Amanda Kasman using a swap to clean the surface of the panel painting, which rests on a black table in the conservation lab.

Uncovering the Restoration History of the "Medical School Portrait"

In 2016 the Gallery’s Conservation Department was asked to evaluate a mysterious panel painting on campus, referred to as the “Medical School Portrait” (fig. 1). Ian McClure, the Susan Morse Hilles Chief Conservator, spearheaded an investigation involving archival research along with ultraviolet florescence imaging, infrared reflectography, cross-section microscopy, scanning electron microscopy–electron dispersive spectrometry (SEM-EDS), and macro-X-ray fluorescence (MA-XRF). Guided by the findings of the extensive technical analysis, Gallery conservators recently cleaned the work of overpainting and old varnish, bringing it closer to the artist’s original vision.

Two views of a painting of a bearded man with his hand on a skull. He gazes at the viewer.

fig. 1: The Medical School Portrait, before treatment, at left; after revarnishing and filling, at right

The unsigned portrait depicts an unidentified man in a white ruff and dark robe, pointing at a skull. An inscription hovers in two sections, one on either side of the man’s head; it reads, “ÆTATIS SVÆ24 / ANNO 1580,” which roughly translates to “at the age of 24 / year 1580.” The portrait is painted on a panel made from two thin oak boards, joined, likely with animal glue, at a seam that runs down the center of the image. Directly atop the wood is a thin preparatory ground composed of chalk and glue, on which the unknown 16th-century artist probably executed an underdrawing. The artist then applied his oil paint and natural-resin varnish to depict the sitter.

What is known about the painting is that the acclaimed Yale neurosurgeon Harvey Cushing acquired it in 1925. Cushing paid Mr. M. L. Walker, a dealer and art appraiser based in Boston, $250 dollars for an “old portrait in oil of a physician.” Cushing had inquired about the painting’s provenance, but to his dismay, the history of the artwork prior to the 20th century was unknown. The painting was added to the Harvey Cushing Collection at Yale in 1939 and, shortly thereafter, was hung in a seminar room in the Yale School of Medicine. There it remained for decades largely unnoticed, until 2015 when Robert Crabtree, the Conkey P. Whitehead Professor of Chemistry, brought it to the attention of Melissa Grafe, the John R. Bumstead Librarian for Medical History, and Laurence Kanter, Chief Curator and the Lionel Goldfrank III Curator of European Art at the Gallery.

In 2016 the wood and dendrochronology expert Ian Tyers examined the growth rings of the wooden-panel support and confirmed that it dates to the mid-16th century, corroborating the inscription. Julianna Ly and I, both Graduate Fellows in Paintings Conservation, subsequently undertook archival research to assess Walker’s 1925 claim that the painting depicts a physician. Although the portrait may have hung in the School of Medicine under the pretext that, because the sitter is shown pointing to a skull, it must depict a doctor, Ly and I have since determined that the skull is better understood as a symbolic motif. Typical of memento mori paintings from the Flemish Golden Age (16th century), skulls served as a reminder of the transience of life and inevitability of death and were frequently represented in both still lifes and portraits to convey this moralizing meaning. Thus, the sitter may not have had any connection to the medical profession at all.

A conservator looks through a microscope while cleaning a portrait of a bearded man.

fig. 2: The author using a swab to clean the surface of the Medical School Portrait

One of the factors that hindered a stylistic analysis to trace the painting to a particular workshop was the severely darkened and thick coating of varnish on the surface. I first cleaned the painting with swabs soaked in a mild citrate solution adjusted to pH 7 (fig. 2), before using solvent solutions to reduce the varnish layers (fig. 3). Instantly, this revealed the vibrant colors, delicate details, and surface textures of the brushstrokes—all of which had been obscured for centuries. However, it also gave rise to further questions about the painting’s history of restoration.

The old paint layers on the portrait did not dissolve when the solvent mixtures were applied to remove the varnish—except for one, which raised suspicions. This soluble layer was semitransparent, pigmented gray, and appeared to cover the entire background. Conservators considered it likely that this was overpaint that had been added by a previous restorer, and thoughtful discussion ensued to determine whether it should be removed. What if the original paint underneath was badly damaged? Did the gray layer have historical value, even if it was not original? What if removing the gray paint also removed the inscription? To gather more evidence to determine the authenticity of the gray paint, assess the condition of the underlying background, and determine the risk to the inscription, Gallery conservators performed a barrage of scientific tests in collaboration with staff at the Technical Studies Lab of the Institute for the Preservation of Cultural Heritage (IPCH).

Detail of two views of a painting of a bearded man

fig. 3: Detail of the reduction of varnish on the Medical School Portrait using solvents in two stages, with the face fully cleaned on the right

Anikó Bezur, the Wallace S. Wilson Director of the Technical Studies Lab, and Marcie Wiggins, Postdoctoral Associate at IPCH, encapsulated in resin extremely small samples of paint from the background (along the edge of the panel) and the green table (in the lower-left corner). Cross-section microscopy and SEM-EDS analyses of these samples showed that, although the gray layer was discrete from the dark-brown one beneath it, the two were not separated by varnish (fig. 4). Had they been, this would have hinted that the brown paint layer was the final one applied by the artist himself, before varnishing his finished work. At the same time, the gray paint contained no synthetic pigments identifying it as modern. More testing was needed.

Microscopic cross-section view of a paint layer showing dots of color and a diagram of what each item is

fig. 4: A cross section is a small sample of paint, usually the size of a period at the end of a sentence or smaller, which is used to examine the microscopic layer structure of a painted surface. This cross section from the background, shown in natural light on the left and ultraviolet light on the right, makes visible the overpaint on top of the original paint, with no varnish in between. Several more campaigns of overpaint separated by varnishes are visible along the top.

MA-XRF is a noninvasive analytic technique that uses a scanning instrument to take elemental maps of the entire artwork. With this technique, Dr. Wiggins was able to determine the location of each pigment and whether it lay on the surface of the painting or deeper in the stratigraphy. To our amazement, the elemental maps demonstrated that the original background was almost entirely intact under the gray paint. Furthermore, the maps showed that the inscription, along with a rectangular space rather carelessly marked out around it, had not been painted over in gray, meaning that removal of the overpaint should not disrupt its thinly painted letters.

SEM-EDS and MA-XRF analyses revealed further surprises. For example, SEM-EDS confirmed that the painting contained copper-green pigment, which conservators had suspected, but it was found in an unexpected place: the sitter’s black robe. There is little reason to suspect that the garments once appeared green, but perhaps the copper-based pigment, applied in the form of a transparent glaze over the black paint or mixed in directly, helped to make the garment appear darker. Meanwhile, on the MA-XRF maps, strange forms not corresponding to known features of the composition were visible in the background. A closer look at the painting itself made clear that these forms were located beneath all of the other layers, including the background. Could this panel have been originally intended for another painting, which the artist started and then painted over? These buried features remain visible to the naked eye because of the loose application of brown paint to the background. Notably, this loose style was typical of Flemish paintings from the 16th century. However, in the centuries that followed, more solid backgrounds became the norm, and it was not uncommon for dealers to repaint backgrounds to make them fit the current style. This might explain the gray layer of overpaint.

A diagram of a painting of a bearded man shown in various colors

fig. 5: Overall MA-XRF maps for iron (Fe), copper (Cu), mercury (Hg), and lead (Pb). For the messy rectangle around the inscription, see the Fe map. For the copper-green pigment in the black garment and table, see the Cu map. For the strange forms in the background that do not correspond to the final composition, see the Hg and Pb maps. Images courtesy of Marcie Wiggins

Based on these analytical findings, close examination of the paint surface, and further cleaning tests, conservators determined that the gray layer was overpaint and that it should be removed. The original paint layer underneath it carried more historical value, having been applied by the artist’s hand. Only if the MA-XRF maps had shown the original paint to be severely abraded would the conservators have opted to retain the overpaint. After fully cleaning the work, I revarnished it with a modern synthetic product that is reversible and more resistant to yellowing than natural-resin varnishes. Lastly, I filled with gesso, inpainted with Gamblin Conservation Colors, and applied a final coat of spray varnish. One can now see the thoughtfully executed, painterly brushstrokes composing the sitter’s face and hands, which may be the key to attributing the painting, though research is ongoing. 

Amanda Kasman
Former Graduate Fellow in Paintings Conservation