Back to

CSI: Conservation Science Indianapolis — Part Four

Our guest blogger today is Jared van der Linden, a clinical laboratory scientist with the Indiana Blood Center. His favorite painting is Vermeer's “The Little Street” in the Rijksmuseum, Amsterdam.

Part Four in a series of blog posts on the ongoing technical examination of a purported 1874 Alfred Sisley Impressionist painting.

Our preceding blog posts have introduced the IMA painting House in a Village, which bears the signature “Sisley 74” in the bottom-right corner and a fuller written name “alf. SISLEY” on the painting’s verso.  This painting lacks a clear history of ownership, and art historians and curators have expressed doubts that it is a genuine work by the Impressionist painter Alfred Sisley. In the previous posts (Part One, Two, Three), we presented the results of technical analyses using imaging techniques and x-ray fluorescence spectroscopy, both of which are nondestructive approaches.  None of these efforts have produced a “smoking gun” that clearly demonstrates that Sisley could not have been the work’s creator.  However, numerous regions of the painting showed curious anomalies that warrant further examination.

In the analysis of paintings not all questions can be answered through noninvasive approaches.  A painting is more than just the two dimensional image observed from a viewing distance of a few yards.  Paintings are three dimensional objects built up of layers including canvas or wood supports, sizing, preparatory grounds, paint layers, glazes, and varnishes.  Analysis of these strata, their composition and mode of application, can reveal much about a painting’s history and construction.  The difficulty is that to expose the underlying strata, a small sample must be taken.

Figure 1. Photomicrograph of the cross section sample location EX3 (at the end of the triangular marker) showing before (left) and after (right) removal of the small paint chip. For reference, the white marker in the lower right corner is 2 mm long.

A conservation scientist does well to follow the physicians’ motto of primum non nocere: first, do no harm.  In practice, the deleterious effect of sample removal is minimized by using microsampling techniques and sectioning from the edge of a pre-existing site of damage (Figure 1.).  This old painting has a pattern of stress-induced fissures over its entire surface.  Combined with small nicks introduced through accidents or normal wear-and-tear, these existing damages give a multitude of potential sampling sites.

Figure 2. Dr. Smith at the stereomicroscope collecting a small cross section sample from “House in a Village.”

Obtaining a cross section is a delicate operation. Using a stereomicroscope, surgeon’s scalpel, and chemically-etched tungsten needle, a sample of approximately a hair’s width can be removed, leaving a lacuna that is invisible to the naked eye (Figure 2).  Although the sample is small, it is purposefully cut so as to contain all the layers of the painting.  Cross section sampling has the advantage of giving us very detailed information about the painting, but the disadvantage of providing that information only for a discrete area of the artwork. Since we can’t take a sample from every square millimeter of the painting without destroying it, it becomes imperative to choose a single representative site for each distinct region of interest.

We chose to examine three areas (marked EX1 to EX3 in the painting’s image in Part Three), one from the black region of the upper-story window and two from different regions of the blue sky.  These sample sites create a vertical transect across the possible underlying composition that was hinted at in the infrared transmittogram image of the painting (see Part Two).  It is hoped that the samples will show the various paint layers used in the uppermost image and the underlying one, thus confirming the presence of an overpainted landscape scene beneath House in a Village.

To prepare our samples for microscopic examination, we must first mount them.  The samples are placed on their sides in a mold to which a clear liquid epoxy is added.  After curing for six hours, the hard resin-embedded samples can be cut and polished to expose the painting’s stratigraphy.    A series of progressively finer grits is used when grinding away the epoxy, terminating in a smooth nap cloth for producing a mirror-like polished surface.  Working with samples that are only a fraction of a millimeter thick requires caution, as one too many strokes on the sandpaper could completely erode the cross section.  When the preparation of the samples was complete, we observed and photographed them using common light microscopy techniques.

Figure 3. Cross section samples for EX1 (far left), EX2 (middle), and EX3 (far right) observed with darkfield illumination. Scale bars are all 100 micrometers.

Figure 3 shows the three samples viewed under a compound microscope.  Sample EX3, excised from the sample site in the sky area shown in Figure 1, is a three-layered cross section.  The upper blue sky layer is over the top of a green layer, which lies over the white ground layer.  This layering clearly suggests that an earlier composition containing green paint underlies the image of House in a Village.  The other two cross section samples each showed only two layers, suggesting that the green passage of paint did not run across the entire vertical transect of the underlying composition.  In those instances, only a single white layer of paint exists beneath the surface image. Moving forward, we can use these polished samples for a multitude of other investigative techniques to learn more about the pigments and the binding media.

Filed under: Art, Conservation, The Collection

Comments are closed.