Back to online format

Visual examination using conventional and ultraviolet light sources led quickly to the conclusion that an opaque black coating had been applied to the entire figure, disguising an intricate patchwork of ancient fragments and modern additions. Once the coating was removed some of the newly exposed surfaces displayed obvious differences: ancient fragments still bearing archaeological corrosion contrasted with the bright, metallic or artificially patinated surfaces of modern restorations. In other cases it was difficult to distinguish between new metal and ancient fragments that had been stripped of corrosion products.

A radiographic study was undertaken to resolve questions left unanswered by these initial investigations. The sculpture is too large to fit in the Fairchild Center's dedicated radiography room and a commercial firm was contracted to carry out the work in a secured area using a portable unit. Because of the presumed thickness of the bronze walls, given the size of the figure and the volume of restoration materials anticipated, gamma radiation generated by an iridium-192 source was chosen over X-rays. The statue was further examined using an Everest VIT PLS 500D videoprobe lent to the Museum by its manufacturer on a trial basis. This long, flexible probe, mounted with a miniature digital camera and high-intensity light source, is able to transmit video and still images, providing invaluable visual access to cavities and interior surfaces of the statue.

Although the head appeared relatively intact even after the black coating had been removed, radiographs revealed a reconstruction from numerous ancient fragments held together on the interior with straps and screws (Figure 2). The most significant fragment preserves much of the face, and with use of the videoprobe, features related to the fashioning of the intermodel that are characteristic of the indirect lost-wax process could be observed on its inner surface (see Indirect lost-wax casting). Finger impressions are present in the metal around the mouth, where separate strips of wax were applied to the inner surface of the model, and additional wax was also placed behind the eye sockets.

Although fragmentary and reconstructed with modern patches, much of the torso is original. The upper portion and left side of the abdomen are ancient, whereas the right side, including the navel, and the genitalia are restored (Figure 1). Radiography was useful for recognizing features that provide further evidence of how the wax working-model was assembled inside the mold. A seam between areas of differing radiopacity, visible along the lower edge of the left pectoral, suggests the joining of two large wax sheets (Figure 3). The porosity in the metal is similar on both sides of this apparent discontinuity, and both an ancient patch and large crack continue uninterrupted across it, establishing the seam as an attribute of the original wax model rather than a metallurgical join or a repair.

The drapery and the upper left arm are modern, as videoprobe examination revealed that they lack archaeological corrosion on their interior surfaces, and both are substantially less radiopaque than the ancient fragments. Excessive porosity and other manufacturing flaws, as well as the angular set-in patches used to repair them, all typical of Roman workmanship and seen elsewhere on the portrait statue, also are absent in the more recently-cast metal (Figure 4). A large opening in the upper chest and back, now covered by the modern drapery, was found not to represent a loss but an original feature of the sculpture, as indicated by the presence of cast edges. Such openings have also been identified on other large Roman bronzes and typically accommodate separately cast garments. It is possible that the original garment was larger and covered the shoulder, more of the torso, and most of the upper left arm, which may explain why no fragments of the shoulder, or of the arm from just above the elbow, are present. Given that the forearm is attached only to the modern upper arm, the accuracy of its orientation cannot be judged from physical evidence.

While the top surface of the right shoulder is a restoration, much of the upper arm itself is ancient, although, like the adjacent torso, it is largely reconstructed from fragments. On the upper interior surface of the intact right forearm several parallel ridges were found, starting from the elbow (Figure 5). Similar ridges, associated with a group of evenly spaced chaplets, have been observed inside the base of a Greek herm (79.AA.138) in the J. Paul Getty Museum. This bronze was examined by former Getty Conservation Institute scientist David Scott and Getty Museum conservator Jerry Podany, who suggested that these ridges are casting fins formed in cracks in the core that occurred when the core supports were introduced. In the case of the Museum's statue of Trebonianus, the absence of a similar arrangement of chaplets in the forearm indicates that an alternative explanation must be found; sequential pourings of wax used to produce the model, or the addition of reinforcing wax strips, replicated in the cast metal, might account for the ridges.

The ancient fragments that make up much of the upper thighs continue uninterrupted from the lower torso, confirming that the positioning of the legs is approximately correct. The feet, wearing ankle-high sandals, seem too small for the statue, although this may be a stylistic peculiarity, and both are joined to the calves using techniques not practiced in antiquity. The presence of impurities in the alloy of the left foot, however, does leave open the possibility that it is ancient, belonging to this or to another Roman sculpture. The right foot is divided fairly neatly into right and left halves by a radiotransparent line. It is not certain whether this feature relates to the use of a two-part mold or to a join between separately produced pieces, but either alternative implies modern manufacture.

Close to seventy-five percent of the Trebonianus figure is original, far more than first expected given the appearance of the bronze when the technical investigation was initiated. This fact takes on added significance since there are very few examples of large-scale Roman bronze statuary of the third century, and the Museum's figure is one of the rare three-dimensional representations of Trebonianus that survive. Still, many questions remain to be addressed. For example, if examination of its microstructure should indicate that the left foot is ancient, the technical and stylistic analysis of other Roman figural bronzes can help to determine whether or not it may belong to this statue. Similar investigation could also resolve uncertainties relating to the proper orientation of the left forearm and its original means of attachment.

Indirect lost-wax casting
Lost-wax casting is one of the oldest and most versatile methods for producing irregular forms in metal. When using the direct lost-wax technique, the artist executes a model in wax, which is invested in a refractory mixture of clay, sand, and usually some organic matter. For hollow casts, a core is modeled from a similar mixture and then clad with wax and invested. Core supports called chaplets, generally made from metal rods, are hammered through the wax layer, where they engage and support the core once the wax is melted away. The indirect method allows the casting of multiples using the same mold. The original model is created from any convenient material, and from it, a piece mold. The mold is broken down and reassembled without the model. A so-called intermodel or wax working-model is poured or applied in sheets and the core poured or pressed inside. The mold is then removed, the model invested, and the core supports inserted. The piece mold can be reassembled and reused to create another intermodel, which, in turn, can be invested and cast.