As curator of the House of the Seven Gables, the 17th-century mansion made famous by Nathaniel Hawthorne's novel, I am especially curious to know, as accurately as possible, the intent of early designers and craftspeople, so our curatorial staff can re-create these finishes for today's public. Just like many of OHJ's owner-restorers, I pace the rooms and also mutter, "If only the walls could talk!"
Happily, with patience, trained conservators, and proper equipment, we've learned the ways that walls can talk, and they are very eloquent, indeed. To explain what I mean, I'll describe a little about why we undertook a paint analysis project at the House of the Seven Gables, and what we learned about the specific paints used in the early 18th century. More important for folks living in historic houses of later eras, our experience will illustrate some of the logic and techniques that can allow homeowners to become their own forensic scientists and track down elusive clues for re-creating the original decorative look of their interiors.
A Deeper Look at an Early Mansion
Over the last few years our curatorial staff has been moving toward a more accurate historical interpretation of the House of the Seven Gables, one based on primary-source documentation and analysis of the actual materials in the building. The house, built in 1668, has been open for tours for 93 years, but the interpretation of the paint through the decades has been more Colonial Revival style than truly colonial. For example, all the rooms were painted in an off-white color that, while conforming to an early 20th-century palette, was neither readily available nor popular for 18th-century homeowners. Since the Georgian paneling added around 1710 was the most drastic change to the building, my goal is to restore some of the rooms to that period based on primary documentation and physical evidence. An inventory of our house from 1742 refers to our parlor as the "Best Room," so this seemed an excellent place to begin the process.
Because early preservationists made many changes to the house during the Colonial Revival restoration of 1910, we approached the room with a lot of questions. First, we was curious about the number of beams on the ceiling 350 years ago,one pair or two? In addition, we wanted to know if the small wood closet to the left of the hearth dated to the 18th century or the 1910 makeover. Above all, we wanted to understand the original Georgian paint scheme so that we could restore it and present a more historically accurate interpretation of the most formal space in the house.
As the investigation began, we gathered early photographs of the room and examined the existing space to see if there was any telling evidence of paint scraping or replaced paneling. One circa-1860 photo from SPNEA shows coffering in the parlor before Caroline Emmerton purchased the house in 1908. Another documents the coffering being there after the purchase. The cross-timbering was removed during the 1910 restoration, raising the question of whether it too should be restored. Was the cross-timbering some original finish from the mid-1600s? Paint analysis would help answer this question.
Examining the room closely, we determined that the decorative display closet bore very early paint that was potentially fugitive in nature. By fugitive, conservators simply mean that the paint is made with highly unstable pigments (common before the industrial era) that can interact with ultraviolet light and change drastically over time. Many fugitive pigments, such as Prussian blue or verdigris (a green pigment made from corroded copper), will eventually blacken in linseed oil. Since our closet was painted in a green hue that was almost black, with some areas articulated in brown, there was a good chance that we were looking at a fugitive pigment.
While other architectural elements in the room were softened by layers and layers of paint, the shell closet near the hearth had crisper features that suggested very little paint build-up. Upon close inspection with a field microscope, we discovered a worn band of gilding at the base of one of the shelves and, over some of the woodwork, details picked out in brown paint. Apparently, this brown "paint" was actually the sizing (bonding agent) for gold leaf that had long since worn away. We took samples of the brown articulation for microscopy analysis and, at 125x, saw that this was indeed gold. Since there was no other treatment beneath it, we assumed we were looking at an original finish. Could it be that the rest of the room once had details picked out in gold? In the excitement of discovery, I imagined the great coffered room, painted with the most expensive paints of the day and dancing with gilded highlights. We hurriedly started the physical analysis.
Scientific Decorative Sleuthing
To establish the chronology of changes in the room, and to determine if the pre-1900 coffering needed to be restored, we began our paint analysis with the casings of the summer beams,the main, central beams in the house. First we determined the spots where beams in the mid-1800s photograph would have intersected. Then we postulated that if we found original verdigris finish missing from these points, we would indeed be looking at a decorative treatment from a first-period house,that is, one from the mid- to late 17th century. No one would have painted under the crossbeams before they were added to create the boxes of the coffering.
The actual investigation to test our hypothesis was very simple, requiring only a scalpel and 20x loupe for magnified viewing. First we carefully excavated several craters or "bullseyes" in the paint at strategic places along the beams, opening a physical window into the paint scheme of the past, called the chromachronology. When we inspected the bullseyes we could see that the green verdigris was indeed visible as the first campaign of paint. To make sure the results of our first bullseyes were not aberrations, we inspected three other places and found the same glaze evidence. The conclusion is that the intersecting beams and coffering must have been a Gothic Revival addition installed by Hawthorne's cousin, Susannah Ingersoll, sometime before the Civil War. So much for my dreams of a singular example of coffered ceilings in an early colonial home.
Another Revealing Test
Next on our list of questions to be probed was the history of the small door covering the wood closet next to the firebox. This closet is a wonderful window into the remains of the monumental 1680s hearth box that still retains a blackened band at the base of the plaster. Since there is a strangely configured "hidden stair" in the house that was designed and built in 1910, I worried that the closet was a similar 1910 concoction. By excavating some quick bullseyes on the front and edge of the door, we learned the answer. We found the verdigris finish on the front, finished face of the door, pretty much as expected.
However, when the bullseyes on the edges also revealed the decorative verdigris treatment, they ruled out the possibility that Emmerton had cut the door out of the existing Georgian paneling. If she had cut the door out of the existing paneling, the edges would have shown only 1910 and later paint colors. This meant that the little closet was truly finished in verdigris during the Georgian period when the huge, post-medieval hearths typical of first-period houses were commonly boxed in to create a sleek paneled fireplace. The additional room left over from the old hearth served as a perfect storage area.
After solving this mystery, our challenge was to search for any gilding on the paneling. We proceeded by very carefully removing layers of paint in channels, using scalpels and magnifying loupes, so as to expose a section across the entire architectural element. This is a delicate and time-consuming process, since a heavy or unsteady hand can destroy the very evidence one seeks. Since our work revealed only verdigris, it proved that gilding had been confined to the cupboard.
Next we opened up a large window of the painted paneling to rule out the presence of any decorative painting, such as marbleizing, and to see the texture of the paint. Had it been stippled? Was it highly striated? We could clearly see thick brush strokes in the finish. Then, when we chose a small section to scrape, in the hopes of clearing away the top layer that had oxidized to black, we found a brilliant grass green was still there.
Re-creating the Verdigris Finish
To confirm our research, our conservator, Christine Thomson from Robert Mussey Associates, took samples of all layers of these first-campaign paints,a grey base coat topped with verdigris glaze,and sent them to a lab. When the reports returned from the lab at the Museum of Fine Arts, Boston, chemical analysis (including gas chromatography and mass spectrometry) showed that our samples were indeed an oil paint ground covered with a glaze of verdigris boiled in oil. The latter was important new information because boiling the verdigris in oil produces a yellower, mellower green than simply mixing the pigment in oil. The absence of any hiding pigments, such as white lead, or bulking agents, such as chalk, meant that we were looking at a finish with a high gloss,much glossier than the common paint of the day.
Christine then contacted Susan Buck, a conservator and paint analyst with extensive knowledge of this early decorative treatment. Susan has worked on several historic house museums, such as Monticello and Mount Vernon, with rooms similarly finished in verdigris paint, and her experience with replication became our guide. Because the historic recipes for verdigris finish vary widely, we had to base the color match solely on our extant samples of unoxidized paint, but Susan's work validated that we were on the right track. We also compared our findings with the verdigris found at the Hunter House at the Preservation Society of Newport County.
After applying a mixture of the traditional materials (verdigris pigment and oil) on a sample board as a standard for what we wanted to re-create, we worked with our decorative painter, Nigel Grace, to mimic the historic effect with modern materials. Because expense and difficulty made using true verdigris out of the question, Nigel matched our sample by first thickening the grey base coat with calcium carbonide so it would simulate the deep, brush-bristle striations we found in the original paint.
Next he formulated the glaze using conservation-grade powdered pigments mixed in oil-based glazing, a liquid used for decorative painting. Nigel applied all coats with the grain of the wood, which meant that he even had to miter the corners as he worked. This was typical practice historically for glaze, and it articulates the architectural elements clearly. After the glaze had dried, Nigel topped the glaze with two coats of oil varnish to approximate the extraordinary high gloss of the finish as well as to protect it. The results are striking and have been greeted with great enthusiasm by the public. In fact, I like it so much, I am thinking about having verdigris glaze in my own home.
Alexandria M. Mason is the curator at the House of the Seven Gables in Salem, Massachusetts.