In the area below Villa Avellino, along via Ragnisco, which retraces the line of the ancient _clivus vitrarius sive vicus turarius_, P. Somella, who had hypothesised the presence of a _via tecta_ here, noted a number of rooms of mid Republican date. Explorative trenches were excavated on the site between March-June 2002. The investigation ascertained that in the Republican period, probably between the end of the 2nd century B.C. and the first half of the 1st century B.C., the south slopes of the original valley were terraced via the construction of a series of walls in _opus quadratum_, of large squared blocks of yellow Flegrean tufa, and _opus incertum_. A number of segments of these structures survived inside one of the rooms of imperial date. It is probable that _opus reticulatum_ walls were already added to reinforce these substructures in the early imperial period. However, it was in the mid 2nd century A.D. that the area underwent a radical transformation with the construction of numerous large rooms in _opus latericius_, characterised by great relieving arches, which incorporated the earlier structures. To date ten rooms have been identified, on an east-west alignment, situated on two levels one on top of the other. These constituted a sort of substructure for the urban terrace of Villa Avellino, probably the site of a large public building. As for their function, they may perhaps have been _horrea_, as suggested by A. Tchernia.