Our team continued excavating the "private" courtyard (space XXV) of the arcaded gallery to the south of it (room XXXIII) and of room XXXIX, to the west of the former two. In the arcaded gallery (room XXXIII), we completely exposed the three solid brick piers. They rest on a socle made of limestone blocks and were heavily damaged by the mid-seventh-century earthquake. Paola Pesaresi's team had to intervene immediately and partially rebuild the central pier to prevent its permanent loss. Only after this intervention could we remove a destruction layer, which covered the gallery's floor. It was composed of a dark, gravel-like and moist soil, containing many rubble blocks and bricks from the walls, as well as tiles from the roof and painted stucco. Several of these colorful stucco fragments contained nice floral motifs. We also found a coin of Honorius (A.D. 393-423) in this layer, which covered a stone bench running along the south wall. We discovered that the floor of the arcade was composed of cobbles set in a mortar floor.

Room XXXIX was also exposed down to its floor level. Its east wall contained a pier of the same composition and continuing toward the west the row of piers of the adjacent southern gallery (room XXXIII). This suggests that the private courtyard had originally been wider toward the west and that at a later date, its western section was incorporated into the newly constructed room XXXIX. A similar pillar in the latter's southeast corner could suggest that there had also been a brick arch closing of this south arcade toward the west. The destruction layer contained a large number of tesserae, as well as chunks of a white-and-black floor mosaic. The discovery of a concentration of locally produced ceramic fragments, belonging to phase 9 of the local production (ca. A.D. 550/75-650), was remarkable. Whereas the pottery finds in the southeast corner of the large mansion's ground-floor thus far had stopped during the sixth century (see report July 20-26), these finds now indicate that at least room XXXIX was still occupied until the seventh century. Toward the end of the week, we removed the topsoil in the area south of room XXXIX. We found several fragments of smooth limestone column drums, which must have belonged to a colonnade around the mansion's large central courtyard.

We also exposed the very (at least 15 m) long room XL, with its alternating curved and rectangular niches, over the total length of the trench until its floor level was reached (see report, July 20-26). The latter was composed of mortar, the substratum of a lost mosaic floor. In front of the large semicircular niche we discovered a concentration of painted stucco fragments. As a large red painted section of the niche's back wall, which was still in situ, had already been removed last week (see Restoration and Conservation, July 20-26), these fragments must have belonged to the decoration of the north wall itself. The destruction layer held a coin datable to the reign of Diocletian (A.D. 284-305).