Full front view of the 56 cm polychrome Balinese wooden architectural carving with standing female figure and lower acrobat

When Green Was Not Corrosion: A Polychrome Balinese Wooden Carving Under Magnification

August 29, 2026

The first microscope images from this carving were deceptive. A reddish-brown ground crossed by dark lines, together with granular green material, looked at first glance like the surface of an oxidised copper alloy. That interpretation did not survive contact with the complete object. In worn areas the substrate resolves clearly into longitudinal wood fibres. The green material is therefore not evidence of bronze corrosion; it belongs to a much more complicated painted wooden surface.

That correction became the useful part of the examination. It forced us to separate what the surface actually demonstrates from what colour, patina and familiarity can make us assume.

The object under examination

The object record identifies this as a Balinese carved wooden architectural element, 56 cm high and weighing 1.2 kg, with surviving polychromy. The front contains a crowned standing female figure above a smaller, compressed acrobatic figure. The carving is pierced in several places; the reverse is comparatively plain and follows the outline of the front without comparable modelling or painted detail.

The measurements and identification belong to the object record. The present article concerns what can be established from direct visual and microscopic examination of the carving itself. It is documented on buddha-statues.com as a research case and is not offered for sale here. View the object record and valuation.

Reverse of the Balinese polychrome wooden architectural carving showing the comparatively flat undecorated back — Reverse of the carving. The back is substantially flatter and less worked than the front, an important constructional clue when considering an architectural rather than freestanding function.

Why the construction matters

The narrow vertical format, openwork around the principal figure and comparatively flat reverse are consistent with an element intended to be seen from one principal direction and incorporated into a larger wooden structure. A temple strut or related architectural bracket is therefore a reasonable functional description. The photographs do not, however, preserve the original building context, fixing points or adjoining timber, so the exact architectural position cannot be reconstructed.

This is more precise than calling the object simply a wooden statue. The reverse was never developed as a second viewing surface. The sculptor concentrated labour on the frontal silhouette, the standing figure, the transverse architectural band and the lower figure.

Lower register of the Balinese wooden carving showing the transverse band and compressed acrobatic figure — The lower register. A compact acrobatic or supporting figure is fitted beneath the transverse carved band; the relationship between figure and structure is more informative than trying to identify the figure from posture alone.

The upper figure: description before identification

The principal figure is female in appearance, richly ornamented and crowned. One arm is raised while the body shifts through the waist and hips within the narrow vertical field. Jewellery or ornamental bands are indicated around the neck, arms and wrists. Deep carving around the body creates a strong silhouette and separates parts of the figure from the surrounding foliage-like ornament.

Close detail of the crowned female figure showing surviving green polychromy, red facial accents and surface wear — Head of the upper figure. Green surface colour, red facial accents, darkened recesses and exposed wood are all present within a very small area, showing how discontinuous the surviving decorative system has become.

The object record describes the upper figure as a goddess. That is a plausible catalogue description, but the surviving attributes visible in our photographs do not allow a secure named deity. We therefore retain crowned female figure as the observation and goddess as the attribution.

There is no Buddhist mudra to identify. The arm and hand positions belong to the figure's pose and to the architectural composition; treating them as a Buddha mudra would import an iconographic system that the object itself does not demonstrate.

The lower acrobat

The smaller figure beneath the main register is one of the most distinctive parts of the carving. The body is compressed into the available space and appears to bend or invert beneath the horizontal member. Earlier descriptions considered terms such as attendant, demon or supporting spirit. None can be demonstrated from the visible attributes alone.

For the present research record, acrobatic or supporting figure is the least speculative description. A more specific identification would require a convincing Balinese architectural parallel in which the same pose, placement and attributes occur together.

A painted surface, not a single patina

The front retains several visually distinct materials: dark green to blue-green areas, yellow or ochre passages, small red accents, dark brown to nearly black deposits and reddish-brown exposed wood. They should not be collapsed into the single word “patina.” What survives is a layered surface altered by loss, abrasion, dirt and ageing.

The distribution is especially revealing. Raised surfaces have lost much more colour than recesses. Material remains in protected cuts, around carved boundaries and within depressions. This differential survival is compatible with prolonged handling, cleaning or environmental exposure, but it cannot be used as an independent clock. Artificial ageing can also concentrate dark material in recesses and remove colour from high points.

Nor can we assume that every colour belongs to one decorative campaign. Repainting and local renewal are common possibilities in long-lived architectural woodwork. No cross-section was taken, so the number and sequence of paint layers remain unresolved.

The microscope correction

The decisive photographs were those taken where the coloured surface had become thin or had disappeared. At magnification the reddish-brown areas show elongated structures running in a broadly consistent direction. Dark material sits in narrow channels and small pale particles lie across the surface.

Microscope view of exposed reddish-brown wood showing longitudinal grain, dark channels and embedded surface particles — Microscope view of an exposed reddish-brown area. Longitudinal fibres and channels follow the grain of the wood; particulate material is embedded across the worn surface.

These features are consistent with exposed wood grain and contradict the first copper-alloy interpretation. The initial error arose because the microscope images were considered before the complete object context had been properly integrated. A red-brown surface plus green material can resemble cuprite and secondary copper products when viewed in isolation. Here, the substrate shows that this visual analogy was wrong.

Second microscope view of exposed wood fibres and particulate deposits on the polychrome Balinese carving — A second exposed area under magnification. The recurring linear fibre structure is more useful for identifying the substrate than the colour of the surface coating.

Can the wood be called teak?

Teak was considered during the examination because it is a familiar carving and architectural timber in South and Southeast Asia and because the exposed areas appear compatible with a dense, strongly grained hardwood. That is not enough for species identification.

Surface photographs do not provide the transverse, radial and tangential anatomical features normally required for reliable wood identification. The responsible description is therefore carved hardwood, species unconfirmed. Teak remains a hypothesis, not a conclusion.

The dark zones

Some magnified areas are almost black, yet underlying brown linear structure remains visible through or beneath the dark material. Small pale particles are also embedded at the surface.

Microscope view of a darkened surface zone with underlying brown linear wood structure and small pale particles — Darkened surface under magnification. The photograph records a deposit or altered coating over the wood rather than a uniform intrinsic colour of the timber.

Several processes could account for this: aged coating, accumulated dirt, handling, smoke or combinations of these. A temple context makes soot or incense an obvious possibility, but visual appearance alone does not prove ritual smoke. No chemical or particulate analysis was performed, so the dark layer remains descriptively identified rather than chemically assigned.

The green material: what we know and what we do not

The green areas were the main source of the original misinterpretation. Under magnification, pale green and blue-green particles appear irregularly around worn boundaries and local losses. Once the wooden substrate was recognised, these could no longer be presented as corrosion products of the object itself.

Microscope detail of green surface material at a worn coating boundary with exposed wooden substrate — Green material at a worn coating boundary. The photograph demonstrates a granular altered surface and exposed substrate, but it does not identify the pigment chemically.

Second microscope detail of granular green surface deposits around a local loss in the polychrome layer — A second green area. Concentration of coloured material beside a loss is useful for studying layer survival, not for naming the pigment without analysis.

A copper-containing green pigment remains possible; so do alteration products within an old paint system, later surface contamination and, in limited areas, biological material. Microscopy at this scale does not distinguish these possibilities. Non-destructive XRF would be an appropriate next step to determine whether copper, arsenic, chromium or other diagnostic elements are concentrated in the green passages. XRF has not been performed on this object in the present study.

Yellow does not mean gold

Yellow and ochre colour survives at several points, but no metallic leaf has been demonstrated in the photographs. We therefore do not describe the carving as gilded. The yellow may be pigment, an altered decorative layer or part of a more complex paint system. Until metallic material is directly observed or gold is analytically detected, yellow/ochre polychromy is the defensible wording.

This is another consequence of correcting the substrate identification. Techniques such as fire gilding belong to metal sculpture and are irrelevant to the wooden body of this carving. If gold leaf was ever applied to the wood, it would represent a different decorative technology and would need separate evidence.

Wear, tool marks and the problem of dating by surface

Close photographs show short linear and curved marks across the modelling, especially around the lower figure and projecting details. Some may be remnants of carving and finishing; others are likely later abrasion. Controlled raking light would be more useful than ordinary photography for separating deliberate tool marks from subsequent damage.

The broader wear pattern is clearer. High points are more heavily abraded, while colour and dark material survive in protected recesses. Exposed wood has itself darkened and accumulated material. These observations are compatible with an object that has had a substantial working life, but they do not provide a calendar date.

The object record calls the carving antique. Our surface examination does not contradict that description, but neither microscopy nor patina gives us a defensible exact century. No dendrochronology, radiocarbon analysis, pigment dating or securely documented provenance was available in this examination. An exact 18th- or 19th-century attribution would therefore go beyond the evidence currently in hand.

What the Balinese attribution rests on

The current object record identifies the carving as Balinese. The frontal architectural format, polychrome carved wood, crowned female figure and lower acrobatic element are compatible with that working attribution. The microscopic examination, however, addresses material and surface history; it does not independently prove geography.

For that reason we distinguish two statements. Observation: this is a carved and pierced wooden architectural element with polychromy, a crowned female upper figure, a lower acrobatic figure and a comparatively plain reverse. Attribution: Bali, Indonesia, based on the object record and stylistic reading. Further narrowing by region, workshop, temple type or named deity would require documented comparative material.

Research status

The examination has established several things with reasonable confidence. The substrate is wood, not a copper alloy. The surviving surface is layered and discontinuous rather than a single homogeneous patina. Green, yellow, red and dark materials survive in different states of preservation. The reverse and frontal construction support an architectural function. The wood species, exact pigment chemistry, presence of any original gilding, precise date and exact identity of the figures remain open questions.

The most important result is methodological. A microscope can enlarge a surface without explaining it. In this case the complete object corrected the microscope, not the other way around. That correction is now part of the research record, because it shows exactly where observation ends and attribution begins.

Back to blog