In Sites 1 and 2, important remains of the Roman period have been discovered, which can be dated between the end of the 1st century BC and the late imperial era.
In Site 1, a construction (A) facing east-west was discovered, partially destroyed by later building and in particular by the Angevin walls (Fig.1). It is a brick wall, its interior is covered by a thick layer of mortar over which there is a thin layer of grey plaster which also covers the bottom. At the eastern end, it terminates in a wide apse, while the long walls are animated by semicircular niches, of which five have so far been found.
The construction is of regular proportions: each niche is 1.44/1.46 metres wide, and there is the same distance between one niche and another. As far as its function is concerned, the most plausible hypothesis is that it is a long pool or a canal (euripus), as suggests the comparison
with similar pools found in peristyle in the gardens of houses and villas. The use of pools or basins of various dimensions, from a simple rectangular shape to more complex ones with niches or a mixed, straight-curvilinear plan, seems to have been common in Campania and elsewhere, and above all from the late Republican period and for the whole Imperial period (some examples are the Villa of Diomedes, the House of Meleager, the House of Octavius Quartio, the praedia of Giulia Felice in Pompei, the House of Galba in Herculaneum, the School of Traiano in Ostia).
Works probably dating back to the same phase as the basin-pool are a white-and-black floor of "signinum" (a cement material used to waterproof constructions), and the remains of cement foundations (B) found in Site 2. The latter probably joined at right-angles, forming a substructure that levelled the incline of the plateau on which the construction was erected.
There was a radical transformation of the area during the 1st century AD. In Site 1 the basin, which appears to be cracked a the result of an earthquake or because the land gave way, was filled with a strata of earth that has helped to conserve (apart from older fragments of residual ceramics) materials mostly datable to the 1st century BC, with a few fragments from the mid-1st century AD (black paint of Campania A production, amphora fragments of the Dressel 1 and 2/4 type).
It is probable that, when the basin was destroyed, a wall going from north to south (D) was built; this is of mixed material, using a technique alternating tufo blocks and brickwork.
The construction has two pillars at a distance of 2.35 m from each other; presumably they indicate the boundaries of a room acting as a passageway. At base course level of the wall there was a white plaster flooring that sealed the filling of the construction with the niches, occupying a large part of the eastern zone of the excavation site.
In Site 2, new rooms were built in place of the previous constructions. There is a wall in reticular work with brickwork door-posts, running from north to south, just like wall D found in Site 1, a foundation which is perpendicular to it and rests on a signinum floor, and two mixed-material pillars that mark the edges of three passageway rooms.
Further building works were carried out in the late imperial period. In Site 2 the previous openings were walled up and the floor level raised; this floor was in "opus sectile" (H) with small-dimension square tiles (15x15 cm) of various colours. In the western sector of Site 1 was constructed a basin (C ) of which only the easterner and western sides remain.
The walls and bottom of the tank, which to the east is backed by the mixed-material wall (D), were lined with slabs of white marble, some of which are still in place. Its shape were quite complex and on its eastern and western sides there were apses with a small pillar backing on to them. In the middle position between the two apses there was another small quadrangular pillar. The lead pipes under the floor and inside the pillars indicate that there must have been some kind of fountains.
These constructions probably belonged to a villa built in a suburban area at short distance from the south-west boundary of the Roman town; it may have been on a road made of the trachyte paving-stones which were found as filling-material between the Angevin and the Aragonese walls of the castle.