Back to overview

Roman panel painting rendering a representation of a man, a winged griffin, and wheel of Nemesis

Ny Carlsberg Glyptotek


This is a Roman panel painting depicting a half-length representation of a man alongside a winged griffin with the wheel of Nemesis. The right part of the painting consists three well-preserved boards of the same width (wider variation.) It is suspected that the original panel may have included two more boards to the left, indicated by the preservation of the dowels alongside the left side and the incomplete transition. On the right and upper and lower sides of the painting is a narrow unpainted band, which corresponds to the original wooden frame, which is now lost.

Description of object

The panel depicts a half-length frontal representation of a man on a creamy-white background.

He has short, curly, black hair and beard, and a small black moustache. His dark-brown eyes, although only one is fully preserved, has a frontal gaze. The details of his face are finely rendered with shading, including eyebrows, nose, red cheeks, lips as well as two horizontal wrinkles across his upper forehead.

He wears a white tunic and a white mantle, draped over his left shoulder in the typical manner of male dress of the period. The mantle is decorated with a purple gammadion (a decorative figure composed of the Greek capital letter gamma), visible on his lower left arm. The left part of the gammadion is further ornamented with two white parallel, vertical, narrow lines.

He wears gold finger rings with red gemstones on his left little and ring fingers, and his fingernails, only visible on his right hand, are finely rendered in a light neutral colour. In his left hand he holds a pink flower wreath, a so-called “Totenkranz” is depicted. His right hand holds a brown pinecone clenched between his thumb and index finger.

In the upper right corner are the faint remains of an unreadable inscription in black. Below the inscription and to the right of the man’s head is a winged griffin, facing towards the left. The griffin rests its right front paw on the wheel of Nemesis. The body of the griffin is light brown, and the wings black, while details such as the eye, beak, ears?, and tail as well as its outline is in a darker brown paint. The wheel is yellow with six spokes.

Finally, at the very bottom of the panel, below the funerary wreath, is a small, undefinable motif, rendered in yellow paint. There is no doubt that it is intentional, but for now, it is impossible to interpret its original appearance.

The portrait renders a deceased man. It has stylistic similarities with the contemporary mummy portraits, and the wreath as well as the pinecone are well-known funerary attributes, also used in the mummy portraits. Yet, this panel was due to its large size and shape not used to adorn a mummy, but most probably functioned as a funerary portrait (“Totengedenkbild” or “Totenporträt”), commemorating the deceased in the picture (Parlasca 1966; Sörries 2003, 43). The main difference between these two portrait categories lies in their function and context, rather than motif or style: although both funerary, the commemorative portraits were never intended to adorn a mummy and were probably made posthumously. While the mummy portraits evidently were used in the tomb, the commemorative portraits served exclusively to commemorate the deceased in the domestic sphere, probably in a shrine in the tablinum, where the portrait could be displayed (Sörries 2003).

Similar funerary commemorative portraits, rendering half-length portraits of men, wearing white garments and holding funerary wreaths are known.9 The man’s white mantle is decorated with a gammadion (Cumbo 2019, type 90), which is so far unique in Egyptian panel painting as well as in mummy portraits. Mantles with gammadiae are often worn by members of the early Christian and Jewish communities, depicted e.g. in the wall-paintings of the early Christian catacombs in Rome as well in the synagogue in Dura Europos (Cumbo 2019).

Choice of methods

Visual examination

  • Macroscopic

Technical imaging

  • UV
  • VIL
  • IR
  • Raking light


  • Cross section
  • Microscopy
  • XRF
  • FT-IR
  • UV-VIS
  • Raman
  • GC-MS
  • Radiocarbon (14C) dating, wood analysis, shotgun liquid chromatography tandem mass spectronomy.

Technical imaging

Multispectral imaging was performed with a modified Canon EOS 5D Mark IV camera body and a Canon EF 50mm f/2.5 Compact Macro lens. The following filters were used for image acquisition: XNite CC1 from for visible light photography (VIS), XNite CC1, PECA 916, and Tiffen Haze 2E for ultraviolet induced visible fluorescence imaging (UVF), as well as a Schott RG830 for infrared photography (IRR) and visible light induced infrared luminescence imaging (VIL). A Midwest Optical Systems BP324 filter in combination with XNite CC1 yielded ultraviolet reflectance (UVR) images. The subtraction of an image at 635 nm shot with a Midwest Optical Systems BP635 filter from one at 735 nm (Midwest Optical Systems BP735) uncovered information about the presence of indigo as described by Webb et al 2014 and Bradley et al. 2020. For this imaging mode the abbreviation MBR (multi band reflectance image subtraction) is used. Light sources employed were incandescent tungsten lamps in combination with soft box diffusers, Excled LED RGB lamps (470 nm, 525 nm, and 629 nm), and Hoenle UVA SPOT 400/T lamps filtered with a Schott UG2A glass. The X-Rite Color Passport 2, Target-UV™ from UV Innovations, and a 99% Spectralon® diffuse reflectance standard were included in all images.

Other types of investigation

Fourier transform infrared (FTIR) spectroscopy

FTIR spectra were recorded with two different instruments. At the Center for Advanced Bioimaging (Copenhagen University), the samples were flattened onto a diamond window and analyzed in transmission mode between 600 and 4000 cm−1 using a Nicolet Continuum infrared microscope (Thermo Fisher Scientific Inc., Waltham, Massachusetts, U.S.) equipped with a liquid nitrogen cooled mercury cadmium telluride detector. The collected spectra were compared with reference libraries for compound identification.

Raman spectroscopy

For Raman measurements two spectrometers were also involved. In Copenhagen, at the Center for advanced Bioimaging (University of Copenhagen), the spectra were acquired on scrapings with an alpha300R Confocal Raman Microscope (WITec Oxford instruments, Ulm, Germany) equipped with two lasers emitting at 532 nm and 785 nm, respectively. The excitation laser power was on the order of 1 to 5 mW. A Zeiss EC Epiplan-Neofluar 50x lens and 300 and 600 g/mm gratings were used. The data was processed with WITec Project Five software, version, and compared with reference spectra. Raman spectroscopy on the cross sections was carried out in Boston. The different layers in several sections were examined with a DXR3 Raman microscope (Thermo Fisher Scientific Inc., Waltham, Massachusetts, U.S.), operated by Thermo OMNIC for Dispersive Raman software version 9.12.1002. All analyses from Boston were carried out with a 785 nm laser, 25 μm pinhole aperture, and 50x objective giving a spot size of approximately 1.0 μm. The standard grating that was utilized acquires data in the range 3374-32 cm-1, with resolution between 2.3 and 4.3 cm-1. Number of scans, scan time, and laser power were adjusted for each acquisition.

X-Ray Fluorescence (XRF) spectroscopy

XRF spectra were acquired with a handheld Tracer 5i XRF spectrometer equipped with a Rhodium tube (Bruker, Billerica, MA, USA). Two measurements at 15kV, 15 µA, with no filter and 40 kV, 7 µA, with a Ti/Al filter were taken for each location optimizing the detection of low and high Z elements, respectively. The data was processed with the Bruker Artax software Spectra, Version

X-Ray Diffraction (XRD)

XRD data were collected using a PanAlytical Empyrean diffractometer equipped with focusing mirrors for Cu Kα radiation (λ = 1.541 Å) and a capillary spinner 8 (Malven Panalytical, United Kingdom). A Ni beta filter and a pair of 0.04 rad soller slits were used. The sample was ground in an agate mortar, placed in a 0.3 mm diameter quartz capillary, mounted on a rotating stage and measured in transmission geometry. Evaluation of the spectra and phase identification was carried out with the software X’Pert Highscore Plus v.4.8 (Malven Panalytical, United Kingdom) using the ICDD PDF 4+ database. Rietveld refinement was performed using the software TOPAS v.6 (Brucker AXS, Karlsruhe, Germany) with reference structures for Gypsum (ICSD 000210816), Anhydrite (ICSD 000371496), Bassanite (ICSD 010742787), Calcite (ICSD 000050586), Alunite (ICSD 010759141). Atomic positions and stoichiometry were fixed, while lattice parameters, average crystallite size and scale factors were refined.

Cross sections and microscopy

Samples for cross sectional analysis were mounted in Technovit 2000 LC resin (Kulzer Technik, Wehrheim, Germany) cured with blue light (Technotray Power, Kulzer Technik, Wehrheim, Germany). A two-step process was used for including a label resulting in 8 mm diameter cylinders. The final polishing was executed on an EcoMet 30 manual twin from Buehler (Lake Bluff, IL, USA) with 6 μm and 1 μm diamond suspension. Images of the polished cross sections were recorded with a Leica research microscope (Leica Microsystems GmbH, Wetzlar, Germany) in bright field mode with crossed polarizing filters and under UV illumination detecting the induced visible fluorescence.

Scanning electron microscopy (SEM) coupled to energy dispersive x-ray spectroscopy (EDS)

Cross sections of the paint samples were examined without carbon coating in a JEOL JSM-IT500 low-vacuum SEM (JEOL, Tokyo, Japan). The instrument was operated at 20 kV, with a beam current of ca.1 nanoampere. The chamber pressure was set at 50 Pa. EDS analyses were carried out with an Oxford Instruments X-MaxN spectrometer with 80 mm2 detector area, operated by Oxford ‘Aztec’ software, version 4.2 (Oxford Instruments, Abingdon, United Kingdom). The measurements included back-scattered electron images, individual point analyses, EDS maps, and line scans.

Shotgun liquid chromatography tandem mass spectrometry

The sample was processed, along with a protocol blank, following the protocol described in Mackie et al. (2018)1. Briefly, protein residues were extracted from the sample using a lysis buffer, and enzymatic digestion was performed with LysC and Trypsin. The resulting peptides were immobilised on C18 Stage-Tips and analysed by nano-liquid chromatography coupled with tandem mass spectrometry (nanoLC-MS/MS). The MS/MS spectra were identified with the MaxQuant software, matching them against a reference database containing all the publicly available sequences of proteins contained in the most common proteinaceous artistic materials: collagens, egg proteins, and milk proteins. In order to investigate the presence of protein residues originating from other sources, the spectra were then matched against a larger database (SwissProt, from UniProt), containing all publicly available and manually reviewed protein sequences. The matches were against fully tryptic peptide sequences, with no taxonomic restriction.

Proteins are considered confidently identified if at least two unique non-overlapping peptides are observed, unless otherwise specified. Peptides were considered species-diagnostic when, after BLAST search against the entire nrNCBI protein database, they were assigned to a single species, or to a limited number of species among which only one can be considered plausible, based on: (i) the nature of the samples, (ii) the geographic origin of the sample, and (iii) the dating of the sample. Peptides assigned to recurrent contaminant proteins were filtered out and not considered further. Contaminants include primate keratins (likely from the laboratory space or through human handling of samples), excess trypsin, and Bovine Serum Albumin (a common laboratory reagent)

Wood analysis

Because of the three-dimensional nature of wood anatomy, each tiny wood sample, irrespective of its size, was fractured manually to show transverse, radial longitudinal, and tangential longitudinal sections (TS, RLS and TLS). Each TS, RLS and TLS wood section was then mounted, uncoated, onto aluminum stubs. Examination of the wood samples and comparative reference specimens prepared and mounted using the same method was undertaken in a variable pressure scanning electron microscope (VP SEM), Hitachi S-3700N, using the backscatter electron (BSE) detector at 15 kV, with the SEM chamber partially evacuated (40 Pa). Magnifications ranged from x65 to x650. The preferred working distance was circa 14 mm but was raised or lowered from 11.1 mm to 15.7 mm as required. With the BSE detector, 3D mode (rather than Compositional) was preferentially selected to maximize the opportunity to reveal diagnostic features for identification. These details, along with the scale bar in microns (μm) can be seen on the data-bar on each of the three SEM images in the Results section. Further details on wood identification methods and techniques can be found in Cartwright 2015 and Cartwright 2020.

Radiocarbon (14C) dating

14C measurements of the panels and the organic binder were conducted at the Laboratory of Ion Beam Physics (LIP) at ETHZ (Switzerland). The wood samples underwent first a Soxhlet extraction (Bruhn et al. 2001) for removal of conservation treatments, which consisted of consisted of a round of 3 solvents starting with hexane (60°C, 1h), acetone (55°C, 1h) and ethanol (65°C, 1h). After which, the cellulose content of the wood was extracted by applying a modified base-acid-base-acid-bleaching method for cellulose extraction (Němec et al. 2010; Brehm et al. 2021) as follows: 1M NaOH (60°C, overnight), 1M HCl (65°C, 30’), 1M NaOH (65°C, 1h), 1M HCl (65°C, 30’), before being bleached with NaClO2 5% +0.1 ml 0.5M HCl (70 °C, 45’) and freeze-drying overnight. The selected paint material was chosen following an already established workflow (Hendriks et al. 2018), which highlights the need to characterize all paint components and ensure the presence of inorganic pigments only. The selected paint samples were cleaned by an acid wash (0.5M HCl, 60°C, 30’).


Schmidt 1908, 640, no. E815; Sörries 2003, 121, no. 22; Parlasca 2004, 323, no. 22. Parlasca 1966, 67, pl. 22.3.

  • ÆIN 686-687
  • Fragment
  • 129 CE to 240/245 CE
  • Roman Imperial
  • Wood
  • Acquired in Egypt by the Danish orientalist Valdemar Schmidt (1836 - 1925) in 1892
  • Height: 65.7 cm. Width: 13.8 / 9.7 cm. Thickness: 2.4 cm.