The excavation of the villa situated in the locality of Starza della Regina at Somma Vesuviana began after the chance find of a large group of walls which was hindering agricultural work. These structures were first studied between 1932 and 1936 by M. Della Corte (during the period when A. Maiuri was the Superintendent) and a substantial architectural complex emerged, preserved intact on its lower level to a height of 3m and above. The complex was characterized by square pilasters surmounted by arches and a colonnade with Corinthian capitals (in the 1930s the existence of a second storey was hypothesized with a colonnade surmounting pilasters). The ancient ground level was at circa 10m below the present level and the structures extended on an east-west alignment for at least 12m. The portico of pilasters met a brick built structure at a 90° angle, this structure was at least 9m long and decorated with three niches. At the time of the excavation, mosaic floors were noted. However, the investigations undertaken in the 1930s did not identify the typology or function of the building, whose monumental construction and topographical position in agro nolanum, support the hypothesis, formulated by Della Corte, that this was the villa where Augustus died. At the time it was thought the complex had been buried by the eruption of Vesuvius in 79A.D. During 2002 the first geophysical investigations aimed at the definition of the probable extension of the archaeological complex began. (MiBAC)

In the 2003 excavations both the extension of the large room with arches and pilasters and the plan of the spaces around it were defined. The construction, circa 18m long in all, originally comprised a total of 4 arches resting on three square pilasters in opus quadratum surmounted by Tuscan capitals, an architrave and projecting half-pilasters at the ends. The structure was delimited to the west by a wall in _opus vittatun mixtum_ with alternating rectangular and semicircular niches, surmounted respectively by a _tympanum_ and a round arch. The structure was symmetrical with that existing on the eastern side. In one of the rectangular niches a white marble statue of a woman wearing a peplos was found in situ, largely intact and datable by its stylistic characteristics to the 1st century A.D. Only the zone at the base of the eastern and central arches was paved in _opus signinum_, whilst the rest of the surface was paved with a mosaic of regular white _tesserae_. A fragmented white marble male statue was also found, identifiable with _Dionysius_ and of a similar date to the statue with _peplos_. The villa was largely obliterated by the eruption of the so-called “Pollena”, datable to around 472 A.D., a period in which the villa was already in ruins, as attested by the traces of climbing plants sealed below volcanic material. The removal of the collapse led to the partial definition of the plan of the area south of the arches, bordered on the south side by an _opus vittatum mixtum_ wall. At the centre of the wall was an access bordered by painted stucco decoration, flanked by two niches with round arches. This wall, together with the other structures joined to it, delimited the hexagonal hall with arches and pilasters. At the western edge of the latter was a structure which seemed to belong to a rectangular kiln. In the sector to the west, behind the wall with niches, a room paved in opus sectile came to light. The pottery finds from this room datable to the 5th century, sealed by the eruption of 472 A.D., attest the continuous occupation of the complex. (MiBAC)

During 2004 the upper part of a colonnade was uncovered at the northern edge of the excavation. The east-west colonnade was of African marble columns (originally 6 but only 4 are visible) topped by white marble capitals of Asiatic Corinthian type on which were positioned brick platbands, faced with polychrome stucco work, probably alternating with arches. New fragments belonging to the statue of Dionysius, found in 2003, also emerged. Excavation continued in the area south of the arched building, delimited on the south side by a wall of small tufa blocks on an east-west alignment. This wall was joined to the structure with arches and pilasters by two transverse walls forming a sort of three sided exedra. Each of the three sides had a large door at the centre. Of the two side walls only the limestone thresholds remain; of the central wall, surmounted by a tympanum and flanked by two semicircular niches, part of the polychrome stucco decoration was preserved. In the area of the exedra and the area south of it the excavation defined the chronology of the collapse of the walls which overlay an abandonment layer which produced pottery datable to the second half of the 5th century. This attests that the sequence of abandonment-collapse-re-occupation-burial of the complex must fall within the period 450-472. The discovery of traces of floor levels above the rubble layers suggest there was an intermediate phase between the collapses and the burial of the site, thus providing a more comprehensible sequence. The exedra was paved with white mosaic, cut in several places when a number of tanks delimited by low walls were built. The brick pavement and the tract of _opus signinum_ to the north and south of the eastern pilaster also belong to this phase, as does the tank positioned at the bottom of the pilaster itself. These structures probably relate to the installation in this part of the complex of a craft working activity involving the use of water. The excavation undertaken in the western sector, behind the wall with niches, where a rectangular room was uncovered in 2003, revealed an opening on the south side. This was blocked by the construction of an apse, up against the south face of the wall, which was decorated with an _opus sectile_ pavement and marble veneered walls (partly removed). This structure was later closed by a blocking which was linked to a transverse wall which divided the space in half. (MiBAC)

The 2005 excavation involved four areas. In the area of the colonnade and the strip immediately to its north a large collapse of _opus vittatum mixtum_ incorporated in the uppermost levels of the late antique eruption was uncovered. This belonged to the tract of masonry on top of the colonnade, probably made up of platbands alternating with round arches, following a scheme similar to that of the adjacent walls and niches. The removal of the volcanic layers down to the ancient pavement level uncovered the column bases in white, Attic type marble, positioned on a stylobate made of grey lava blocks joined by metal gudgeons. The floor level north of the colonnade was 1.60m lower than the white mosaic pavement in the area between the structure with pilasters and the colonnade itself. Two flights of stairs in rectangular lava blocks, each with six steps, placed one at each end of the colonnade overcame this change in level. An underground structure with a square plan, defined as a cistern, built of regular courses of lava stone faced with _opus signinum_ was also uncovered. Of the same phase but not connected to it was a small channel on a south-east/north-west alignment, below the level of the _opus signinum_ floor, on a tangent to the cistern in the north-east corner. This evidence suggests the presence of a system for water collection, perhaps linked to the use of the tanks found in the southern area of the excavation (in 2003 and 2004). There were no elements which gave certain dating. However, on the basis of the pottery found in the abandonment layers in the area that were sealed by the eruption of 472 , a generic _terminus ante quem_ may be set at the second half of the 5th century. The investigation of the square room (called room 2) at the western edge of the excavation brought to light a further aperture to the west. This was later blocked by a small oven, found intact, built in the last occupation phases of the complex. The floor of the room was originally in _opus sectile_, of which only the impressions remain. These showed that the marble decoration had been reserved for the central part of the room, whilst the rest was paved with _tesserae_. Numerous bronze coins of the first half of the 4th century A.D. were also found. The excavation of the area immediately north of room 2 revealed a wall on a north-south alignment, parallel to the room’s northern perimeter wall. In the area to the west and south-west excavation was limited to the removal of the volcanic deposits which uncovered a stretch of wall on an east-west alignment, in line with the southern wall of room 2 to which it was attached. (MiBAC)

In 2006 the investigation looked at the north-eastern corner of the excavation (Trench 1) and the south-western corner, that is the area to the south of and up against the apse of the western square room (Trench 2). In Trench 1, after the removal of the volcanic deposits, what came to light was an area of floor in a square room (6). This was delimited to the south by the _podium_ wall of lava blocks and brick blocking on which the colonnade stood. To the north the room was delimited by the western stretch of a structure in _opus vittatum mixtum_ with pilasters. The use of _opus signinum_ was probably linked to the cistern and tanks found in previous years. On the eastern edge of the area an opening was identified thanks to the impression left by the threshold of the large door of a sizable rectangular apsidal hall. Of this structure only the northern edge was found, constituted by a wall in _opus incertum_, on which stood two pilasters in _opus vittatum mixtum_ (of the second only traces remained). The pilasters supported three round arches, of which only the eastern one remained, linked to the transverse wall in which the apse opened. The dome of the apse was decorated with a fresco showing sea monsters (tritons and seahorses) ridden by Nereids, immediately above the impost of the apse vault so as to form a sort of continuous frieze on a blue background and recreate a marine environment. The upper part of the dome was decorated with a shell motif. These frescoes were in phase with the walled structures and can be dated on a stylistic basis to the second half of the second century or the first half of the third century A.D., a period which also suits the architectural style of the apsidal hall itself. The excavations in Trench 2 defined a stratigraphic sequence that went down to 3m below the ancient floor level: Period I was characterised by the formation of geological deposits of volcanic material containing no anthropological material, of uncertain date, perhaps post 79 A.D.; period II, the pottery from this period (ARS A, coarse wares and Italian sigillata) fix its date to around the 2nd century A.D.; period III, characterised by an abundance of material interpretable as waste from some manufacturing activity and ARS A pottery, can be dated to the 2nd century A.D., the _terminus post quem_ for the construction of the stone paving; period IV, in a later period, datable to the 4th century A.D. on the basis of the pottery found in the apse foundation trench ( several fragments of ARS C, no earlier than the 3rd century), the stone paving next two room 2 was removed and an apse built; period V, saw the abandonment of the area, pottery such as ARS D, coarse wares and lamps date this period to the 5th century, that is immediately prior to the eruption of “Pollena”. (MiBAC)