Skip to Main Content

Notes to the Reader

The paintings cataloged here are arranged alphabetically by artist. For each artist, there is a short biography with bibliography, followed by individual entries on paintings. Each entry begins with the National Gallery of Art’s unique accession number, the title of the work, execution date, medium, dimensions, credit line, and a transcription of signatures and / or inscriptions. This heading is followed by a technical summary, provenance, exhibition history, scholarly entry, notes, and bibliography.

The following conventions are used for dates:

1250 Executed in 1250
c. 1250 Executed sometime around 1250
1250–1265 Begun in 1250, finished in 1265
1250/1265 Executed sometime between 1250 and 1265
c. 1250/1265 Executed sometime around the period 1250–1265

Dimensions are given in centimeters, height preceding width preceding, for certain dimensions, depth, followed by dimensions in inches within parentheses. For all paintings at least two dimensions are given, first that of only the painted surface, and then that of the overall support. Sometimes depth dimensions for paintings on panel are also given; these are at times approximate, when a cradle prevents a precise measurement. Parenthetical phrases provide precise descriptions of additional measurements taken. Framed dimensions are also included when frames are integral to the panel. The measurements are so exact in order to aid scholars in making and evaluating reconstructions or the ensembles to which a number of the panels once belonged.

If present, signatures and / or inscriptions on the paintings are recorded as accurately as possible. In these transcriptions, a slash with a space on either side indicates a new line; a slash without these spaces indicates that the slash is included in the original text. All directional references to signatures and inscriptions are given from the viewer’s perspective.

The Provenance section gives the names of all known owners. A semicolon between two names indicates a direct transfer of ownership of the painting, whereas a period indicates uncertainty about the chain of ownership and the whereabouts of the object between two documented owners. The names of dealers, agents, and auction house sales are given in parentheses. Notes provide sources, details of research, and discussion of outstanding questions.

The majority of the paintings cataloged here, three-fourths of them, were acquired by Samuel H. Kress himself or by the Samuel H. Kress Foundation in New York, which has kindly provided copies of its acquisition records to the Gallery. These records are held in Gallery Archives, with copies of relevant material in NGA curatorial files. Nearly a third of the paintings passed through the hands of Duveen Brothers, Inc., a company headed by Joseph Duveen with offices in London, New York, and Paris. The firm’s records, now held by the Getty Research Institute in Los Angeles and available to researchers both on microfilm and online, are an important source of information and are referenced frequently in notes to the Provenances. Copies of the referenced material are in NGA curatorial files. Seven paintings, including Duccio’s The Nativity with the Prophets Isaiah and Ezekiel, were given by Andrew Mellon. The A. W. Mellon Educational and Charitable Trust in New York has made copies of its records available to the Gallery, where they are held both in Gallery Archives and, as appropriate, in NGA curatorial files. Every effort has been made to provide the exact location of the records pertinent to each painting.

Exhibition histories through 2014 record the presence of the paintings in special exhibitions or as individual loans from an owner to another institution. They are as complete as possible and include exhibition catalog numbers, unless it is indicated that there was no catalog or the catalog was unnumbered, and a note if the painting was reproduced. 

The references for each entry lists with full citations and in chronological order, only those texts that specifically discuss the National Gallery of Art painting in question. An exhaustive Bibliography on the general subject of thirteenth- and fourteenth-century Italian paintings is provided at the end of this volume.

Unless otherwise indicated, all altarpiece reconstructions are by Miklós Boskovits. Comparative figures of panels that once belonged to the same altarpieces as Gallery panels use dates proposed by the author.

Each entry includes a Technical Summary, which discusses the materials and techniques used by the artists in the creation of the paintings, as well as any changes and documented treatments. They are based on the contents of the examination reports prepared by members of the National Gallery of Art department of painting conservation. 

Each painting was examined unframed, in visible light, front and back. The paintings were examined with a stereomicroscope with magnifying power up to 100x and under ultraviolet light. X-radiographs were taken to answer questions regarding the paintings’ construction and condition. Infrared photography and / or infrared reflectography was used for each painting to reveal underdrawing, compositional changes, and condition. The results of these examinations are discussed only when they yield information considered essential to interpretation of the painting.

The medium of the paint has not been analyzed unless stated in the Technical Summary. The medium is estimated or known to be egg tempera paint for all paintings in this catalog. For the panel supports, the type of wood is specified only if the wood has been analyzed. Any scientific analysis that was used to help understand the paintings is cited. The procedures and equipment used for this analysis are described below.

The conditions of the paintings vary. Many paintings cataloged here were originally created as components of altarpieces with engaged frames, most of which were removed when the altarpieces were dismantled in earlier centuries. Where present, the barbe, the lip of gesso remaining after an original engaged frame was removed, is noted as a measure of the painted design’s original dimensions. The separate panels were often later cradled. The cradling process usually included shaving down the sides and thinning the back of the original wood panel, marouflaging it to a backing, and attaching wood shims to the sides. Some of the panel paintings were transferred from their original supports, a process that retains the paint and ground layers but substitutes a fabric support for the original wood one.

Treatments performed by Gallery conservation staff after acquisition are described briefly in the Technical Summaries. Occasionally, records of earlier treatments are included in the Technical Summaries. Damages such as paint losses should be assumed to have been repaired and inpainted. Significant areas of inpainting are discussed in the Technical Summaries. The varnishes are all later replacements and impart no information about the artist’s choice of finish.

Instrumental Methods of Analysis

A variety of techniques and instruments were used to examine and analyze the paintings in this catalog. The equipment is described below: 

Energy dispersive spectrometry (EDS): Small samples were examined with energy dispersive spectrometry using an Oxford Inca 300 spectrometer with a Super ATW Si(Li) detector on a JEOL 6300 SEM.

Fourier-transform infrared (micro)spectroscopy (FTIR): A Thermo Nicolet Nexus 670 instrument was used, fitted with a Continuum microscope. Spectra were collected in transmission mode at 4 cm-1 resolution. The samples were compressed between two windows of a Diamond Cell (Spectratech).

Gas chromatography / mass spectrometry (GC / MS): The samples were methylated with TMTFTH (TCI America, 0.5M in MeOH) or hydrolyzed using 6N HCl for 24 hours under vacuum. After removal of the HCl, the amino acids were silylated with MTBSTFA / TBDMCS. Samples were examined by gas chromatography / mass spectrometry on a 30 meter DB-5 column, a Varian Saturn CP3900 gas chromatograph, and a Saturn 2100T ion trap mass spectrometer.

Infrared examination: When infrared examination is designated as “Vidicon,” a Vidicon camera system was used, which included a Hamamatsu c / 1000-03 Vidicon camera fitted with an N2606-10 or N214 lead sulphide tube and a Kodak Wratten 87A filter. When infrared examination is designated by microns, one of four cameras was used: a Kodak 310-21X PtSi camera configured to 1.5 – 2.0 microns, a Mitsubishi M600 PtSi camera configured to 1.2 – 2.5 microns, an Indigo / FLIR Alpha Visible-­InGaAs camera, and a Santa Barbara Focalplane SBF187 InSb camera. The latter two cameras were configured with various band filters between 1.1 and 2.5 microns. For the Kodak and Mitsubishi cameras, the images were captured using a ScionPCI framegrabber card in a Macintosh computer with Scanalytic’s IP-­Lab software. For the Indigo / FLIR camera, the images were captured onto a Dell computer with an IMAQ capture board housed in a Magma external PCI box, and IRVista software. For the Santa Barbara Focalplane camera, the images were captured using a Windows Empower tower computer and WinIR software. Nikon 55mm macro, 50mm macro, and 35mm lenses were used with the various cameras, as were Astronomy J, H, and K filters. The infrared reflectograms were automatically mosaicked and registered to a reference color image using a novel registration algorithm developed by George Washington University and the National Gallery of Art. For more information see Damon M. Conover, John K. Delaney, and Murray H. Loew, “Automatic Registration and Mosaicking of Conservation Images,” in Optics for Arts, Architecture, and Archaeology, vol. 4, ed. Luca Pezzati and Piotr Targowski, Proceedings of SPIE, vol. 8790 (Bellingham, WA, 2013). 

Optical microscopy of cross-sections: Small paint samples (c. 0.25mm2) were removed using a scalpel and were mounted in polyester-type resin blocks. The samples were cut at right angles to the layer structure and polished using silicon carbide papers and examined using a Leica DRMX microscope.

Polarized light microscopy (PLM): Transmitted polarized light microscopy of dispersed samples was conducted using Leitz Orthoplan and Leica DMRX microscopes. Particle identification was accomplished by comparing characteristic features — including particle size, color, refractive index and relief, birefringence, extinction characteristics, pleochroism, and anomalous polarization colors of the unknown — to those of reference materials in the Forbes Pigment Collection and other reference collections.

Scanning electron microscopy (SEM): Small samples were prepared for optical microscopy and examined with a JEOL 6300 scanning electron microscope at magnifications 100× – 10,000×. A Tetra backscatter electron detector was used to obtain BSE images.

X-radiography: X-radiography was carried out with equipment consisting of a Eureka Emerald 125 MT tube, a Continental 0-110 kV control panel, and a Duocon M collimator or a Comet Technologies XRP-75MXR-75HP tube. The image was captured on Kodak X-­OMAT film or digitally using a Carestream Industrex Blue Digital Imaging Plate 5537. The scanned x-ray radiograph films or digital x-radiograph captures were automatically mosaicked and registered to a reference color image using a novel registration algorithm developed by George Washington University and the National Gallery of Art. For more information see Damon M. Conover, John K. Delaney, and Murray H. Loew, “Automatic Registration and Mosaicking of Conservation Images,” in Optics for Arts, Architecture, and Archaeology, vol. 4, ed. Luca Pezzati and Piotr Targowski, Proceedings of SPIE, vol. 8790 (Bellingham, WA, 2013). 

X-ray diffraction (XRD): The Philips x-ray generator XRG 3100 was used with a tube with a copper anode and nickel filter. The paint sample was mounted on a glass fiber in a Gandolfi camera. Data were collected on film, and line spacings and intensities were estimated using a calibrated rule.

X-ray fluorescence spectrometry (XRF): This is a noninvasive analytical technique, which was carried out using one of two systems. The first system used a secondary emission Kevex 0750A spectrophotometer equipped with a rhodium tube with either a barium chloride secondary target or a molybdenum secondary target using a variety of excitation conditions and a silicon lithium Si(Li) detector with a resolution of approximately 155eV @Mn Kα. For this system the range of Rh tube excitation was 40kV – 60kV and 0.4mA – 2mA. The second system used a Bruker ArtTAX Pro μXRF spectrometer, which uses primary excitation and is equipped with a helium (He) flush, a rhodium (Rh) x-ray tube, and a capillary optic lens with an analysis area of approximately 75μm. In this system the x-ray tube voltage was 50kV, the current was 200μA, and the accumulation time was 200 seconds.