In order to determine how the test samples compared in color, texture and thickness to the Madison-era washes, seven areas of surviving red wash were investigated. After reviewing the existing washes, the historic colors of the surviving washes were found to be fairly intact. The intact colors allowed for a comparison to be made to the samples on Dr. Buck’s recent test panels (all test samples were found to be more red then the original washes). However, same areas of surviving washes do not retain their period textures and were also found to be so degraded that it is impossible to reliably determine their period thickness. Therefore, the best way reliably recreate the texture and thickness of Montpelier’s historic red wash may be to attempt to replicate the composition and viscosity of the period washes and to apply it using period craft techniques.
To determine the original red wash color, Dr. Buck analyzed three red wash samples SB-EX-2, SB-EX-22 and RC-4) with a chroma meter to develop a color match (link to report).1 Specifically, she analyzed the samples using the Munsell color system (which specifies hue, value and chroma) and more accurate CIE Lab color system. She then selected two commercially available Benjamin Moore paint chips that most closely matched the three different samples based on the CIE lab color system analysis. As a result both Benjamin Moore #HC-49 and Benjamin Moore #2173-30 were specified. HC-49 was found to be a closer match and 2173-30’s hue runs much deeper into the red end of the spectrum then either the original sample or HC-49. (For a more detailed description of how Dr. Buck matches colors please click here).
To check Dr. Buck’s the initial color matches, seven samples of surviving red wash found in situ on the Mansion were matched to chips from the Munsell color books.While these chips do not have as wide a range as Dr. Buck’s Minolta Chroma Meter, they were still able to accurately define the color range seen in the original washes. Interestingly, very little variation was seen across the house’s four elevations. As a whole all of the samples tended towards a reddish-orange color and all were found between the 2.5 YR and 5 YR hue range. All surviving samples also tended to have the same values (4 or 3) and close to the same chromas (6 to 8). The most common Munsell notation was 2.5 YR 4/6 (three samples). The second most common notation found was in the 5 YR 3/8-4/8-5/6 range. No pattern was identified for the samples found on the circa 1765-1797 main block and those found on the circa 1809 wings, suggesting that the range in colors is not associated with different construction periods. Instead, the color range is probably associated with the inability of the period painter to mix the same proportion of pigments in each batch of red wash.
Three of the seven surviving patches of red wash sampled.
When Dr. Buck’s Benjamin Moore color matches are compared to the Munsell chips, HC-49 is shown to be a fairly close match to 2.5 YR 4/6 (the most common Munsell match). When the Benjamin Moore and Munsell color chips are compared to the mocked-up samples produced by Susan Buck, it is clear that the original wash was slightly more orange. Of the three test samples, the closest color match to the surviving period wash samples is the “cream” consistency lime/casein wash found at the top layer of the brick panel. The color still needs adjustment, but Dr. Buck says the oranger color can be produced by changing the ratios of the pigments and lime. Interestingly, a wash with a higher lime-to-pigment ratio should carbonate more completely and make for a more stable lime wash.
Comparison of Benjamin Moore, Munsell and Wash Sample Colors
Thickness and Texture Findings
According to Dr. Buck’s analysis, the approximate thickness of the color washes she sampled was 10 microns (or .000393 (1/3,000th) of an inch). Because these samples had been exposed to the weather for approximately 30 years, and because they had multiple generations of stucco applied over them, the thickness of the samples does not represent the thickness of the wash when it was first applied. Similarly, evidence for the historic texture of the washes is also hard to determine because of their deteriorated condition.
However, since the thickness of the coating usually determines how thick the cured wash is and how well it holds the texture of the brush strokes, a quick search of 18th and early 19th century lime wash recipes for was made using Google Books (www.books.google.com). Specifically references that related how viscous traditional exterior lime washes were in the late-18th and early-19th century were sought. Oddly enough, the thickness of whitewash mixes seems to have been almost universally known during this time period and it is often cited in reference to mixing other liquids. However one 1809 English reference was found that compared the thickness of the whitewash to a thin paint.2 Because typical oil paints were relatively thick in 1809, it is possible that the above reference refers to a wash with the viscosity close to that of a modern paint. Another English reference, this time from 1804, describes a white wash for exterior work that consisted of one part water, ½ part “fresh-burnt” lime and ½ part sand (sand being, at least in the agricultural journals of the period, a common additive for exterior white washes).3 Although the actual thickness of this recipe can not be known until it is mixed, the wash also appears to be thicker then skim milk (which is about what Wayne Mays from Price Masonry is currently applying to the interior plaster walls) but thinner then a pancake batter.
Exterior Red Wash Specification
Four elements are needed to specify the red-wash for the Montpelier Mansion; composition, color, texture and thickness.
Through extensive testing of the existing samples found at Montpelier, as well as other houses in the region surrounding Montpelier, it has been determined that the wash originally applied to the mansion was a lime-based wash. Other then a trace reading of a protein component found in one of the Montpelier samples, no additives have been found. While there were worries about a pure lime-pigment wash not being durable, test samples of this mixture have been found to be just as durable as the samples that incorporate casein as an alternative binder.
As noted above, the best guide for the color of the reconstructed red wash are the surviving areas of washes found on the mansion. While these colors range slightly in hue (from the 2.5YR to 4YR in Munsell notation), they are fairly consistent in chroma and value. Therefore, the color of the new wash will be specified to fall, visually, between Munsell 2.5YR 4/6 and 5YR 4/8.
If a pigmented lime wash mixed to a “thin paint” consistency described above is determined to be acceptable and accurate for the period, then the final reconstructed wash will be mixed to that viscosity. By applying a wash mixed to the period viscosity, an accurate thickness should also be recreated on the exterior the mansion. Similarly, the texture and amount of brush marks retained in the cured wash can also be replicated by applying the wash using brushes and application techniques similar to those used in the early-19th century. However, additional research will need to be done to determine if the wash was applied in straight, long strokes or with random strokes of irregular length.
1 See James Madison’s Montpelier Exterior Paint Study. Susan Buck. Files of the Montpelier Architectural Research Department (12/2004).
2 Education of the Poor; Being the First Part of a Digest of the Reports of the Society for Bettering the Condition of the Poor. London, W. Bulmer and co. 1809.
The full reference is a good description of how white was produced in the early 19th century and the relevant parts are transcribed below:
“She therefore determined to lay out in lime the sum of four-pence, adding four-pence more for the loan of a brush and hair sieve, from the man of whom she bought the lime. She had contrived, without her sister’s observing it, to get a tub, in which to slack the lime. On this she had poured the water by degrees, and stirred it with a stick that was broad at one end. When the lime and water were well mixed, and of the consistence of mud, she took it out with a scoop, and strained it through the sieve into another vessel, where it settled to the bottom in a mass of white-wash. After skimming off the little water that remained at the top, she set it by, ready for the work of next day.
…as soon as her mother was set off for market, she began mixing her stuff with cold water to the consistence of a thin paint, and the operation of white-washing began.”
3 A General View of the Agriculture of the County of Norfolk. London, B. McMillan. 1804