Download as pdf or txt
Download as pdf or txt
You are on page 1of 20

Figure 1 Cupboard, Hadley, Massachusetts,

ca. 1715. Oak, maple, chestnut, and yellow pine.


H. 61!÷ *", W. 50", D. 21!÷$". (From the Collections
of The Henry Ford Museum.)

42 susan l. buck
Susan L. Buck ▼ O N C E I N A W H I L E newly discovered artifacts or inno-
vative interpretations of previously known objects change our perceptions
of how people lived centuries ago. The Hannah Barnard court cupboard
from Hadley, Massachusetts, serves as a case in point (fig. 1). Although long
Early Polychrome recognized as an important example of early-eighteenth-century New
England case furniture, the Hannah Barnard cupboard has more recently
Chests from Hadley, been viewed as the product “of an isolated craft and patronage system” and
as a dower object that says as much about gender roles and the expression
Massachusetts: A of female identity in the Connecticut River valley as it does about regional
stylistic preferences. To develop these new interpretations, scholars like
Technical Investigation Philip Zea, Suzanne Flynt, and Laurel Thatcher Ulrich adopted method-
ologies drawn from art history, traditional history, social history, and
of Their Paint and women’s studies. They also incorporated information brought to light
through paint and finish microscopy, analysis, and treatment, which is the
Finish subject of this article.1
The Hannah Barnard cupboard is an exception among contemporaneous
joined case pieces from western Massachusetts, which typically have two-
dimensional carved decoration and framing members painted in dark red
and black and rarely bear the full name of their original owner (fig. 2). Such
an elaborate piece of furniture, emphatically marked with Hannah
Barnard’s name, made a clear statement of her significance in her home.
Despite having a murky surface and a nearly illegible inscription, the Han-
nah Barnard cupboard was recognized as an important object when Henry
Ford purchased it from antique dealer Israel Sack in 1936. By that date, two
Hadley chests with similar painted designs were already in notable collec-
tions. George Sheldon, curator of the Pocumtuck Valley Memorial Associ-
ation in Deerfield, Massachusetts, acquired a two-drawer chest bearing the
initials “SW” in about 1870 (fig. 18) and donated it to Memorial Hall a few
years later. When Henry Francis du Pont established the Winterthur
Museum, his collection included a related chest he had acquired in 1926.
The restoration of that chest (fig. 3) was the subject of an article in Antiques
published in September of that year.
The chest of drawers pictured in this month’s Frontispiece stood for some
years . . . in a slaughtering barn in Chester, Vermont. That it might make
claim to antiquity was obvious; but it was a cumbersome thing, whose
somewhat dour massiveness was not materially ameliorated by an
enshrouding layer of heavy green paint. . . . none viewed it with compre-
hending aVection until Mark LaFontaine of Springfield, Vermont hap-
pened along. Perceiving its ancient flavor and the stout integrity of its

43 early polychrome chests


Figure 2 Photograph taken in the early twentieth
century showing the SH and WA chests in a bed-
room at Memorial Hall, Deerfield, Massachusetts,
ca. 1910. (Courtesy, Memorial Hall, Pocumtuck
Valley Memorial Association.)

construction beneath its green overlay, Mr. LaFontaine purchased the


piece and carried it home, where by the grace of special providence, he per-
sonally undertook the task of reconditioning his find. Instinctively he felt
that some pattern might be concealed beneath the modern coat of paint.
Accordingly, he proceeded slowly, and most carefully, with the aid of paint
solvents, until, having removed the last vestige of green paint, he was able
to reveal in its entirety one of the most interesting polychrome decorations
which has come to light on any American furniture of the period.

The results of this paint removal will be discussed in more detail below, but
even a cursory examination of the chest shows LaFontaine’s paint removal
eVorts wore and abraded the paints on the drawer fronts, and there is still
an uneven, grayish green film on the original decoration.2
When, in 1991, curators Philip Zea of Historic Deerfield and Suzanne
Flynt of the Memorial Hall Museum were planning their exhibition of
Hadley chests, the Hannah Barnard court cupboard had a dark brown sur-
face, its compass-generated, inscribed decoration was partially obscured,
and no traces of the bright colors and delicate, whimsical, wavy patterns
found on the two other chests were visible (fig. 4). The proposed exhibit
spurred Michael Ettema of the Henry Ford Museum to allow me to exam-
ine the Hannah Barnard cupboard and remove tiny samples from protected,
accessible areas of all representative painted surfaces.3
To characterize variations in the painted surfaces, I embedded the sam-
ples in polyester resin, ground each sample at a right angle to the surface
plane to expose the stratigraphy, and examined the cross sections with a
fluorescence microscope. Using biological fluorochrome stains, I was able
to characterize the organic binder components in the paints and varnish lay-
Figure 3 Hadley chest illustrated and discussed
ers. This investigation and subsequent research into the paints on the other
in Antiques 10, no. 3 (September 1926): 189–90 two chests revealed that the brown surfaces of the Hannah Barnard cupboard
(Courtesy, Antiques.) concealed the same type of exuberant polychrome designs. Moreover, the

44 susan l. buck
results of the cross-sectional analysis suggested it would be possible to safely
and responsibly remove the substantial accumulation of later plant resin var-
nishes, shellac, layers of overpaint, oily maintenance dressings, and grime to
reveal the original oil-bound painted decorations.4
Last year, I revisited the research conducted during the early 1990s using
the latest methods and equipment. Access to a new-generation fluorescence
microscope with digital imaging, more choices of ultraviolet light filters, and
other features allowed me to examine the original samples with greater clar-
ity, cross-check previous characterization of the stratigraphy, and select the
most informative samples for further analysis. Winterthur scientist Cather-
ine Matsen performed scanning electron microscopy energy-dispersive
spectroscopy (SEM-EDS) to help me identify the specific inorganic com-
ponents in each layer and more accurately assess the composition and dis-
persion of the pigments. The 1991 and 2008 investigations provide insights
into the relations between the paints used on this group of objects, as well
as more information about how the brilliant colors of the original deco-
ration have irrevocably changed over time. This new research also facilitates
comparisons between the paints used on the Hannah Barnard cupboard and
related chests, as well as those on a carved Hadley chest with a two-color
scheme and the initials “RA” (fig. 22).5

The Painted Surfaces of the Hannah Barnard Court Cupboard


At the time of the 1991 investigation, the Hannah Barnard cupboard had
been languishing in storage at the Henry Ford Museum for many years. Its
nearly monochromatic dark brown surface (fig. 4) resulted from an annual
rubdown with a linseed oil, turpentine, and vinegar dressing, a detrimental
practice followed by many museums and collectors of Ford and du Pont’s
era. A large container of this concoction still remained in the Ford
Museum’s conservation lab in 1991, complete with a long brown drip of
polymerized oil hanging from its spout as a warning of what could happen
if it was applied to any more objects.6
The first phase of the initial investigation involved sampling and analyzing
all the various design components on the front of the cupboard as well as
the sides and top, to determine the comparative coating stratigraphies. The
paint and varnish history on the cupboard was described in several reports.
The cross-sections showed that the drawer fronts and upper panels were
first painted with a white paint layer, and then the intricate floral and geo-
metric patterns were formed of green, bright red, blue, maroon and black
glazes and paints. The cross-sections also showed that the earliest layer on
the moldings and case sides was a deep red wash or stain, and that the two
turned columns were painted a surprising robin’s egg blue. Pigment analy-
sis using polarized light microscopy and microchemical testing was con-
ducted on paint samples from selected areas, and the most interesting
finding was that the columns were painted with a combination of white
lead and Prussian blue—an artificial pigment first synthesized in 1704. This
intense, high tint strength blue would have been quite a rare and expen-
sive pigment in the Colonies at the time the cupboard was made.
The early polychromy was completely obscured by at least eight alter-
nating layers of degraded plant resin varnish (such as copal, amber, mastic

45 early polychrome chests


Figure 4 Preconservation view of the cupboard
illustrated in fig. 1.

or sandarac), dirt, shellac, and two layers of dark repaint sandwiched


between the varnish layers. On top of this thick accumulation of darkened
material was a distinct sludge layer of degraded oil—the result of decades
of applications of the linseed oil, vinegar and turpentine dressing.

So, before embarking on the cleaning, the most important information


coming from the initial analysis was that the original polychrome decora-
tion was mostly intact below many later coatings. Additionally, the presence
of numerous layers of degraded, grimy, plant resin varnish and shellac, with
the two later generations of overpaints sandwiched between, suggested that
it might be possible to employ a cleaning system that softened and solubi-
lized the later varnishes without aVecting the original oil-bound paints.7
The uppermost drawer, which was much shinier than the others, had the
most areas of active paint loss. Comparisons of the samples taken from all
three drawers showed that the top one had an additional layer of compara-
tively fresh shellac (identified based on its characteristic orange autofluores-
cence in reflected ultraviolet light). It is likely that this layer of shellac was
applied in an attempt to counteract flaking of the paint, but instead, it exac-
erbated loss because the coating shrank as it dried. During this phase of
analysis, 35mm photographs of all the cross-section samples were taken with

46 susan l. buck
a camera mounted to the top of the microscope. These photographs were
compared and annotated in an attempt to identify the variations in the paint
and coating histories on all the finished surfaces of the cupboard (fig. 5).
The Hannah Barnard cupboard arrived at the conservation center of the
Society for the Preservation of New England Antiquities in late 1991. After
determining the original paint scheme, I took additional samples to exper-
iment with the eYcacy of diVerent solvents (figs. 6, 7). These cross sections
showed that it was possible to safely remove the later coatings but still retain
the original paints and the earliest plant resin varnish layer, using a high-

Figure 5 Cross-section sample taken from a red-


painted design on a drawer front of the Hannah
Barnard cupboard at 200X in reflected ultraviolet
light. This sample was taken before conservation
in 1991.

Figure 6 Trial cleaning on a drawer front with


sample locations noted to assess the level of var-
nish and overpaint removal.

viscosity isopropanol gel applied with cotton swabs and wadding. As the
cleaning slowly progressed, the brightly colored decoration that was
revealed proved how closely the designs and palette of this cupboard related
to the other two chests (figs. 8, 9). The worst areas of damage on the court
cupboard were primarily owing to use, particularly on the edges of drawers,
on the bottom rail of the paneling of the upper case, and the area around
the door lock for the upper panel, which had been replaced and filled in. The
original paints were still remarkably intact under their brown mask.8

47 early polychrome chests


Figure 7 Cleaning the upper case of the Hannah
Barnard cupboard using isopropanol gel and
swabs.

Figure 8 Detail of the center panel during con- The biggest surprise to come to light during the cleaning process was the
servation of the cupboard illustrated in fig. 4. original color of the columns. The slow process used to remove the varnish
enabled me to stop quickly if an original painted area appeared to be more
readily soluble, but as the cleaning progressed, it became clear that, unlike
all the other painted areas of the cupboard, the maple columns had lost most
of their decoration. Fortunately, enough original paint remained in the
recessed areas of the turnings to verify that the columns had originally been
painted a distinctive robin’s egg blue, a color not found anywhere else on
the cupboard. This paint was composed primarily of white lead with a small
proportion of the artificial basic copper carbonate pigment blue verditer
and the newly invented pigment Prussian blue.9
After consultations with the advisory committee organized to oversee the
cleaning process, I isolated the original areas of blue with a protective barrier
and inpainted the surface around them with a blue glaze of modern gouache.
This glaze could never be mistaken for the original because of the modern
acrylic resin barrier coat that separates it from the original oil-bound blue
paint and because the gouache contains titanium white, a pigment not in

48 susan l. buck
use before 1910. Clearly, the stripped wood columns, with their traces of
original blue paint, did not reflect the original intent of this totemic object.
The expense of the blue pigments used to decorate them also suggested they
were intended to stand out against the paints on the upper case.10
Pigment analysis using polarized light microscopy techniques showed
that the white background paint on the court cupboard is composed pri-
marily of white lead and calcium carbonate; the red-orange areas are red
lead and vermilion; the maroon areas are primarily red ochre and lamp-
black; and the blue-green areas (now degraded to almost black) are almost
pure blue verditer (fig. 9). The characteristic white autofluorescence of the

Figure 9 35mm photograph of a cross-section


sample taken from a degraded blue-green design
on a drawer front of the Hannah Barnard cup-
board at 500X in reflected visible light. This sam-
ple shows a coarsely ground layer of blue verditer
pigment in a resinous binder on top of the white
background paint. This sample was taken from a
cleaning test area in a blackish design area of the
compass-drawn decoration on a drawer front.
Remnants of the varnish layer above the pebbly
blue paint reveal that some of the earliest film
finish was left intact in the cleaning process.

binding media surrounding the rounded blue verditer particles documents


the presence of a plant resin varnish, which would have made this blue-green
layer quite glossy and slightly translucent. This resinous binder also con-
tributed to the significant darkening of this originally brilliant decoration.
The iron oxide–based reds and the carbon-based black are very stable pig-
ments, and those areas have darkened primarily owing to the degradation
of the oil-binding medium and embedded grime. Vermilion, like that used
for the orange triangles and half circles in the compass-drawn designs on the
drawer fronts, can darken from prolonged exposure to light, although the
mechanism of this phenomenon is not fully understood. The presence of ver-
milion is significant, as it was an expensive pigment in the early eighteenth
century and typically would have been confined to use on easel paintings,
especially in a rural area like the Connecticut River valley of Massachusetts.
Thus, although it was possible to successfully remove the later brownish
coatings from the painted surfaces of the court cupboard, some of the col-
ors, particularly the red-orange and blue-green areas, have permanently
darkened.11
I deliberately did not clean a 4-by-6-inch area of the upper case (behind
the right column) so as to preserve the complete finish history of the cup-
board as it transpired from the date of manufacture until 1991 and provide

49 early polychrome chests


a test site for future analysis. Samples taken from that site in 2008 confirm
that some of the more protected areas of the upper-case designs were not
overpainted brown and maroon like the drawer fronts and were not var-
nished as frequently as the lower case.12
The stratigraphy in a cross section (sample HB-1) from an area of red-
painted decoration reveals that the red-orange paint was thinly applied on
the white base coat and that both layers are now fractured and disrupted. In
reflected ultraviolet light it is apparent that the first layer on top of this red-
orange paint is a degraded plant resin varnish layer, identifiable by its
characteristic bright white autofluorescence under reflected ultraviolet
light. There is no dirt separating the red-orange paint and the plant resin
varnish directly on top of it, suggesting the first varnish is the original coat-
ing. The second coating is pigmented shellac, identifiable by its pale orange
autofluorescence and the tiny particles of pigment in suspension. The third
coating appears to be a grayish pigmented varnish, followed by two more
layers of clear plant resin varnish. The uppermost coating is a thick layer of
pigmented shellac. Distinct films of dirt separate each application of shellac
and plant resin varnish, and there is a layer of oil on the surface of the upper-

Figure 10 Digital photographs of a


cross-section sample (HB-1) taken from a
red-orange design behind the right col-
umn of the Hannah Barnard cupboard at
200X in reflected visible light (right),
reflected ultraviolet light (lower left), and
with the flurochrome stain DCF to tag
saturated and unsaturated lipids (lower
right). This sample was taken in 2008
from a site intentionally left uncleaned
during conservation in 1991.

most coat of pigmented shellac (fig. 10). SEM-EDS analysis of this same
cross-section sample confirmed that primary elements in the white back-
ground paint are lead with calcium and silicon, and the primary elements in
the red-orange paint are mercury and lead (fig. 11). This corroborates the
initial identification of the white background paint as white lead and cal-
cium carbonate and the red-orange paint as vermilion and red lead.13

50 susan l. buck
Figure 11 SEM-EDS analysis of the sample illus- The Painted Surfaces of the Winterthur Chest
trated in fig. 10. These images show the distribu- In the early phase of investigating the paints on the Hannah Barnard cup-
tion of elements in the white background and
red-orange decoration. Vermilion in the red-
board, the Hadley chest acquired by du Pont in 1926 was examined and
orange layer is confirmed by the presence of mer- sampled to identify how the paint schemes on the two objects related
cury which appears as green in the elemental (fig. 12). This research was revealing, as it showed that in some areas two
map. (Photo, Catherine Matsen.)
layers of finely ground grayish green overpaint remained on top of a frac-
tured and degraded plant resin varnish coating (fig. 13). Additionally, where
the leaf and vine decoration had degraded to almost black, cross-section
analysis confirmed that the same type of pure blue verditer paint on the
Hannah Barnard cupboard was used to create what must have been a jewel-
like blue-green color when it was first applied to the fronts of the stiles and
drawers of the chest.

Figure 12 Chest, Hadley, Massachusetts,


ca. 1715. Red and white oak, and chestnut with
pine. H. 42!÷$", W. 43!÷@", D. 20!÷@". (Courtesy,
Winterthur Museum.)

51 early polychrome chests


Figure 13 35mm photograph of a cross-section
sample taken from a deep red area of a compass-
generated motif on a drawer front of the chest
illustrated in fig. 12 at 250X in reflected ultraviolet
light. Two layers of finely ground overpaint are
visible on top of a cracked, degraded plant resin
varnish and a damaged red paint layer on the
white background paint. This sample was taken
in 1991.

In 2008 I selected areas of the painted decoration on the Winterthur


Museum chest for sampling and reexamination. This small group of sam-
ples shows even more strongly the relationship between the paints on the
Hannah Barnard court cupboard and the chest at Winterthur. It also reveals
some significant new information about the first layer of opaque overpaint
on the latter object. In two samples (W-1, W-3) there is a worn layer of
degraded plant resin varnish on top of the first decorative paint layer, which
was likely the original varnish applied to this chest. Above this varnish coat-
ing is a notably coarse and unevenly ground grayish blue paint with excep-
tionally large Prussian blue particles unevenly distributed throughout the
layer (fig. 14). SEM-EDS analysis of a sample (W-1) from a bright

Figure 14 Digital photographs of a cross-section


sample (W-1) taken from a red-orange area of a
compass-generated motif on a drawer front of
the chest illustrated in fig. 12 at 250X in reflected
visible light (left) and reflected ultraviolet light
(right). A coarsely ground blue-gray paint sits
directly on top of the original plant resin varnish,
which is thin and degraded. The varnish covers
the red-orange paint and white background paint
beneath it. This sample was taken in 2008.

orange–painted design confirms the use of vermilion and red lead, and it
also proves that the coarse overpaint is composed of white lead and very
large particles of Prussian blue (fig. 15). The irregular dispersion of chunky
pigments in the grayish blue overpaint is characteristic of an early hand-
ground paint, and it strongly suggests that the polychrome designs were
first painted over before the end of the eighteenth century.14
In cross section the original blue-green layer of blue verditer suspended

52 susan l. buck
Figure 15 SEM-EDS analysis of the sample illus- in a resinous binder in sample W-3 is identical to samples from comparable
trated in fig. 14. These images show the distribu- designs on the court cupboard (fig. 16). This was confirmed with SEM-EDS
tion of elements in the layers. Vermilion in the
analysis (fig. 17). Another sample (W-4) taken from an area that now
red-orange layer is confirmed by the presence of
mercury which appears as green in the elemental appears maroon in color must have originally been a deeper, richer red, as
map. (Photo, Catherine Matsen.) it is composed of red lead and red ochre, much like the comparable areas
sampled on the court cupboard. The similarities in composition, thickness,
pigment dispersion, and binding media components suggest the Hannah

Figure 16 Digital photographs of a cross-section


sample (W-3) taken from the painted vine on the
left front stile of the chest of drawers illustrated in
fig. 12 at 200X in reflected visible light (left) and
reflected ultraviolet light (right).

Figure 17 SEM-EDS analysis of the


sample illustrated in fig. 16. These images
show the distribution of elements in
the layers. Blue verditer is confirmed by
the presence of copper which appears as
blue-green in the elemental map. (Photo,
Catherine Matsen.)

53 early polychrome chests


Barnard court cupboard and the Winterthur chest were painted using the
same paint materials, and possibly by the same hand.15
Regrettably, the cleaning in 1926 appears to have caused the ghostlike
appearance of most of the polychrome designs on the drawer fronts. The
white lead–based background paint that can be seen in the cross-section
samples is also barely discernible on close examination of the surfaces.
Examination of the surface also revealed that the white paint was mostly
removed from the exposed background areas surrounding the brightly col-
ored compass-drawn designs. Thus, the background now appears more tan
in color, rather than its original bright white.

The Paint Evidence on the SW Chest


The chest at the Memorial Hall Museum now has the brightest palette of
this group of three polychrome case pieces (fig. 18), but a considerable
amount of the current paint scheme is not original. On the front feet the
orange-painted triangular decorations extend down beyond where the orig-
inal white background paint has been completely worn away, and the bright
blue outlines on the drawer fronts cover dents and abrasions on the exposed
outer edges. There are also bright blue areas of the compass-drawn designs
on the drawer fronts that seem to relate to the fresher blue paint on the
edges of the drawers, and there are flat black areas of the compass-drawn cir-
cular decorations that do not appear in similar areas of the other two pieces.

Figure 18 Chest, Hadley, Massachusetts,


ca. 1715. Oak and pine. H. 43!÷$", W. 45", D. 19!÷@".
(Courtesy, Memorial Hall, Pocumtuck Valley
Memorial Association; photo, Amanda Merullo.)

54 susan l. buck
Figure 19 Digital photographs of a sample
(SW-1) taken from a repainted blue area of a
compass-generated motif on the right front panel
of the chest illustrated in fig. 18 at 400X in reflected
visible light (left) and reflected ultraviolet light
(right). The cross section shows that this area was
originally red-orange on a white base coat and
that it was repainted after the red-orange layer
had become worn and cracked.

Cross-section analysis of a small group of samples taken before the 1993


Hadley chest exhibition revealed that the blue paint around the edges of the
drawers is much later than the object’s date of manufacture. The blue is
emulsion paint, whereas the original paint on the rest of the chest is oil-
bound, and the former extends over dirty surfaces. The cross section visible
in figure 19 (sample SW-1), which was taken from the blue rectangular dec-
oration on the right front panel, indicates that this finely ground artificial
ultramarine was applied on top of the original worn and cracked red-orange
decorative layer and white base coat. SEM-EDS analysis further demon-
strated that the blue overpaint contains sodium, aluminum, silicon, and
potassium (components of artificial ultramarine), whereas the remnants of
red-orange paint contain mercury and lead (fig. 20).16
The identification of artificial ultramarine in the blue overpaint is signifi-
cant, since that pigment was not commercially available until about 1826.
Equally important, the presence of vermilion and red lead in the red-orange
decoration below the blue paint proves that the same combinations of pig-
ments were used for the red-orange decoration on the SW chest, the Han-
nah Barnard cupboard, and the Winterthur Museum chest.17

Figure 20 SEM-EDS analysis of the


sample illustrated in fig. 19. These images
show the distribution of elements in the
layers. Vermilion in the red-orange layer
is confirmed by the presence of mercury
which appears as green in the elemental
map image. (Photo, Catherine Matsen.)

55 early polychrome chests


Recent analysis also revealed more about the bright blue-green paint
originally used for the leaf and vine decorations, as well as one of the alter-
nating colors in the compass-generated motifs. Many of the areas of black
overpaint on the SW chest cover the original, now quite degraded, blue-
green paint, which is composed primarily of blue verditer in a resinous
binder. It is likely that by the late nineteenth century, when George Shel-
don purchased the SW chest, the formerly blue-green decorative areas were,
in fact, quite black owing to degradation and the accumulation of grime
layers. Identical passages of blue-green, using the same type of pure blue
verditer bound in plant resin and oil, are visible in samples taken from the
geometric and leaf and vine motifs of both the Hannah Barnard court cup-
board and the Winterthur chest.18
Other painted designs on the SW chest simply look questionable, such as
the orange triangles at the bottoms of the front legs. The cross section from
the orange-painted triangle on the left front leg shows a surprisingly coarsely
ground paint applied over worn, dirty wood fibers (fig. 21). This unevenly

Figure 21 Digital photographs of a sample


(SW-6) taken from a repainted triangular design
on the worn area of the left front foot at 200X in
reflected visible light (left) and reflected ultraviolet
light (right).

mixed, deep orange paint could be easily mistaken for a hand-ground, eigh-
teenth-century paint. However, the original red-orange paint, composed
primarily of red lead and vermilion, is finely ground and evenly mixed (char-
acteristic of paint composed of inherently tiny particles of pigment). This
allows us to conclude that not only were many areas of the SW chest painted
over, but certain colors and passages, such as the original red-orange com-
pass-drawn designs and the black drawer borders, were radically altered.
This overpainting was likely done in 1893, based on the chronology
described by Zea and Flynt in their catalogue Hadley Chests. They noted that
George Sheldon donated the chest to the Memorial Hall Museum in
December 1892, and the following November the Pocumtuck Valley
Memorial Association’s assistant curator Janitor Mary P. Wentworth
(1835–1901) received $1.50 “for painting and cleaning [a] dresser.” Zea and
Flynt surmised that the “dresser” was the SW chest because it is the only
conspicuously overpainted object in that institution’s collection, and the
process of cleaning and painting would have taken about a week.19

The Paint Evidence on the RA Chest


Probably made in Hadley or Hatfield, the RA chest was donated to Memo-

56 susan l. buck
rial Hall by Chester Graves Crafts in 1887 (fig. 22). Although relatively con-
ventional in form, it survives with most of its original paint scheme intact.
To compare the composition of its paint with that of the Hannah Barnard
cupboard and related chests, I examined three samples taken by Historic
Deerfield architectural conservator William Flynt.20

Figure 22 Chest, Hadley or Hatfield, Massachu-


setts, 1695–1715. Oak and yellow pine. H. 33#÷$",
W. 47!÷$", D. 20#÷$". (Courtesy, Memorial Hall,
Pocumtuck Valley Memorial Association; photo,
Amanda Merullo.)

Figure 23 Digital photographs of a sample taken


from a beveled portion of the left stile of the chest
illustrated in fig. 22 at 400X in reflected visible
light (left) and reflected ultraviolet light (right).
This sample shows that the chest initially received
a very thin coat of red-pigmented paint. The un-
even layer of degraded plant resin varnish directly
above the paint is early and probably original.
The white haze on the surface is accumulated
paste wax.

The sample from a beveled portion of the left stile, which was protected
by the edge of the bottom drawer, indicates that there is only one layer of
finely ground, oil-bound black paint on top of the wood. This cross section
contains a remnant of a plant resin varnish coating on top of the original
black paint, but it is not possible to tell from the layer sequence if this
varnish is original. The sample from the lower right corner of the left panel
provides more intriguing information. In this cross section there is a thin
layer of finely ground red-pigmented paint on top of the wood, followed by
an equally thin layer of plant resin varnish (fig. 23). There is no evidence of
a film of dirt trapped between the red paint and the degraded varnish,

57 early polychrome chests


suggesting that it is an original or very early protective coating. There is also
an amorphous, translucent whitish layer on top of the varnish, which is
likely a later accumulation of paste wax. Polarized light microscopy pig-
ment analysis and SEM-EDS indicate that the pigments in the red layer are
primarily red ochre with some red lead.21
The original black and deep red paints on the RA chest are much thinner
than the paints on the three polychrome pieces, suggesting that these coat-
ings were not meant to obscure the figure of the oak completely. The use of
lampblack for the black paint and the combination of red ochre and red lead
for the red paint are typical of inexpensive paints made of stable pigments
that might have been used by a joiner or housepainter. The overall palette
of the paints on the RA chest has darkened owing to the accumulation of
dirt, the yellowing of the plant resin varnish, and the oil-binding media, but
the colors are still readily identifiable as a deep red and black.

Paint Analysis Past and Future


Analytical work on the paint on the Hannah Barnard cupboard, subsequent
conservation performed in the early 1990s, and more recent cross-section
microscopy and SEM-EDS findings have helped clarify relationships
between the original paints used on that object and the two related chests.
These investigations reveal that it is possible to analyze the coatings applied
to furniture made centuries ago, confidently determine how the surfaces
may have looked when new, and ascertain how those coatings have
degraded and changed over time.
One of the intriguing questions arising from this study is: How long did
the extraordinarily bright and whimsical decorations on the three poly-
chrome pieces remain in place? The exceedingly coarsely ground, grayish
blue paint directly on top of the polychromy on the chest illustrated in
figure 12 suggests that it might have been painted over before the end of the
eighteenth century (fig. 14). By contrast, there is no physical evidence to
suggest that the decorations on the SW chest were ever completely painted
out (fig. 19), although some areas were selectively repainted. And although
the Hannah Barnard cupboard had turned an overall brown by the early
twentieth century, the comparative cross-section evidence suggests that the
later touch-ups were intended to enhance the by-then obscured decora-
tions, rather than cover them completely.
Another question pertains to survival: Why have only three polychrome
case pieces from the Hadley-Hatfield area come to light, whereas more than
250 conventional carved and joined chests have survived?22 The former were
undoubtedly expensive objects, so it is likely that fewer were originally
made. In addition, the brilliance and the distinctly identifiable style of the
decoration on the Hannah Barnard cupboard, the SW chest, and the
Winterthur chest may have made that class of object less desirable when
tastes began to change. One can only hope that other examples survive with
their decoration intact below later coatings. As this article has shown, the
technology for understanding and conserving such painted forms exists and
continues to improve.

58 susan l. buck
a c k n o w l e d g m e n t s For assistance with this article, I would
like to thank Mark Anderson, Amelia Bagnall, Edward A. Chappell, Wendy
Cooper, Clara Deck, Mary Fahey, Suzanne L. Flynt, William Flynt,
Natasha Loeblich, Catherine Matsen, and Stephanie Rabourdin-AuVret.

1. For recent interpretations of the Hanna Barnard cupboard, see Philip Zea and Suzanne
L. Flynt, Hadley Chests (Deerfield, Mass.: Pocumtuck Valley Memorial Association, 1992),
pp. 20–24; and Laurel Thatcher Ulrich, “Furniture as Social History: Gender, Property, and
Memory in the Decorative Arts,” in American Furniture, edited by Luke Beckerdite and
William N. Hosley (Hanover, N.H.: University Press of New England for the Chipstone
Foundation, 1995), pp. 39–68.
2. Antiques 26, no. 5 (November 1934): 168. Zea and Flynt, Hadley Chests, p. 20. Henry Francis
du Pont purchased the chest from Mark LaFontaine. Antiques 10, no. 3 (September 1926):
189–90.
3. The samples (about 200 microns in size with paints attached to a few wood fibers) were
removed with a scalpel from the edges of losses and joins to avoid intruding into intact paint
surfaces. They were permanently cast in polyester resin in mini ice cube trays and polished with
sandpapers (grits 220, 400, and 600) and MicroMesh silica-embedded polishing cloths (grits
1,500–12,000) to expose the cross sections and provide smooth, flat surfaces for analysis and
photography.
4. This initial research was coordinated with Henry Ford Museum curator Michael Ettema
and conservator Ralph Wiegandt. The samples were examined at the Society for the Preserva-
tion of New England Antiquities (SPNEA, now Historic New England) Conservation Cen-
ter in Waltham, Massachusetts. Biological fluorochrome characterization was conducted using
an Olympus BH-T Series II fluorescence microscope and the fluorochrome stains triphenyl
tetrazolium chloride for the presence of carbohydrates, fluorescein isothiocyanate for proteins,
and Rhodamine for oils. See Richard C. Wolbers, Cleaning Painted Surfaces: Aqueous Methods
(London: Archetype, 2000), pp. 167–83. An advisory committee for the conservation of the
Hannah Barnard court cupboard was convened to discuss and oversee the conservation treat-
ment. This committee included Philip Zea, Suzanne L. Flynt, Michael Ettema, William Hosley,
and Richard Nylander. I undertook this work while a furniture conservator at SPNEA.
5. Catherine Matsen, unpublished SEM-EDS report for Susan L. Buck, November 21, 2008,
p. 1: “All six cross-section samples were reduced in width so as to provide the proper working
distance for better imaging once mounted on a carbon stub and placed in the SEM chamber.
The excess casting resin was cut from the side opposite the cross-section with a jeweler’s saw
and mounted to a carbon stub with double-sided carbon tape adhesive. Carbon paint was
applied over the side and top surfaces of the casting resin, without covering the sample itself.
The samples were examined using the Topcon ABT-60 scanning electron microscope with an
accelerating voltage of 20kV, working distance of 26mm, and sample tilt of 20°. The EDS data
was analyzed with the Bruker X-flash detector and microanalysis Quantax model 200 with
Esprit 1.8 software. Back-scattered electron (BSE) images, energy dispersive spectroscopy
(EDS) spectra and elemental maps were gathered of all samples.”
6. This practice of oiling the furniture was described by Ralph Wiegandt, who discontinued
it when he joined the staV of the Henry Ford Museum.
7. Analysis was conducted using an Olympus BH-T Series II fluorescence microscope with
a polarizing light base and a dedicated 35mm camera. Susan Buck to Philip Zea and Suzanne
L. Flynt, July 21, 1992.
8. The cross-section evidence suggested that a prudent cleaning approach would involve sev-
eral steps. The first would remove the oily dressing residues with a targeted water-based
enzyme cleaning system using the enzyme lipase in a water-based gel. After this cross-linked
oily material was gone, it appeared that it would be possible to slowly remove the two over-
paints, the pigmented shellac layers, and the plant varnish coatings with thickened solvent gels
that would restrict the cleaning eVect to the uppermost layers. Ultimately, the controlled clean-
ing tests on one drawer front and one upper panel showed that it was possible to slowly under-
cut the uppermost oily residues and the brown overpaints using an isopropanol gel, which
softened and solubilized the plant resin varnishes and shellac layers below the overpaints. This
process required numerous applications of the gel, wiping away the residues with cotton, and
then clearance with quick applications of first isopropanol and then odorless mineral spirits as

59 early polychrome chests


free solvents. This made it possible to get down to the varnish layer that remained directly on
top of the original paints without removing either it or the original paints. The isopropanol
gel was made with isopropanol, Carbopol 954, Ethomeen C-25, and water.
9. Blue verditer is noted as a housepainter’s color in the seventeenth century (Rutherford
J. Gettens and Elizabeth West Fitzhugh, “Azurite and Blue Verditer,” in Artists’ Pigments: A
Handbook of Their History and Characteristics, edited by Ashok Roy, 2 vols. [Washington, D.C.:
National Gallery of Art, 1993], 2: 31). Prussian blue is described as “the earliest of the synthetic
colors” first made in Berlin in about 1704 (Rutherford J. Gettens and George L. Stout, Paint-
ing Materials: A Short Encyclopedia [New York: Dover, 1966], pp. 149–50).
10. Gettens amd Stout, Painting Materials, p. 161. Titanium white was put into regular pro-
duction between 1916 and 1919.
11. Rutherford J. Gettens, Robert L. Feller, and W. T. Chase, “Vermilion and Cinnabar,” in
Roy, ed., Artists’ Pigments, 2: 167.
12. Henry Ford Museum objects conservator Clara Deck took three additional samples for
this analysis in September 2008. This work was coordinated through that museum’s head of
preservation, Mary M. Fahey. The cupboard is now permanently exhibited in the furniture sec-
tion of the museum.
13. The conclusion that oil components existed in the paint layers and on the surface of the
uppermost pigmented shellac layer was based on the positive reactions for unsaturated lipids
with the fluorochrome stain 2, 7 dichlorofluorescein. SEM-EDS analysis provided solid evi-
dence of the elemental composition of the white and red-orange layers: “Calcium (Ca) and lead
(Pb) are the primary elements in the ground and suggests the presence of chalk (calcium car-
bonate, CaCO3) and lead white (basic lead carbonate, 2PbCO3·Pb(OH)2). A large, translu-
cent particle—likely chalk—is present in the ground. With spot EDS analysis of the red paint
layer, calcium, lead and mercury (Hg) are the strongest elements detected to suggest a com-
position of chalk, red lead (Pb3O4) and vermilion (HgS)” (Catherine Matsen, unpublished
SEM-EDS report, p. 5).
14. Wendy Cooper, Mark Anderson, and Stephanie Rabourdin-AuVret helped provide
access to the Winterthur chest. “Spot EDS analysis of the first-generation red paint of Sample
W-1 detects mercury and sulfur to indicate the presence of vermilion. Elemental mapping
shows the presence of silicon also in the first-generation red paint (likely present as silica,
SiO2). With elemental mapping at 900X magnification, low levels of lead and calcium detected
in the red paint suggest red lead (Pb3O4) and chalk (CaCO3) are present. Low levels of alu-
minum and sodium suggest a clay is present, potassium may be present in the paint’s organic
binder. The paint layer over the first red paint generation is a lead-containing paint (likely lead
white, 2PbCO3·Pb(OH)2) with iron-containing (likely Prussian blue, Fe4[Fe(CN)6]3) blue
particles suspended within” (Matsen, unpublished SEM-EDS report, p. 3).
15. Cross-section microscopy and polarized light microscopy analysis conducted by me and
SEM-EDS analysis confirm the similarities in paint composition. SEM-EDS analysis of sample
SW-1 showed that “Copper and oxygen are detected in association with one another in the blue
particles of the lowest paint layer. This combination suggests the blue pigment is azurite or a
synthetic blue verditer (both 2CuCO3·Cu(OH)2). A calcium-containing layer above the blue
pigment particles is not expressly seen in visible light but is evident under ultraviolet light. This
may be an oil varnish layer (that contains chalk) applied to protect the blue paint. The white
paint above this layer contains predominantly lead, likely as lead white. Small blue particles are
seen within this paint layer and correspond to iron detected with elemental mapping; this sug-
gests the presence of Prussian blue pigment” (Matsen, unpublished SEM-EDS report, p. 4).
16. Cross-section microscopy with biological fluorochrome staining was used to identify the
comparative paint layer stratigraphies and to characterize the organic components in each
layer. SEM-EDS analysis helped distinguish the original paints from the blue overpaint: “Spot
EDS analysis and elemental mapping indicate lead and mercury are the main elements detected
in the orangish-red paint, again likely present as red lead and vermilion. Aluminum, potassium,
calcium and iron x-ray lines are weakly detected likely as clay, chalk and iron oxide red—either
as intentional additives to the paint or due to contamination from polishing the cross-section.
The second generation blue paint is composed of sodium, aluminum, silicon, lead, potassium
and calcium. The blue color and combination of elements detected suggests the blue pigment
to be ultramarine blue (Na8-10Al6Si6O24S2-4). Calcium may be present as chalk, lead as lead
white and potassium may be present in the paint’s organic binder” (Matsen, unpublished
SEM-EDS report, p. 3).

60 susan l. buck
17. Joyce Plesters, “Ultramarine Blue, Natural and Artificial,” in Roy, ed., Artists’ Pigments,
2: 37.
18. Polarized light and cross-section microscopy analysis I conducted of samples taken from
darkened blue-green areas of all three objects showed the blue-green paints to be nearly iden-
tical in pigment particle size and distribution and the thickness of the layers.
19. Zea and Flynt, Hadley Chests, p. 22.
20. Suzanne L. Flynt allowed access to the chests for sampling.
21. SEM-EDS analysis provides important information about the deep red layer and the
accumulation of whitish material above the surface: “Detection of elements in the thin, red
paint layer was not possible with spot EDS; instead, the elemental composition is best under-
stood with elemental mapping. At 1500X magnification iron (Fe) is the primary element
detected in the thin, red paint layer, likely as iron oxide red (Fe2O3). Sulfur is detected in asso-
ciation with iron though the phase of this combination of elements is unknown. The appear-
ance of the discontinuous layer above the red paint and detection of sodium (Na), silicon (Si),
calcium (Ca), potassium (K), and aluminum (Al) suggests a clay or dirt composition; this may
simply be the accumulation of furniture polish from over the years” (Matsen, unpublished
SEM-EDS report, p. 3).
22. Zea and Flynt, Hadley Chests, p. 5.

61 early polychrome chests

You might also like