The most important find from the excavation of the eastern area of the sanctuary of Demeter was the Hellenistic sacred well. Formed by a ring of clay (diam.90cm, height 50cm), it had a lid with a central hole. Close to the well numerous votive vases had been deposited during ritual celebrations: to the north were miniature drinking vessels and a cup with a perforated base; to the east a votive deposit composing of small _hydria_ and jugs of various sizes; to the south and west were balsamari, miniature bowls and a votive of Demeter with a cruciform torch. Trench 3/95 revealed the late archaic remains of a ritual banquet with a fragment of _obolos_, pig's teeth, a cup with omphalos, drinking vessels, a Black-figure _lekythos_, a cleaver, an iron knife, jugs and a spearhead (whose presence in this context is still unclear). During the period 433-432 B.C. the sanctuary of Demeter was situated inside the walls of Herakleia, in the centre of the city near the agora. It was composed of two small temples both facing north, the megaron A, in the upper part and the _megaron_ B lower down the slope near the springs. When Herakleia became part of the Italiot League (347 B.C.) the sanctuary was enlarged. On its upper part a wall was built, reinforced by orthostats alternating with niches. In front of this, beside megaron A, a terrace was created, closed by a retaining wall at the bottom. A central courtyard with a semicircular altar (_bothros_) and a smaller rectangular courtyard, both having ritual functions stood on the terrace. From here, a flight of steps, of limestone blocks, lead down to the area of the springs. On the slope behind _megaron_ B, _oikos_ C was built. The central area of the sanctuary was delimited by two porticoed structures.

Excavations at the sanctuary of Demeter were undertaken to discover how far it extended to the east and west. To the west of the "sacred well" a votive hoard dating to between the 5th to 3rd centuries B.C. came to light. It compriesed miniature _skyphoi_, drinking vessels and figurines of Demeter. At the southern edge of the sanctuary, above the terrace with the semicircular _bothros_, a small sacrificial area was uncovered, dedicated to _Artemis Bendis_. It comprises a pit with traces of burning, bone and pottery and a fragment of a terracotta depicting the goddess. In the western area the discovery of new walls and tile collapses attests the enlargement of the sanctuary in the direction of the agora.

Excavations have been undertaken in the western part of the sanctuary of Demeter by a team from Innsbruck University. The new trenches revealed that the sanctuary extends further to the north and west than had been thought to date. The remains were uncovered of walled structures on a north-south alignment. These rooms can be dated to the 4th to 3rd century B.C. In the northern area of the sanctuary (trench 6/01) a tank (2.40 x 2.50m) came to light. The walls and floor constructed with squared limestone blocks, the floor sloping slightly from west to east with an aperture for drainage. The structure is cut by a small wall, of later date, which continues beyond the edge of the tank, attesting the latter's obliteration. This wall is crowned by a series of terracotta pepes which channel water above the structure. Tanks placed within sacred areas, as part of purification rituals, are well documented in southern Italian sanctuaries of the same date, both in Greek and indigenous contexts within the Lucanian territory.

This campaign aimed to investigate the extension and layout of the Hellenistic sanctuary of Demeter, particularly its western area, study the typologies of the votive deposits and identify the various phases of the pre-Heraklean sanctuary, specifically the Sirian phase and that following the Achean conquest. In the western part of the sanctuary the walls of two Hellenistic rooms were excavated, their link to the rest of the sanctuary needs further clarification. During the Hellenistic period the sanctuary must have extended for at least 90m from east to west, almost reaching the area of the so-called agora and the temple of Dionysius, situated at only 70m to the west. In the central area of the site a votive deposit of the 4th century B.C. was revealed by the presence of a lens of burnt earth, charcoal and bone, surrounded by a circle of cobbles and small stones. On the north side of the deposit several votive vessels were found together with a terracotta figurine of a worshipper or priest accompanied by a sacrificial animal, a dog. This figurine is similar to another found at the sanctuary which represents a worshipper or priest holding a dove. Dogs and doves are known to have been sacrificed to Aphrodite, who, as attested by the Aphrodite "tondo" found at the sanctuary, was worshipped here together with Demeter. Trenches excavated in the central area of the site revealed an occupation level dating to the first half of the 6th century B.C. In one trench the remains of a votive deposit came to light, resting on a layer of cobbles and gravel. This was composed of a krateriskos and a one-handled cup in an upside down position and fragments of iron, perhaps the remains of a grill and a knife. The excavation needs to be deepened further in order to bring to light the foundations of these structures, which were not examined during the previous excavations carried out by Neutsch.

Previous excavations conducted in the sanctuary of Demeter, in the area of the eastern temenos, had brought to light a late Archaic occupation level. Subsequently, several terracotta votives of the so-called 'armed goddess' type were found within the hollows in the foundations of its E-W orientated wall. These votives are well attested in the Achaean colonies and this find has enabled the foundation date for this structure to be fixed within the period when Siris was conquered by Achaean colonists. This year work on the evidence for the Siritan phase of the sanctuary continues to be one of the focal points of excavations by the University of Innsbruck. To this end trench 1/98, containing various Archaic layers with architectural terracottas of varying dates, was re-opened. At a depth of 4m, a layer came to light that was rich in architectural terracotta fragments (geison, sima, tiles and imbrices) which date to the first half of the 6th century B.C. thus placing these finds within the phase of the Achean conquest.

The excavations undertaken by the University of Innsbruck at the spring-water sanctuary of Demeter at Herakleia in 2004-2006 made a significant contribution to the understanding of the topography of the cult area, the division of the structures, the architecture of the sanctuary and the rituals carried out there. In antiquity the area occupied by the sanctuary was more extensive than previously thought, in the light of the new excavations it is now thought to have covered an area of circa 110 m x 60 m. Several trenches were opened with the aim of reaching an understanding of the topographical relationships within the parts of the sanctuary that had remained unexplored, between the central zone, the recently uncovered rooms in the western zone and the ritual “tank” to the north-west. The discovery of a paved open space near the ritual “tank” and of several stretches of a probable internal road network, provided evidence for the organisation of the infrastructures within the sanctuary in its different phases. The discovery of fragments of terracotta polychrome geison and sima attest the sanctuary’s architecture was already monumental in the 6th century B.C. The excavation of several votive deposits provided clear evidence of how rituals took place. Various levels of votives suggest that several rituals took place at different times but always in the same place. Two deposits of iron votives and a pair of slave shackles were discovered which may have been offered on the occasion of the large scale re-organisation of the cult area in the 4th century B.C. It appears that this ritual was dedicated to a particular deity (Artemis Bendis?) who was to protect the passage from the old to the new, thus consecrating the restructuring of the sanctuary. The offering of the slave shackles could attest the gratitude of the supplicants to a divinity of the sanctuary. It cannot be excluded that Demeter at Herakleia is seen not only as the liberator of slaves but also from slaves.

The stratigraphic excavations begun in the eastern sector in 2006 continued. To the northwest an extensive deposit was excavated. On the rim of a vase the letters _epsilon_ and probably the remains of a _ni_ were still legible, and the traces of other letters suggest that this was a dedication (\"anetheken\"?). The finds ranged in date between the 5th- 4th century B.C., with a predominance of later material. Below this deposit was another containing fragments of terracottas representing Demeter, miniaturistic vases, single-handled cups and banded cups, Black-figure _skyphoi_ and large animal bones. Below a thin layer of clay and cobbles another cult deposit came to light. As the deposits were one above the other, separated only by thin layer of clay and cobbles they probably represent diverse layers of a single deposit. As there was no clear chronological differentiation between the layers, the question as to whether this was a votive deposit created during a single ritual, or whether these were offerings collected over time (in a sort of dump of votive material) remains unanswered. In the central area a number of sterile sandy layers below the layer of fill revealed the upper interface of the natural slope. The paleo-ground surface sloped steeply from south to north, towards the area of the springs. During the restructuring of the sanctuary in the 4th century B.C. the substantial drop in height down to the springs was filled, thus drastically reducing the natural slope. This operation made it easier to reach a part of the springs that was otherwise always damp, and facilitated the construction of the terraced late Classical-Hellenistic sanctuary complex. No architectural remains dating to the archaic period were found on the slope. Due to the steepness of the natural slope it is probable that no buildings were present in this zone prior to the construction of the late Classical-Hellenistic sanctuary. The excellent state of preservation of many of the vases suggests that they were carefully deposited and then covered with earth, stones and tiles. It was probably in this way that the earliest ex-voto were deposited in a sort of “ritual burial”. To the west of the “via sacra”, which climbed towards the south, two low terracing walls were revealed. The stones, which formed a straight line down the slope, whilst to the south they protruded in an irregular manner in the surface of the terracing, delimiting a terrace circa 1.90 m wide and 0,20 m high. The remains of a floor surface suggested the terraces were paved with cobbles. In the same way, at circa 1.90 m to the south of this wall, at circa 0.40 m higher, a low wall with the same function came to light in trench 3/07.

In the previous year a stratified votive deposit comprising three levels came to light in trench 6/04. The single levels of the sacred deposit were separated by a thin layer of clay and cobbles. In 2008, another deposit was found below a further layer of clay and cobbles. The extension of the deposit became smaller towards the bottom. This shows that the votive offerings were deposited in a ditch/pit. The material from the deposit ranged in date from the 6th-3rd century B.C., although there was a predominance of later material. Therefore the material was deposited in the proto-Hellenistic period in an area were earlier offerings had already been deposited. A deposit was uncovered and excavated in several levels in trench 1/08. The levels, on the contrary to the stratified deposit in trench 6/04, were not separated by intermediate layers, but formed a single deposit, probably the result of a single ritual event. To the south the extension of the deposit visibly decreased towards the bottom, so that in this case the existence of a ditch/pit may also be presumed. The remains of two terracotta figurines of archaic date were of particular interest. They reflected the vast stylistic horizon covered by the votive offerings from the sanctuary. Beside the choroplastic of severe pose from an area of Sybaritic influence, there was a terracotta female head of Ionic-oriental style. The finds in the votive deposit dated, as did most of the material, to the period between the 6th and the 3rd century B.C. Next to these was a terracotta loom weight, an iron lance head and a bronze coin from Taranto, with an amphora on both obverse and reverse, datable to between 275-200 B.C. An unusual find was a terracotta figurine of a draped young man found inside a deep cut (h. 14 cm), showing a clear Ionic-oriental influence and datable to the first half of the 6th century B.C.,. In 2008 the remains of the “via sacra”, which climbed slightly from north to south in its central part, and its substructure were removed. The finds dated from the end of the 5th to the second half of the 4th century B.C. In the western zone a row of stones preserved in a single course was uncovered in 2007. The stones were placed so that only the north, finished, side was visible, and the wall was interpreted as a terracing structure. The surface of the supposed terrace was paved with cobbles. Together with the charcoal fragments within this relatively compact layer of cobbles a body sherd from a Red-figure _skyphos_ was found. This dated to the late 5th or the beginning of the 4th century B.C. and showed Hermes sitting with the _kerykeion_ and the remains of a graffito _[…] ENA_.

In 2007 in the central zone of the sanctuary the upper interface of a layer of fill and of terracing comprising large tile fragments mostly in a horizontal position, large stones, occasional bone fragments, a little earth and some small cobbles was uncovered. The entire layer was removed. The material clearly diminished as the excavation went down. At about the centre of trench 5/04 the termination of a sort of steep step sloping down towards the north emerged. Between this steep “step” and the two low walls at right angles there was a cult deposit. The walls at right angles probably originally belonged to a rectangular structure (2.5 x 1.00 m), further remains of which appeared in trench 5/04. The walls comprising a single course of stones and tile were only finished on the exterior. The line of the wall was very irregular on the interior and suggested the presence of a massive fill. In this way a structure is delineated with a height of at least 0.40 cm, whose shape resembles a block altar. No remains of a presumed slab for the altar top were found. The tops of the walls excavated to date were all at the same level and which would seem to attest that they were part of the upper and final part of the structure. The height of the walls decreased to the north in steps, following the north-south slope. Among the small-medium stones in the fill inside the “altar” there were also miniature vases (beakers, plates and _hydriskai_ ) and fragments of large containers of 4th-3rd century B.C. date. Up against the southern edge of trench 1/07 a cult deposit was identified (4 x 0.80 m), excavated in several layers although it was clear that this was the result of a single ritual event. Within the deposit there was evidence of diverse types of ritual. Two medium sized cobbles were inside a _skyphos_ that was placed upside down, attesting the offering of stones also attested in the _Thesmophorion_ of Bitalemi/Gela. The perforated necks and bases of vases inserted into the ground like funnels indicated the pouring of libations. A pig’s tooth and several animal bone fragments attest possible sacrifices. The find of a silver incuse drachm with a backwards looking bull, datable to around 520 B.C. documented that as well as the usual ex-voto, offerings of coins were also made within the sanctuary. An unusual find was constituted by two small hand modelled terracotta feet belonging to a miniature table.

In 2010 the investigation of the spring-water sanctuary of Siris-Herakleia concentrated on two distinct areas of the sanctuary. In the western part, trench 3/07 produced Hellenistic material attesting the occupation of the sanctuary in that period. In the eastern part of the trench an extensive layer of cobbles constituted a floor surface of Hellenistic date. To its west was a stone structure whose function is still unclear. In trench 5/04, in the central part of the sanctuary, bedrock was reached. Materials datable to between the 4th and 3rd century B.C. were found directly above the sterile gravel. Among the later material there was also Archaic material, but an actual Archaic horizon was not identified. The later finds were mainly concentrated close to an alignment of stones about 2 m long, probably what remained of an altar. Excavation of trench 1/07 was interrupted due to the presence of the water table. During the 2010 campaign a zone to the north of the springs was investigated for the first time (Trench 1/10). Here a large building came to light (Oikos H). From east to west it was 5.30 m long, whilst its northern limit has yet to be uncovered. The walls were 0.45 m thick and built of various stone materials, tile fragments and reused earlier architectural fragments. The building was on a much lower level than the other structures found within the sacred area. Therefore, it was seen that the late Classical sanctuary extended not only over the three terraces hypothesised to date, but occupied at least four terraces, and the presence of another terrace to the north cannot be excluded. Numerous small votive deposits were found to the east and west of the building. Building H was probably abandoned at the end of the 4th century-beginning of the 3rd century B.C. as attested by a very extensive votive deposit found overlying the walls of the structure. This votive deposit (US295) produced a vast assemblage of pottery forms which constituted the typical ritual banqueting service for the 4th century B.C. sanctuary. The deposit contained at least four cooking pots whose use was demonstrated by traces of burning, eight hydriai, some with painted crossed torches and two with deliberately-made holes in the bottom, two hydria necks with dedicatory inscriptions from male individuals following the model: “….. OS DAMATRI ANETHEKEN“, one mortar, one lamp, a few fragments of busts of Demeter, miniature vases, black-glaze and Gnathian table ware dating to the end of the 4th century B.C. The implements used for the banquet were placed together with the animal bones (at least one adult bovine and one pig) in a pit probably surrounded by the tiles that were found.

In 2011, excavations continued on the fourth terrace in the northern part of the sanctuary, where the layout of the structures was further defined. Trench 1/10 on the slope was extended downwards to the north in order to gain an overall idea of the aspect and actual dimensions of the building known as “oikos H”, discovered in 2010. Building H measures 5.30 x 3.10 m. Adjacent to building H, to the west, was another rectangular structure of similar size (“oikos I”). An alley, 80 cm wide ran between the buildings, almost on the same alignment as the so-called via sacra and the steps leading to the upper terrace. A significant find in this area was a dump of ceramic table and kitchen wares in association with animal remains and shells. The great quantity of the shells suggests that the eating of seafood was an important part of ritual banqueting. A certain variety of edible bivalve marine molluscs were documented, including venus verrucosa (warty venus), glycimeris glycimeris (bittersweet clam), arca noae (Noah’s ark), _donax trunculus_ (tellina), _pectin jacobaeus_ (scallop) and even marine gastropods such as the common murex. Work also continued on the removal of US 293, which covered the entire terrace. This layer contained finds dating from the 4th century B.C. until well into the second half of the 3rd century B.C., including cooking pots, tablewares, Gnathian style vases and Demetriacal protome-busts. New dedicatory inscriptions to Demeter by Zopyriskos («ZOPYRISKOS DAMATRI ANETHEKEN») were documented on the rims of two _hydria_, a name already known from other inscriptions on vases from the sanctuary. An entrance to structure “H” (1.60 m wide) came to light in the middle of south wall USM 297 below the earth layer US 320. This entrance, later blocked, was paved with cobbles and reused elements dating to the end of the 6th-4th century B.C. A concentration of stones and fragmented building materials was found at the northern edge (of the northern part) of _oikos_ H. This was probably a dump of materials in secondary deposition. Numerous 4th century B.C. deposits were identified within layer US 310 in the sector west of the _oikos_. These produced the usual table and kitchenwares as well as miniature vases, animal bones and edible marine mollusc shells. An interesting find was the skeletal remains of dogs’ paws. North of deposit US 319, a wall 40 cm wide came to light, on an east-west alignment (USM 324). It reused earlier building materials and 4th century pottery B.C., for example a louterion fragment. A long trench (16 x 2 m) was opened in area 3/07 situated between the two zones of the sanctuary, where work on the eastern sector continued. Layer US 288 comprised five layers of cobbles, among which were fragments of _tegulae_, archaic pottery, Ionic cups, one-handles cups with banded decoration and skyphoi dating to the second half of the 6th century B.C. and beginning of the 5th century B.C.

In 2012, excavations in the sanctuary continued on the fourth terrace, where the investigation was extended towards the north. Intense building activity was documented in this area situated just outside the humid zone of the springs. The foundations of an earlier building were found in a level below “oikos I”. The foundations of its perimeter wall, made up of small stones, were very wide measuring 0.75 m. Unlike the other buildings in the sanctuary, this structure was on an east-west alignment. To date about 3.50 m of its length have been uncovered, mostly taken up by a room whose internal measurements are approximately 2.00 x 3.40 m. The room had a well-preserved and well-made cobblestone floor. There may have been another room to the east, entered from the south. A drain covered by a tile was exposed in the external wall of the so-called antechamber, situated lower down on the slope to the west. A new trench (1/12) was opened to the north of trench 1/10, of the same length and an average width of 3.30 m. This revealed a wall that extended for its entire length. At several points, the wall showed breaks and reinforcements, despite this it is to be considered a single structural element. The stratigraphy and its construction technique, date this wall to the latest building phase on terrace 4. Another three sections of wall uncovered this year also belong to this phase. The majority of the finds from this zone were ceramic cooking wares and banqueting wares, including Red-figure vases and good quality Gnathian ware, datable to between the second half of the 4th and the first decades of the 3rd century B.C. There were also transport and storage vessels such as amphora and _pithoi_, plus a repertory of miniature vases characteristic of the sanctuary of Demeter. Alongside the Demetrian type bust-protomi, depicting donors and carriers of the sacrificial basket (_kanoun_), rarer coroplastic types were also present. These included statuettes of the Dioscuri, or the squatting god “Bes” and even terracottas of juveniles in cloaks with the attributes of musical instruments or toys, known from the excavation of the so-called sacred Agora of Herakleia, which G. Pianu suggests are linked to initiatory rituals for adolescents. The faunal remains, analysed by a team from Vienna University directed by Gerhard Forstenpointner, included bones from pigs, sheep/goat and bovines, and also dog, some with butchers marks. The sample included a high percentage of pig mandibles, with a predominance of female individuals.