Today's guest blogger is Jacquelyn N. Coutré, the Allen Whitehill Clowes Curatorial Fellow, European Painting and Sculpture before 1800.
In the last posting on the Rembrandt self-portrait in the Clowes Collection (Fig. 1), we considered how art historians evaluated its status according to characteristics visible on the picture’s surface. But we can also gather scientific data to support this stylistic analysis.
In the early 1980s, IMA conservator David A. Miller examined the surface of the painting with a stereomicroscope and looked below its surface using X-rays (Fig. 3). The high magnification showed the “RHL” monogram to be contemporary with the painting, which means that it was applied while the painting was still wet. The x-radiograph, in turn, provided important insights into the artist’s creative process. It illustrates, in fact, two significant changes below the surface: the beret was originally poised more squarely on the head, and the contour of the proper left shoulder had previously extended further to the right. In other words, the artist had made changes to his painting while working on it, changes that would not have been visible to a student in his workshop or a later artist making a copy. The best of the other versions of this painting, the one in Atami, Japan, shows a strong correlation between the surface and underlying layers – telling evidence for the Atami version being a copy after the Clowes original! (It also omits those pesky pimples.)
But could the Clowes panel have been done by a later artist in order to look like a painting by the 17th-century master?
The investigations of Peter Klein, a wood biologist at the University of Hamburg, in 1999 help us to understand more about the panel upon which the painting was executed. It is made of oak and comes from the Baltic region, a profile typical of panels used by 17th-century Dutch artists. Relying on the facts that tree rings grow at different rates in different years and that trees of the same species in a particular region will show similar growth patterns, Dr. Klein has determined that the youngest growth ring in our panel dates to 1581. Add on a few years for the panel to dry and become less porous, and the painting could have been executed as early as 1598. While this may seem quite a few years before our estimated date of c. 1629, it confirms that the panel was ready to be used during Rembrandt’s lifetime.
Combining the stylistic and technical evidence yields the conclusion that our painting is indeed a self-portrait by Rembrandt. What was first supported only by connoisseurship is now augmented by scientific study – a wonderful demonstration of the important role that science plays in the museum.