Archaeological investigations undertaken below the 19th century palazzo Donnaregina, prior to the creation of the Naples Museum of Contemporary Art (“Madre”), produced important results. On via Settembrini the palace façade followed the line of the northern fortifications dating from the Greek period, which functioned as a buttress to the scarp of the deep valley below the present via Foria. A short stretch of the fortifications was uncovered, revealing the earliest curtain wall, probably of the 5th century B.C., in blocks of yellow Neapolitan tufa laid as orthostats joining to the scarp directly behind with transverse tie courses. At circa 3.00 m further north a later curtain wall, of yellow tufa blocks laid in horizontal levels, doubled the width the existing curtain wall towards the exterior, abutting it with new tie courses. The second wall appeared to have been constructed during the 4th century B.C. It was only possible to see the interior face of the new curtain wall, characterized by the presence of numerous quarry marks. In the various points investigated the ancient ground levels from the Graeco-Roman, proto-historic and in part the prehistoric period were missing. Only those covered by the Agnano-Montespina (4100 b.p.) eruption were preserved. An agricultural surface emerged below the material from this eruption, characterised by the presence of plough furrows. This data is an important addition to knowledge regarding the original settlement phases on the plateau on which _Neapolis_ was built centuries later. The eruptive stratigraphy found at the southern end of the palace was cut by circular wells or middens of late Hellenistic date and large pits containing building materials. The latter were datable to a period between the 1st and 2nd century A.D. and probably related to the destruction attributed to the earthquake of 62 A.D.

In 2005 the rescue excavation inside Palazzo Donnaregina was concluded. The excavation undertaken prior to the creation of the Naples Museum of Contemporary Art (“Madre”), brought to light various structures. In the central part of the south courtyard a cuniculus on a north-south alignment was identified as probably belonging to a water supply system, perhaps a secondary branch of the Serino aqueduct. The cuniculus, excavated as a tunnel, was later closed by a barrier in its southern part. In the north-eastern corner of the same courtyard a cistern with two small niches with depressed arches came to light. Access to the cistern was to the east via a ladder placed up against the south wall. The first installation of the cistern was dated to the 1st century B.C. by the presence of an _opus reticulatum_ facing on the dividing wall between the ladder and the cistern space. There was no link between this cistern and the cuniculus described above, however, it must be imagined that this reservoir belonged to a _domus_, which may have been occupied until the 6th-7th century A.D. The large pits containing dumps of building materials date to this period: fragments of frescoed walls, _opus signinum_ floors and architectural elements which probably came from the destruction of the building which occupied the _insula_. In a later period the cistern was substantially modified: the entrance to the east of the space was closed by a wall of small quadrangular blocks, whilst pipes connected to a manhole fed into the reservoir. These transformations may have related to a new form of occupancy documented by a number of Byzantine structures. From this moment onwards the sector of the southern courtyard to the east of these structures was occupied by a large cemetery area which seemed to have been in use until the 12th century. Sixty-four superimposed burials on several levels were identified, all were in earth graves and there were no grave goods.

Archaeological investigations continued inside Palazzo Donnaregina. A new trench was excavated prior to the installation of a goods lift serving the new Museum of Contemporary Art. Important data was gathered regarding the geomorphological reconstruction of the north side of the hill, where the ancient town stood, and for its layout. Part of the natural stratigraphy of eruptive formation dating to the prehistoric period was uncovered. Part of the terracing system and the fortifications along the north side of the hill were also revealed. The foundations of the double curtain wall of this stretch of the defences were identified, with the _émplekton_ layer behind filling the space between the hillside and the buttress. The foundations of a tie course toothed into the buttress also came to light together with a single block of yellow tufa from the wall of orthostats. The presence in at least three places of the quarry mark ΔΗ on blocks in the curtain wall foundations and those of the tie course, together with the few materials found in the preserved patches of émplekta, suggest a date of the 3rd century B.C. for the fortifications-terracing. This find, on the one hand confirms what partially emerged from the excavations in the 1960s, that is the presence on the slope of more than one alignment of walls at different heights. On the other hand it proved definitively that the town’s curtain wall was built up against the scarp at mid height and did not run along the hill summit.