The Northwest Heroon
The 2004 season of the Northwest Heroon anastylosis project, financed by the Group Arco, will last last three months. It is directed by Ebru Torun, for the moment assisted by Tom Verbist. The first month of the season is now over and works continue as planned. During the first weeks, three upper rows of the famous dancing girls frieze were temporarily replaced, giving a very good idea about the scale of the monument and its prominence over the site (see Restoration, July 4-8). The intention of this trial was to determine the geometric relationship among the various rows and the structural implications of possible deflections. Therefore, Ebru and Tom have spent several days documenting each stone and its relationship to the rest of the structure. Once all levelling and geometric information was recorded, the stones were dismantled again in order to be prepared for their final positioning. Each block was examined once more for cracks, and, when necessary, was either stitched by using fiberglass rods or sealed with lime mortar. On the bottom surfaces of the stones, fiberglass dowels embedded in epoxy mortar were placed and are now ready to make the vertical connection with the row below. Meanwhile the stone carvers have been busy preparing the few missing blocks of the monument. Two supplemental pieces have been successfully carved using the pantograph technique.

The architects have been testing a new method for the vertical connections of the stones. The new method would save time by reducing not only the number of times a block needs to be lifted before being definitively fixed but would also reduce all risks of damage to the original stone that comes with it. WIth this method, once the dowels in the row above are fixed and placed in their final position, new channels are carved leading to the lower dowel hole through which epoxy can be injected from outside, but only once the position of the stone is perfectly determined (in antiquity that molten lead filled the holes in the lower row of stones). If this technique proves satisfactory, the vertical connections of several blocks (or even rows of blocks) can be completed at once without lifting the stone a third time in order to fill the lower dowel hole.

Last week on the north facade, two rows above the orthostats of the frieze have already been installed, which compensates for all levelling problems. As this facade can be rebuilt using only original pieces, these rows provide a very good reference for the geometry and the precise levelling of the rest of the building. Before placing stones of the other three sides, a precise levelling of the new orthostats with the dancing girls replicas must be completed. Levelling will be based on measurements taken from the original frieze blocks that are now at the Museum of Burdur as well as recordings on the site. In fact, it has become clear that each individual stone adjusted the levelling of the rows in order to stop the building's leaning outwards toward its top. A very careful recording of every single inclination and its consequences for the rows above must be checked first. The past month has provided a good basis for the rest of the season because potential construction problems in the coming weeks have been recognised and corresponding solutions been devised.

The Antonine Nymphaeum
This week we concentrated on the study of the arches and architraves of the nymphaeum. The fourth or easternmost arch had already been placed back during the previous week. As the right wing's voussoir (first arch block) was missing, we made a gypsum mold and cut it into the exact shape and dimensions on the nymphaeum itself. Then it was taken down to be copied in new Afyon marble using a pantograph. The new voussoir was almost brought to completion. For the study of the second arch, 1/1 scale drawings of the original three voussoirs were made. Working from those, six new voussoirs were designed on the platform to complete the arch. The missing voussoirs will be carved from new Afyon marble.

We began carving the missing right pilaster capital of the second arch. The last step in the copying process was the transfer of the decorative band from the symmetric capital onto the new block. For this purpose the decoration of the original block was photographed, then a 1/1 scale drawing was made from the rectified photo. The drawing was perforated with pins and paint was dabbed over it. After this transfer, the decorative band was actually carved.

The first stages of carving a Corinthian capital as used in Roman Imperial times (after N. Asgari).

To study the first arch, we assembled the architraves of the first niche on the platform with the crane. Since the architraves had been carried by this arch, the study would give clues about the span and height of the arch. After the assembly, the architraves were documented at 1/20 scale. The information was compared with drawings of the back wall, giving the exact dimension of the second part of the back wall.

The addition to column 16 was carved, glued with epoxy, and fixed with a fiberglass rod. The column is now ready to be placed on the nymphaeum. The addition to column 9 was also carved and its joint with the remaining original part checked. The two parts are ready to be joined together. Some broken parts of column 8 were also glued to the original.

Five original Corinthian capitals are missing in the nymphaeum. Stone carver Eva Leplat (Academy Antwerp) has started to carve one of them from a new block of Afyon marble. She is following the ancient step-by-step approach used for carving Corinthian capitals as identified by Nusin Asgari (former director of the Istanbul Museum).

Conservation
Many of the conservation activities supervised by Paola Pesaresi (Italy) are made possible through a generous grant from the S.H. Kress Foundation, which supports especially the work carried out within the team by Nathan Fash (Tufts University).

Three arches of the palatial mansion's later eastern courtyard (uncovered during the 2002 season) were severely out of balance because of the tilting of the portion above the springing of the arches proper. The arches offered an interesting opportunity to preserve the physical evidence of the architectural structure while making its history legible. The seventh-century earthquake had partially toppled two arches above the springing point. One remained intact, while the others leaned outward and had become disjointed from their supporting piers. Our intervention here aimed to keep the arches standing while preserving the undulating form caused by the earthquake. During the last two weeks, the south (right) arch was mended from above, which involved the replacement of brick, pointing, and a generous grouting of the south haunch. After the unobtrusive pointing was completed, a slow process of dismantling and reconstruction began. In addition, the deteriorated stone courses on top of the keys of the arches were set in new mortar gauged with white cement. Three stainless steel cables were embedded in this capping to create a wall head of reinforced masonry capable of resisting some tension. Afterward, a sector of the west side was excavated in front of the pier carrying the leaning arches, down to floor level and a buttress was built in brickwork with sloping courses. This buttress increased the structural stability of the arches while showing that it is a modern repair. The 2003 conservation effort allowed the excavation to continue down to the floor level, exposing the arches' supporting piers and a stone trough. Meanwhile, the springing point of the final arch was repaired so as to maintain the deformation caused by the earthquake.

Another major intervention involved the complete cleaning of mosaic floors in the vestibule and waiting lounge in front of the mansion's second-floor reception hall. According to the various types of stone tesserae that were used, different proportions of ethyl silicate were used to preserve them, before the whole floor is lifted.

In a dining hall (?) on the ground floor of the villa, the surface of an early Christian (probably early sixth century) wall painting in a vaulted niche, discovered last year (see Field notes 2003, July 13-19), was suffering from lack of cohesion. Valentina Lini (Italy) treated the surface and studied the composition of its mortar and paint. The wall painting was probably executed by applying pigments, mixed with lime milk for additional binding, to a wet or dry plaster surface (intonachino). A simple decoration represents a Chi-Rho monogram, executed with two pigments: a red decorative element, based probably on red iron oxide, and a light grayish blue in the background, which probably consists of black carbon pigment, also mixed with lime milk. A definitive lab analysis is suggested to confirm the chemical composition of the pigments, and to exclude Egyptian blue as an alternative. The final plaster (ca. 4 mm thick) may be composed of a mortar with a high lime ratio and very fine-grained aggregates that are probably derived from limestone and quartz. The layer in contact with the brick masonry (also known as "arriccio") is probably composed of lime, aggregates, such as sand and limestone, and a high presence of vegetable fibers, perhaps including wheat straw or husk. The presence of such fibers in the mixture ties the work to the Byzantine tradition. The Benedictine monk, Theophylos, describes this practice in his well-known work, "Schedula diversarium artium" (ca. 1122). The first layer of mortar's disrepair is attributable to lack of adhesion and cohesion, combined with the loss of mechanical strength. Previous investigations on the wall paintings and plaster fragments from the same area indicate that the growth of roots between masonry and plasters was the primary cause of the detachment between one or more plaster layers and the masonry. In preparation for the next intervention of stabilization, a pre-consolidation of both layers has been executed. The pulverized pictorial layer, which was extremely sensitive to both organic solvents and water, called for the application of ethyl silicate by brush, pipette, and low-pressure spray, until full saturation.

The early Christian wall painting after its treatment by Valentina Lini