IV. Technical Study of a Tibetan Five-Panel Initiation Crown

Introduction

In the spring of 2021, a Five-Panel Initiation Crown (2018.112) (Fig. 1), believed to date to the eighteenth century, was brought to the Sculpture and Decorative Arts Conservation Lab at Virginia Museum of Fine Arts (VMFA) for examination and treatment prior to display. The crown entered the VMFA’s collection in 2018 as part of a large bequest from the estate of Dr. Mary Shepherd Slusser, a preeminent Western scholar of Himalayan art with a research focus on Nepalese material culture. During her time in Nepal, she acquired a large collection of paintings, works on paper, and three-dimensional objects, but unfortunately among the records accompanying the bequest, there was no information on where or when she purchased this particular object.

The VMFA’s Five-Panel Initiation Crown consists of five flat (~2 mm thick) panels, each shaped as an elongated pentagon with a scalloped edge along the top, which is a shape that mimics the negative space under the torana, an arched gateway that leads to Buddhist temples or shrines.¹ The front of each panel depicts a seated figure at the center representing one of the Five Cosmic Buddhas. The panels are primarily decorated with red paint and gilding, though there are large areas on the fronts that appear black. Fine details on the figures are rendered in black ink and the panels are each coated overall in a thick reticulated orange coating. The edges of the panels are smooth and coated continuously in red paint from the front to the back side. The back of each panel is painted red with a gilded symbol at the center. The backs of the panels are partially coated, and the coating is most visible over the area of gilding. At one point the panels were joined with a long tie made of folded silk which was stitched with silk thread through holes in the bases of the panels. The tie has since been cut or otherwise broken, separating the panels.

A significant amount of treatment was anticipated for the Initiation Crown. The five panels, once joined together along their bottom edge with a ribbon, had become separated and were adhered to a fabric-wrapped board. Additionally, a translucent orange coating covering the front of the panels had developed a distracting craquelure pattern, making it difficult to read the designs (Fig. 2). Initial examination also revealed the panels’ substrate material, which had been misidentified in the acquisition documents as leather, was made of laminated textile. With limited and incorrect information in the museum database and acquisition records, a literature review was conducted to gather information on traditional fabrication methods used for this category of object. While a plethora of literature is available on Tibetan religious and secular material culture, such as thankga paintings, wooden furniture, Vajracharya crowns, and various other ritual objects, very little technical information was found to have been published on this type of Five-Panel Initiation Crown.² Therefore, an in-depth technical study was carried out to more fully characterize the materials and techniques used to construct the VMFA crown and to help inform its treatment.

Analysis was carried out using a variety of techniques available at the VMFA and the analytical lab at the Colonial Williamsburg Foundation (CW). At VMFA, the surface of the crown was examined under the stereomicroscope, and samples (fibers and pigment) were removed for analysis with polarized light microscopy (PLM) and Fourier-transform infrared spectroscopy (FTIR). Cross-sections were also taken for microscopic examination, and elemental maps of each panel were generated using a scanning micro-X-ray fluorescence spectrometer (µ-XRF). At CW, cross-sections were re-photographed and further analysis of decorative layers was carried out with scanning electron microscopy – energy dispersive spectroscopy (SEM-EDS).

An effort is made in this article to provide the Tibetan terminology (in transliterated Lhasa Tibetan) wherever possible for materials and techniques. When a material or technique is introduced, the Tibetan term is used first, with the English in parentheses. After that, where an equivalent or more specific English term is available, it will be used in order to more clearly communicate the findings of this research with the audience of this publication. The authors would like to acknowledge that their understanding of Tibetan culture and language is imperfect and incomplete but an effort was made to learn as much as possible over the course of this research and to include the Tibetan terminology in order to show respect for the culture of origin. The Tibetan words also provide specificity in relation to the artist’s materials and practices. A glossary of applied terminology is provided at the end of the article. Most translations are from Tibetan Thangka Paintings: Methods & Materials by David P. and Janice A. Jackson and The Rangjung Yeshe Tibetan-English Dictionary of Buddhist Culture, compiled by Erik Pema Kunsang.

Five-Panel Initiation Crowns

This type of Initiation Crown depicts the Five Cosmic Buddhas, often also called the Buddhas of the Five Families, the Five Symbolic Buddhas, the Five Jina Buddhas, and the Dhyani Buddhas. Such crowns are worn during specific rituals and initiations by lamas (Tibetan monks) who wish to embody the perfect wisdom and positive qualities of the Cosmic Buddhas (Fig. 3).³ Each panel on the crown represents a different Cosmic Buddha, Akshobhya, Vairocana, Ratnasambhava, Amitabha, and Amoghasiddhi, each of which can be identified according to his associated implement or emblem, mudrā or hand gesture, and throne adornments.⁴ Centered on the back of each panel is the seed syllable that corresponds with the deity represented on the front (Fig. 4). The seed syllables are sacred Buddhist syllables written in Lantsha script that represent the core attributes of each Cosmic Buddha. Practitioners aspiring to enlightenment connect with the spiritual principles of each syllable through visualization, meditation, and recitation.⁵

Comparable crowns can be found in many different museum collections in various states of preservation. While these crowns range in age and material, they share the same basic form of five torana-shaped panels joined at their bottoms with ribbon or tape, each panel depicting one of the Five Cosmic Buddhas. These are often made from tshag ‘debs (hammered metal sheets), tshem drub (embroidered fabric), lugs shog (papier mache), shing brkom rgyag (carved wood), or so hrug (leather), and would often be decorated with two dar dpyangs (hanging ornamental sashes) that drape from each side of the crown on the sides of the wearer’s face.⁶ These crowns are not considered a ritual object on their own, but only when worn during a ritual context.⁷ They are similar in ritual use to the Vajracharya Crowns used in Newar Buddhist traditions in Nepal, but differ drastically in form (Fig. 5). The Vajracharya Crowns also depict the five Cosmic Buddhas, but take the shape of helmets, and are typically made of metal with inlaid stones and bone ornaments.⁸

There are a number of similar five-panel initiation crowns currently in US-based collections that are reported to be made from a variety of different materials. The VMFA collection contains one other example of a five-panel initiation crown (89.25), however, this second crown is made from silk embroidery mounted on cardboard panels and secured at the bottom with a long black silk tie (Fig. 6). There are several crowns that are described in their museum’s collections database as watercolor on paper board, such as those in the collection of the Philadelphia Museum of Art (1963-63-5D) and the Met’s ritual crown (1985.391). An example from the Walters Art Museum (86.39) is embroidered silk with gold and silver leaf on paper board. Another example in the Minneapolis Institute of Art (Mia) collection (98.52), dated to the eighteenth century, appears most similar to the subject of this technical study. Both are decorated with raised gesso and have an aged and darkened varnish. Although the Mia collections database website describes their crown as leather with polychrome lacquer, if the object originated from Tibet then Asian lacquer is unlikely to have been used.⁹ All of these examples are brightly colored, in keeping with the Tibetan aesthetic preference, which contrasts with the relatively subdued color scheme of the VMFA’s crown that is the focus of this technical study.

Most of the crowns found in US-based collections are dated to approximately the eighteenth and nineteenth centuries. The oldest Tibetan five-panel crown currently known to the authors in a US museum collection is the crown mentioned above at the Metropolitan Museum of Art, which has been dated to the late fourteenth or early fifteenth century. However, it is often unclear how historical research or technical investigations informed these dating estimations.

Relationship to other Tibetan art forms

The construction and materials of the VMFA’s Five-Panel Initiation Crown (2018.112) are strongly related to other forms of Tibetan art and material culture, in particular the traditions of thangka painting and furniture decoration.

In many ways, the paint colors and method of application in the crown resembles that of painted Tibetan furniture from the same period. While the substrate of the crown panels is fabric rather than wood, the ground and raised decoration are built up in a similar manner to furniture and decorative arts objects. For example, a technique common in furniture decoration that was also implemented in the creation of this crown is the use of kyungbur, a thick gesso-like material that is piped onto a surface of the ground to create raised lines, dots, and patterns, most commonly found on furniture.¹⁰ The background of the object is painted in red, which was also a popular color choice in the region for many pieces of furniture.¹¹ This color was achieved using cinnabar/vermilion, which was available in Tibet, China, and India.¹² Vibrant colors were preferred, and furniture, architecture, and other decorative arts reflected this aesthetic. Such items were typically painted in shades of bright red, blue, green, yellow, orange, purple, and often with gilding to emphasize applied designs or embellishments.¹³

The crown also bears material resemblance to thangka paintings, which have been well documented in English language art historical and conservation literature. While many thangka painters were, and remain to this day, highly specialized and extensively trained artisans, and likely would not have been involved in the manufacture of this crown, the object shows some material similarities to thangka paintings such as pigments and substrate. However, the underdrawing on the crown panels, where visible, does not show the extensive layout and mapping that thangka paintings are known to have.¹⁴ This suggests that the crown was likely not made by the same highly-specialized artisans who created thangka paintings.

Substrate

The substrate of the VMFA’s Initiation Crown, originally described in the acquisition documents as leather, is constructed of plain woven textiles that are laminated together. The laminated fabric was cut into torana form shapes to form the panels. The exposed edges that are visible are too frayed to accurately assess how many layers of textile have been utilized (Fig. 7). Small samples of the textile fibers were taken from the exposed proper left corner of Panel 2,¹⁵ and were then identified through polarized light microscopy (PLM) as cotton fibers. Cotton was readily available and used in Tibet for the decorative arts, such as painting thangkas, as well as other utilitarian purposes. Historically, the cotton used in Tibet was imported from India, though cotton imported from China has also been found.¹⁶ Cotton imported from India before the nineteenth century was expensive, only becoming cheaper and more accessible later on.¹⁷

Results from FTIR analysis indicate that the textiles were likely adhered together with gelatin, which was readily available and in common use in the eighteenth century as an adhesive, size, and paint binder.¹⁸ The textile substrate was found to fluoresce a bluish-white under ultraviolet light, but there was little accumulation of adhesive on the surface or visible agglomeration between layers at the areas of loss.

Ground and Kyungbur

The ground layer, ‘dam, is painted thinly over the laminated fabric substrate. It has a dark cream color consistent with some thangka painting traditions. As Jackson and Jackson point out in Tibetan Thangka Painting,

It should be noted that some painters, particularly those from Eastern Tibet, preferred a gesso that had a slightly ochreish [sic] hue. To achieve this they added a small amount of ochre or yellow pigment to the gesso…. It was particularly effective there because these techniques used thin washes and the minimal application of paint. The ground itself showed through in places and a plain white ground would have been glaring and unpleasant.¹⁹

The primary bulking agent in ‘dam is ka rag, meaning either chalk or kaolin, both of which can be mined in Tibet.²⁰ Several instances of ro nye (lead white) as a bulking agent have also been recorded, generally on objects produced after the mid-nineteenth century.²¹ In the tradition of furniture making, ‘dam and kyungbur were both generally bound with animal glue with potential additives like bu ram (sugar), zar khu’i snum rtsi (linseed oil), and ka pi ta (resin) to modify the working properties and prevent cracks when drying.²²

In the thangka painting tradition, the ground layer(s) would be both wet and dry sanded, then burnished to achieve a smooth and uniform surface.²³ Exposed areas of the ground layer on the crown have a somewhat glossy appearance, which suggests that the surface was smoothed significantly (Fig. 8).

On thangka paintings, the image was drawn or transferred onto the blank fabric panel after the ground application was finished. It is unclear if designs were transferred onto this crown or drawn directly on the panels, but the design was reinforced using snag tsha (black ink), visible in areas of loss and in cross-section (Fig. 9). This ink drawing was then traced with the raised kyungbur, outlining the figure, architectural elements, decorative patterns, and emblems to prepare for painting. Kyungbur decoration came into fashion in the eighteenth and nineteenth centuries in Tibet and similar decoration methods are not seen on contemporaneous Indian or Nepalese objects, possibly indicating that it was a uniquely Tibetan technique at the time.²⁴ Kyungbur is traditionally applied using a waterproof sheep intestine bag with a metal cone tip which is squeezed to draw narrow lines of ‘dam that has been further bulked to a paste-like consistency.²⁵

Two small samples of the VMFA crown were taken for cross-sectional examination to gain a better understanding of the layers used to build up the ground, kyungbur, and paint layers. Sample 1 was taken from an area of kyungbur and ground at the damaged corner of Panel 4 just below the hole used to tie the panel to the adjacent panel, and Sample 2 was taken from an area of loss in the kyungbur and painted border just to the proper right side of the figure’s head on Panel 5 (Fig. 10). While both samples include the kyungbur and paint layers, only Sample 1 contains the ground layer. Sample 1 revealed that the ground is applied very thinly and it appears that only one layer of ground was applied over the textile substrate. Also in Sample 1, there is a thin black line between the ground and the kyungbur that appears to be the black ink underdrawing. The kyungbur is applied over the black line, though is somewhat offset. The paint layers, gilding, and coating will be discussed in their respective sections. The ground and kyungbur both are fairly homogeneous, having an overall deep yellow color with finely-ground beige, brown, and white particles evenly dispersed throughout.

Results from SEM-EDS mapping indicate that the ground and kyungbur in Sample 1 and the kyungbur in Sample 2 all have relatively high levels of silicon and aluminum along with iron, indicating the presence of an iron-rich clay, such as yellow ochre, possibly mixed with additional yugs rdo (kaolin) (Fig. 11). Little to no calcium was detected in any areas of ground or kyungbur, indicating that chalk was not a significant part of the bulking material in this object.

Paint and Gilding

Paint traditionally used in the production of thangkas and furniture is generally bound with refined protein glue called spyin.²⁶ The binding media could be a variety of different qualities, the lowest being nearly unrefined animal glue, with the highest quality being called lha spyin (“deity glue”) made from the highest quality skins that were free from fat, hair, and other impurities.²⁷ While the paint binder on the VMFA’s crown was not analyzed at this time, it is likely to be proteinaceous, as it seems to be the most common paint binder in Tibet during the time period attributed to the object.²⁸ Furthermore, it has been observed that the paint remains water soluble where the coating has worn away, so it is unlikely to be a drying oil or resin. Further characterization was not performed at this time due to a limited time frame and lack of local availability, though the option exists to perform analysis such as Py-GC-MS in the future.

The pigments used in traditional painting in Tibet (Table 1) have been extensively studied in the context of thangka painting.²⁹ Almost all these pigments could be obtained from local sources within Tibet or were readily available through trade with India and China.³⁰

Table 1. List of common pigments in literature, their Tibetan names

Color	Tibetan Name	Pigment	Chemical Compound
Red	mtshal	Vermilion/Cinnabar	⍺-HgS
	btsag	Red ochre	Fe2O3
Red-Orange	li khri	Minium	Pb3O4
Orange	btso ma	Realgar	⍺-As4S4
Yellow	ba bla	Orpiment	As2S3
	ngang pa	Yellow Ochre	FeO(OH)·nH2O
Green	bod ljang	Malachite	Cu2CO3(OH)2
	sbyar ljang	“Compound Green” (mixture of indigo and orpiment)	mixture
Blue	bod mthing	Azurite	Cu3(CO3)2(OH)2
	rams	Indigo	C16H10N2O2
Pinks/purples	Various names	Lake dyes on inorganic mordant	Various
	zi hung	Pale Mauve (color, not pigment)
White	ka rag	Chalk	CaCO3
	kham sa	Kaolin	Al2(OH)4Si2O5
Black	sol ba	Charcoal	Organic

The object appears to be painted in a relatively muted color scheme of red, black, and gold, but when examined closely, it is possible to see bright blues, greens, and yellows that have been obscured by either degradation or restoration (Fig. 12a, b). The primary color is red and is emphasized by raised gilded designs. The figure at the center is surrounded by a black outline that continues into the lotus throne to outline the petals. Under magnification these black lines appear to be obscuring bright blue and green paints that are visible through cracks in a black coating (Fig. 13). The petals of the thrones alternate in color themselves between orange and red. The innermost outline surrounding the figure is a pale purple only visible at high magnification under bright direct light, as the darkened coating over it almost entirely obscures the color (Fig. 14). Lastly, the deities’ faces and the animals supporting their thrones are delineated by thin black ink lines.

Elemental mapping of the individual panels of the crown with a scanning micro-XRF helped to characterize the types of pigments used and their distribution (Fig. 15). Each panel appears to have approximately the same elemental layout with minor variations based on the symbols present and the order of the alternating lead and mercury-based pigments in the lotus petals (Fig. 16).

The presence of mercury on the elemental maps in most areas of red paint indicates the use of mtshal (vermilion) or cog la ma (cinnabar). PLM confirmed the red pigment as vermilion, the synthetic analogue of cinnabar, due to its finer particle size distribution. The exceptions to this finding are the alternating petals of the lotus throne. Elemental mapping revealed these areas to contain higher levels of lead, indicating the presence of li khri (minium), a red-orange lead-based pigment, common in Tibetan art and material culture.³¹ The differences between the minium and the vermilion paint colors are difficult to ascertain through the darkened coating at first glance, but are apparent on close examination (Fig. 17).

The areas of gilding correlate with the presence of arsenic in the elemental maps, which could correspond to a ba bla (orpiment) preparatory layer for the gilding commonly used to disguise any gaps in the gilding or painting so they wouldn’t stand out as bright white.³² Cross-sectional examination shows that the area of gilding that was sampled has a bright yellow paint layer preceding the sizing that adheres the gilding to the surface (Fig. 18). SEM-EDS analysis found that this yellow paint contained high levels of arsenic and sulfur. Dispersed pigment samples of the yellow preparatory layer were not taken, as it would have disrupted the gilding, so the composition of this paint could not be further characterized with PLM.

Elemental mapping of the gilded areas with the scanning micro XRF detected gold, but not any silver or copper, except in areas where the gilding covered the blue and green pigments. This indicates that the gold used is a fairly pure alloy. In cross-section, the gold is visible as a very thin metallic layer characteristic of gser shog (gold leaf) rather than gser rdul (gold powder) or grang gser (gold paint) which would appear as visible metallic flakes suspended in an organic medium.

Green and blue stripes, now obscured by darkened material, are painted around each Cosmic Buddha, while the outline of each lotus petal alternates between blue and green paint under the black. The L-shaped brackets above the figures and the vertical throne supports were also painted blue and green, though gold leaf has been applied over the paint in these areas (Figs. 19-20). It is unclear at what point in the object’s history these areas were gilded, whether it was during the initial fabrication, a change made during its life as a ritual object, or a later restoration campaign. Compared to the surrounding areas, the coating appears to be darker where there is gilding over copper-based pigments.

All areas of blue and green paint mapped strongly for copper, and PLM identified coarsely-ground transparent blues and greens, with features consistent with bod mthing (azurite) and bod ljang (malachite). Azurite and malachite are the most commonly documented blue and green pigments in thangka paintings of this time period; several studies of thangka painting have found these pigments along with copper-based minerals that are commonly associated with mined azurite and malachite, such as brochanite and antelerite.³³ Both the blue and green pigments are coarsely ground, consistent with the traditional preparation of azurite and malachite as pigments, due to their mineral origins and tendency to desaturate as they become more finely ground.

Initially, upon discovering that there were pigments underneath the black stripes, conservators at the VMFA suspected that the black material on the surface of the panels was the result of dark degradation products of copper-based pigments, such as tenorite. However, such degradation would only have formed in extreme conditions, which was unlikely to have occurred on this object because it remains in relatively good condition.³⁴ After further consideration, it was postulated that the blues and greens were modified to suit later tastes by overpainting with black paint. Both theories were dismissed when a third cross-section was taken from an area of dark material over blue pigment to the proper left side of the figure on Panel 4 (Fig. 21). Viewing this area under magnification, the layer of blue copper-based paint was found to be in good condition with the pigment particles retaining their intense blue hue throughout. Instead of an opaque black paint layer, only one distinct layer of a dark brown coating, displaying the quenched autofluorescence typical for oils, was present on top of the blue paint. The black coloration over areas of copper-based pigments is consistent with literature on fifteenth to the seventeenth century European paintings exhibiting darkening of the linseed oil binder in areas with high levels of copper.³⁵ The coating material on this crown is primarily a drying oil, which has darkened over areas containing copper-based pigments. This could explain why the coating over areas of gilded copper-based paint has darkened, but not entirely blackened as in areas of drying oil applied directly over the copper-based paints, as the gold acted as a barrier between the two incompatible materials.

On the inside of the now-black border around the figures are thin lines of a lighter toned purple. This is covered with a discolored orange coating, which has become nearly indistinguishable from the red background. It is visible only with strong direct light and in cross-sectional examination (Figs. 22-23). Cross-sections of paint in these areas show what appears to be a purple/mauve pigment rather than a combination of red and blue pigment particles. This is confirmed by dispersed pigment samples that show minimal additions of red or blue particles, which could be contamination from the blue and red paint layers just below the purple. Pinks and purples were often made using lake dyes mordanted onto various salts. In SEM-EDS analysis the area showed strong peaks for magnesium. While most literature refers to dar mtshur (alum), organic red colorants on magnesium carbonate substrates have been detected on some eighteenth to nineteenth century thangkas.³⁶

Coating

Historically, Tibetan furniture and decorative objects could be coated with la chu (shellac),³⁷ though by the late eighteenth century, it was common to finish furniture pieces with a mixture of a drying oil and resin.³⁸ There are accounts of craftspeople in Lhasa who used a varnish formulation called siling trakzi, a mixture of boiled linseed oil and a resin, though the type of resin was unspecified.³⁹

The cross-section taken from the kyungbur at the proper right side of Panel 4 showed a single layer of thickly applied transparent coating material (Fig. 18). It appears that the coating material was also used as a size for gilding over the orpiment paint layer. Traditionally, thangka paintings would be emphasized using gold powders applied as paints, but when gold leaf was used, the leaf would often be applied using boiled linseed oil as a mordant.⁴⁰ Both the coating and sizing materials were thickly applied over the surface of the panel causing it to pool in the recesses. Under long-wave ultraviolet illumination, the coating has a dull whitish autofluorescence, possibly from the resin component in the coating mixture that has been quenched by the oil component (Fig. 24).

FTIR analysis was performed on samples of the coating using a benchtop ATR. The spectra were found to closely resemble those of various drying oils, possibly tung oil, with additional peaks matching with a natural resin, which is consistent with the literature on traditional Tibetan coatings.⁴¹ Further characterization using combined chromatography and spectroscopy techniques could be performed in the future to more fully understand the mixture and identify any potential modifiers in the coating.

The darkening of the coating over the areas of copper-based pigments is likely a reaction between the copper-based pigments and the drying oil component of the resin coating. This is a well-known phenomenon and has been studied most frequently in reference to fifteenth to seventeenth century European paintings.⁴² At this time, the interaction between the drying oil binder and the copper-based pigments has not been proven to be the reason for the darkening. However, the lack of apparent pigment particles within the darkened material and the presence of the dark coating directly over areas of copper suggests that it is likely the cause. This could not be confirmed through analysis, as FTIR did not identify any copper salts in the dark coating, nor did SEM-EDS detect any appreciable copper levels in cross-section. The coating has darkened significantly over the vertical throne supports and the L-shaped brackets above the figures heads; both are areas where gold leaf has been applied over the azurite and malachite (Figs. 19-20). The coating in these areas has not blackened completely, but it is noticeably darker than the surrounding areas of coating over vermilion or orpiment-backed gilding (Fig. 20).

Tie

The fabric fragments which line the bottom of the panels once formed a longer tie that held the crown together at the base, which could then be tied at the back of the wearer’s head to fasten it in place. At some point in time, this tie was ripped or cut, allowing the panels to be mounted on a flat board within a frame. Such ties are composite objects themselves, being a complex brocade weave structure with multiple colors and types of thread (Fig. 25).

The tie was made from a strip of textile, cut along the straight grain of the fabric and folded to form the tie with the raw cut edges encased within the folds, resembling double fold bias tape. The top edges of the folds were backstitched in order to keep the fabric from unfolding and the edges from fraying. The ties were then sewn onto the panels using three silk threads alternating green and white, through three holes bored in each panel, and are knotted on the reverse.

Small samples of the ties were taken from Panel 2 for PLM in order to characterize the fibers used in their manufacture. The primary fiber in the ties was identified as dar (silk). Not all of the warp and weft fibers were sampled as there is significant degradation, and further loss was undesirable. In the weft are supplementary structures, less than 1 mm in width that are coated on one side with a dark brown material. These appear to be metal threads that were manufactured by coating and gilding parchment or paper then cutting them into thin strips to weave directly into the fabric.⁴³ A small sample was taken for PLM and was consistent with reference samples of Asian papers, though further identification was not performed at this time. FTIR was also performed on a sample and the resulting spectrum included the characteristic peaks of a cellulosic material. Visual examination showed no remnants of gilding, but one of the elemental maps that captured a section of the textile showed a higher level of silver than the background. No significant peaks for gold were found in this area. There are additional structures in the weft that have not been identified yet which appear to be fibers embedded in a colorless crystalline matrix.

Silks were not often produced in Tibet in the eighteenth century, but rather imported from China, Europe, and even Russia. The Tibetan Plateau was a hub of trade along the Silk Road and had access to goods from around Europe and Asia.⁴⁴ Indian silks were also imported from Varanasi, where Chinese-style silk brocades were woven specifically for Buddhist regions in the Himalayas.⁴⁵

Condition

The crown was in poor condition when it arrived at VMFA. The silk and paper tie had been either cut or broken, separating the crown into its five panels, and adhered to a fabric-wrapped acidic paper backing board with hot melt adhesive. The ties at the bottom of the crown are frayed and grimy with evidence of prior insect activity. The paint and ground over the laminated fabric substrate had flaked away on many of the corners and at the apexes of the crown panels, particularly around the holes at the shoulders of the panels where ties would have bound neighboring panels together. The paint has also worn away on many of the high points on the kyungbur, and in areas of significant abrasion, the apex of the kyungbur is now flattened.

The paint and substrate have suffered from insect damage despite many of the pigments used being toxic to insects. The paint binder, likely being proteinaceous, would have been attractive to pests. No active pest infestation is suspected.

The coating, which may have once been clear or only moderately tinted, has darkened to a deep orange-brown and has reticulated in a way that makes the image depicted on each panel difficult to read. The entire surface of the crown had become somewhat grimy and the coating especially trapped particulate soiling. The black discoloration of the coating that obscures the brightly colored borders around the figures was discussed in the previous section.

Treatment and Digital Recoloration

The panels have been removed from the fabric-wrapped paper board backing using a 5% w/v Laponite RD in deionized water gel, applied locally over the adhesive (Fig. 26).⁴⁶ Once the adhesive was softened, excess Laponite RD gel was removed with a spatula and the fabric was slowly peeled off of the adhesive at an oblique angle. The remaining adhesive was poulticed with 5% w/v Laponite RD in deionized water, applied under magnification using a blunt syringe and being careful not allow the gel to come into contact with the paint below. The Laponite RD gel was removed using a microspatula and the adhesive was shaved down with a scalpel under magnification. Once the bulk of the adhesive was removed in this manner, the surface of the panel was masked with silicone solvent to prevent water penetration into the water soluble binder and the remaining adhesive was softened and scraped as before.⁴⁷ The paint under the adhesive is significantly lighter in color possibly from grime that became embedded in the adhesive or due to a masking effect preventing the paint from oxidizing at the same rate as the exposed paint.

Future treatment steps will include stabilization of the silk ties, cleaning the surfaces to remove accretions, and working with the Curator of South Asian and Islamic Art and the VMFA mount maker to determine an appropriate display configuration for the panels. The coating, while disfiguring and darkened, will not be removed at this time. The coating material is consistent with those that would likely have been used during initial fabrication or possible application during the object’s use as a religious object in its original context, though the materials do not necessarily exclude the possibility of later application. Removing this material is not crucial for the ongoing stability of the crown, as it does not appear to be contributing to the active degradation of the object, and removal of the coating may in fact damage the paint or gilding below. Cleaning tests have been done using several chelating solutions at a range of pH values. These have each been successful in reducing surface grime, and will be pursued as an avenue of treatment. While significantly improving the appearance of the panels, the designs are still obscured by the reticulation pattern.

As the imagery is difficult to interpret due to the reticulated and darkened coating material, it was decided to perform a digital restoration using Photoshop to allow viewers to better understand what the crown depicts and its originally intended aesthetic (Figs. 27-28). Additionally, while microscopy and scanning µ-XRF revealed an entirely different color scheme present under the uppermost layers of darkened coating and gilding, it was difficult to visualize from analysis alone. While scanning XRF was able to map elemental copper in areas that currently appear black, it was unable to differentiate between different copper pigments. And while microscopy revealed two different types of copper-based pigments were present, one appearing bright blue and the other bright green, digital recoloration was needed to more clearly visualize the alternating colors on a macro scale.

Conclusion

Through microscopic examination and a range of analyses, the materials and techniques used to make the VMFA’s Initiation Crown were more fully characterized. The crown is composed of five panels sewn together at their bases onto a strip of multicolored silk and metal thread brocade fabric. The crown was found to be constructed from a substrate of layered cotton fabric adhered together with a proteinaceous adhesive, over which a thin ground layer made from yellow ochre and clay was applied and smoothed. Fine ink underdrawings were added on top of the ground and kyungbur decoration made from the same material as the ground was piped over the underdrawing to create a raised design. Areas were then blocked out in mtshal (vermillion) and ba bla (orpiment) to prepare for gilding and painting. Areas painted with orpiment were gilded with gser shog (gold leaf) using a mordant. Fine details were then painted in li khri (minium), bod mthing (azurite), bod ljang (malachite), zi hung (pale mauve) pigments, and detailed with snag tsha (black ink). Lastly, the front of the panels were varnished with a coating material made up of a drying oil and natural resin. Severe degradation of this coating over the copper-based pigments as well as the overall reticulation and darkening of the coating have worked together to obscure much of the imagery. Digital recoloration combined with creative mounting techniques will allow visitors to the museum and scholars to view the object as it may have once appeared. Treatment is ongoing to stabilize the panels for display, and mounts will be created in consultation with the Curator and Mount Maker to safely display the panels upright in a configuration similar to how it would have been worn.

Acknowledgements

The authors are grateful to Dr. John Henry Rice, E. Rhodes and Leona B. Carpenter, Curator of South Asian and Islamic Art at VMFA for their support of this research, contributions to the discussion on terminology, and review of the paper. This research was primarily carried out within The Susan and David Goode Center for Advanced Study in Art Conservation at the VMFA and many thanks go to Interim Chief Conservator Debbie Linn and the conservation staff for their assistance and support. Particular thanks go to Samantha Sheesley, Senior Conservator and Head of Paper Conservation for assistance with paper fiber identification and to Meredith Watson, Assistant Conservator of Paintings and the Paul Mellon Collection, and Josh Summer, Louise B. and J. Harwood Cochrane Assistant Conservator of Paintings for their help with understanding the darkening phenomenon in the coating. The authors are also indebted to Patty Silence, Director of Conservation Operations at the Colonial Williamsburg Foundation, for supporting our collaboration and permitting the use of instrumentation in the Colonial Williamsburg Materials Analysis Lab.

Author Bios

Leila Sabouni is Assistant Objects Conservator in the Sculpture and Decorative Arts Conservation Department at VMFA. She recieved her M.S. from the Winterthur/University of Delaware Program in Art Conservation with specialization in Objects Conservation. She previously completed internships at the Smithsonian American Art Museum, the Asian Art Museum of San Francisco, and the Winterthur Museum.

Ainslie Harrison is Objects Conservator and Head of Sculpture and Decorative Arts Conservation at the Virginia Museum of the Fine Arts. She previously held positions at the National Museum of the American Indian, and the Smithsonian’s Museum Conservation Institute. She has also carried out archaeological conservation work at sites in Panama and Turkey. She received her MA in Art Conservation from Queen’s University, specializing in artifact conservation.

Kristen Travers Moffitt is the Conservator and Materials Analyst for the Colonial Williamsburg Foundation in Virginia, where she works with conservators, curators, and historic interpreters to conduct scientific research for all collection materials. She specializes in historic paint and pigments, particularly those related to objects and in architectural settings. She received her M.S. from the Winterthur/University of Delaware Program in Art Conservation with a specialization in painted surfaces and is adjunct faculty at that program, where she teaches Polarized Light Microscope and Cross-section Microscopy courses to conservation graduate students.

1. The Editors of Encyclopaedia Britannica, “torana,” Encyclopedia Britannica, November 13, 2009, https://www.britannica.com/topic/torana. ↩︎

2. Robert E. Fisher, Art of Tibet (Thames and Hudson, 1997); Michael Henss, Buddhist Ritual Art of Tibet: A Handbook on Ceremonial Objects and Ritual Furnishings in the Tibetan Temple (Stuttgart: Arnoldsche, 2020); Valrae Reynolds, From the Sacred Realm: Treasures of Tibetan Art from The Newark Museum (Prestel Verlag, 1999); John Henry Rice and Jeffrey S. Durham, Awaken: A Tibetan Buddhist Journey Towards Enlightenment (New Haven, CT: Yale University Press, 2019). ↩︎

3. Henss, Buddhist Ritual Art of Tibet, 413. ↩︎

4. Crowns of this type and the Vajracharya type are traditionally laid out with Vairocana to the far left, Akshobhya left of center, Ratnasambhava at the center, Amitabha right of center, and Amoghasiddhi at the far right. This is consistent with how the panels were laid out on their backing board. See Alexander Von Rospatt, “Vajracharya Crowns and Diadems,” Awaken: A Tibetan Buddhist Journey Toward Enlightenment (New Haven, CT: Yale University Press, 2019), 171-179. ↩︎

5. Takumi Fukuda, “Bhadanta Rama: A Sautrantika Before Vasubandhu,” Journal of the International Association of Buddhist Studies 26, no. 2 (2003): 255-286. ↩︎

6. Henss, Buddhist Ritual Art of Tibet, 413. ↩︎

7. Henss, 413. ↩︎

8. Henss, 416. ↩︎

9. Luca Corona and Camilla Hulse Corona, “Tibetan Furniture: Construction, Form and Function,” Wooden Wonders: Tibetan Furniture in Secular and Religious Life, ed. David Kamansky (Chicago: Serindia Publications, 2004), 38. ↩︎

10. Chris Buckley, Tibetan Furniture (Warren, CT: Floating World Editions, 2005), 17; Corona and Corona, “Tibetan Furniture,” 36. ↩︎

11. Corona and Corona, “Tibetan Furniture,” 34*.* ↩︎

12. Cinnabar could be mined in Tibet, though rich sources of cinnabar and vermilion were available in India and China where it was mined and produced alchemically. See Corona and Corona, “Tibetan Furniture,” 34. ↩︎

13. Buckley, Tibetan Furniture, 17. ↩︎

14. Corona and Corona point out that “Thangka painters had to undergo a rigorous and demanding training…” See Corona and Corona, “Tibetan Furniture,” 33. ↩︎

15. Panels were numbered 1-5 from left to right based on their location on the framed board. ↩︎

16. David P. Jackson and Janice A. Jackson, Tibetan Thangka Painting Methods & Materials (London: Serindia Publications, 1984), 16. ↩︎

17. Corona and Corona, “Tibetan Furniture,” 34. ↩︎

18. Corona and Corona, “Tibetan Furniture,” 34-35. ↩︎

19. Jackson and Jackson, Tibetan Thangka Painting, 20. ↩︎

20. Corona and Corona, “Tibetan Furniture,” 36; Jackson and Jackson, Tibetan Thangka Painting, 20. ↩︎

21. Jennifer Mass et al., “Thangka Production in the 18th - 21st Centuries: Documenting the Introduction of Non-Traditional Materials into Himalayan Painting Practice,” Proceedings of the Forum on the Conservation of Thangkas Special Session of the ICOM-CC 15th Triennial Conference, New Delhi, India, September 26, 2008, eds. Mary W. Ballard and Carole Dignard (Rome: International Council of Museums, Conservation Committee, 2009), 116; Paola Ricciardi and Anuradha Pallipurath, “The Five Colours of Art: Non-invasive Analysis of Pigments in Tibetan Prints and Manuscripts,” Tibetan Printing: Comparison, Continuities, and Change, eds. Hildegard Diemberger, Franz-Karl Ehrhard, and Peter Kornicki (Leiden: Brill, 2016), 492. ↩︎

22. Corona and Corona, “Tibetan Furniture,” 36. ↩︎

23. Jackson and Jackson, Tibetan Thangka Painting, 20-21. ↩︎

24. Corona and Corona, “Tibetan Furniture,” 37. ↩︎

25. Corona and Corona, “Tibetan Furniture,” 36. ↩︎

26. Jackson and Jackson, Tibetan Thangka Painting, 15; Corona and Corona, “Tibetan Furniture,” 34. ↩︎

27. Jackson and Jackson, Tibetan Thangka Painting, 18. ↩︎

28. Buckley, Tibetan Furniture, 17; Corona and Corona, “Tibetan Furniture,” 34; Jackson and Jackson, Tibetan Thangka Painting, 75. ↩︎

29. Publications include Jessica Brocchieri et al., “Combination imaging techniques in Buddhist thangka paintings,” X-Ray Spectrometry 50, no. 4 (2021): 320-331; Robert Bruce-Gardiner, “Realizations: Reflections on Technique in Early Central Tibetan Painting,” Sacred Visions: Early Paintings from Central Tibet (New York: Metropolitan Museum of Art, 1998), 193-205; Buckley, Tibetan Furniture; Corona and Corona, “Tibetan Furniture”; Jackson and Jackson, Tibetan Thangka Painting; Mass et al., “Thangka Production in the 18th - 21st Centuries.” ↩︎

30. Corona and Corona, “Tibetan Furniture,” 34. ↩︎

31. Brocchieri et al., “Combination imaging techniques,” 326; Bruce-Gardiner, “Realizations,” 197. ↩︎

32. Jackson and Jackson, Tibetan Thangka Painting, 20, 87. ↩︎

33. Brocchieri et al., “Combination imaging techniques,” 326-327; Jackson and Jackson, Tibetan Thangka Painting; Mass et al., “Thangka Production in the 18th - 21st centuries,” 115; Ricciardi and Pallipurath, “Five Colours of Art,” 485-500. ↩︎

34. Alessia Coccato, Luc Moens, and Peter Vandenabeele, “On the stability of mediaeval inorganic pigments: a literature review of the effect of climate, material selection, biological activity, analysis and conservation treatments,” Heritage Science 5, no. 12 (2017): https://doi.org/10.1186/s40494-017-0125-6. ↩︎

35. Marion Atler et al., “Photochemical Origin of the Darkening of Copper Acetate and Resinate Pigments in Historical Paintings,” Inorganic Chemistry, no. 58 (2019): 13115-13128; Michèle Gunn et al., “Chemical reactions between copper pigments and oleoresinous media,” Studies in Conservation 47, no. 1 (2002): 12-23; E. Ioakimoglou et al., “Thin-Film Study on the Oxidation of Linseed Oil in the Presence of Selected Copper Pigments,” Chemistry of Materials 11, no. 8 (1999): 2013-2022. ↩︎

36. Mass et al., “Thangka Production in the 18th - 21st centuries,” 115. ↩︎

37. Buckley, Tibetan Furniture. 17-18. ↩︎

38. Corona and Corona, “Tibetan Furniture,” 38. ↩︎

39. Corona and Corona, 38. ↩︎

40. Jackson and Jackson, Tibetan Thangka Painting, 87. ↩︎

41. Corona and Corona, “Tibetan Furniture,” 38. ↩︎

42. Atler et al., “Photochemical Darkening”; Gunn et al., “Chemical reactions.” ↩︎

43. Norman Indictor and Mary Ballard, “The effect of aging on textiles that contain metal: implications for analyses,” Conservation of metals: problems in the treatment of metal-organic and metal-inorganic composite objects: International Restorer Seminar, Veszprém, Hungary, 1-10 July, 1989, ed. Marta Jaro (Budapest: National Centre of Museums, 1990), 67–75. ↩︎

44. Dale Carolyn Gluckman, “A Multifaceted Relationship: Textiles and Tibetan Painted Furniture,” Wooden Wonders: Tibetan Furniture in Secular and Religious Life, ed. David Kamansky (Chicago: Serindia Publications, 2004), 72. ↩︎

45. Gluckman, “A Multifaceted Relationship,” 73. ↩︎

46. Determined by solubility testing, the adhesive swelled sufficiently in water to loosen the adhesive in order to slowly remove the backing fabric. ↩︎

47. Cyclomethicone D5 (decamethylcyclopentasiloxane) purchased from Kremer Pigmente. ↩︎