The Italic sanctuary of San Pietro di Cantoni at Sepino stands at 665 m a.s.l. dominating the wide valley of the Tammaro. The sacred area is enclosed by megalithic polygonal walls forming an irregular triangle with sides a few hundred metres long. The cadastral map and aerial photograph clearly show the conformation of the enclosure with its apex pointing towards the plain below. The interior, obviously levelled, developed over a large artificial rock-cut terrace along the steep slope which from Terravecchia (953 m a.s.l.) descends sharply to Altilia (548 m a.s.l.) and the river Tammaro. It stands in a very pleasant position, not only because the sanctuary is constantly exposed to the sun, but also, and above all, because this position constitutes a balance, not least topographical, between summit areas destined for defence (Terravecchia) and valleys destined for trade and production (Republican phases at Altilia) within the community and district of the _Saepinates_. The sanctuary played a unifying role, situated at the centre of a settlement structure constituted by a diffusion of small and specialised (even if in an elementary way) nuclei. Although the north eastern side of the Matese appears to have been occupied some time earlier, the sanctuary is documented only at the beginning of the 3rd century B.C. From this date the presence of artefacts becomes more conspicuous, more “structural” as they begin to show a precise destination of use, with common formal and dimensional characteristics which can be related to cult use. The 3rd century B.C. artefacts, mainly imported with some of local production, constitute an important document for the sanctuary of San Pietro di Cantoni, but are also a tangible sign of the new prosperity of the community of the _Saepinates_. Most of the materials from the excavation (from the few fragments of antefix to the appliqués in black glaze ware depicting pregnant women, from the anatomical votives to the shells, from the statuettes of Eros to the ring settings, from the loom weights to the _balsamaria_), are evidence of a standard of living that is in strong contrast with the period and the brutality of the war in progress. The cult seems to be that of a female divinity, documented with certainty in 2006 as Mefite (and the statuette dedicated by _Trebis Dekkiis_ should reveal her features and attributes), who watched over maternity, the family, procreation and, more generally, related work activities, thus the fertility of the fields, the pastures, the herds. Finds of fragmentary statuettes of Hercules suggests the existence of two or more cults in the sanctuary, as often seemed to occur. In clear contrast with the increasing prosperity of the settlement in the valley (Republican phase of Altilia), the 2nd century saw the sanctuary’s progressive decline. There were still a substantial amount of finds for this period but the material did not seem to indicate that the sanctuary shared in the economic opportunities of the emporium of Altilia, situated along the _tratturo_, and its probable related production activities. The last century of the Republic and the early Imperial period were marked by the constitution of the _municipium_ of _Saepinum_. Doubtless the town, as the capital of an extensive territory, took on all civil and religious roles. In this situation the sanctuaries in the _ager_ soon lost all importance and, often, their original function. The structures were often used as quarries and left to fall to ruin. The excavation at San Pietro, especially in recent years has, however, began to produce material dating to the early and mid empire in a quantity suggesting that the area was still occupied. The finds, although mainly functional artefacts (spinning implements, lamps, coarse ware pottery), but also glass balsamaria, rings and coins, suggest a possible, although contracted, form of cult life on the site. At the beginning of the 6th century there was a recovery, with renewed occupation which occurred at the moment in which _Saepinum_ lost all importance in the centralised administration of the territory. The community in part dispersed, often returning to the high lands and within a few generations re-proposing systems for territorial occupation and control that were of ancient date and tradition. A large ecclesiastical complex grew up on the ruins of the temple podium, which seems to have had a short life (perhaps due to repeated seismic events) between the 6th and 7th century. The finds dating to the subsequent centuries seem to be the result of sporadic occupation. The small amount of material was very fragmented, unhomogeneous and from widely differing periods. The excavation, undertaken by the Cattedra di Topografia e Urbanistica del Mondo Classico dell’Università degli Studi di Perugia began in 1991 following a survey in the municipal territory of _Saepinum_ carried out between 1987-1989.

The 2007 campaign excavated and removed the rubble collapse on the podium structure, clearing a substantial area of the presbytery of the imposing remains that obliterated almost the entire surface. A series of interventions for the recovery and consolidation of large stretches of the south wall and the restoration of a number of bases and columns in the nave were undertaken. In particular, one of these was put back together in its original position. Below the temple complex to the north, the excavation, from 2003 onwards, had uncovered at some distance from the podium, a curved wall almost forming a semicircle, constructed with a double facing. It was preserved to the top of the foundations and was built of both homogeneous and reused materials. To the north-east there was a stretch of wall on the same alignment which was presumably a continuation. They may have belonged to the perimeter wall enclosing the area in which the church stood. In the intermediate area a large elongated platform came to light at a right angle to the podium. It appeared to join the front of the podium to the staircase at the northern end of the sanctuary enclosure. Built of large cobbles bordered with large irregular limestone and travertine slabs laid with the short sides fitting closely together, almost as if forming a continuous frame precisely delineating the structure on the ground, it ran right across the sanctuary. This may have been to mark a direct relationship with the exterior and the valley below. Therefore, the end of the structure, could terminate by forming an entrance to the sacred area, as yet not uncovered by the excavation. The structure is to be related to urban and structural interventions undertaken on the sanctuary in antiquity. In front of the rectangular structure, in front of the podium, a number of earth graves were excavated. Orientated on an east-west alignment, they had no grave goods and were covered by large irregular limestone slabs. The alignment was the same as that of the majority of the graves excavated to date in the area, all datable to the latest occupation phase and, therefore, to the period of the church’s life. The covering slabs and lining of the graves was also identical.

Continuation of the excavation down hill from the podium removed a large part of the rubble from the surface, but has still not defined the chronological or functional relationship between the paved structures delimited within the sacred area and the perimeter walls. Neither were the route of the paving within the sacred area and its points of entry and exit (downhill and towards the north) defined in a satisfactory way, either in relation to the life of the latter or, more generally, to the continued use of the area when the church complex was founded. The pottery recovered from the rubble covered a wide chronology suggesting that the accumulation of the rubble occurred slowly and gradually in a fairly late period. The rubble appeared to overlie the original perimeter of the sanctuary in this section. At the same time the area in front of the rectangular structure situated in front of the podium’s short side was investigated. In this sector (eastern edge of the excavation) digging was temporarily halted at a contained, but extensive, series of layers relating to roofing and walls that were in ruins, collapsed or had slid. All of this constituted a single horizon. The simultaneous removal of this material was mostly undertaken during the 2007 and 2008 campaigns. In the western part, behind the podium, excavation exposed the sanctuary’s original ground level. The surface sloped from south to north, situated between two distinct terraces and was very uneven due to the presence of rocky outcrops. The unusual form of the podium foundations on this side reflected the state of the ground surface. The wall was the highest at the downhill corner of the north-western side. Travelling uphill exploiting the natural ground slope, the wall was built directly on the bedrock and without any sort of foundation, gradually becoming lower and finally disappearing. At the same time a campaign of recording was undertaken in order to check the existing graphic documentation and a survey was made of the majority of the architectural and construction elements found during the excavation.

Continuation of the excavations on the site of San Pietro di Cantoni at Sepino uncovered the stylobate of the ecclesiastical complex. It appeared deeply sunken in parts under the pressure and weight of the collapsed structures above it. Many parts were missing due to the action of tree roots from the woods which had gradually grown over the ruins of the podium and the church. At the same time specific interventions of consolidation and restoration were also carried out. These involved patches of _opus signinum_ floor, walls and in two cases, abutting the church apse and along the inner facing of the southern perimeter wall, small patches of monochrome plaster and also relatively large areas of polychrome and figured plaster. The presbytery was definitively freed of the enormous pile of large sized rubble covering it. The paving of large slabs was exposed, arranged around a monumental staircase built on the same axis as the church, more or less abutting the apse structure. The 2009 season also brought to light the central section of the front (east side) of the rectangular structure in front of the podium (structure which is appearing increasingly like a cult structure, probably the latest in date within the sanctuary). The rectangular structure formed a path leading to the podium, marked on the ground, in the excavated stretch, by an _opus signinum_ paving bordered by a regular alignment of stones. In the northern sector work continued on the slow ‘deconstruction’ of the rubble overlying the early perimeter of the sanctuary along the side facing the Tammaro.

La campagna di scavo 2010 nel sito di San Pietro di Cantoni di Sepino è valsa a restituire allo stilobate del complesso ecclesiale, manomesso in parte, sul declinare dell’Estate 2009, da interventi di clandestini, l’integrità dimensionale e icnografica faticosamente raggiunta nel corso della precedente campagna 2009. In particolare si è definitivamente liberato il presbiterio dagli ultimi residui di materiale lapideo di grande pezzatura e in disordinato accatastamento che ne obliteravano per porzioni limitate e circoscritte ancora la superficie. Si è così rimessa in luce per intero la pavimentazione a grandi lastre commesse in piano, originariamente delineata a risparmio attorno alla monumentale gradinata realizzata sull’asse mediano della chiesa sulla fronte e a ridosso del catino dell’abside. Contestualmente si sono realizzati specifici interventi di consolidamento e di restauro di elementi architettonici fuori sito, tuttavia riconosciuti come pertinenti alle strutture del presbiterio, e ancora di superfici d’intonaco policromo e figurato disteso lungo il paramento interno del muro perimetrale di fondo all’intersezione con le murature dell’abside. Lo scavo 2010 è poi valso a riportare in luce un più esteso tratto del percorso ortogonale di avvicinamento, realizzato in cocciopesto e definito marginalmente da una sequenza regolare di pietre di contenimento, tracciato sulla fronte (lato Est) della struttura rettangolare antistante il podio (struttura che va sempre più configurandosi come edificio di culto, verosimilmente quello di più alta cronologia nell’ambito del santuario). Nel tratto Nord è proseguita l’azione di lenta scomposizione della macera perimetrale verosimilmente sovrappostasi e ispessitasi sul primitivo limite santuariale lungo il lato che prospetta la valle del Tammaro. E si è attuata, ancorché a tutt’oggi parzialmente, una sistematica azione di messa in luce e di rilievo delle strutture e delle cortine paramentali in opera poligonale. Nel frattempo lo scavo si va sempre più estendendo alla periferia delle strutture monumentali, in direzioni opposte, riguadagnando progressivamente quote originali di calpestio. Ad Ovest, ad esempio, un acciottolato stradale connesso e ortogonale al grande lastricato delineato a terra trasversalmente da Nord a Sud nell’ambito dell’area santuariale e volto alla gradinata di accesso al podio sembra documentare l’esistenza di un ulteriore accesso verosimilmente dischiuso e in fase lungo il tratto occidentale della recinzione. Due tombe bisome di inumati, a fossa con copertura e foderatura realizzate con lastre di pietra, estendono l’area cimiteriale, già nota per rinvenimenti occorsi negli anni passati, che costeggia da Sud, in risalita di versante, l’edificio ecclesiale.

The 2012 campaign set out to investigate the routes within the sanctuary area. The excavations uncovered an even longer stretch of the orthogonal approach pathway, built of opus signinum bordered with stones, laid out along the east side of the rectangular structure in front of the podium, the earliest cult building in the sanctuary. At a certain distance from the podium, work began on the removal of a high linear dry-stone wall built across the approach route. This wall was a modern property boundary. In the northern sector the gradual removal of the dry-stone perimeter wall continued. This structure overlay and widened the original line of the sanctuary enclosure, along the side facing onto the Tammaro valley. Work also continued on exposing and documenting the structural components and the polygonal facing of the enclosure wall. Excavations on the porticus, situated inside the enclosure and running parallel to it, revealed a new sector and several bases, mostly covered by a substantial collapse. Excavations continued around the monumental structures with the aim of reaching the original ground levels. To the west, the cobbled road, at a right angle to the large paved surface on a north-south alignment within the sanctuary area and leading to the steps providing access to the podium, was seen to continue. This discovery suggests the existence of another entrance of the same phase in the western stretch of the enclosure wall, although at present there is only scarce evidence for the latter. Nearby, close to the late structures abutting the long northern side of the podium, the excavation of the tomb identified at the end of the 2011 campaign was completed. Created within the layers of abandonment and collapse of the religious complex, the tomb contained three adult individuals.

A number of trenches were opened this season at San Pietro di Cantoni. These were situated at the front of the temple podium, behind the podium itself and along the western edge of the paved road leading south to the temple steps from the north enclosure. At the front of the temple, the removal of a row of trees meant that in depth-excavation could finally start, with the aim of exposing the continuation of the eastern processional way running at a right angle to the so-called rectangular structure situated at the front of the monumental podium. This structure represents the earliest monumentalisation of the cult within the sanctuary area. To the rear of the podium, the excavation concentrated on checking several contexts relating to votive hoards and pottery deposits, carefully arranged between the rocky outcrops. The excavation was then extended to the adjacent area located at the intersection between the podium’s short and long sides, where the sloping terrain meets the levelled ground north of the podium. Here, a partial and superficial intervention uncovered fragments of black gloss vessels, some of which were miniatures, clearly destined for cult use. The third trench was opened on the west side of the road leading up to the monumental podium. A low surface, probably for water collection, was abutted by a sort of tank. This was used by the faithful, but also had an important cult purpose. In fact it was a large votive deposit, one of the largest found among the many excavated in the sanctuary of San Pietro di Cantoni. Not only the quantity and quality of the artefacts and their excellent state of preservation makes this deposit important, but also its unusual components. In addition to the pottery (mainly black gloss, but also coarse wares, both functional and miniatures), the excavation recovered tens of loom weights, carefully placed one on top of the other to form a very neat deposit rather than a jumbled assemblage. The absolute importance of the discovery lies in its position in the ground, in the open air. The relationship with the deities, unknown but certainly distinct one from the other, was not celebrated inside the temple structure or structures, but outside in the open sanctuary area. In other words, the already complex layout of the sanctuary area with double cult structures, double processional routes and multiple entrances along the curtain wall has now produced evidence of a further peculiarity in its use.

In the 2014 season excavations took place in the sanctuary area. Trees were removed in order to excavate the continuation of the pathway leading to the so-called rectangular structure at the front of the large podium and to investigate a second structure in the immediate vicinity, neglected for years at the edge of the excavation area. The latter was circular in plan, the walls made of small-medium cobblestones, and some reused materials roughly placed one on top of the other within the wall. The excavation also showed the widespread presence of decayed bonding material characterising a thin band of soil along the internal edge of the structure and a central nucleus, also circular, formed by a closely-packed pile of large shapeless stones. At present, its function remains uncertain. Perhaps connected with water supply for a productive activity, it is clear that the structure dates to the later occupation phases in the area. A trench opened at the front and east side of the final stretch of the road that joined the ramp of steps leading up to the stylobate. This revealed the connecting walls between the rectangular structure and the podium behind it. The facing was made of medium sized squared blocks in regular courses that provided protection for the flight of steps. This type of construction technique is already known in the sanctuary and is generally considered to have been used for repairs and restoration, as for example in the podium structure. The structural and architectural similarities between the rectangular building and the structures relating to the podium front and its use seem in this case to confirm the late date of the intervention. Nearby, excavations continued on the final stretch of the paved road mentioned above. This removed all the structures built on its sides and a hearth. A miniature terracotta mask was found, so far unique among the numerous types of votive discovered at the sanctuary. A Survey and 3D Modelling Summer School was organised by the Universities of Venice and Perugia at _Saepinum_ and San Pietro di Cantoni, during which an aerial and ground photogrammetric survey was undertaken of the entire excavation area and a substantial area of the surrounding environment using GPS and UAV devices. The sanctuary’s main structures were surveyed and documented using a laser scanner for the preliminary acquisition of points for the subsequent creation of a 3D model.

Excavations took place in several trenches in the sanctuary area. At the border between lots 295 and 299, the excavation was extended in order to uncover the continuation of the path leading to the small rectangular structure situated in front of the large podium. The path was edged with rectangular stone blocks and had a fine earth and gravel surface. It has now been uncovered for a considerable length that has shown it was perfectly orientated in the direction of the access into the enclosure in the eastern part of the sanctuary. A second trench was opened on the front and along the eastern side of the terminal stretch of the path that joins the flight of steps climbing up to the stylobate. This completed the excavations begun in 2014 of the walls linking the rectangular structure to the podium behind it. As noted in the past, the facing, made up of regular courses of medium sized roughly-worked elements positioned to protect the steps, appears similar to other known interventions in the sanctuary that are usually considered to be restoration and repairs (as for example in the podium structures). However, the structural and architectural similarities between the rectangular building and the structures forming the podium front and its use would seem to suggest the late chronology of this intervention. The late date also appears to be confirmed by the materials found, in 2014 and 2015, in the associated stratigraphy. Nearby, excavation continued of the final stretch of paved road mentioned above. This completed the removal of all residual overlying structures. Excavations along the northern section of the enclosure uncovered the continuation of the monumental porticoed structure, which had at least seven column bases. It is now clear that this was a structure of some importance, providing a perimeter to the sanctuary area but above all a significant and scenographic architectural and urban definition to the sanctuary area itself. The Survey and 3D Modelling Summer School was again organised this season by the Universities of Venice and Perugia at _Saepinum_ and San Pietro di Cantoni, during which the aerial and ground photogrammetric surveys made in 2014 of the entire excavation area and a substantial area of the surrounding environment using GPS and UAV devices were updated. Further survey and recording work was carried out on the sanctuary’s main structures using a laser scanner for the preliminary acquisition of points for the subsequent creation of a 3D model, and for two-dimensional and three-dimensional vector and raster products.