By Donna Strahan and Blythe McCarthy
Not much is known about early Buddhist lacquer technique because so few sculptures exist. By studying the buddhas along with a lacquer bodhisattva head made about one hundred years later, we have been able to trace the technique across three centuries. Secrets of the Lacquer Buddha provided an opportunity to study the construction, repairs, and internal features of these four lacquer sculptures. Our goal was to determine how and of what these unique sculptural images were made.
We thought that such sculptures were expensive and fragile; through scientific analysis, we have confirmed this. But many questions remain. Though they were created over a two- to three-hundred-year period, between the sixth and eighth centuries, were the same traditions and similar techniques and materials used in all four sculptures? Could the materials used shed light on the images’ use as reliquaries? And could the Metropolitan and Freer buddhas have been made in the same workshop but by different hands? Or was each one the work of a single artist?
While we were not able to answer all these questions, we did uncover substantial new insights.
How were they made?
Lacquer sculpture is a very difficult, time-consuming, and expensive craft. The first step is to create a core onto which the lacquer can be sculpted. Once the core is completed, layers of cloth strips wet with lacquer are applied, helping to create a more detailed, three-dimensional form.
The four sculptures studied are similar in many aspects, but each has its own special features. Only the Walters buddha was constructed on a wood core, carved from a solid, joined-wood structure, covered with cloth and lacquer, and then painted and gilded. It is the earliest preserved example of a wood-core lacquer sculpture. The Freer and Metropolitan buddhas and the bodhisattva are hollow-core lacquer sculptures, formed over clay cores that were removed just after the sculptures were completed.
Although all four sculptures were fabricated in the same lacquer method, also called the dry lacquer technique,[i] only the Walters buddha is constructed on a wood core. It is considered a wood-core lacquer sculpture. The other three sculptures were formed over a clay core that was removed before the sculptures were completed. They are considered to be hollow-core lacquer sculptures.
Asian lacquer is a resin. This tacky substance is essentially a water in oil emulsion that sets by polymerization to become tough, flexible, and durable. It is produced from the sap of a number of trees belonging to the Anacardiaceae family. Regardless of the type of sap used, the refining and fabrication techniques of lacquer are similar. The main tree species that produces the resin used to make lacquer, known as Toxicodendron (formerly Rhus vernicifera) vernicifluum and in the cashew family, is grown in China, Japan, and Korea. Raw lacquer tree resin is toxic and causes a poison ivy-like contact dermatitis that may be severe and even fatal.
The tree is usually tapped in summer and the sap purified and condensed to form a paste (fig. SB). Excess water present in the sap is evaporated by slow warming. Lacquer is then divided into processed and unprocessed grades. The processed lacquer is further divided into plain lacquer, to which additives such as extenders and coloring materials are added, and black lacquer, to which color from iron filings or soot is added.[ii]
Once purified and applied to a surface, lacquer requires a temperature of 20–28 degrees Celsius and high relative humidity (more than 65 percent is best) to harden into an impervious film. Lacquer objects are usually made of multiple layers of lacquer, and many weeks of work, even up to a year, may be required to complete a single object. Each layer must cure before the next layer can be applied. Once set, lacquer produces a tough, resistant coating.
When lacquer is applied to an object, the first or lowest layer is often bulked with fillers to seal and fill flaws in the support. Fillers vary depending on the support: examples include raw or fired powdered clays, sawdust, starch, ground-up dried lacquer, seashells, bone ash, and wood ash.[iii] Then, thinner layers are applied. The lacquer sculptures we studied were completed with brightly colored paint and gilding.
There are numerous Chinese classical texts and historical documents produced over hundreds of years that mention lacquer, but few of them contain in-depth fabrication details.[iv] In addition, there are some major pitfalls encountered when reading and interpreting historical Chinese literature. These include literature identification, language complications, changes of definition, alterations of measurement, and reliability of the information.[v] Fortunately, recent developments in analytical techniques have made it possible to identify many of the materials included in Asian lacquer objects, including the four we researched.[vi]
Methods of study
The life-size sculptures and head were studied and analyzed between 2014 and 2017. To answer questions concerning their structure and composition, we applied multiple analytical methods.
X-radiography and computed tomography (CT) scans were used to study the sculptures’ construction, joining methods, repairs, and other features not visible on their exteriors (fig. S1).[vii]
Binocular and plane-polarized light microscopy (PLM) were employed to study minute samples removed from the sculptures (fig. S2). Using known reference materials for comparison, these techniques can identify unknown substances. They helped us identify the cloth on the sculptures, as well as to examine the lacquer, lacquer layers, lacquer inclusions, and pigments. Cross sections and dispersed samples of paint and lacquer were examined using reflected, transmitted, polarized, and ultraviolet light microscopy. Stains were applied to selected lacquer samples and viewed under the microscope in both visible and ultraviolet light to help identify components in each lacquer layer.[viii]
X-ray fluorescence (XRF) analysis is a nondestructive analytical technique used to determine the elemental composition of inorganic materials. It helped confirm the identification of pigments and the material used for the eyes, which could not be sampled (fig. S2a).[ix]
Scanning electron microscopy (SEM) allows magnification far beyond that attainable with a light microscope. When the SEM is equipped with energy dispersive X-ray spectroscopy (EDS), it is capable of identifying individual elements. It was used to isolate inorganic elements present in the lacquer layers. The information obtained confirmed many of the findings from polarized light microscopy. In addition, elemental distribution maps were produced to confirm the location of specific elements within cross-section layers. SEM images were an additional aid in identifying morphological features of various inclusions, such as bone in the lacquer.[x]
Fourier transform infrared spectroscopy (FTIR) obtains an infrared spectrum of a material for identification. It was used to understand the components of the individual textile layers and their relationship with the clay core.
Pyrolysis gas chromatography mass spectroscopy (Py-GCMS) was used to identify all the organic components mixed in each lacquer layer. The National Institute for Standards and Technology’s AMDIS (Automated Mass Spectral Deconvolution & Identification System) was used for deconvolution and identification of the data; and the expert system RAdICAL, developed at the Getty Conservation Institute, was used for interpretation (fig. S3).[xi] [xii]
Proteomics is a field of molecular biology that studies the proteins in organisms. In proteomics, proteins in a material (in this case, bone) are broken up into their component peptides. The molecular weight of the peptides together with the sequences of amino acids that they contain can be compared to databases of known protein sequences to identify animal species.[xiii]
As a wood-core lacquer sculpture, the Walters buddha’s core is made entirely of wood, with clay filler used in the lips, ears, textile folds, and possibly the eyelids. It is composed of twelve pieces of solid, carved wood pegged together with wood dowels and iron nails.[xiv] The back was hollowed out, providing a large cavity, which probably held dedicatory material. The wood panel that originally covered the cavity is now missing. There is no visible coating sealing the inside of the open wood cavity in the back of the Walters sculpture.
Both the Freer and Metropolitan images are hollow-core lacquer sculptures. They are hollow from the top of the head through the torso cavity and open on the bottom. Wood pieces are used for support in both buddhas, although not always in the same location. Both sculptures have narrow wood pieces enclosed in textile running around the interior edges of their bases, providing structural support to a location that receives much wear. Originally, the bottoms of both the Freer and Metropolitan buddhas were closed, but there is no evidence remaining of how this was achieved. These wood pieces may have aided in securing a cover over the open bottom. Some wood pieces are later replacements, but those wrapped in textile are original.
Additionally, imbedded in the textile of the Metropolitan image are three wood boards running vertically up the buddha’s back, acting as a spine. The Freer buddha does not have any wood in his back. However, the Freer buddha’s forearms are formed from two planks of wood butt-jointed just below the elbow. No wood was used in the arms of the Metropolitan image; they are hollow.
The bodhisattva head does not contain any wood, but it is only a fragment of a much larger sculpture—nearly 7.5 feet tall if it was a seated image.[xv] No doubt wood was used to support certain areas of such a large sculpture.
The Freer and Metropolitan buddhas have vertical wood slabs in the back of their heads, incased in the textile layers (fig. S4). In X-radiographs, two large, horizontal nail holes can be seen penetrating the boards of both sculptures. The Metropolitan has the remains of two large iron nails in the holes. Two similar holes are present in the same location of the Walters buddha’s wood head. These were probably used to hold now-missing halos. All three buddhas have repairs covering the holes. The back of the head is missing from the bodhisattva; therefore, we could not determine whether a halo ever existed.
The hands of all three buddha sculptures are missing. They would have been attached separately and made of either wood or lacquer with wire armatures in the fingers.
Both the Freer and Metropolitan sculptures have irregular holes (approximately 12 centimeters in diameter) in the middle of their backs (fig. S5a–c). These are probably robbing holes for the removal of dedicatory material. The hole in the Metropolitan buddha is off center, toward the right, avoiding the wood armature running up the back.
In the past, carbon 14 dating was carried out on wood from the Freer and Walters buddhas.[xvi] The results are broad (Walters: range of 420–645 CE; Freer: range of 474–574 CE) because the size of the tree and the location where the sample was taken do not provide a date for when the sculptures were made, but rather for some point when the tree was growing. Additional carbon 14 dating on the textile or lacquer would give more accurate dates.
Clay core material
The development of hollow-core lacquer sculptures was an improvement over the wood-core lacquer technique. Without wood, objects became insect impervious, since lacquer is toxic. They were also very lightweight—the Freer and Metropolitan buddhas each weigh only about thirty pounds (13.6 kilograms)—with their hollow bodies providing space for depositing consecratory material. The Walters buddha, on the other hand, weighs more than twice as much.
To begin a hollow-core lacquer sculpture, an artist makes a clay core in the shape of the desired image. In the buddhas and the bodhisattva head, an internal armature likely was needed to support the clay during fabrication. Once a sculpture was nearing completion, the clay core and armature were removed, leaving a hollow shell.
The bodhisattva head is essentially a mask: it is completely open in the back, providing access for study. To determine the level of detail in the head’s original clay core, Smithsonian Exhibits made a 3-D scan of its interior and exterior (fig. S5).[xvii] Then, a positive print of the interior scan was created (fig. S6). It revealed that a surprising amount of detail was carved in the original clay core. Examination of the interior of the Freer and Metropolitan sculptures revealed that their cores were not as detailed, perhaps because of their smaller size, but it was not possible to scan the tight space.
All four sculptures have clay in between drapery folds and/or in noses, ears, lips, and eyelids to form fuller dimensions. Only the bodhisattva head still contains accessible clay clinging to the interior of the face, in the recesses of the lips, chin, and nose. A sample revealed it to be a gray-tan, unfired clay with a fine texture. The clay is uniform in particle size with no added organics visible in SEM images; however, a large carbon peak was present in EDS spectra. The carbon is from the lacquer that was used to adhere the textile onto the clay (fig. S7a–b).
Textile and fibers
Once the clay or wood core of the sculptures was completed, it was covered with cloth strips wet with lacquer. The textile provided the dried brittle lacquer with more flexibility and strength. This made it more durable, helping to prevent loss of the lacquer, especially when the wood core expanded and contracted during changes of relative humidity. However, the movement of the wood does cause the lacquer to crack or fracture, especially over wood joins.
Strips of plain woven textile were dipped in lacquer and placed over the core piece by piece (fig. S8). Small individual strips allowed for more control over shrinkage or stretching of the textile than a larger piece of fabric would have. Cloth layers were built up and, when needed, additional strips were used to create folds in the drapery and enhance other details over the core. Where strip ends could be identified, no selvage was used. The ends and sides of the textile were often already frayed and unraveling when they were applied. Some locations have as many as six layers of cloth; others only have two.
All four sculptures incorporated strips of plain weave textile of varying lengths. All sculptures have an S twist to the fibers. The thread count for each sculpture is:
- Walters: 12 to 16 threads per square centimeter
- Freer: 10 to 15 threads per square centimeter
- Metropolitan: 10 to 12 threads per square centimeter
- bodhisattva head: 8 to 12 threads per square centimeter
The fibers from all four sculptures were identified by polarized light microscopy as bast fibers with crystalline nodes. The fibers’ colors further identified them as hemp when examined under polarized light using the first-order gypsum plate and compared with known samples (figs. S9–S10).[xviii]
In all of the sculptures, the textile was bulked with clay to form naturalistic folds in the robes. This is especially evident in the X-radiographs and CT scan of the Freer buddha (fig. S11). In the Freer buddha, EDS found aluminum between the textile fibers in the weave at the base of the lacquer, confirming that clay was used on the textile.
The structure was built in two phases for the hollow-core lacquer sculptures (Freer and Metropolitan buddhas and bodhisattva head). In the first phase, textile and lacquer were applied to the cores. The tops of the heads were either left open or cut off after this phase cured. Once the textile layers with lacquer were cured, the clay core was removed from the head through the top, providing access to the interior of the face. At this point, the eyes—including the third eye, now missing and filled in on all four sculptures—were set in. The bodhisattva has an additional cloth square applied over the back of the eyes. After work on the interior was complete, phase two began. The top section of the head was reattached or made separately and attached. The separate attachment is clearly seen on the interior of the sculptures and in X-radiographs but well-hidden on the exteriors. Then, more textile was added and, finally, lacquer layers (fig. 4a–c, fig. S12).
The interiors of all three hollow-core lacquer sculptures, originally in contact with the clay core, have a reddish-brown color to their textiles. In the bodhisattva, the reddish-brown material was applied as a liquid and pooled in some areas around the neck. A sample was taken from one of these areas (fig. S13) and the cross section imaged in both binocular and scanning electron microscopes (fig. S7a–b). GCMS identified lacquer and a protein glue in the sample. We found that the sample comprises three layers, which we then analyzed individually with Fourier transform infrared spectroscopy.
The first layer of the sample from the interior of the bodhisattva, the layer closest to the clay core, was the sole source of the protein seen in the GCMS analysis. This layer contains mostly residual clay from the core and soil, as confirmed by EDS, which found elements from clays: aluminum, silicon, magnesium, potassium, calcium, and iron, the source of the sample’s red color.[xix]
The bubbles in the third layer were formed when the lacquer was applied to the textile and air was trapped in between. The small amount of quartz, clay, and other particles found in this layer may have adhered to the lacquer coating before the textile was placed against the clay core. As there is interpenetration between the clay and the lacquer coating layer, the coated textile must have been applied wet.[xx] Certainly, it would have needed to still be wet to flexibly follow the clay core’s contours.
Lacquer on the sculptures
Separate from the lacquer used to apply the textile to the clay core, lacquer was used to form the surface of the sculpture. All four sculptures we studied were made with Toxicodendron vernicifluum lacquer applied alone or with fabric in multiple layers. The lacquer structure in all four sculptures includes five types of layers: first, a lacquer coating on the textile to attach it to the core; second, a thin lacquer layer above the textile coating to seal and smooth it; third, a thick area of coarser material to provide bulk, composed of one to many layers; fourth, a dark lacquer layer; and fifth, final finish layers, either dark or light in color.
A cross-section sample from the Freer buddha shows five lacquer layers above the textile (fig. S15); the same number is present in the Metropolitan buddha (fig. S16top, fig. S16bottom). In the sample from the bodhisattva (fig. S17), five lacquer layers are also present above the textile (layer C), followed by a layer of paint (layer I) and of soil (layer J). The Walters sculpture’s lacquer is more complex, with seventeen layers above the wood core, the most of all the sculptures studied (fig. S14).
But if you look closer at the Walters cross section, there is an unusually uniform and sharp interface between layers F and G (fig. S20top, fig. S20bottom). It is also in a cross section studied in 1993[xxi] that comes from a different location on the buddha. Particles at the interface appear cut, suggesting that it was polished before applying the next layer—a standard practice in lacquer production today. However, if polishing occurred between layers, why is this the only place where we see such a sharp line? There are a few possible explanations. It could indicate that the artists took a break during production, giving layer F time to become harder and causing it to keep its edge. This also would mean that the materials and techniques did not change in the layers above F, which is in fact the case: both the lower layers (F and below) and upper layers (above F) contain Toxicodendron vernicifluum lacquer, cedar oil, tannins, protein, and bone. An exception is the use of plant fibers or sawdust in the thick layer immediately above F; however, similar plant fibers are seen in the other sculptures. There is a difference in technique in the upper layers. Thin, alternating light and dark layers were applied, and these account for the high number of layers on the Walters buddha.
A second explanation for the sharp interface is that it was polished to even out areas of loss and damage and prepare for a later reworking of the surface, with the layers above added during a restoration. The alternating light and dark layers in this case could indicate a restoration. If the upper layers were added later, the three or four layers below the interface (layers D–F, assuming no layers were lost) are fewer than the five layers seen in the other sculptures we studied.
While all four sculptures feature the five types of lacquer layers, there are many differences in how the lacquers were layered, as well as in the added components found in individual layers. The additives and inclusions in each layer are listed in tables 1–4 (table 1 | 2 | 3 | 4 ) and explored below.
Samples were removed from areas of loss on the four sculptures. Sampling location may have affected the results: different materials may be present depending on the area from which the sample was taken. For the Walters buddha and the bodhisattva, the samples were taken from areas of “flesh,” such as the upper back, while the samples removed from the Metropolitan and Freer buddhas were removed from the left drape edge.[xxii]
Lacquer additives: bone
When exploring the components of the lacquer, the most prevalent inclusion in all four sculptures is ground, partially burnt bone (fig. S19). It was clearly used as a filler to bulk up the lacquer and to form a paste. The bone particles were visible using microscopy, and their identification was confirmed by EDS in a scanning electron microscope.
X-ray maps, collected with EDS by scanning over the sample, show the distribution of different elements and were used to determine which layers contained bone. Figures S18a–d show X-ray maps for calcium, phosphorous, and silicon for the bodhisattva head sample. The overall brightness of the calcium and phosphorous maps indicate that finely ground bone is present throughout the cross section, and the large bright areas in the middle and near the surface of the maps are large bone fragments (up to 20 micrometer diameter) included in layers E and G. The sharp-edged particles of ground bone vary from light to dark in color depending on the degree to which the bone was burned.
In all of the sculptures, the bone fragments near the textile substrate are generally small, as would be needed to allow the lacquer to fill in holes and flaws in the textile and form an even surface. In the middle layers, the bone fragments are larger, as their primary purpose is to bulk up the lacquer. Toward the top surface, the bone fragments again decrease in size and amount as necessary to form a smooth surface that could be polished. The bone near the surface tends to be more uniformly burnt and black in color, possibly to blend with the darkened color of the lacquer that was colored by tannins and/or soot.
Ground burnt bone, although not common, has been used to bulk Chinese lacquer at least since the Warring States period (475–221 BCE): it was found in the lacquer layers of a cart in a tomb burial dating to that time.[xxiii] Bone could be ground into a variety of sizes, from coarse to fine particles. Bone powder particles are nonabsorbent, lightweight, and nonreactive to the lacquer resin.
Bone is composed of roughly 75 percent inorganic material and 25 percent organic material. Burning the bone would remove much of its organic components. However, residual proteins or other organics could help bond the bone to the lacquer matrix. It is clear that as one of the main ingredients, bone helped add body to the lacquer, creating a dough-like paste that made it easier to apply to vertical surfaces.
What kind of bone was used in these sculptures? Was it animal or human—perhaps the cremated remains of a monk? Attempts were made to answer these questions with DNA analysis. A sample of the Freer buddha was given to Robert Fleischer, research zoologist at the Smithsonian’s National Zoo’s ancient DNA laboratory. Unfortunately, he could not get a result, as the lacquer has its own DNA and the bone was partially burned, destroying most of the organic remains.
The protein in the lacquer was then analyzed via proteomics to determine the bone species. Timothy Cleland, physical scientist at the Smithsonian’s Museum Conservation Institute, ran the analysis and determined that the major source of bone protein in the bodhisattva lacquer is equid (horse or donkey), not human. However, since it is not possible to separate the bone from the remainder of the lacquer, further research needs to be done to determine whether other protein materials in the lacquer, such as animal glue, affected the proteomics results. Cleland also analyzed the source of bone protein in the Freer buddha and found it to be bovid (cow).[xxiv]
Lacquer additives: blood
A combination of cholesterol, protein markers, and trimethyl phosphate has been found in blood additives to lacquer.[xxv] Blood protein markers and trimethyl phosphate were seen in two of the lower layers of the Metropolitan buddha’s lacquer and in the upper layers of the Walters buddha. Protein markers for blood were also seen in a ground layer of the bodhisattva and possibly of the Freer buddha. It was also seen in two interior layers of the Walters buddha.
Blood may have been introduced as part of the bone or as a separate intentional additive. It was probably used as a binding medium. Blood is mentioned in several Chinese texts from as early as the Yuan dynasty as an additive for ground layers, and DNA analysis has revealed both pig and cattle blood in a ground layer of a tea box dated to 1820–50.[xxvi] [xxvii]
Through proteomics, Timothy Cleland found human blood in a sample from the bodhisattva. The sample contained all layers, so we do not know how the blood was added. This was unexpected and is still being explored.
Lacquer additives: tree resins
One organic additive to the lacquer formulation was a resin from trees from the family Cupressaceae or other fir trees, termed cedar oil here. This has been found in most lacquer layers of the Walters and two layers of the Metropolitan buddha. Several sources and forms of the material fit the chemicals identified by GCMS, so a specific source cannot be determined. There are several reasons why it may have been added: to act as an extender to cut the cost of the lacquer; to affect the physical properties—either the working properties, such as ease of application or drying time, or visual properties, such as increased gloss; or possibly as a preservative in the raw lacquer, as cedar oil has antimicrobial properties.[xxviii] A second resin, gum benzoin, was found with the cedar oil in the textile layer of the Metropolitan buddha.
Lacquer additives: oil
Oil was present as an additive to the lacquer in all four sculptures. It was not possible to identify specific oils used because, with the exception of tung oil, the fatty acids from the lacquers, bone, and waxes interfere in oil identification. However, the Metropolitan buddha’s lacquer likely has heat-treated (heat-bodied) oils.[xxix] Either cold-pressed or heat-bodied oil has been found in all periods of China. An early mention of heat-bodying can be found in a Northern Song dynasty document that mentions high and low temperature heating of tung oil.[xxx] Heat-bodying partially polymerizes the oils prior to use, resulting in a thickened oil that has lower shrinkage and is more durable after drying.[xxxi]
Lacquer additives: plant materials
Plant materials (such as sawdust) as well as small amounts of quartz and other silicates also are found in the lacquer layers.
In all four sculptures, several lacquer layers stained positive for starch, and starch was identified by GCMS. In some cases, positive staining for starch occurred specifically at the edges of fibers. This occurred in the textile layer of the Freer buddha (layer C) as well as interior layers with fiber pieces in the other sculptures. In the textile, starch at the fiber edge may have resulted in increased stiffness, as seen in a starched shirt today. More relevant for the cut fiber pieces, the starch may help strengthen the bond between fiber and lacquer, reducing the possibility of cracking. Starch could also have been used to thicken or provide more tackiness to the lacquer paste, or it could be coming from fibers or rice husks used as bulking materials in the layers. Plant materials weigh less than bone or silicates and would result in decreased weight—a plus for a self-supporting hollow sculpture, especially if portability was a goal.
Tannins, brown to black colorants that can come from several plant sources, were found in the Walters and Metropolitan buddhas and the bodhisattva. In addition, one marker compound for soot was found in upper layers of both the Walters and Metropolitan buddhas. At times, markers for compounds associated with specific plants used as dye sources (young fustic and old fustic) were seen in the GCMS results; however, further research is needed to connect these markers solely with these plants.
Lacquer additives: glue?
The bodhisattva, the Freer buddha, and possibly the Walters buddha include layers, often lighter in color than the lacquer layers, where no Toxicodendron vernicifluum lacquer was found, but that contain proteins, most likely acting as an adhesive. Early texts advise the addition of protein glue to increase the adherence and durability of so-called ground layers. Markers for protein glue are also found in the lacquer layers of all four sculptures but the Freer buddha. However, while the bone fragments that are visible in the cross sections in many cases stained positive for proteins and are one source of the compounds seen with GCMS, glue cannot be ruled out as a component of the lacquers. Starting in the Northern Song dynasty, there are references to mixing lacquer and glue.[xxxii] Further research is needed to distinguish markers for bone from those for glue, as protein glue can be made from either boiled skin or bone.
Lacquer additives: wax?
We found remnants of prior conservation treatment in the analysis of the Walters buddha. Every layer contained elemi resin and beeswax. These materials were used along with paraffin wax on the entire sculpture to secure the flaking painted surface decoration. The museum’s records of the treatment[xxxiii] were critical in allowing us to separate the conservation materials from the original lacquer components. Still, it is not possible to tell if any waxes were among the original materials due to the presence of waxes from conservation treatment. The presence of the wax also interferes with our ability to identify the type of any oils added during production.
In the Freer buddha, wax was found in two interior layers as well as the textile layer. Since it is only in some of the layers, and those are on the interior, the wax may be original and not from a later conservation treatment.
Lacquer additives: miscellaneous materials
In addition to the materials discussed above, we identified a few miscellaneous materials in the sculptures. Cellulosic materials were found in the Walters buddha, the Freer buddha, and the bodhisattva, likely from chopped fibers or the hemp textile. Indigo was found in the Walters buddha.
Py-GCMS allowed us to identify many of the components added to the lacquer. However, there are others that could not be identified, as too little is present or their marker compounds are still unknown.
FINDINGS: MISCELLANEOUS MATERIALS
The eyes of all four images are similar in that they are white with black, shiny pupils. The Walters buddha’s eyes have both been replaced with restored eyelids. XRF analysis confirmed that they are calcium-rich, but not glass.[xxxiv]
The eyes of the Freer buddha consist of black pupils set into the head with a white surround of clay or lacquer putty. The right eye was repaired at one time, but both pupils look to be the same material. It is not completely opaque to X-rays. XRF on both eyes determined the eyes are glass with copper and iron present but no significant barium.[xxxv] The eyes have a dense ring around them, but are darker in the center. The ring is caused by the X-rays going through the curve of the sphere.
The Metropolitan buddha’s left eye is original; the right eye is restored, as confirmed by XRF analysis. The left eye is made of leaded glass with some copper and iron present and no significant barium, similar to the eyes of the Freer and the bodhisattva head. It also has a stem out the back, similar to one of the bodhisattva’s eyes. The left eye’s white pigment was identified as primarily muscovite and quartz.[xxxvi]
The bodhisattva’s eyes are solid, dark-green spheres, one of which has a stem extending out the back. Bubbles help confirm they are made of glass. The bubbles are round and not stretched, so they were not made from a pulled glass rod (figs. S21–S22). XRF analysis confirmed they comprise leaded glass with some copper and iron present with no significant barium, similar to the Metropolitan buddha’s one original eye.[xxxvii] There is no evidence of restoration; therefore, both of the bodhisattva’s eyes must be original.
The ears of all four sculptures are constructed the same way. The earlobes are formed of U-shaped iron wires with a square cross section. Although the Freer buddha has lost his earlobes, the remains of iron wires can be seen in X-radiographs. The Walters buddha has also lost his earlobes, but while there is no visible or radiographic evidence, wire likely was used in them as well. The Metropolitan buddha still has earlobes, and the wires are visible in X-radiographs. The bodhisattva also still has his earlobes. From the interior, the ends of the thick iron wires can be seen extending into the head (fig. S4a–c). Clay was applied over the wires of both the Metropolitan buddha and the bodhisattva, with multiple lacquer layers covering the clay.
After the multiple layers of lacquer were applied and cured, the sculptures were painted in realistic colors and designs. All four sculptures were originally painted following the same color scheme and design. Today, much of the pigment that remains is barely visible, and the designs are highly fragmented. Often, sculptures were restored or repainted multiple times during their service; therefore, it was not possible to characterize the remaining pigment on each sculpture to the same extent as other components.[xxxviii]
The flesh areas of all four sculptures were gilded over a pink paint layer on a white ground. The ground has been identified as lead white on the three life-size sculptures, and the pink flesh color is a mixture of red lead and vermilion. Azurite blue was found on the hair of all four sculptures and identified by PLM and XRF analysis. The lips of all four have traces of vermilion red.
The Metropolitan buddha retains the most color. An altered image illustrates his reconstructed appearance. His shawl is a patchwork in red squares, with wide, dark-purple edging. The robe has an elaborate cloud pattern border in alternating blue and green on a red background. A number of the colors were analyzed by X-ray micro-diffraction and XRF. The purple on the shawl has been identified as hematite, the orange-red as a mixture of red lead and cinnabar, the green as malachite, and the blue as azurite.
The Walters buddha’s shawl has a painted red and green patchwork design. The robe is very indistinct. At the midriff, the remains of a central circular pattern can be detected. A polychrome band on the upper edge of the robe is painted orange, red, white, green, and blue (indigo). The background of the robe is blue and green, as is the central circular design. The skin has a kaolin and calcium-rich ground with a pink tone mixed of white and red lead. Traces of bright-blue azurite mixed with lead white, together with some tenorite (black), were found around the hairline (fig. S23).
The Freer buddha’s shawl also has a painted red and green patchwork design; however, its color is much less distinct than the Metropolitan buddha’s. The robe also has orange-red squares with black bands around the edges. The blue pigment was identified microscopically as azurite by Elizabeth Fitzhugh.[xxxix] A ground layer under the gilding of the flesh was identified as lead white.[xl] Visual evidence confirms at least three repaints.
The bodhisattva has traces of pigment similar to those on the three life-size sculptures. There is blue azurite in the hair, identified by PLM. The red in the lips was identified as vermilion by XRF. Traces of gilding remain on the “flesh” around the ears over a white ground.
We have confirmed that the sculptures were made of expensive materials and that they were time-consuming to create. Though they were fabricated over a two- to three-hundred-year period, between the sixth and eighth centuries, the same traditions and similar techniques and materials were used in all four sculptures. While the design styles of the sculpture may have changed over time, the techniques did not.
We found that the techniques used were much more complicated than those used to create utilitarian lacquer objects. Along with the high quality of the materials and the rarity of the objects, this suggests that the four sculptures were fabricated in workshops that received imperial patronage or support from an elite group. We do not know the lacquer workshops’ structure, however, or whether workshops operated out of monasteries or were private industries.
Three of the four sculptures originally had clay cores. All four have a similar plain woven hemp textile for the base and support of the lacquer layers. All use the same species of lacquer, with no imported lacquer species from elsewhere in Asia. And all of their lacquer layers begin with a coarse lacquer layer, followed by finer layers and, finally, several layers of the finest lacquer on top.
Partially burnt bone was used in all four sculptures to bulk up the lacquer. The bone analyzed in two sculptures, the Freer buddha and the bodhisattva head, was revealed to be not human but, as was found earlier in China, animal.
All the sculptures used the same pigments, regardless of variations in designs. The flesh of all four was gilded.
Future studies should be carried out on later period sculptures to extend the research and knowledge gained in this project.
Summary of contents: Toxicodendron vernicifluum. Bone, a drying oil, a wax, starch, soot, and blood (possible identification by GCMS confirmed by proteomics) were found.
|Layer||Name||DescriptionMaterials identified (GCMS)|
|F||Dark brown lacquer 3||Dark brown lacquer. Contains a few bone particles, mostly black, up to 5 µm in size; a few white.||Toxicodendron vernicifluum. Oil, indeterminate protein.|
|E||Fine brown lacquer 2||Fine brown lacquer. Contains few bone particles. Crack goes through center. Possible positive staining results for oil/protein.||Toxicodendron vernicifluum. Oil, indeterminate protein.|
|D||Dark brown lacquer 1||Thick, dark brown layer, approx. 30–40 µm thick. Contains many white and black bone particles, small (2–5 µm) and large (approx. 10 µm). Contains fibers or second textile layer. Positive staining results for oil/protein in particles. Positive staining results for starch in edge around fibers.||Toxicodendron vernicifluum. Oil, wax. Indeterminate protein, starch, tannins.|
|C||Ground layer||Light colored layer, 10 µm thick. Contains many white bone particles approx. 5 µm thick. Positive staining results for oil/protein in particles.||Wax. Indeterminate protein.|
|B||Dark brown layer||Dark brown lacquer layer, approx. 15 µm thick. Contains black bone particles, small (2 µm) and large (10 µm). EDS: Ca, positive staining results for starch.||Bone or glue, possibly blood.|
|A||Textile and lacquer||Textile identified as hemp with fiber microscopy. Lacquer in between fiber bundles contains black particles approx. 5 µm thick, some of which appear white or gray in UV. Positive staining results for oil/protein in the textile and a few small particles in the lacquer. Positive staining results for starch in edge around textile fibers.||Toxicodendron vernicifluum. Oil. Wax. Indeterminate protein. Starch, carbohydrates from lacquer, and 3 unverified markers for “paper.”|
Summary of contents:
Toxicodendron vernicifluum. Bone, gum benzoin, oil (possibly heat-bodied), protein from bone or glue, wax, starch, cedar oil, tannins, and possible blood were found.
|Layer||Name||Description||Materials identified (GCMS)|
|J||Soil||Dark surface layer, darker than layer H. EDS: Al, Fe, K, Mg, O, Si, Ti, low C.||H–J: Toxicodendron vernicifluum (from layer H). Indeterminate protein. Fatty acids from lacquer and possibly bone and/or oil.|
|I||Paint layer||Thin layer appears colorless under visible and UV. EDS: Al, K, Si.|
|H||Upper lacquer 1||Dark, thin layer; no visible inclusions. EDS: S.|
|G||Upper ground 2||Porous brown layer up to 10–20 μm thick. Appears gray in UV. Contains bone particles that are dark in visible light, but some appear white to yellow in UV light. Particle sizes are smaller than in layer E or F, and many have an elongated shape. Many particles in 5–10 μm range, but smaller are present. Some calcium sulphate present. EDS: Ca, P, S. Positive staining results for protein (weak staining of matrix and discrete particles), starch (weak staining), oil/protein (particles and possibly matrix), and possibly oil/lipids.||Indeterminant protein. Monocarboxylic fatty acid.|
|F||Dark brown lacquer layer||Dark and gray brown layer that ranges from 10–20 μm thick. Most particles are dark in visible light, with one large white one at right. Under UV, some particles remain black, some are white to light gray; the large particle appears yellow. EDS: higher Si. Positive staining results for oil/protein (in particles only).||Proteins from bone or glue. Fatty acids from bone and/or oil.|
|E||Upper ground 1||Reddish brown layer up to 40 μm thick, contains black and gray bone particles with rounded edges that remain dark in UV, gray bone particles that lighten in color in UV, and gray to white bone particles that appear white to yellow under UV. Many particles are very large, with several up to 20 μm longest dimension. EDS: Ca, Positive staining results for protein , starch (darkening visible in UV), oil/protein (green cast and blue particles in VIS) and oil/lipids.||Proteins (indeterminate). Fatty acids from bone and/or oil.|
|D||Brown lacquer 2||Dense brown layer with white and black particles ranging in size from a few micrometers to 15 μm in largest dimension. There are many voids in the layer, and it appears to have the remains of fiber cross sections that have been filled with the lacquer. Positive staining results for protein (weak staining), oil/protein, and oil/lipids (layer appears a uniform brown color except for fiber cross sections, which remain yellow).||Toxicodendron vernicifluum. Drying oil, possibly heat-bodied. Protein from bone or glue. Tannins.|
|C||Textile||Textile, bundles of fibers from 10–20 µm diameter are white under visible light and yellow under UV. Identified as hemp via fiber microscopy, EDS: O. Positive staining results for starch, oil/lipids. High oil/protein content delineates individual fibers. Oil/lipids seem to be in the space between fibers, possibly in lacquer saturating the fibers.||Oil, starch and paper markers likely from textile. Tannins. Indeterminate protein.|
|B||Brown lacquer||Dense brown layer with approx. 10 µm particle that is dark in both visible and UV light. Layer appears to fill interstices of woven textile. Positive staining results for oil/lipids.||Oils. Protein from bone and/or glue. Tannins. Minor amount Toxicodendron vernicifluum.|
|A||Ground||Not in cross section.||Toxicodendron vernicifluum. Oil. Protein from bone or glue, possible blood. Tannins.|
Summary of contents:
Toxicodendron vernicifluum. Bone, gum benzoin, oil, starch, cedar oil/pitch, tannins, blood, and possibly soot were found.
|Layer||Name||Description||Materials identified (GCMS)|
|H||Upper lacquer layer 3||Approx. 20 µm black surface layer of lacquer mixed with white and black bone particles of size 5–8 µm. White particles are fewer and smaller than black. It appears brown in UV. Surface is rough, and upper portion may be missing. Layer stained positive for oil/protein. A few particles stained positive for starch.||Inconclusive|
|G||Upper lacquer layer 2||Thin (5 µm or less), light colored layer appears red-brown in UV with very fine black and white bone particles (1 µm). Positive staining results for oil/protein, starch, and oil/lipids. Positive staining results for starch.||Toxicodendron vernicifluum. Oil, indeterminate proteins, possible tannins, starch, possible soot.|
|F||Upper lacquer layer 1||Black lacquer mixed with white and black bone particles of size 5–8 µm. More white particles than in layer G. It appears brown in UV.||Toxicodendron vernicifluum. Cedar oil/pitch. Oil and lacquer fatty acids. Protein from bone or glue.|
|E||Fiber and lacquer||Very thick layer (>100 µm) of reddish-brown lacquer with textile. Some of the textile fibers are solid, some appear filled with lacquer. Large number of pores throughout. Positive staining results for starch in layer around fibers.||Toxicodendron vernicifluum. Cedar oil/pitch. Oil. Bone or glue protein. Possible blood. Marker for starch (possibly from textile). Carbohydrates from lacquer. Tannins.|
|D||Lower lacquer layer 1||Thin brown layer with small black bone particles a few micrometers in size. Sharp edges and elongated shape as well as size range indicate grinding. A few light to white-colored particles are more equiaxed in shape. Some of these remain white in UV, while some appear dark. Possible positive staining results for oil/lipids. EDS: Ca and P.||Toxicodendron vernicifluum. Oil, possibly heat-bodied. Blood. Protein from bone or glue. Carbohydrates from lacquer. Tannins. Indeterminate resin, possibly cedar.|
|C||Textile||Multithread woven textile appears tan to brown in visible light and white in UV. Weave is visible as fibers in sample are in two perpendicular directions. Impregnated with brown (yellowish-brown in UV) material. Positive staining results for starch (weak staining) and oil/lipids.||Toxicodendron vernicifluum. Oil. Protein. Tannins. Possible blood. Carbohydrates from lacquer. Gum benzoin.|
|B||Black lacquer||Thick brown lacquer, not in cross section.||Toxicodendron vernicifluum. Indeterminate resin, possibly cedar. Oil. Tannins. Possible blood. Starch. Protein from bone or glue.|
|A||Underlying layers||Not sampled||Not analyzed.|
Summary of contents: Toxicodendron vernicifluum. Oil, tannins, starch, cedar oil, possible soot, “paper,” possible blood, and indigo were found. Protein occurred in all layers as determined by staining the cross section. Soot was possibly found in the surface layer, either from soil in layer T or a pigment in layer R. Starch and paper markers were found in lower lacquer layers, consistent with the chopped fibers seen in the cross section that added bulk to the layers. Elemi resin and beeswax from a previous conservation treatment.
|Layer||Name||Description||Materials identified (GCMS)|
|T||Surface dirt||Dark, uneven, sporadic, <1 µm. Positive staining results for oils/lipids.||N–T: Toxicodendron vernicifluum. Elemi resin, oil. Starch, possible soot. Bone or glue. Beeswax, carbohydrates from lacquer.|
|S||Coating 3||Translucent white layer, approx. 2 µm thick. Possible weak staining results for oils/lipids.|
|R||Red pigment||Red pigment in white binder, approx. 2 µm thick. Possible weak staining results for oils/lipids.|
|Q||Coating 2||White dense layer, approx. 2 µm thick. Possible weak staining results for oils/lipids.|
|P||Coating 1||White porous, thin layer, approx. 2 µm thick. Possible weak staining results for oils/lipids.|
|O||Lacquer 8||Dark colored layer, approx. 10 µm thick. Dark bone particles, some appear gray or white in UV (5–20 µm). Positive staining results for oil/protein in particles.|
|N||Upper ground 4||Light colored porous layer, approx. 5 µm thick. Contains many 5 µm white bone particles that remain white in UV and occasional large dark particles. Positive staining results for starch; positive staining results for oil/protein in matrix and white particles.||M and N: Toxicodendron vernicifluum. Elemi resin with some unassigned compounds. Oil. Blood. Beeswax . Tannins.|
|M||Lacquer 7||Dark colored layer, 2–3 µm thick. Black and white bone particles, small (2 µm) and large (10 µm). Some darken, some lighten in UV. Positive staining results for oil/protein in particles; positive staining results for oils/lipids.|
|L||Upper ground 3||Light colored porous layer, approx. 5–10 µm thick. White and black bone particles (2–5 µm).||Toxicodendron vernicifluum. Elemi resin. Oil. Beeswax . Indeterminate proteins.|
|K||Lacquer 6||Dark colored layer, 2–3 µm thick. Dark bone particles in two size ranges: approx. 2 µm and 10–15 µm. positive staining results for oils/lipids.||Toxicodendron vernicifluum. Elemi resin and cedar oil/pitch. Oil. Beeswax. Bone or glue proteins . Tannins. Carbohydrates from lacquer. Indigo.|
|J||Upper ground 2||Light colored layer approx. 10 µm thick. White and black bone particles (2–5 µm and occasionally up to 20 µm).||K and J: Toxicodendron vernicifluum. Cedar oil and elemi resin. Oil. Beeswax . Indeterminate protein. Tannins. Carbohydrates from lacquer.|
|I||Lacquer 5||Dark layer approx. 10 µm thick. Mostly dark, bone particles appear black to gray in UV. Positive staining results for oil/protein in particles; positive staining results for oils/lipids.||Toxicodendron vernicifluum. Elemi resin and cedar oil. Oil. Beeswax. Tannins. Indeterminate protein. Carbohydrates from lacquer.|
|H||Upper ground 1||Thin, light layer approx. 2 µm thick, occasional bone particles approx. 2 µm. Positive staining results for oil/protein in particles; positive staining results for oils/lipids.|
|G-3||Lacquer 4||Dark layer, 10 µm thick. Contains randomly oriented fibrous material (likely sawdust). Large bone fragment approx. 20 µm extends through G1–3.||Toxicodendron vernicifluum. Elemi resin and cedar oil/pitch. Oil. Tannins, trimethyl phosphate. Bone and/or glue protein. Supplemental protein markers include one for blood. Beeswax. Carbohydrates from lacquer. Glucoside compound, possibly from sawdust.|
|G-2||Ground 2||Thin light colored layer, 2 µm thick, in left of cross-section, small white and black particles 5 µm. It is part of one larger layer (G). Area at surface of fiber(likely sawdust) positive for starch.||Toxicodendron vernicifluum. Elemi resin, cedar oil/pitch. Beeswax. Indeterminate protein. Starch. Tannins.|
|G-1||Lacquer 3||Dark layer at left of cross section only, less than 5 µm thick, contains randomly oriented plant fibers (possibly sawdust). Area at surface of fiber stained positive for starch.|
|F||Lacquer 2||Dark layer, 15–20 µm thick, first layer under horizontal line that indicates a span of time occurred between layers F and G. Positive staining results for oil/protein in particles and around particles.||E and F: Toxicodendron vernicifluum. Elemi resin, cedar oil/pitch. Beeswax. Bone or glue. Supplemental protein markers include one for blood. “Paper”. Tannins.|
|E||Lacquer 1||Lighter, more porous lacquer layer; has fine, reddish layer at top that may be a thin, separate layer. Positive staining results for oil/protein in particles and around particles.|
|D||Ground 1||Large particles.||Not analyzed|
|C||Textile and lacquer||20 µm thick layer. 5–20 µm particles, white and black. Appear white and shades of gray in UV.||Not analyzed|
|B||Lower lacquer||Textile identified by fiber microscopy as hemp. Lacquer intrudes into textile and contains black particles that appear white and shades of gray in UV.||Not analyzed|
|A||Textile 1||Not analyzed|