During 2005 the excavations at _Urvinum Hortense_ covered two separate areas, indicated as Trench A (opened in the years 1998-2004) and Trench B (2001-2204). Along the eastern edge of the main road crossing the urban area, which follows the line of the watershed on the ridge occupied by the city, the excavation uncovered, almost in its entirety, a cistern complex. Along the western edge of the same road an outcrop of sandstone, eroded over time by cultivation and cut by a trench and a ditch filled with ancient material, were brought to light (Trench A). The presence of _post_ -Classical artefacts suggests the cistern was in use or re-used in a late period: in fact, the fill contained many closed-form pottery vessels with handles used for carrying water. Trench B, along the opposite side of the ridge, revealed the residential area. A building which appears to be a prestigious private domus with frescoed rooms is of particular interest. The plan of the house is not completely recognisable at present, however, it seems certain that the structure developed on several levels linked by stairways. The different widths of wall, the frequent renewals and blocked or obliterated annexes suggest that the complex was in use for a long period. The extension of the excavation to the north and west uncovered a well that had been dug previously. (MiBAC)

The 2006 and 2007 excavations were mainly concentrated along the western edge of the town of Urvinum Hortense (Trench B) and on the domus, partially exposed in previous campaigns. The excavation was extended to the north which revealed four new rooms (“D”, “E”, “F”, “G”), and considerably increased the overall dimensions of the domus, however, the complete perimeter has yet to be defined. In 2006 three previously known rooms (“A”, “B”, “C”), were excavated down to the floor levels. The frescoes were left in situ, consolidated and roofed-in. In particular, in 2007 room “G” was partially excavated. This exposed an opus signinum floor in which there was the opening of a well-head. At a short distance from the well, the base and part of a column were found. The column was built of triangular bricks and mortar and had a stucco finish imitating a grooved marble column. This find suggests that this was an atrium. In Trench B (eastern side of the settlement) the excavation inside the second cistern exposed a thick, irregular containing wall, dry-stone built with reused elements of large dimensions, across the cistern at the rear, parallel to its south side. In fact, it appeared to be the wall marking the edge of excavations carried out by the abbot Giuseppe Giustino Di Costanzo inside the cistern in 1805. The surface of the terrain behind the wall was characterised by distinctive colouring and large fragments from marble and limestone sculptures, architectural decorations and inscriptions. The evidence suggests that the southern and deepest part of the cistern was not touched by the 19th century excavations.

This campaign centred on Trench B (eastern side of the settlement) and uncovered the structures adjoining the well and adjacent _opus signinum_ floor, a substantial area of which was exposed. The 2007 excavation had found the base and part of a column resting on the floor. A second base and column segment, also constructed with bricks and faced with stucco came to light during this season. A monumental travertine threshold at the centre of the room’s north side, in relation to a wide corridor ( _fauces?_ ), also paved in _opus signinum_, was uncovered. The columns would seem to suggest the presence of a portico on the same axis as the threshold. This evidence supports the identification of this space as a large _atrium_ with a _compluvium_ supported on four corner columns. The layer covering the floor surface was characterised by the presence, at times highly concentrated, of minute fragments of stucco from the walls. Most of them belonged to linear moulded dividing elements which occasionally presented traces of colour. Along the south-eastern edge of the trench the excavation completed the removal of a dump of pottery from outside room “B” and completely exposed a small room, entered to the north via a monolithic brick threshold. The _opus spicatum_ floor presented a number of small channels at right angles to each other which may indicate this was a latrine. On the south-western side of the trench work continued on the removal of a dump outside room “C”, partially revealing a series of large ashlar blocks on a south-north alignment. These were certainly foundation elements and support for the wall. However, they may have been components of a walled structure pre-dating the building of room “C”, which were incorporated into the new construction.

The 12th excavation season at _Urvinum Hortense_ (2009), mainly concentrated on recording the finds and completing excavation documentation rather than digging. As regards Trench A (eastern side of the town) the relationship between the structures facing onto the main road crossing the town was investigated and long stretches of the paved road surface were exposed. At the same time gaps in excavation records were filled and new recording was undertaken in areas where residual stratigraphic units were removed for the purpose and, above all, through the renewed stripping back of surfaces that for decades had not been cleared of weeds. This was carried out mainly with the aim of checking the reliability and precision of the plan made by Giovanni Bizzozzero following his excavation, undertaken around the great temple on the summit, between 1932 and 1938. However, the 2009 campaign also aimed to continue investigation of the cistern complex situated by the road in order to create a correct, although by necessity partial, plan. In the case of trench B on the opposite side of the town, the 2009 season also mainly concentrated on survey to check recording and the making of new records to fill in gaps in the existing documentation. Excavation was limited to the tidying up of the north trench edge, rendering it regular in order to better appreciate the continuity of the floor surfaces and the perimeter walls of the domus’ _atrium_, by removing residual portions of collapse.

La XIII campagna di scavo di _Urvinum Hortense_ (2010), si è finalmente di nuovo caratterizzata per azioni di scavo e per interventi di rilievo sul campo. Per quanto riguarda il Saggio A (versante Est dell’impianto municipale), si è proceduto all’impostazione di nuovi rilievi delle strutture a vista mediante interventi di rinnovata e più estesa scarnitura di superfici da decenni non sottoposte più ad alcuna azione di sistematico diserbo. E ciò soprattutto allo scopo di verificare l’attendibilità e la veridicità del rilievo planimetrico di scavo redatto, all’indomani dell’intervento attuato attorno al grande tempio di sommità, da Giovanni Bizzozzero, fra 1932 e 1938. La campagna 2010 ha, dunque, rivolto specifica attenzione al grande tempio posto sulla sommità del pianoro e al più tardo complesso della cosiddetta “basilichetta”, impostata ortogonalmente rispetto alla principale arteria di attraversamento urbano. Ai fini, ancora una volta, di una corretta e compiuta, ancorché di necessità parziale, verifica e riproposizione grafica di entrambi gli edifici. L’attività di scavo ha interessato esclusivamente il Saggio B (versante Ovest dell’impianto municipale) e segnatamente il grande atrio e uno dei vani interni (Vano A) della domus terrazzata che si estende per largo tratto lungo lo scosceso pendio che prospetta la valle del Sambro. Si è messa finalmente allo scoperto gran parte della superficie pavimentale in signino dell’atrio, continuando nell’azione di rimozione, iniziata già nel 2008 e proseguita ancora nel 2009, degli estesi e obliteranti accumuli di elementi di copertura e d’alzato in crollo. Ne è scaturita una rinnovata continuità delle superfici pavimentali che sembra aver evidenziato un diverso numero e una diversa distribuzione e disposizione delle colonne attorno al pozzo aperto al centro del vano. Insomma ne uscirebbe invalidata l’ipotesi, avanzata in precedenza a scavo ancora in corso, di una disposizione angolare delle colonne a sostegno di spioventi compluviati. Nel Vano A, manomesso già in antico, si è scavata la fronte di una fornace posta all’intersezione di due murature d’angolo e si è ricercato, per ampio tratto, l’originale livello della superficie pavimentale.

The 14th campaign at _Urvinum Hortense_ continued excavation and survey. The consolidation and restoration of floor surfaces and walls was also undertaken. At the same time, a geophysical survey began an investigation of the entire urban complex. In Trench A (east side of the town) survey of the visible structures continued after vegetation was cleaned back from the surfaces. This was undertaken, as in 2010, with the aim of checking the accuracy of the excavation plans made following the work carried out by Giovanni Bizzozzero around the large temple on the summit, between 1932 and 1938. Therefore, the 2011 campaign again concentrated on the large temple podium situated on the summit and on the later, adjacent complex of the so-called “basilichetta”, standing at a right angle to the main road through the town. The aim was to make a correct (although necessarily partial) check of the dimensions and new graphic documentation of both buildings. Excavation only took place in Trench B (west side of the town), again mainly concentrating on the large atrium of the terraced domus extending down the steep slope overlooking the Sambro valley. The _opus signinum_ floor was almost completely exposed, work continuing on the removal of the extensive accumulations of collapsed roofing and masonry, which revealed large sections of the perimeter walls. Thus, the complete plan of the room was reconstructed. The sandstone column bases and residual tile and stucco column drums, evenly distributed around the _impluvium_, were exposed. The stone components of the latter had been robbed. The wellhead was at some time partially narrowed by the insertion of a _catillus_ from a mill, in order to make drawing water easier and safer. A corner slab of the _impluvium_ cut the well. Most of the exposed floor surfaces, column bases and preserved column drums in the _atrium_ were consolidated and restored.

The 15th campaign at _Urvinum Hortense_ was characterised by excavation, survey and the consolidation and restoration of floor levels and standing structures. At the same time, the geophysical survey of the urban area was completed. In trench A (eastern side of the town), a new survey was made of the visible structures following extensive clearing of the area (the vegetation had not been removed for several decades). This was undertaken with the aim, as in the past, of checking the reliability of the site plans made at the time of Giovanni Bizzozzero’s excavations around the large summit temple between 1932 and 1938. Therefore, in 2012 excavations again concentrated on the large temple podium on the plateau summit, and above all on the later and adjacent complex of the so-called “basilichetta”, situated at a right angle to the main road crossing the town. Excavations were only undertaken in trench B (western side of the town), again on the large atrium of the terraced domus which extends along the steep slope overlooking the Sambro valley. The opus signinum floor was almost completely exposed, continuing the complex operations of removing the substantial accumulations of roofing materials and collapsed walls, which began in 2008. The plan of this room was also defined as sections of the perimeter walls were gradually exposed. The sandstone column bases and residual tile and stucco column drums were also fully exposed. The six bases seem to document a progressive transformation of the room’s plan as part of the architectural and spatial redefinition of the complex. Lastly, excavation showed that all the stone elements of the impluvium had been removed. The only exception, which appears to be deliberate, was the slab into which the circular well mouth was cut. This opening was partially reduced by the forced insertion of a catillus (the cone-shaped upper part of a millstone) to make drawing water from the well easier and safer. The newly exposed floor surfaces, the column bases and drums in the atrium were consolidated and restored.

In 2013, the extension of the excavations in Lots 152 and 153 renewed work in the area of the large baths complex where the Nile mosaic, now in Cannara Museum, was found. The excavation exposed an extensive and regular network of walls, all made of small sandstone fragments. They seemed to delimit a series of large rooms adjacent to the perimeter walls of the bath complex, and appeared to define the access route to the complex itself. At this point in the excavations, the construction techniques, building components and materials, in addition to their well-organised and linear arrangement, seem to suggest that the majority of the structures found do not belong to the baths complex. On the contrary, they probably relate to the late Republican and early Augustan structures incorporated into the construction of the baths themselves, during their construction in the first half of the 2nd century A.D. A separate trench dug in Lot 155 in order to check for the presence of structures associated with the forum area, revealed an unusual pit cut into sterile terrain. It contained a pile of iron tools (pickaxes, scythes, picks, wedges, nails) associated with a bronze _stadera_, a coarse ware jug and coins of late imperial and late antique date. While awaiting the conservation of these artefacts, the find is interpreted as a hoard. During the campaign a preliminary investigation was made of the town wall (in 171, 172, 150, 152, 149, 132/p, 151, 154/p), which uncovered sections of wall, some extensive, built of large sandstone blocks. The intervention demonstrated how, contrary to what is published on the subject, the components used, even though very badly preserved, were fairly regular and neatly placed, where possible in horizontal and parallel rows. The size and shape of the blocks often seemed conditioned by the original stratigraphy from which they were quarried. The stone was all local. The recording, consolidation and restoration of the walls and floors in the area of the _domus_ excavated between 2003 and 2012 were completed during the 2013 campaign. The excavation is awaiting a roof that will not only preserve the already exposed structures, but also facilitate the termination of the excavation of the _domus_ itself. Lastly, the removal of several dumps partially and temporarily restored a more natural appearance to the landscape around the site.

In 2014, the excavations again concentrated on the large bath complex where the Nilotic mosaic now housed in Cannara museum was found. Most of the exposed walls were faced with minute sandstone elements, sometimes alternating with travertine or tile inserts. The repetitive construction methods and serial design of the plan suggest the structures were all built at the same time. This would seem to exclude that they originally belonged to the baths. Indeed, the walls delimited a series of rooms that only seemed contiguous to the perimeter walls of the baths. They appear to be later in date and probably related to a residential structure. In the first decades of the 2nd century A.D., the careful planning and construction in this area of the baths must have redesigned the earlier layout of the neighbouring areas in order to permit maximum practicality and use of the complex, defining and channelling the access into the common spaces, but also selecting and englobing stretches of wall that seemed useful for the new building. All this must have led to tangible transformations in the role and use of the spaces, in particular those immediately adjacent to the construction site of the baths. This may be the reason why the 2014 excavations uncovered extensive rubble collapse used to fill all areas between the old and new walls, in order to guarantee effective static support and levelling, but also drainage through a generalised reuse of the earlier building materials. All the walls were razed to the same level. The ground around them appeared to be artificially flat, and continues to be subjected to invasive agricultural practices. The latter are probably the cause of the dense dumps of material found between 2013 and 2014, containing late and jumbled material culture from the surface layers. The 2014 campaign exposed the perimeter and internal walls of the bath complex and extensive areas of surfaces, in particular the mosaic floor of the so-called peristyle. The aim was to check the previous plans given that they were not uniform in their graphics or in the conventions used, and the topographic reference points were uncertain. The objective was to create a single, uniform georeferenced plan of the structures. Using the same methods and subsequent processing, the state of the wall facings was also documented. A specific photogrammetric and topographic survey was made in order to obtain a textured 3D model of the baths. In addition, this season a new stretch of the city wall was identified close to the baths complex. It was decided not to complete the consolidation and restoration of the walls and floors in the area of the _domus_ excavated between 2003 and 2012 until the area has been roofed. For the same reasons it was decided not to intervene on the cistern structures linked to the baths.

The 2015 excavations once again took place in the urban area immediately adjacent to the bath complex where the Nilotic mosaic now in Cannara Museum was found. Most of the walls uncovered in the area (as seen in 2014) were organic in plan and presented uniform construction characteristics, which showed they were not originally part of the baths complex. They were of earlier date and had a separate function, in many cases probably residential. The 2015 excavations documented sections of wall that were repaired with robbed material indicating the transformation and reconstruction of old, deteriorated structures, evidence for the long period of occupation and use of the area. Dating of the structures and the reused material or associated finds was made difficult by the radical alteration to the related layers, which produced finds covering a wide-ranging chronology. It is certain that the planning and building of the baths complex in the first decades of the 2nd century A.D. redesigned at least part of the previous urban layout in order to create links between pre-existing structures and the new complex and create access to the baths themselves. The 2015 excavations uncovered even more extensive sections of collapsed wall often placed to fill gaps between old and new walls, indicating that static support, levelling and drainage were all efficiently dealt with, redefining the new topographical, environmental, and architectural layout of the area. The uniform and generalized razing of the wall crests and the formation of at least some of the dumps of materials found between 2013 and 2015, is the result of recent widespread agricultural activities. The results of the 2015 were documented using the same methods as in 2014 with the aim of creating a single texturised 3D model of the baths and surrounding area using a photogrammetric survey linked to safe topographical points. During the 2015 campaign, the excavation area was gradually extended to reach the eastern edge of the ancient city, not far from the baths complex. The structures of a modest house were uncovered. The walls, with a double facing made of small sandstone blocks bonded with mortar, were preserved to a low height. The presence of reused materials in the facings highlighted the precarious and rundown appearance of the building. The rooms that were exposed appeared to have a regular plan, were small and had floors formed by a thin surface of beaten lime laid at the level of the foundation offset. The exploration also continued of the city wall in the area of the baths complex and a short stretch was exposed.

In agreement with the local administration of Cannara and the Archaeological Superintendency for Umbria, Perugia University, with the participation of students from other Italian and foreign universities, renewed excavations at Collemancio di Cannara, site of the Roman _municipium_ of _Urvinum_ _Hortense_. Excavations took place in three distinct sectors: the first (trench 2, divided into three areas: a, b, and c) was on the summit plateau, immediately south of a large temple structure; the second sector (trench 1), the largest, was situated between the temple and a medieval church, and occupied an area of c. 210 m2. The third sector (trench 3) was opened to the north-east of the preceding one, on a nearby plateau. A small cemetery area was uncovered in trench 1, characterised by well-made funerary structures, adjacent to the medieval church and probably associated with it. This confirmed the continued use of the site from the Roman to medieval period. A small channel was also identified in this area. In trench 2, area A, a secondary road on a north-south alignment was uncovered. It lay on a tangent east of the temple structure, whose foundation trench (investigated in area C) seemed to have been damaged by excavations in the modern period (probably those undertaken by Giovanni Canelli Bizzozzero). Trench 3 revealed a substantial wall on an east-west alignment, built in an unusual fashion. The upper course, built in large horizontally placed sandstone blocks, rested on small lumps and chips of sandstone placed in an irregular dry-stone construction. All three sectors provided new evidence regarding the plan and urban development of _Urvinum_ _Hortense_, and it is hoped to continue this research in 2018.

L’Università degli Studi di Perugia il nei mesi giugno, luglio e settembre 2018 ha ripreso le ricerche archeologica a Urvinum Hortense. Alle attività progettuali, di scavo e di documentazione hanno preso parte circa 100 studenti dell’Università di Perugia e di altri Atenei nazionali e internazionali. Sono stati prescelti per le indagini, cinque settori: uno sul lato occidentale del tempio Saggio 1B-est); uno a ovest della strada presso il tempio e a nord della necropoli cristiana scoperta nel 2017 (Saggio 1B-ovest), una sul terrazzamento a ovest della cisterna (Saggio 4), già indagata nel 2017, una presso la strada a ovest della cisterna (Saggio 3), una presso le terme (Saggio 5). Le indagini svoltesi nel 2018 sono state eseguite al fine di esplorare nuove significative porzioni urbane del Municipio di Urvinum Hortense. L’attento prelievo stratigrafico ha permesso anche di recuperare materiale archeologico diagnostico, il cui studio potrà restituire dati cronologici utili per la datazione di strati e strutture. In particolare, le indagini hanno permesso di mettere in luce: Saggio 1B-est: una struttura circolare costituita da grandi lastre di calcarenite accuratamente levigate, disposte in modo da formare una piattaforma. Gli interstizi tra le lestre sono riempiti da cocciopesto. In cocciopesto è anche il bordo della struttura. Saggio 1B-ovest: un declivio roccioso privo di strutture. Il banco di roccia individuato presenta anche tracce di litotomia indicanti lo sfruttamento della collina come cava. Saggio 3: una struttura muraria in opera incerta lunga circa 30 m e due file costituite da cinque pilastri equidistanti. Saggio 4: un ampio tratto di basolato stradale a ovest della cisterna. Saggio 5: due ambienti in prossimità dell’impianto termale una dei quali con un portico con tracce di pittura parietale di colore rosso e articolati sistemi di canalizzazione

This season, Perugia University’s excavations at _Urvinium_ _Hortense_ concentrated on four sectors: one on the north side of the temple (Trench Temple-N); one on the artificial plateau leading the area of the baths (Trench 8); two at the baths: one east and one west of the peristyle (Trenches 5 and 6). The aim was to explore significant new urban portions of the _municipium_ of _Urvinium_ _Hortense_. Trench Temple-N: a substantial section of a basalt road was exposed, running parallel to the main road and tangent to the north of the temple’s rear perimeter. Trench 5: a substantial layer of collapse from the roof of the vestibule leading into the peristyle was uncovered. Its removal exposed a floor surface of terracotta slabs that covered a channel providing water to the nearby _frigidarium_. Trench 6: a new room was discovered to the west of the peristyle’s western perimeter, paved in _opus_ _signinum_and with plastered walls. Part of the floor of the peristyle’s walkway was removed in order to date it and investigate the construction technique. Below the floor there was a water channel built of tile with a covering formed by gravel and lime. The associated inspection well came to light on the south side of the trench. Trench 8: a layer was uncovered constituted by dumped material containing abundant fragments of coloured wall plaster. This obliterated a rectangular structure with an apse in the rear wall and channel on the façade, probably a nymphaeum facing onto the main road of the _municipium_.