During the first three centuries of the Christian era, funerary artisans in the Roman province of Egypt produced a unique genre of mummy covering which reflected the multicultural society that existed there. These mummies exhibited a meld of ancient Egyptian religious beliefs, Hellenistic social norms and Roman Imperial fashion. The body of the deceased was first embalmed and then encased in intricately wound linen bandages or enclosed in a cartonnage casing.¹ Cartonnage is made from layers of plaster and linen to create a thin shell. Then either a wooden panel was attached to the front of the mummy, over the position of the head, or a linen shroud was wrapped around the body. On the panel or shroud a ‘portrait’ ² Within the confines of set types, the portraits range from highly individualised examples to those that were clearly produced for a mass market (see L. Corcoran, Portrait Mummies from Roman Egypt (I–IV Centuries A.D.): With a Catalog of Portrait Mummies in Egyptian Museums, Chicago, 1995, pp. 68–74; D. Montserrat, ‘The representation of young males in ‘Fayum Portraits”, in Journal of Egyptian Archaeology, vol. 79, 1993, pp. 215–25 for a discussion on types and the portrait’s role in establishing the deceased’s social position). of the deceased was painted using either encaustic wax or tempera and very occasionally, a combination of both. The mummy proper is clearly identified with Egyptian culture as are the funerary symbols and motifs. The portrait panels, however, follow the classical Greek style of portraiture perfected in Athens during the fourth century BCE. Hellenistic influence is also found in the figures decorating the casings and shrouds, which are often depicted in naturalistic poses, unlike the stylised poses normally associated with Egyptian art. The portraits show the deceased wearing Greek dress, decorated with clavi (stripes) in the Roman manner and sporting fashionable Roman Imperial hairstyles and jewellery. The Roman Imperial family exerted just as much fashion influence throughout the Roman Empire as royalty and film stars do today. Affluent Egyptians were never far behind the latest trends in Rome because of the constant transfer of Imperial iconography to Egypt in the form of official statues and coins, and the exchange of ideas facilitated by officials and commercial travellers moving between the two countries. Dating portrait mummies has always been difficult because, unlike many other Egyptian objects, portrait mummies have no dates inscribed upon them.³ The term ‘portrait mummy’ refers to a complete example of the genre (consisting of both mummy and portrait), whereas the term ‘mummy portrait’ or ‘portrait’ only refers to the portrait itself. Nor have they been found with any significant grave goods to help determine their date of burial. Therefore, Egyptologists have been forced to develop dating criteria, both relative and absolute, to help determine the period of production for each mummy portrait. Trends in Roman Imperial hairstyles and jewellery are a fundamental dating tool because styles displayed in the portraits can be compared with well-dated Imperial marble and bronze busts and coins, thereby establishing an approximate production date.

It is twenty-five years since the National Gallery of Victoria’s Egyptian mummy portraits have been critically examined and the results published in English.⁴P. Connor, Roman Art in the National Gallery of Victoria, Melbourne, 1978, pp. 18–28.Since then the study of Egyptian portrait mummies has enjoyed a resurgence, re-igniting the debate relating to their dating criteria.

When Peter Connor published his study in 1978 he largely followed the dating system developed by Klaus Parlasca, a scholar at the forefront of mummy-portrait research. His system of chronology, starting from the reign of the Emperor. Tiberius (14–37 CE), used the hairstyles of the Roman Imperial family as well as stylistic differences in painting technique to date the portraits. By following Parlasca, Connor dated the portraits from mid first century CE to mid fourth century CE, encompassing almost the whole period of portrait-mummy production.⁵ ibid.

A recent study by classical archaeologist Barbara E. Borg has established a new set of dating criteria based upon minute variations in hairstyle, clothing and jewellery.⁶ B. E. Borg, Mumienporträts, Mainz am Rhein, 1996. She asserts that since none of the portraits display hairstyles known exclusively from the second half of the third century, portrait production must have ceased around this time.⁷ Borg, ‘Problems in the dating of the mummy portraits’, in The Mysterious Fayum Portraits, E. Doxiadis, London, 1995, pp. 229-33. In comparison, Parlasca has long held that production ceased towards the end of the fourth century, after the Edict of Theodosius (392 CE) forbade the practice of pagan rituals.⁸ K. Parlasca, ‘Mummy portraits: Old and new problems’, in Portraits and Masks: Burial Customs in Roman Egypt, ed. M. L. Bierbrier, London, 1997, p. 127. Borg’s chronology is gaining favour over Parlasca’s,⁹ S. Walker, ‘A note on the dating of mummy portraits’, in Ancient Faces, Mummy Portraits from Roman Egypt, (exh. cat.), ed. S. Walker, British Museum, London, 2000, p. 36. resulting in the contraction of portrait-mummy production by at least a century. Therefore, it would seem timely to re-examine the NGV mummy-portrait collection in light of the revised dating criteria.

The NGV is fortunate to possess seven Egyptian mummy portraits dating from the period of Roman occupation.¹⁰ Original inventory numbers: 1499.5, 1500.5, 1501.5, 1502.5, 1503.5, 1504.5, 1505.5. Changed in 1970 to D42/1970, D37/1970, D40/1970, D41/1970, D38/1970, D39/1970 and D43/1970 (in corresponding order). Reverted to the original inv. nos. in 2002. As is the case with the majority of mummy portraits in collections today, the NGV portraits were probably torn from their mummies immediately upon discovery. The Viennese art dealer Theodor Graf originally owned them, along with hundreds of others. ¹¹ Parlasca, Mumienporträts und verwandte Denkmäler, Wiesbaden, 1966, p. 29. He was particularly interested in portrait mummies, exhibiting and selling them to most major European and American museums. Although the provenance of Graf’s acquisitions cannot be proven, most of them are thought to have come from the cemeteries around er-Rubayat, in the Fayum, where his agents are known to have worked. After his death in 1903 Graf’s family gradually sold off the remainder of his collection. A number of these pieces were sold at auction to Dr Neudörfer, another Viennese collector. In 1935 he auctioned them, with twenty-one other portraits, but it wasn’t until 1940 that they were finally acquired by the NGV through the Felton Bequest.¹² ibid., pp. 254–5.

The portraits represent five males and two females, with the males further subdivided by age into the categories of an old man, two men in their prime, a youth and a boy. The most tantalising question of all is: who are these people? In a multicultural society such as Egypt during the Roman period, the answer is not so easy to determine. Without even assigning cultural identity, one factor is patently clear: the people whose faces peer back at us belonged to an elite class within Egypto-Roman society. Even the poorest example would have cost much more than the annual income of an Egyptian labourer.¹³ Montserrat, ‘Death and funerals in the Roman Fayum’, in Portraits and Masks: Burial Customs in Roman Egypt, ed. M. L. Bierbrier, London, 1997, pp. 36. Because over half the find-sites and personal names associated with discoveries of portrait mummies are found in or around the Fayum,¹⁴ Walker, in Ancient Faces (see p.8 for a map showing find-sites).it is reasonable to assume that the people of the portraits could trace their origins back to this area. They were probably produced for a group within Egyptian society known as the katokoi.¹⁵ R. S. Bagnall, ‘The people of the Roman Fayum’, in Portraits and Masks, pp. 7–15 These people were the descendants of the original Macedonian and Greek settlers who were given land in rural Egypt by the Macedonian rulers, the Ptolemies. This group, while retaining their Hellenic heritage, assimilated many Egyptian beliefs and customs during 300 years of Ptolemaic rule as they interacted with the population, even marrying native Egyptians. This explains the mixture of Greek, Egyptian and Roman features found within the genre.

The first image under discussion is the better-preserved Portrait of woman (fig. 1). Parlasca dates her to mid first century, during the Claudian era.¹⁶ Parlasca, ‘Ritratti di mummie’, vol. 1, in Repertorio d’arte dell’Egitto Greco-Romano, series B, vol. 1, ed. A. Adriani, Palermo, 1969, pp. 32–3. (Parlasca no. 23). There are a number of defining chronological indicators in this portrait. The woman’s hairstyle is the most indicative of all. It is representative of a style made popular in the mid-late Antonine period by the Empress Faustina Minor, wife of Marcus Aurelius (145–175 CE), and Lucilla, wife of Lucius Verus (161–169 CE). Her hair was parted in the middle and fell in stylised waves set into a structured, corrugated ripple. It was then pulled into a loose knot at the base of the neck. This portrait is fairly unusual within the genre as it shows the bun at the side of the neck.¹⁷ There are two other portraits, of which I am aware, that show the bun in this position, one (81.AP.29), in the J. Paul Getty Museum, Malibu; reprod. in D. L. Thompson, Mummy Portraits in the J. Paul Getty Museum, Malibu, 1982, cat. no. 9 (Parl. no. 643); and another (5750) in the Pushkin Museum, Moscow, in Parlasca, ‘Ritratti’ vol. 1, pl. 6, no. 2. (Parl. no. 21). Parlasca believes the appearance of the bun to one side indicates an asymmetric hairstyle; ¹⁸ Parlasca, ‘Ritratti’, vol. 1, p. 33. however, it is more likely that the bun actually sat at the nape, but the painter chose to give it more prominence than reality dictated.¹⁹ Borg also sees this supposed asymmetry as a manifestation of artistic licence (Mumienporträts, p. 53, note 231).

The woman appears to wear two necklaces, and although badly damaged,²⁰ Parlasca claims that in an old photograph the necklaces were clearly distinguishable, but after restoration they became barely recognisable (‘Ritratti’, vol. 1, p. 33). it is possible to identify them within the corpus of mummy-portrait jewellery. The upper one probably represents a style made up of oblong tablets of precious or semi-precious stones, strung together on a gold chain; the lower one was probably a simple twisted gold chain. Many mummy portraits depict this combination of necklaces as both styles were popular throughout the entire period of portrait-mummy production. Fortunately, the earrings worn by the woman give us an opportunity to support the time frame proposed by her hairstyle. They belong to a very popular three-post style which evolved over time. Originally they consisted of a horizontal rod from which two posts hung, with most terminating in a pearl. Later the number of posts was increased to three and then four. The woman in this portrait wears the type with three posts although, because of surface damage, not all of them are visible. The three-post style came into fashion during the Trajanic period (98–117 CE) and continued in popularity for the next 100 years,²¹ Borg, Mumienporträts, P. 171. confirming a post-Trajanic time frame.

Changes in costume styling and colour combinations help to date portrait mummies and evenpinpoint provenance. For example, a decorative border around the neckline would indicate a date some time after the start of the second century; or the appearance of a dalmatica (tunic with sleeves) would indicate a date after the middle of the third century, and possibly a provenance of Antinoopolis. The woman’s clothing in this portrait is typical of Egyptian costume during the early to middle period of Roman rule. It consists of a single layered chiton (belted robe) with decorative clavi running down each side; black and gold clavi were popular on women’s clothing in the Antonine period.²² ibid., p.161. Her hairstyle is the strongest indicator of date, being very easily identified as mid-late Antonine. This time frame is also supported by her costume and to a lesser extent by her jewellery. Therefore, it is possible to conclude that this woman’s portrait was painted some time around 160–190 CE.

The Portrait of old man with a grey beard very clearly falls within a well-represented subset of portraits from er-Rubayat (fig. 2). These portraits depict men with varying degrees of grey in their hair and beards, and sometimes showing signs of balding. Typically they wear a cream chiton, with thin brown or pink clavi. A seam line, represented by dots, is shown across the shoulders and a cloak is draped over the left shoulder. The lips are drawn very full, often with the lower lip larger than the top one. This is quite different from the women of this particular group who are shown with Cupid’s-bow mouths. The background is typically painted in shades of grey. This subset of portraits belongs to the Antonine period, with the women of the group displaying hairstyles similar to the woman described earlier

The hairstyle of this portrait is clearly Antonine, which is mainly discernible from the beard. The hair is curly, but without the volume of hair shown on busts of Marcus Aurelius or Commodus (180–192 CE), indicating a possible date during the reign of Antoninus Pius (138–161 CE). The beard is cut short and combed fairly straight, with the hair trimmed away from the cheeks. These are all indicators of Antoninus Pius’ style, whose beard was not as long as his successor’s, nor was it coaxed into luxuriant ringlets like those of Marcus Aurelius, Commodus and Lucius Verus. Parlasca places this portrait within the last thirty years of the third century;²³ Parlasca, ‘Ritratti’, vol. 2, series B, vol. 2, Rome, 1977, p. 88. (Parl. no. 483). however, this date cannot be supported because of the old man’s stylistic similarity with portraits that are clearly dated to the Antonine period. Borg believes that this portrait belongs to the late Antonine period because of the curly hair,²⁴ Borg, Mumienporträts, p. 84.though I believe the portrait is representative of the early Antonine period (140–160 CE) because of its resemblance to the busts of Antoninus Pius.

The Portrait of young man with cloak (fig. 3) is quite distinctive in several respects. First, it is the only mummy portrait in the NGV collection where the deceased is shown wearing a cloak. It is draped over his left shoulder and caught with a fibula (pin) on his right. Unfortunately, the portrait has suffered surface damage where the pin would have been, but there are some vestiges of red paint to indicate its existence. Cloaks shown draped in this manner are quite rare, with only twenty-one instances appearing in Parlasca’s corpus of mummy portraits.²⁵ ibid., p. 157. A feature of the cloak, which is even more interesting, is the fringe shown above the wearer’s left shoulder. This is extremely rare within the genre of mummy portraits. Some show the fringing across the top of the shoulders, as in Portrait of woman in the British Museum (fig. 4), ²⁶ ibid., pl. 42, no. 2. (Parl. no. 560). whereas others show the fringe edging the sides of the cloak, from shoulder to chest (Martin von Wagner Museum, inv. no. H 2196).²⁷ ibid., pl. 50, no. 1. (Parl. no. 497). Cloaks were worn across the chest and caught with a pin from the time of Hadrian and by Antonine times were shown with a fringe.²⁸ ibid., p. 157. The third interesting feature is the man’s chiton which has a thin, coloured border around the neckline. This fashion was introduced for both men and women in Hadrianic times and is the only panel in the collection that exhibits a decorative border. The hairstyle and beard of this man fall within the Antonine period: the hair is fairly full and curly, arranged around the top of the head, with the ears either partially or completely covered by hair. The beard is short and trimmed away from the cheeks and chin, with only a slight moustache. Parlasca dates this portrait to the end of the third century, ²⁹ Parlasca, ‘Ritratti’, vol. 3, series B, vol. 3,1980, p.35. (Parl. no. 546). with Connor estimating its production to be between 220 and 240 CE.³⁰ Connor, P. 24. However, the clothing and hairstyle in this panel place it solidly within the period of Antonine portrait production, with a tendency towards the early Antonine period. The style of beard is typical of those seen in portraits from the Hadrianic period (117–138 CE), yet the hair is more reminiscent of the style adopted by Antoninus Pius or even Marcus Aurelius. This would place its manufacture some time around 140–160 CE.

The Portrait of young man (fig. 5) is technically far better than the Portrait of young man with cloak (fig. 3), yet they are contemporaneous. The hair is shown in loose curls around the head and the beard is trimmed, with the cheeks and chin clean-shaven. The moustache does not extend out to meet the beard. Other than being a well-executed portrait, there is not much that is remarkable about it. It should be noted that the underscoring of the lower lip on the NGV portrait is intended purely to represent a shadow, not the tuft of hair under the lip so typical of images of Lucius Verus. This portrait indicates an early-Antonine production date also around 140–160 CE. Originally Parlasca dated it to the first thirty years of the third century.³¹ Parlasca, ‘Ritratti’, vol. 2, p. 80. (Parl. no. 448).

The Portrait of a youth exhibits slight, downy hair along the upper lip, with no evident beard (fig. 6).³² See Montserrat, ‘Representation of young mates’, pp. 215–25, for a discussion on the iconography of males in mummy portraits. Youths are notoriously difficult to date because they haven’t developed the full accompaniment of facial hair necessary to distinguish different periods. Therefore it isnecessary to establish stylistic links with portraits whose dates have been firmly established. The encaustic portrait in the British Museum (fig. 4), shows facial modelling very similar to the NGV youth. The nose and mouth especially are similar. Walker has dated this portrait to the late-Antonine-early-Severan period (180–200 CE) and Borg to the late-Antonine period.³³ Walker, Ancient Faces, 2000, p.82; Borg, Mumienporträts, p. 57. Another indicator of date is the black, painted border around the left edge. Bands of this type appear on other panels from er-Rubayat from late-Antonine to early-Severan times, although the majority of those portraits are painted in tempera. Parlasca also dated this portrait using comparison; however, he concluded that it was executed in the second half of the third century.³⁴ Parlasca, ‘Ritratti’, vol. 2, P. 91. (Parl. no. 495). The portrait of a youth is stylistically close enough to the British Museum example to be able to confidently date it to the late-Antonine-early-Severan period, with further re-enforcement being provided by the painted border.

Because of the poor condition of the Portrait of woman with pin (fig. 7), historically it has been extremely difficult to date with confidence; however, there are enough indicators to place it within a specific period. The defining feature is a long, gold hairpin – similar to a hatpin – projecting from the right-hand side of her head. This pin would have held a thick braid of hair that had been coiled high on the back of the head. Unfortunately, the coil of hair itself cannot be discerned, but the hairpin indicates its original existence. There are two small, tight curls against the woman’s cheek, just above her right ear. Portraits exhibiting this hairstyle invariably have tight little curls framing the face and sometimes have ringlets falling in front of the ears. Ringlets are definitely absent in this portrait, although I would expect that little curls (kiss curls) originally framed her face. This particular hairstyle is extremely well represented in portrait mummies and was very popular in the Trajanic period (98–117 CE), sometimes with the owner creating tiers of curls by using hairpieces, as did Marciana, the sister of Trajan.

The woman is wearing a mauve chiton with plain black clavi; a colour combination that was perennially popular with women in Roman Egypt, reaching the height of popularity in the Trajanic and Hadrianic periods. Her jewellery is hard to discern because of surface damage. Fortunately, some sections of her necklace are clear enough to be able to identify the links of a simple, rope-like gold chain. If she wore earrings, they are now completely undetectable. The clothing and hairstyle both indicate a date in the first two decades of the second century. It is a pity this portrait is so badly damaged because her features are modelled quite successfully, giving the viewer an impression of individuality that is missing from the other portraits.

The final portrait in the collection is perhaps the most challenging to date, the Portrait of boy (fig. 8). Parlasca dates it to the second half of the fourth century, making it, in his opinion, one of the last mummy portraits produced.³⁵ Parlasca, ‘Mummy portraits’, p. 129. He aligns this portrait with the Constantine period because of the frontal pose of the sitter and the appearance of folds of flesh forming rings around the neck. Both traits are regarded as precursors to early-Byzantine Christian iconography. Even though the hairstyle is typically Trajanic (250 years earlier), Parlasca supports his argument by pointing to the revival of Trajanic style with the advent of Constantine the Great. Connor supports Parlasca’s dating by pointing out the trend to be clean-shaven in the Constantine era,³⁶ Connor, P. 27. but it must be remembered that this is the portrait of a boy; a fact borne out by the diminutive size of the portrait when compared to the others in the collection.

Nevertheless, this portrait gives us the opportunity to test the revised portrait-mummy chronology because its traditional production date now falls outside the new time frame. So, how does it fit into the new chronology? The most compelling evidence against a Constantine date is found in its similarity to a set of tempera portraits, said to come from er-Rubayat, which is dated to the Trajanic period. They are all of women whose hairstyles are identical to the NGV Portrait of woman with pin (fig. 7), but what is surprising is their frontal pose and fleshy rings around the neck. One portrait in the Ashmolean Museum (fig. 9), exhibits such close parallels in the rendering of the facial features that, I believe, they can be assigned to the same workshop, if not to the same painter. The crescent pendant worn by this woman is well dated to the first and second centuries,³⁷ C. Johns, ‘Gold necklace with crescent pendant’, in Ancient Faces, p.150. further emphasising a Trajanic date. The boy’s own hairstyle is also typically Trajanic, being combed forward around the face, with the fringe reaching mid-forehead and cut short around the ears. In fact, every aspect of this portrait compares favourably with Trajanic chronological indicators.

Another portrait, which may have a bearing on the dating of this one, is a portrait in the British Museum (inv. no. EA 63395).³⁸ S. Walker & M. Bierbrier (eds), Ancient Faces, Mummy Portraits from Roman Egypt, (exh. cat.), British Museum, London, 1997, pp.87–8. (Parl. no. 629). Its substrate has been described as ‘coarse stuccoed linen’, which is an accurate description of the support of the boy’s portrait. Within the genre of portrait mummies, this substrate is unusual and was clearly meant to imitate the more conventional, expensive, wooden panels. Coincidentally, the British Museum portrait also comes from the former Graf Collection and belongs to the sub-set of portraits from the Trajanic period described above. Possibly this substrate is a feature of a particular place and time, namely er-Rubayat in the Trajanic period; at this time there is not enough evidence to argue the case convincingly. Stylistically and materially, this NGV portrait can be placed confidently within the Trajanic period (100–120 CE) and not over 200 years later as once thought.

After re-evaluating each of the NGV mummy portraits, it is clear that some have experienced a small dating shift, whereas others have received a major realignment within the chronology of the genre (see table 1). Without exception, every portrait has been assigned a different Roman Imperial period from that given by Parlasca. Only the better-preserved Portrait of woman (fig. 1) has changed from an earlier to a later date; a shift of a little over a hundred years. All the other portraits have had their dates put back from later to earlier dates. Debate over the termination of portrait-mummy production impinges most significantly upon the dating of the Portrait of boy (fig. 8). His portrait has received the greatest shift, with the production date being reversed by over two hundred years. The portrait least affected is the Woman with pin (fig. 7) whose date has only altered by less than fifty years.

The technique of dating mummy portraits is a continual process of refining the dating criteria through infinite comparisons and careful observations until an approximate date is arrived at. This study has aimed to correct the chronological misplacements of previous publications and will allow the NGV mummy-portrait collection to become integrated within the restructured corpus of mummy portraits from Roman Egypt.

Table 1. Dating History of NGV Mummy Portraits

(image pending)

Joylene Kremler, Information Officer, National Gallery of Victoria (in 2004).

Acknowledgements

Translations to the Italian texts were provided by Dr Carla Gallorini and Rita Vitri-Anastasio. I thank Dr Barbara E. Borg and Dr Colin A. Hope for providing valuable comments on the preliminary version of this article.

Notes

1      Cartonnage is made from layers of plaster and linen to create a thin shell.

2      Within the confines of set types, the portraits range from highly individualised examples to those that were clearly produced for a mass market (see L. Corcoran, Portrait Mummies from Roman Egypt (I–IV Centuries A.D.): With a Catalog of Portrait Mummies in Egyptian Museums, Chicago, 1995, pp. 68–74; D. Montserrat, ‘The representation of young males in ‘Fayum Portraits”, in Journal of Egyptian Archaeology, vol. 79, 1993, pp. 215–25 for a discussion on types and the portrait’s role in establishing the deceased’s social position). 

3      The term ‘portrait mummy’ refers to a complete example of the genre (consisting of both mummy and portrait), whereas the term ‘mummy portrait’ or ‘portrait’ only refers to the portrait itself.

4      P. Connor, Roman Art in the National Gallery of Victoria, Melbourne, 1978, pp. 18–28.

5      ibid.

6       B. E. Borg, Mumienporträts, Mainz am Rhein, 1996.

7       Borg, ‘Problems in the dating of the mummy portraits’, in The Mysterious Fayum Portraits, E. Doxiadis, London, 1995, pp. 229-33.

8      K. Parlasca, ‘Mummy portraits: Old and new problems’, in Portraits and Masks: Burial Customs in Roman Egypt, ed. M. L. Bierbrier, London, 1997, p. 127.

9     S. Walker, ‘A note on the dating of mummy portraits’, in Ancient Faces, Mummy Portraits from Roman Egypt, (exh. cat.), ed. S. Walker, British Museum, London, 2000, p. 36.

10     Original inventory numbers: 1499.5, 1500.5, 1501.5, 1502.5, 1503.5, 1504.5, 1505.5. Changed in 1970 to D42/1970. D37/1970, D40/1970, D41/1970, D38/1970, D39/1970 and D43/1970 (in corresponding order). Reverted to the original inv. nos. in 2002.

11     Parlasca, Mumienporträts und verwandte Denkmäler, Wiesbaden, 1966, p. 29.

12     ibid., pp. 254–5.

13     Montserrat, ‘Death and funerals in the Roman Fayum’, in Portraits and Masks: Burial Customs in Roman Egypt, ed. M. L. Bierbrier, London, 1997, pp. 36.

14      Walker, in Ancient Faces (see p.8 for a map showing find-sites).

15     R. S. Bagnall, ‘The people of the Roman Fayum’, in Portraits and Masks, pp. 7–15.

16      Parlasca, ‘Ritratti di mummie’, vol. 1, in Repertorio d’arte dell’Egitto Greco-Romano, series B, vol. 1, ed. A. Adriani, Palermo, 1969, pp. 32–3. (Parlasca no. 23).

17      There are two other portraits, of which I am aware, that show the bun in this position, one (81.AP.29), in the J. Paul Getty Museum, Malibu; reprod. in D. L. Thompson, Mummy Portraits in the J. Paul Getty Museum, Malibu, 1982, cat. no. 9 (Parl. no. 643); and another (5750) in the Pushkin Museum, Moscow, in Parlasca, ‘Ritratti’ vol. 1, pl. 6, no. 2. (Parl. no. 21).

18      Parlasca, ‘Ritratti’, vol. 1, p. 33.

19      Borg also sees this supposed asymmetry as a manifestation of artistic licence (Mumienporträts, p. 53, note 231).

20     Parlasca claims that in an old photograph the necklaces were clearly distinguishable, but after restoration they became barely recognisable (‘Ritratti’, vol. 1, p. 33).

21      Borg, Mumienporträts, P. 171.

22      ibid., p.161.

23     Parlasca, ‘Ritratti’, vol. 2, series B, vol. 2, Rome, 1977, p. 88. (Parl. no. 483).

24     Borg, Mumienporträts, p. 84.

25       ibid., p. 157.

26      ibid., pl. 42, no. 2. (Parl. no. 560).

27      ibid., pl. 50, no. 1. (Parl. no. 497).

28      ibid., p. 157.

29      Parlasca, ‘Ritratti’, vol. 3, series B, vol. 3,1980, p.35. (Parl. no. 546).

30      Connor, P. 24.

31      Parlasca, ‘Ritratti’, vol. 2, p. 80. (Parl. no. 448).

32      See Montserrat, ‘Representation of young mates’, pp. 215–25, for a discussion on the iconography of males in mummy portraits.

33     Walker, Ancient Faces, 2000, p.82; Borg, Mumienporträts, p. 57.

34     Parlasca, ‘Ritratti’, vol. 2, P. 91. (Parl. no. 495).

35     Parlasca, ‘Mummy portraits’, p. 129.

36     Connor, P. 27.

37     C. Johns, ‘Gold necklace with crescent pendant’, in Ancient Faces, p.150.

38     S. Walker & M. Bierbrier (eds), Ancient Faces, Mummy Portraits from Roman Egypt, (exh. cat.), British Museum, London, 1997, pp. 87–8. (Parl. no. 629).