146 was finely produced and decorated, with its shiny gilded face (fig. 5). The face is surrounded by a simple blue wig enhanced with a gilded winged scarab pushing a sun disc on top, a yellow band around the head, and strips of colours at the extremities of the lappets. Eyes and eyebrows are delicately painted in white and black, with small red dots at each corner of the eye. Between the lappets, the stylised broad collar is made of the usual motifs showing alternating flowers, petals, and dotted and geometrical patterns. The wig is separated from the face by a thin white band, with small red dots along the forehead, and a thin red band. Mask EA 51146 (© Trustees of the British Museum).
The context in which this mask was discovered is relatively well known: it was unearthed in tomb 422 of cemetery E, at Abydos, during an Egypt Exploration Fund dig led by Thomas Eric Peet.
22
The mask was then donated to the British Museum in 1912. The excavations were soon afterwards published by the archaeologist, who described the tomb as a well preserved vault tomb with seven limestone coffins, all containing mummies with cartonnage, as well as five coffins containing bodies of children and a cat mummy.
23
The publication also illustrates the mask in its original context.
24
The published photograph shows the mask still on the mummy lying inside an apparently uninscribed stone sarcophagus (there is no evidence of a lid in the picture) (fig. 7). The photograph also reveals the presence of other cartonnage pieces. These were acquired along with the mask by the British Museum (fig. 6) and consist of: a broad collar with a winged sun disc on top flanked by two falcon heads and a pectoral showing a shrine hanging in the middle; a rectangular piece depicting the four Sons of Horus separated by columns of texts indicating their names and a central djed pillar; a rectangular piece showing a seated winged goddess; a leg piece with openwork openings and a central hieroglyphic inscription flanked by floral motifs and a row of guardian deities holding knives on top; a foot case with a central inscription, flanked by the two feet in sandals with gilded toenails, topped by a jackal seated on a tomb and mourning goddesses at the sides. The base is decorated with two sandal soles separated by a lotus flower. Rosette motifs run along the bottom edge. Cartonnage trappings EA 51146 (© Trustees of the British Museum).

The five pieces were placed (and some attached, as suggested by the holes at the corners of some pieces) on the mummy’s outer shroud. Every item was painted using a similar colour palette and their uniformity could indicate that the set was produced in the same workshop, although the small variations observable from one piece to the other suggest that they were not necessarily all made by the same hand.
The set was certainly not produced with a specific individual in mind, as mass production is made obvious by the absence of a name in the inscription – a protective spell involving Atum and Khepri. A space was not provided for the potential addition of the deceased’s name, which was replaced by three ‘commas’, a motif that appears consistently on the foot case, the leg piece and the Sons of Horus piece.
Other photographs preserved in the Egypt Exploration Society archives show the mummy out of the sarcophagus, standing against a wall (fig. 7).
25
Unfortunately the mummy was thought to be ‘only moderately preserved’
26
– an incredible statement by modern standards! – and only the pieces of cartonnage were kept.
27
Mask and trappings EA 51146 soon after their discovery (courtesy of the Egypt Exploration Society; AB.NEG.11.020, AB.NEG.11.024, AB.NEG.11.603).
British Museum EA 29474
H. 51 cm; W. 25 cm; D. 37 cm
Dating: Mid to late Ptolemaic Period
Provenance: Unknown
Acquisition: Purchased from R. J. Moss & Co, 1897
Published: J. Mack (ed.), Masks: The Art of Expression (London, 1994), 30, fig. 12; L. Miatello, ‘Ptolemaic mummy masks with spells from the Book of the Dead concerning the head’, JSSEA 39 (2012–13), 51–85.
Unlike the three masks presented above, the last two do not come from known archaeological contexts. Their great similarities are, however, worth exploring as part of this research.
Mask EA 29474 is decorated with an intricate design (fig. 8). The face, ears, neck and collar are gilded. The face has large black eyes, with a naturalistic red dot in each corner. The top of the mask is decorated with a gilded scarab pushing a sun disc, surrounded by a stylised band of triangular motifs and a gilded band of text reproducing the beginning of Spell 151a of the Book of the Dead.
28
The face is enclosed in a striped wig with lappets ornamented by chequered bands and a pattern of multi-coloured lines. Only visible between the lappets, the broad collar consists of geometric and floral motifs in relief. Mask EA 29474 (© Trustees the British Museum).
The decoration at the back is composed of four registers (from top to bottom): four gods, probably Osiris, Horus, Isis and Sekhmet, identifiable by their attributes or by the crowns on top of their heads; a ba bird spreading its wings; a falcon in a very similar position; a series of vertical rectangles of alternating colours.
There is no text on the back panel, but columns seemingly meant to receive inscriptions were prepared. They remained unused and were painted in various colours, perhaps because they were mistakenly reinterpreted as decorative motifs.
The mask shows evidence of heavy conservation work, especially at the back, which was largely remodelled. We do not know when the damage occurred, the context and circumstances of discovery being unknown. The recent history of the mask does not go back further than 1897, when the mask was purchased from R. J. Moss & Co, a British shipping company that E. A. W. Budge, the then Assistant Keeper of Egyptian and Assyrian Antiquities at the British Museum, regularly employed to buy, store and ship antiquities on his behalf. There is no evidence of its provenance or connection to a mummy. When registered in the museum collection, it was associated with the foot case British Museum EA 29475. Both mask and foot case are of similar manufacture with comparable motifs (particularly the chequered motif, the choice of pigments and the gilded inscription in raised relief), suggesting that they were produced in the same workshop. The foot case decoration is centred on red feet with nails and sandal straps enhanced with gilding (fig. 9). The inscription (a formula to be said by Osiris) in-between is also gilded, while Isis and Nephthys are seated on either side and two black jackals on a tomb entrance face each other on top of the case. The base of the case reproduces soles decorated with bound enemies. Neither the inscriptions on the mask, nor those on the foot case were produced with a specific individual in mind, as indicated by the lack of name and title. Foot case EA 29475 (© Trustees of the British Museum).
British Museum EA 29472
H. 44 cm; W. 22 cm; D. 31 cm
Dating: Mid to late Ptolemaic Period
Provenance: Unknown
Acquisition: Purchased from R. J. Moss & Co, 1897
Published: W. Seipel, Ägypten: Götter, Gräber und die Kunst. 4000 Jahre Jenseitsglaube (Linz, 1989), 344; C. Andrews, Eternal Egypt: Egyptian Treasures from the British Museum (Hong Kong, 1998), 186–9, no 57; C. Andrews, Egyptian Treasures from the British Museum (Santa Ana, 2000), 140–1; Stadler, Ägyptische Mumienmasken in Würzburg, 35, 40; J. H. Taylor and N. C. Strudwick, Mummies: Death and the Afterlife in Ancient Egypt. Treasures from the British Museum (Santa Ana, 2005), 70–1; J. H. Taylor, Journey through the Afterlife: Ancient Egyptian Book of the Dead (London, 2010), no. 58; Miatello, JSSEA 39 (2012–13).
With its gilded face, striped wig, band of text running on top and the almost identical four-register panel at the back, the style and design of this mask are globally similar to the previous one, in spite of its smaller size (fig. 10). It differs, however, when scrutinized in detail. The wig for example is not enhanced by additional motifs, but is simply covered with stripes. The impression of a plainer example clashes with the more extensive use of gilding which not only covers the stripes nearer the face, but also those at the sides and back of the wig. The large alert eyes, with their unusual visible pupils, draw attention directly to the face. The mouth, nose, chin and ears are emphasised by an outline in raised relief. On the neck, two strings of beads support an ankh amulet, on top of a broad collar in raised relief, which consists of a succession of motifs very similar to those of mask EA 29474. The top is also ornamented with a winged scarab surrounded by a band of text (as before, an extract from Spell 151a of the Book of the Dead). The scenes at the back are extremely similar, although on EA 29472 the text fills the columns in the lower register. The translation of this inscription has proven difficult, but Luca Miatello suggested that it refers to libations.
29
The difficulty in reading the text on the back panel could indicate that it was not inscribed by the same individual who added the text on the surrounding band. As on mask EA 29474, the other rectangles, possibly meant to receive an inscription, remained empty and were simply filled with a layer of red pigment. The other minor differences which can be observed – for example the style of the ba bird’s head or the use of a chequered pattern – could suggest that the two masks were produced by two different artists, using the same or a similar model. Mask EA 29472 (© Trustees of the British Museum).
The mask was acquired with a foot case (EA 29473), composed in a similar manner to foot case EA 29475 (fig. 11), although like the mask it reveals a more extensive use of gilding. The motifs around the base, and the treatment of the jackal figures and the feet vary as well. The base of the case depicts the two sandal soles with a papyrus in between and a palace façade motif underneath. Foot case EA 29473 (© Trustees of the British Museum).
Emphasising even further the similarities between the two sets, mask and foot case EA 29472 and EA 29473 were also purchased from R. J. Moss & Co and were registered in the British Museum collection in the same year as EA 29474 and EA 29475. Their common history suggests that they were found in the same necropolis, which remains unfortunately unidentified.
Digital technologies as a way to study the manufacture of the masks
X-ray CT scanning is used to provide a volumetric image of the internal and external structure of an object, based on the relative densities (and thus X-ray attenuation) of its component materials. Scan data can be explored as a 3D rendering, or a series of 2D slice images through the object in any orientation. X-ray CT is well-suited to visualising features and objects which are otherwise hidden, for example, a surface underneath a layer of gesso, or the inside of the mask, which it is not practical to scan by optical methods. 30 Limitations of X-ray CT scanning include the maximum object size, as determined by the system geometry (specifically the size of the detector, and the relative distances between the X-ray source, object and detector), and image artefacts caused by dense objects - e.g. metal nails in the mounts supporting the masks. These artefacts manifest as streaks through the volume centred on these dense objects, and give misleading X-ray attenuation values for the surrounding material.
Cone beam X-ray CT scans of the five masks were acquired in the British Museum using an YXLON Access Y.100 system, with images recorded on a PerkinElmer XRD 1621 AN15 ES flat panel digital detector (409 x 409 mm active area, 200 x 200 µm pixel pitch) (fig. 12). For each scan, the positions of the X-ray source and detector were kept fixed, and a series of X-ray projection images was acquired, the mask being rotated a small amount about the vertical axis between each projection. This process was continued until a 360-degree rotation of the mask was recorded. Volumetric CT reconstruction was performed using the VolumeGraphics VGStudio 2.2 software package using the filtered back projection algorithm. The CT scan parameters for the masks are given in Table 1. Mask EA 29474 being CT scanned at the British Museum X-ray facility (© Trustees of the British Museum). X-ray CT scan parameters. The source to object distance refers to the spacing between the focal spot of the X-ray tube and the centre of axis of rotation of the mask.
Due to the size of the masks and the inherent magnification of the X-ray imaging geometry, it was not possible to record the complete mask in the projection images. Instead, region of interest scans were performed, with the face of the mask positioned at the centre of rotation. The resulting 3D reconstructions therefore excluded regions of the mask, which were not in the field of view throughout image acquisition, typically the rear and lower regions of the mask.
Layer by layer: Exploring the cartonnage masks
Combining observation and new technology, this study brings us closer to the masks’ manufacturing techniques, and ultimately to the people who produced and owned them. The five masks described above will be examined layer by layer, following the steps of production, from their structure to their decoration.
Structure: Moulding
The first production step of a mask or any sort of cartonnage piece is to form the general shape. A separate element was usually needed to mould and define the contours of the piece.
31
The peculiarity of Ptolemaic cartonnage masks is the remarkably crisp internal facial features, in comparison with the far less well-defined exterior (figs. 13 and 14).
32
This trait suggests that the masks were modelled on a mould made of wood, stone or plaster. CT scan visualisation showing the internal and external surface of the five masks. View of the inside of mask EA 29474 (© Trustees of the British Museum).

The full-body cartonnage cases of the Third Intermediate Period provide interesting points of comparison. It is supposed that a core of coarse mud was used to shape the body, while the face was moulded from a separate piece, made of a harder material that would provide more precise contours.
33
The various British Museum examples that have been CT scanned indicate that this piece could be made of wood or a dense material similar to plaster, probably suggesting different production methods.
34
While the core in mud was removed, the harder material forming the face was regularly left in situ. The techniques used to produce a full-length cartonnage could vary from those used for a mask, with more layers of fabric attested for full cartonnage cases, although the combination of a core in mud and a mould seems to have been a feature of both processes.
35
As Ptolemaic masks were heavily mass-produced, the reusability of the mould was of importance. Such an object still remains to be firmly identified, and the British Museum sculpted limestone element EA 13352 is worth considering here (fig. 15). Unpublished, it was described on the museum’s database as a female head wearing a wig. The front is delicately carved, but the back of the piece was left rough. While this feature would exclude the piece having originally been part of a statue,
36
it is an interesting feature if we consider its use as a mould. An uneven surface would certainly help its adhesion to the mud core, to which it would have to stay attached throughout the production process, while the smoothness of the face would facilitate a careful separation from the drying cartonnage. The piece was purchased in 1878 from Greville John Chester, who collected extensively for museums including the Ashmolean and the British Museums.
37
Although the head has no known archaeological record, it is said to be from Benha el-Assal in the central Nile Delta. This town is adjacent to the settlement and necropolis of Athribis, an important centre during the Graeco-Roman Period. Limestone piece EA 13352 (© Trustees of the British Museum).
The limestone piece measures 23 cm high, 17.5 cm wide, and 8.5 cm in depth. The volume of the sculpture was measured from a 3D scan of the surface (using an Artec Eva with Artec Studio 15 software). The dimensions and proportions roughly fit the impression left inside the masks discussed here, although unsurprisingly none is a perfect match, confirming that the sculpted piece was not used to produce one of the five masks discussed here. It seems generally slightly smaller than the masks scanned but the volume and details suggest that it could have been used to produce cartonnage masks of smaller size.
38
Comparing the profile of each mask shows that different moulds had visibly been used to produce them. The internal impressions revealed by the CT scan confirm the strong resemblance between the masks, the proportions of which could vary as indicated by the sagittal views (fig. 16). Masks EA 29472 and EA 29474 share, for example, many similarities in the proportion of the nose, the position of the mouth and the angle of both chin and forehead. Conversely, the features of mask EA 51146 appear at first to be in greater contrast with the other profiles. However, the proportions – of the chin in particular – are likely to be not strictly comparable to the original ones, rather a result of a slight distortion due to pressure on the mask. CT scan slices showing the profile of the mould (left) in comparison to each mask. The sections show them through from the middle of the nose.
The delicate carving of the limestone face corresponds to the level of detail visible on the inside of the masks. 39 Perhaps was it necessary to ensure as many visible features as possible internally, as they would be partly lost on the exterior through the layering of material. These layers of textile, gesso, and glue would have been applied on top of the mould and core. Unfortunately, analysis attempted on the limestone sculpture to identify some of the materials layered – textile, gesso and glue for the cartonnage and a possible mixture of mud and straw for the core – remain so far inconclusive.
In spite of all the comparable features, one element of the limestone face remains difficult to explain: it seems that someone attempted to paint the eyes and eyebrows. If the piece had been used exclusively as a mould, there would be no reason for these details to be enhanced. Although we cannot exclude the possibility that these details were added at a later stage, they cast a doubt on the use of the piece. Its identification as a mould remains, however, very attractive. 40
Structure: Applying the textile
Height of each mask compared to the thickness of the cartonnage and the average number of layers of fabric (resolution of acquisitions: 89 µm).
The present study revealed that it is not only the number of layers that could vary, but also their positioning. EA 29472 and EA 29474 were both covered by large pieces of fabric, consisting of two bands, one covering the face and back, and the other crossing transversally from one side to the other. The edges of EA 29472 are visibly thicker that the rest of the cartonnage and the CT scans show that the textile has been folded to form the layers of cartonnage (fig. 17a). The folding was not used for all the masks and the pieces of fabric were in some cases cut to the required size (for example EA 51146). These pieces of textile would have adhered to each other using glue. Analysis of glues on similar cartonnage pieces show that they are often of animal origin, particularly from egg white.
42
CT scan slices of mask EA 29472 showing the folded edge (a) and the additional gesso to form the decoration of the collar (b).
Some of the weaving features of the fabric are occasionally apparent. A double self-band, which would have followed the edge of the original piece of fabric, is visible on the internal side of EA 29474. The same mask was at some point heavily conserved. The repair, which mainly focused on the damaged back of the piece, consists of a patchwork of small squares of textile (perhaps reusing the ancient fabric). This work is not only visible when looking at the mask itself, but also on the CT scan which shows an irregular layering of fabric, differing greatly in appearance from the front.
Over the internal surface of most of mask EA 29474, a thin layer of a brownish substance, probably a residue of the core used to model the mask, can be observed. It would have been made of a mixture of mud and straw, perhaps covered by some finer clay to help to detach it from the mask once the cartonnage had dried.
Structure: Applying the gesso
The layers of textile and glue were then covered with a coating of gesso. It was meant to fill in any undesired cavity and to smooth the uneven surface of the fabric in preparation for the decoration layers. The gesso is a gypsum plaster, a mixture of powdered limestone and a binder. 43 The higher density of the gesso makes it particularly well perceptible on the CT scan (it glows white). 44 Its thickness and application can be observed on top of the solidified fabric. It shows that EA 29472 is covered with a thin and regular layer of gesso, an application similar to EA 50668 – the latter is particularly thin and uniform throughout the surface. EA 29474 differs slightly, showing less regularity in the layering. It is thicker around the nose and the ears, as well as on the transitional corners between the face and the wig. It shows that the gesso was not only used to even the surface, but also in certain cases to remodel some features: the ears were thin and flat, and the gesso would have allowed a more 3D rendering.
Apart from the general shape of the face provided by the mould, the decorative details of EA 29472 and EA 29474 45 were modelled on top of the gesso surface, as an additional layer (fig. 17b). The CT images show two distinct layers, the decorative one probably being added after the first layer had dried.
Decoration: Gilding
Decorating the surface was a crucial step. It would not only enable the mask to acquire its magical function, but would also diversify the appearance of the otherwise uniform pieces. Although these masks were seemingly produced on a (semi-)industrial scale, we will see that decoration, motifs and application could vary. A recurrent technique used is gilding. Thin pieces of gold foil would provide the masks with their impressive appearance, which made them attractive to both early excavators and modern museum visitors. Such gilding on the face was symbolically meant to ease the deceased’s transfiguration as a divine being. 46
The gilded surface of the five masks appears uniform to the naked eye. It is only when exploring the same areas on the CT scan rendering that structural detail is revealed. Mask EA 29472, for example, shows the application of square and rectangular pieces of material, the largest measuring about 4 x 6 cm and smaller pieces filling the gaps (fig. 18, right). A fairly similar treatment can be observed on some of the other masks: EA 51146 is another clear example of the application of gilding foil covering the entire face with the exception of the eyes (fig. 18, left).
47
It differs, however, on EA 50668, a mask which is also set apart by its consistently thinner structure. The layers of gilding are not visible, suggesting once more a divergent treatment: either the gilding is thinner, or its application differed. Square gilding foils covering masks EA 29472 (right) and EA 51146 (left).
The gilding was sometimes burnished, and some masks display traces of the tool used on the surface to allow it to adhere. A red bole (a red clay) – not identified on the British Museum examples – was used in some cases as a preparatory layer, enhancing the colour. 48 This preparatory layer was seemingly not deemed essential since gilding could also adhere directly on gesso. 49
Decoration: Painting
Important variations in the palette and decorative motifs, as well as in the techniques of application, can be observed from one example to another. A detailed study will show that very similar masks such as EA 29472 and EA 29474 could differ in the layering of the colours and pigments (fig. 19). Detail of masks EA 29474 (top) and EA 29472 (bottom) (© Trustees of the British Museum).
The layers were generally applied one pigment at a time. The first layer of EA 29474 is the gesso, also used for its white qualities. A red outline was then sketched on the surface to map out the different features. 50 A more specific examination of the eyes shows that the painter applied consecutively the black, the dark blue, the red, and finally the gilding. The order in which the pigments were applied is made clear when observing paint overlaps, revealing that one was systemically applied on top of the other. On top of the head, the dark blue was applied on the gesso, and was then followed by the red and the green (it is difficult at this stage to identify which came first, as no direct overlap can be observed), and finally the gilding, with the edges of the gold leaf clearly visible. The dark pigments forming the outline of the motifs were usually applied first, and only then would the colouring and gilding be added. On EA 29474, only the first three rows of the striped wig were gilded, while the remaining ones were left bare (i.e., white). The application of the gilding itself is even more surprising: added last, on top of all the other pigments, it overruns some of the painted areas, giving the whole mask an irregular, even untidy, look. This is one of the main differences with mask EA 29472, on which the gilding, which covered the front and back of the mask, was applied at an earlier stage, followed by the dark blue pigment. Because the raw edges of the gold foil are hidden under the blue pigment, the overall aspect of the mask seems cleaner. The eyes of mask EA 29472 also mark an interesting, and rather unusual, difference with the iris painted in two shades of black, a darker one marking the pupil. 51
Mask EA 29472 revealed another unexpected feature. Another pattern is just visible under the gilding (fig. 20). Small squares – possibly red – can be singled out at regular intervals, suggesting that the decoration was originally planned differently, perhaps as a wig with small gilded rings placed at equal distance on the stripes,
52
before applying the gilding and blue pigment and producing a striped wig. Another explanation for this feature could be an overlap of the gold leaf, given the regularity of the pattern that approximately matches the size of the leaves observed on the face. These markings could indicate the overlap locations that would point out in red paint where the gold leaf was to be applied. Detail of mask EA 29472 showing (arrows) the wig and motifs still visible in red underneath the gilding (© Trustees of the British Museum).
Finally, holes can be observed at the bottom edge of masks EA 29472 and EA 29474. Masks and other cartonnage pieces are regularly found pierced with holes, and examples photographed directly on mummies confirm that they were meant for a cord or string to secure the piece on the mummy. 53 This is the case with the cartonnage trappings EA 51146: the panels placed on the body were sewn to the outer shroud, while the mask and the foot case were simply placed over the wrapped body.
Comparing the manufacture with the production of a mask of the Roman Period: The example of British Museum EA 69020
Roman Period masks are recognisable by their more individualised look. They are usually considered as ‘cartonnage’ masks, suggesting that they were produced using the same techniques as the Ptolemaic examples. As part of this project, a mask dating to the first century AD was CT scanned to gain a better understanding of its materiality and manufacture, and to compare it to the Ptolemaic masks discussed above. Although additional examples will be scanned and studied for a more extensive examination, the preliminary comparison of mask EA 69020 challenges preconceived ideas regarding its production and reveals evidence of new manufacturing techniques (fig. 21). Mask EA 69020 and CT scan section crossing the centre of the face (© Trustees of the British Museum).
Mask EA 69020 belonged to a woman named Aphrodite, daughter of Didas. It was excavated at Hawara by W. M. Flinders Petrie. 54 The CT scan data revealed a very different technique used for the fabrication of this piece. The mask consists of a relatively thin layer of clay, which was probably applied on a sort of core, perhaps in mud and straw. Several layers of textile – the data resolution did not allow the identification of the number of layers – were then applied on this surface. The thickness of the relatively coarse fabric varies in places. At this stage very little of the facial features were already modelled. Only the shape of the nose and the general form of the face could be distinguished. On top of the core layer is a much thicker layer of gesso (or a material denser than the core layer). 55 This thicker layer was used for modelling the face, shaping the lips, the ears and earrings, and the nose, as well as the individual curls of hair that were added separately on top. The composition of this layer is not yet known, but could resemble that of mask Fitzwilliam Museum E.103A.1911, also from Hawara. The paste composing the Fitzwilliam mask was analysed and the results showed that it consisted of a variety of local rocks. 56
These preliminary observations show that the techniques and the material used to shape and decorate the mask varied greatly from those of masks produced during the Ptolemaic Period. Instead of an internal mould, it was the external layer of gesso that was used for the shaping of the various features, while the fabric reinforced the structure. Because of the sheer quantity of material, the mask is much heavier than the pieces made of thin cartonnage, hindering its easy handling. Conversely, the possibility of adjusting and re-shaping the face and other features was a substantial advantage sought by the craftsmen to individualise each mask. While the Ptolemaic clientele did not solicit facial likeness, this new trend marked a shift that certainly influenced the Roman manufacturing production. This shift in production does not mean that all masks made during the Roman Period were thought as portraits of their owners. Nevertheless, as the face and extra features (i.e., hair, jewellery) were modelled on top of the core, the craftsmen appeared to have more freedom when creating each mask, either seeking a likeliness with the future owner, or based on tastes, trends or models.
Conclusion
The five Ptolemaic masks examined here were produced following broadly the same technique, starting with the use of a mould, perhaps a piece comparable to the limestone face British Museum EA 13352. The resemblance of this object to the masks’ internal impressions is particularly startling. The discovery of the masks at different sites across Middle Egypt confirms the wide distribution of moulds, models and technology. The layering process could vary slightly, perhaps depending on the material and the resources at hand. 57 Far less material was used to produce mask EA 50668 than the others, and its gilding was thinner, perhaps denoting a cheaper version. It is often during the painting and decoration process that the pieces can be more visibly differentiated, as no two masks are identical. This is clear for the two masks EA 29472 and EA 29474: probably discovered in the same cemetery, their decoration and the pigment layering varied greatly in spite of their general resemblance. More research is needed on Roman Period masks in order to obtain a better understanding of the variety of techniques that were then used, but it seems that, at a time when more lifelike features were sought after, the process eventually changed and the use of a mould became redundant.
Because of the mass manufacture of Ptolemaic masks, we know very little about their owners. The inscription of names and titles is scarce on these objects, which cannot often be attributed to a specific individual, a family or a social group. The owners and their families were certainly wealthy, or at least rich enough to have such masks produced for them. 58 Access to both gilding and funerary texts is another hint toward their socio-economical background. Although study of their manufacture confirms their mass production, it also reveals that they were produced following varying steps of fabrication, and this gives a unique insight into mask manufacturers, their techniques and skills.
X-ray CT-scan technology was used for the first time to explore Ptolemaic cartonnage masks, revealing new insights into their manufacture, such as the varying number of textile layers used to produce each mask (two to four layers were used on these five examples), their positioning and, in some cases, the addition of extra layers to reinforce the most fragile areas, such as the nose. CT scans confirmed striking similarities in the internal faces of masks from the Ptolemaic Period, as well as the particularly uniform application of gesso on top and the interesting application of sheets of gold. The crispness of the internal features suggests the use of a mould. In this paper, we consider the sculpted piece EA 13352 as possible evidence of such a mould. If confirmed, this would certainly provide a new understanding of the production process and, while not yet totally conclusive, would give a first clue to the mass production of masks. Although many questions remain unsolved, a combination of imagery and observation of cartonnage masks allowed this first step towards understanding cartonnage mask manufacture, from initial shaping to final decoration.
Footnotes
Acknowledgments
We would like to thank our colleagues at the British Museum for their help, support and advice: Caroline Barton, Neal Spencer, Emily Taylor, and John H. Taylor in the Department of Egypt and Sudan; Janet Ambers, Joanne Dyer, Carl Heron and Catherine Higgitt in the Department of Scientific Research; Nicholas Lee in the Department of Conservation; Dan Pett (former Senior Digital Humanities Manager) and Amelia Knowlson (visiting PhD researcher from Sheffield Hallam University) for testing the different 3D modelling techniques; as well as Stephanie Boonstra (Egypt Exploration Society) for her kind help and for giving us permission to reproduce archival photographs. Earlier versions of this research have been presented at the British Egyptology Conference 4 (Manchester, September 2018) and at the XIIth International Congress of Egyptologists (Cairo, November 2019).
Funding
The authors did not receive funding for this project.
