The area, situated close to the Romanesque church of San Tommaso, came to be known through a series of sporadic finds attesting the site’s occupation from at least the Roman imperial period onwards. In particular, finds of numerous inscriptions, some of which were dedicated to Jupiter, suggest the presence of a sacellum dedicated to this divinity, situated in the countryside surrounding the church. Furthermore, coins of imperial date and at least one terracotta lamp were found in the same area, although they had already been lost by 1932.

The archaeological investigations looked at the sections in a large rectangular cut on a north-south alignment which had been dug during the construction of an extension to a private building. A series of walls were identified, concentrated in the southern and western sides of the excavated area. There was no evidence of ancient stratigraphy in the other sides. In particular, three structures were identified in the south side and two in the southern part of the west side.

The vicinity of the walls suggests they belonged to a single rural building of Roman date. However, due to the limited nature of the investigation it was not possible to reconstruct its extension or plan. Subsequent trenches were positioned to the south-west of the construction cut. The two trenches (circa 4 × 4 m) were positioned next to the walls identified in the cut sections. In trench A, positioned to the south, at a depth of about 35-40 cm below ground level, two almost parallel walls were identified. On an approximately north-east/south-west alignment they were only preserved at foundations level. Situated at about 1.3 m from each other, they were built of evenly sized cobbles and reused tile and brick fragments, irregularly arranged and bonded with clay.

In trench B, placed to the south-west of the construction cut, another two walls were identified. They converged to form a right angle beyond the trench edge. The first, on a north-west/south-east alignment, was characterised by regular facings of cobbles arranged in horizontal rows and a core of very small stone fragments. The wall was also visible in the south section of the construction cut where it was preserved to a height of three courses. At the southern edge of the trench a second structure came to light, positioned along what would be the ideal line for the continuation of the north wall found in trench A.

The limited extension of the trenches left many questions open, especially in relation to the structure’s plan and phases of use, which could however be suggested on the basis of the different building techniques and alignments. The absence of occupation layers and associated stratigraphy also hampered the interpretation. The few fragments of pottery recovered, mainly body sherds from coarse ware containers suggest a generic dating within the late Roman period.