In 2001 work was undertaken in three trenches: trench P, situated in the pars postica of temple B; trench G, in the area in front of the temple; trench V in the southern excavation area. Trench P, which had already been excavated in the 1980s, was re-opened with the aim of finding, below the wall of a late medieval tower, the northeastern face of the wall which originally constituted the front of the late archaic temple and was subsequently (end of the 3rd century B.C.) incorporated into temple B as a closing wall of the cella. This hypothesis was formed on the basis of the orientation of the medieval wall which diverged from that of contemporary structures and was aligned instead with the Etruscan buildings. It was also so supported by the fact that several blocks were incorporated into the south-west face of the same wall which do not conform to medieval norms and are comparable with those in the foundations of one of the perimeter walls of the late archaic temple. The excavation identified the part of the medieval tower below which was the structure relating to one of the long sides of late archaic temple A. Excavations in trench G, in the area in front of temple B (where earlier campaigns had identified an external floor level relating to the temple and a step from the stairway), revealed a stratigraphic sequence which from the mid 3rd century B.C. went back as far as an artificial fill datable to the end of the 5th century B.C. This fill was probably part of the re-organization of the area around the late archaic temple. Trench V was situated in the area in front of a wall belonging to a service building on a north-south alignment in the southern area of the excavation. Its foundation relates to the cult pools of the 3rd century B.C. found in previous years. The deposit, situated outside the building, revealed a stratigraphy relating to layers of collapse from the structure dating to the 2nd century B.C. and relates to the construction in this part of the sanctuary of the late temple A. An external floor level dating to the end of the 3rd century B.C. was also uncovered. (MiBAC)

The 2002 excavations were undertaken in trench Z in the area south-west of trench V. These aimed to examine the upper stratigrapic layers relating to the later phases of use of the sanctuary and the area in general. Below a series of late medieval structures there was evidence for open-air occupation datable to the 2nd century A.D., below were the remains of a kiln. (MiBAC)

TIn 2003 trench Z revealed a large ditch, of Imperial date, containing Hellenistic material. This was probably part of the levelling and re-organisation of the terrain undertaken during the sanctuary’s abandonment. This deposit, in the absence of any structures, confirms the abandoned state of the sacred area. Below emerged the Hellenistic stratigraphy whose first level consisted of a layer of obliteration directly covering the razing of the walls in trench V. The layer, which contained 2nd century B.C. material, relates to the construction of nearby temple A and the subsequent creation of a large square. The excavation of trench Y uncovered a probable fragment of wall from the temenos(?). (MiBAC)

In 2004 work continued in trench Z and a new trench was opened linking this with trench T, which had been excavated in the 90s. The new area, denominated J/K, was situated in correspondence with trench T to the south-east of the temenos(?). The removal of a rubble fill datable to the 1st century B.C. revealed the razing of a wall which seemed to close the room that came to light. The occupation level in the room had a sandy matrix and contained 3rd-2nd century B.C. material. In contrast to the preceding campaigns in the north-west part of the enclosure (trench T), where numerous bothroi and votive pits had come to light, no similar elements were found. Thus the site must have been used as some sort of vestibule with respect to the area destined for cult practices. Below this level was the make up for a pavement and then that for the actual construction of the building. (MiBAC)

In 2005 excavations were carried out in a new trench, X, situated between the wall enclosing the cult area that was in phase with temple B and the wall relating to another enclosure of slightly earlier date. Worthy of note was the find of a clay settling tank. Its south side was missing and was probably removed when the wall of the sacred enclosure was built. Outside this structure an occupation level dating to the 3rd century B.C. came to light. (MiBAC)

In the 2006 campaign the field work was undertaken in the area of the chamber subsequently added to enclosure III. The excavation results definitively confirmed the building chronology of the room itself, to be dated to around the mid 2nd century B.C. in concomitance with the obliteration of enclosure I, situated in front of temple B. Exploration of the internal stratigraphy also highlighted the fact that this space functioned exclusively as a vestibule and did not house cult structures. The exploration was extended to the south-western vicinity of enclosure III, to a site which was recently shown to include a fourth enclosure. The most notable find in this trench was constituted by a group of metal artefacts which had been dumped. The group included two bronze objects (a situla handle and the handle of an infundibulum) and numerous iron objects including; a key, a sword and scabbard, a bundle of javelins and two javelin points. This find can probably be linked to a military episode (82 B.C.) involving Volterra when the partisans of Marius clashed with those of Sulla. It is likely that on that occasion metal artefacts belonging to the sanctuary’s reserves were dispersed and looted. A small kiln was found below the level of the deposit described above. Rectangular in plan with walls formed by bricks fixed into the ground, it dated to the second half of the 2nd century B.C. Together with other similar structures found in previous campaigns, the small kiln attests the production of ceramic objects, within the sphere of the sanctuary, probably for use by the devotees such as cooking jars, small balsamarii, torches for cult use. All this material was abundantly documented in both primary and secondary deposition within the sacred area.

The excavated site is situated on the south-western edge of the acropolis overlooking the Cecina valley and the coast. The main aim of the investigation was to define the plan of a late Hellenistic building (mid 2nd century B.C.) where, between 1989-1991, a room frescoed according to so-called I style models, had been found. At the end of the excavation the building was revealed to be a small temple with a rectangular aula, the façade facing to the south-east, similar in structure and contemporary with temple A. On the front of the temple there was a flight of steps, partially destroyed by a robber trench of post-medieval date. The interior was divided into two spaces: a closed frescoed cella, accessible only from the front and probably where the sacred image of the cult of Demeter was housed and a room which could also be accessed from a door in the temple’s long side over looking the valley. From the data relative to the cults practised at the sanctuary it may be supposed that during the annual feasts similar to the Thesmophoria of the Greek world, the temple gave hospitality to the women initiated into the rites of Demeter, who, having spent the night inside the building, then exited via the backdoor and went in procession towards the enclosure area. The back wall of the temple and the area immediately outside it presented some structural peculiarities linked to the performance of the rituals described above. In fact, the back wall of the building had an unusual opening with a two columned porch, and a paved pathway led from the door itself into the nearest cult enclosure (enclosure IV). At least two overlying levels of this basalt paved road were uncovered, datable to the 2nd and 1st centuries B.C. This evidence provided further confirmation of the continuity of cult life between the “Etruscan period” and the “Roman period”, which increasingly appears to be a characteristic trait of this sanctuary. Also of note the fact that, a substantial quantity of small pyriform balsamarii and torches showing signs of use lay on the same level as the pathway.

This excavation campaign concentrated on the zone adjacent to the back wall of the tempietto of Demeter, uncovered during the 2007 campaign. An unusual cult feature emerged in this area. It comprised a small shallow sloping channel (a sort of chute) which emerged in a slightly lowered circular area (diameter circa one metre) surfaced with large dolia fragments so as to constitute a sort of filter. This was probably a ritual press used for vegetal products (grapes and similar). The mechanism foresaw that the vegetal products were crushed inside the small channel and the liquid produced by the compression of these products flowed towards the filter where the solid waste (peel, pips etc) was deposited. It goes without saying that the liquid produced by the press was destined to reach the Chthonic divinities dwelling below ground. In the area inside enclosure IV excavations uncovered an intact stratigraphic sequence which produced important data for the reconstruction of the later phases of the cult site which to date were little known. Below a series of medieval dumps a midden emerged relating to the abandonment of the enclosure itself and datable to the 2nd century A.D. The fill of the midden contained numerous votive and cult objects both of Hellenistic and Roman date. Of particular interest were several lamps of the type decorated with raised points and the _Firmalampen_ type. These objects are of particular importance for the interpretation of the site as they attest that nocturnal processions in honour of Demeter took place and therefore the long lasting continuation of Hellenistic cults. Below the layer described above a level of hearths emerged datable to the post-Sullan period and overlying a layer of rubble from a building which collapsed during the known military clashes of 82 B.C. The building may be identified with the nearby temple of Demeter. The earliest occupation level of enclosure IV was identified below the level of rubble of Sullan date, presenting a number of in situ votive features.

Excavations were undertaken both inside enclosure IV and outside its eastern part. Inside the enclosure the continuation of the perimeter wall exposed to the south was uncovered, heavily damaged due to destructive episodes during the medieval period. Outside the enclosure, beneath a massive dump of almost sterile earth of post-medieval date, an occupation level dating to the 1st century B.C. came to light. The level presented _in situ_ votive material and was partially damaged by a large robber trench datable to the 2nd-3rd century A.D. The line of the robber trench probably represents the outline of the structures belonging to the forepart of the monumental entrance to the enclosure area. On the eastern edge of the excavation area an irregular oblong pit emerged (2.60 x 0.80 m) filled with a substantial quantity of animal bones (intact crania and long bones), to date identified as male and female sheep, ox, pig and a dog’s femur. The presence of species such as sheep and dog, which have attributes associating them with the underworld, indicates in all probability the presence in the sacred area of a cult of Persephone, daughter of Demeter. At the centre of the pit was a quadrangular well, with walls of stone slabs, covered by a large fragment of antefix of a type pertinent to the roof of the nearby tempietto of Demeter. The removal of this fragment revealed an intact, partially glazed, amphoretta datable to the 1st century B.C. A preliminary interpretation suggests this may be an abandonment deposit relating to the temple building mentioned above, whose collapse was securely dated, in the previous campaigns, to the Sullan period. Lastly, below the layers and structures described, the excavation uncovered an occupation level which included a paved make up datable to the 2nd century B.C. and interpreted as part of an access pathway to enclosure IV. In the zone to the west of the enclosures an occupation level was brought to light which presented hearths within small holes in the ground. These contained ash, charcoal and skeletal remains of birds, and can be generically dated to the Augustan period. The finds again attest the long life of the sanctuary, as attested by numerous inscriptions in the Latin language and alphabet recovered on the excavation and to be published in the _Rivista di epigrafia di Studi Etruschi_.

The 2010 excavations concentrated on two different areas, both situated in the south-eastern sector of the sanctuary. A series of cult pits of various types (partially excavated in 2009), situated a short distance from each other and cut into a floor level, were found outside of the southernmost of the open-air enclosures dedicated to the cult of Demeter (trench R4). The first was a circular pit (diam. 83 cm) lined with small stones, containing small animal bones, ash, charcoal and a few fragments of a small impasto jar. This was clearly a small hearth that had been filled, when it was abandoned, with the residue from animal sacrifices. The fill in a second pit (1.50 x 1 m) revealed a more articulated stratigraphy and was also constituted by the remains of animal sacrifices, an iron nail and three small pyriform balsamaria. Below, there was a layer of burnt animal bones and a series of drinking vessels placed upside down around a pyriform balsamario, that was cut in half and contained the bones of small animals. The large _bothros_ attributed to a cult of an underworld divinity, probably Persephone, and already identified and partially excavated during the previous campaign, also opened in this occupation level. Continuing with the removal of the remaining fill, below a layer of small and medium sized stones, a layer of charcoally soil was exposed which contained a black glaze cup datable to towards the end of the 2nd century B.C. and a fragment from a mould for an architectural terracotta. A structure interpreted as a small conduit relating to a ritual during which liquids, water or other types of libation, were fed into the bothros from the far northern end of the pit. Part of a collapsed wall was uncovered in the western sector of the complex (trench Z1), where previous campaigns had revealed a series of service rooms on a north-south alignment, and so differing with respect to all of the cult buildings excavated so far. This belonged to the service structure, which it is known fell into disuse when temple A was built. Below a substantial layer of post-medieval in-fill, one room in particular was investigated in which the collapsed roof was still in situ, as attested by the overlapping arrangement of tiles and imbrices abutting the above mentioned perimeter wall. The excavation also ascertained that this system of rooms, dating to the mid 3rd century B.C., abutted what was probably an earlier building. To date, a section of wall on an east-west alignment has been uncovered and the rest of this building will be explored during the coming campaigns.

This campaign concentrated on the western sector of the sanctuary. At the south-eastern end, below a massive modern earthwork (18th century), the original line of the cliff came to light from which the landslide fell away, partially dragging the area in front of the tempietto of Demeter downhill. A complex structure must have originally stood in this area, comprising a staircase with two lateral wings. Traces of the original plaster were present. In the area to the north-east of trench AB, a circular cistern was uncovered. Restructured in the 18th century, when first constructed it was in phase with the adjacent temple. During cleaning, it was ascertained that the cistern was built (2nd century B.C.) in a position abutting an archaic terracing wall (7th-6th century B.C.), on a north-east/south-west alignment. In the section, it was seen that this wall was abutted to the west by a rock bench, artificially cut into shape, and that it had been built with large pseudo-parallelepiped blocks wedged with small thin slabs and stone chippings. The wall was preserved for about one metre and will be excavated during coming campaigns. In trench AA, situated next to the temple’s rear perimeter wall, the closing wall of another enclosure (enclosure V) was found. This represents the last building operation in this quarter. This courtyard appeared to have been created in order to fill the empty space between enclosure IV and the tempietto of Demeter, which were both pre-existing. Inside the courtyard, largely damaged by the construction of a medieval building, a layer characterised by the presence of small hearths containing votive balsamari and iron nails, was uncovered. The western sector of the sanctuary dedicated to the cults of Demeter, therefore resulted as comprising a sequence of five open-air enclosures, ending with a temple building situated towards the downhill area.

The excavations concentrated on the south-east section of the sanctuary, in the area between the façade of the tempietto of Demeter and the open-air enclosures IV and V, identified during previous campaigns. The aim was to investigate the vicinity of the complex arrangement of cult buildings in the southern part of the sacred area, as georadar surveys undertaken here in 2012 indicated the presence of probable structures at a shallow depth. The excavations identified a square building (c. 6 x 6 m), well-preserved on the north side, while the southern part seemed to have been effected by a landslide from the south slope of the acropolis, whose remains were found in 2011, that had also cut part of the square in front of the tempietto of Demeter. Below a layer containing late medieval pottery, the collapse of the building’s roof was exposed, made up of tegulae and imbrices mixed with fragments of archaic majolica pottery. Once this rubble was removed, the room was seen to be completely paved with a mortar floor . This was a carefully made mosaic floor in _opus_ _tessellatum_ with travertine type light-coloured stones, decorated at the centre with a square _emblema_. This was made of bluish stone, probably gabbro, and enclosed a floral motif with six lanceolate petals. Underneath the mosaic was a make-up of lime mixed with stones and crushed brick/tile, about 10 cm thick. A preliminary analysis of the mosaic suggests it was laid in the second half of the 2nd century B.C., thus the building must have been part of a scenographic terracing together with the adjacent tempietto of Demeter. A mortar floor surface was uncovered to the west of the building with the mosaic, abutting its perimeter wall. The fragments of black glaze pottery with grey fabric and early Italian sigillata that were present date the mortar floor to the mid 1st century B.C. The presence of a late medieval roof collapse directly in contact with a late Hellenistic floor, suggests that the building had a long life and underwent numerous alterations during the course of time. A section of wall about two metres long was identified (but not excavated) close to the eastern edge of the trench. It was built of medium-large stone blocks and probably belonged to the sanctuary’s _temenos_.

The excavation investigated the area situated immediately east of the room with mosaic floor found in 2012. The eastern edge of the trench was extended by 7 m and the accumulation of late medieval material removed. An occupation layer relating to the paved building was exposed that extended across the entire excavation area and was characterised by the presence of several hearths. One of the latter, situated at the north-west trench edge, was bordered by a circle of stones and contained pottery datable to between the late 2nd and early 1st century B.C. (Campania B black glaze, _balsamaria_, a lamp and fragment of a terracotta statue) and numerous fragments of pig and sheep/goat mandibles. In the southern part of the trench, this level was cut by late medieval buildings. In particular, the remains of a square construction, probably a small tower, and a stretch of a small paved road made up of bricks, broken tile and some archaic majolica fragments. In the eastern part of the trench, a section of wall, identified the previous year, was completely uncovered, and was preserved to a length of 5 m. A bronze coin and an almost intact cooking pan datable to the late 2nd century B.C., were recovered from the earliest layer in phase with this structure, which abuts its west side. Likewise, the excavation to the west of the wall revealed an occupation level made up of a layer of grey clay with charcoal fragments, in which there was a hole filled with soil and numerous fragments of moulds, architectural terracottas, metal slag, and black glaze pottery dating to the late 2nd century. A preliminary interpretation of the wall, prior to extending the trench to the north to check whether it continues, suggests it was a section of an enclosure relating to a late phase when the sacred area was enlarged.

The work took place in the central part of the excavation area where, in 199-2001, a complex of four rooms was identified. They were on a north-south alignment that differed from the cult buildings (temples and open courtyard, arranged on a north-east/south-west axis) and were interpreted as service buildings. The excavation took place in two adjacent trenches: trench V investigated the immediate area outside enclosure I, and trench VI, investigated the interior of the service building’s north room. The exploration of the area outside enclosure I (trench V) confirmed that it was built in the late 3rd century B.C. at the same time as temple B. Overlying the level in phase with its construction there were a series of occupation layers in the courtyard containing abundant charcoal remains. Datable to the first half of the 2nd century B.C., they were characterised by the presence of two large hearths in shallow hollows cut into the _opus_ _signinum_ floors. The hearths contained numerous examples of cooking pots, storage and table-wares, in particular plain ware and black glazed cups, and abundant faunal remains. Two pottery fragments with inscriptions relating to deities worshipped at the sanctuary were found in these levels. One is a jar fragment, datable to between the late 4th century and the early 3rd century B.C., inscribed with the theonym _uṇ[i]_ and the other is a fragment of a small pre-sigillata plate with the inscription _ap_, abbreviation of _apa_. These inscriptions attest the presence in this area of the sanctuary of the goddess named _apa/papa_, founder of the sanctuary, accompanied by her _paredros_ _uni_ (Iuno). The excavations in trench VI provided a more precise chronology for the construction of the service room, which must have occurred at the beginning of the 3rd century B.C. The presence of the remains of a kiln suggests that the room was used as a production area relating to the sacred area and therefore was open-air. It is possible, and this remains to be confirmed, that in a secondary phase the room was roofed and provided with an entrance with a threshold. The levels underlying this structure, only partially excavated, can generally be attributed to the external environs of the late archaic temple as occupation surfaces and depositions of cult material and votives. In this regard, a very interesting discovery was a deposit, datable to around the mid 4th century B.C., comprising about 30 small plain cups deliberately placed upside down, accompanied by two jars and a large fragment of a disc acroterion, from the roof of the late archaic temple. This was probably a collective libation ritual perhaps to be associated with the deactivation of the temple itself, symbolically represented by the disc acroterion.