The 2013 campaign on the site of San Giovanni, situated at the eastern end of the coastal plain of the same name, continued excavation of the _pars_ _rustica_ of the villa delle Grotte, begun the previous year. The first campaign at San Giovanni intended to check for the presence of ancient kilns for the reduction of iron ore that are attested by a conspicuous literary tradition and archive documents. However, the excavation did not reveal the remains of any such kilns, which may still be buried below the occupation layers that were unexpectedly revealed. During the 2012 campaign, two trenches, 12 x 6 and 6 x 8 m, were opened in areas that had produced the most anomalies. The 2013 season continued work in the first trench where the remains of a large building had emerged, with a substantial layer formed by the collapse of the roof and pisé walls. Five rooms were uncovered. In room III, interpreted as a sort of vast open courtyard, five large _dolia_ _defossa_, used for completing the fermentation of wine or for its conservation, were excavated. At the same time, two smaller trenches were dug to investigate the site’s extension, which revealed substantial masonry remains and collapsed walls. Based on the study of the pottery finds and the tile and _dolia_ stamps, the building at San Giovanni appears to date to the first half of the 1st century B.C. and was abandoned following a fire in the 1st century A.D. that baked the clay structures, thus preserving them. This large complex was the _pars_ _rustica_ of the Villa delle Grotte, situated on the adjacent promontory, excavated in the 1960s, and thought to be part of the estates of the _gens_ _Valeria_.

The site of San Giovanni is situated at the eastern end of the coastal plain of the same name. The 2014 campaign aimed to continue the excavation of the pars rustica of the Villa delle Grotte. This season’s results can be summarized as follows: The large quantity of carpentry nails found in the area with the _dolia defossa_ (Room III) indicates it was a roofed area rather than a courtyard. The _dolia_ were probably reached via intermediate floors resting on the structure’s perimeter walls, paved in _opus signinum_ with inserts of limestone tessera. A sixth _dolium_ was located in the north-west corner of Room V. It lay on its side: perhaps it was being cleaned shortly before the fire broke out. The building’s walls seemed to “abut” the terrain inside a large cut in the bedrock (serpentino) and the structure’s stability was thus derived from the effect of the counterthrust from the geological formation in the zone. A series of important finds were made inside Room V: -a smaller _dolium_; -an iron element, found between the head of wall USM149 and dolium US188; -an unusual concentration of black glaze askoi lying on the floor surface a few centimetres east of _dolium_ US2254. The building situated inside area 1 appeared to have only one construction phase. There is little doubt that the structure was built sometime around 100 B.C., however the date of its destruction remains problematic. This event was previously thought to have occurred around about the mid 1st century A.D., but this no longer appears to be the case given that no materials post-dating the first half of the 1st century B.C. have been found. In area 2-6 several striking structures suggest that part of the pars rustica was a high quality structure. This may have been an intermediate stage between the farm-wine cellar, datable to around 100 B.C., and the villa situated on the promontory, datable to no earlier than 50 B.C. Area 5 was opened with the aim of identifying ancient metalworking structures possibly lying below layers formed by accumulations of metalworking waste and fragments of furnaces for the reduction of iron minerals created by intense activity to recover ancient iron slag (in the 1930s and 40s). As the mechanical excavation proceeded, it became increasingly clear that this activity had severely disturbed the entire stratigraphy in the area. An accumulation of metal waste finally emerged at a level characterised by large-scale emergence of salt water that made it impossible to continue the excavation. This large complex can certainly be interpreted as the pars rustica of the Villa delle Grotte, situated on the adjacent promontory and excavated in the 1960s. The work for the construction of the _pars rustica_ also represented the end of the metalworking phase that has yet to be identified below the late Republican buildings. In fact, the metalworking waste, which the builders must have found in large quantities on the site, was used as the make up for the _opus signinum_ floors.

The site of S. Giovanni is situated at the eastern end of the coastal plain of the same name. This season’s excavations continued the investigation of the _pars_ _rustica_ of the Villa delle Grotte, which began last year. The results can be synthesised as follows: The area in which the five _dolia_ _defossa_ were situated (Room III) must have been roofed rather than being open-air, as attested by the large quantity of carpentry nails found. Among other things, the identification of five square stone plinths at the edges of the area suggests the existence of a covered courtyard that opened to the north. Here, access to the _dolia_ was probably provided by raised timber planks that rested on the building’s perimeter walls. The floors were made of _opus_ _signinum_ with inserts of limestone tesserae. The hypothesis suggesting the building situated in area 1 had only one construction phase appears to be confirmed. If its construction can be dated to around 100 B.C., perhaps two decades earlier, the date of the building’s destruction is rather more controversial. The earlier hypothesis, that dated it to around the mid 1st century A.D., now needs to be revised in consideration of the fact that there are no materials dating to any later than the mid 1st century B.C. The analysis of the biological material inside a Dressel 1C amphora found inside Room 1 seems to indicate that it contained cider. Other palaeo-botanical analyses are adding considerable information to the picture of the plant landscape on the island of Elba in the late Republican period. At this stage of the research, the interpretation of the building at S. Giovanni needs to be revised. The complex, whose actual size has yet to be determined, cannot be considered as the _pars_ _rustica_ of the Villa delle Grotte situated on the promontory above. It precedes the construction of the large villa by at least 50 years and should be considered as a settlement built at least two generations earlier. It was probably a villa that was still of the Catonian type, inspired however by models completely different to those of the Villa delle Grotte.

The 2017 excavations on the site of San Giovanni, situated at the eastern end of the coastal plain of the same name, aimed to continue the investigation of the rural villa begun in 2012. The Roman villa is thought to have belonged to the _latifundium_ of the _gens_ _Valeria_. The building had two floors and a wine cellar with six large _dolia_ in which the wine was fermented. The villa also had a room for the storage of cider in amphorae. It was immediately clear that the villa’s structures were in a very good state of preservation: it was built in the late 2nd century B.C., mainly in unbaked clay, and destroyed in the 1st century A.D. by a fire which, “baking” the structures guaranteed their survival. During the previous campaigns, only one construction phase was identified for the building located in area 1. Based on this season’s results it can be suggested that a first construction phase, datable to between the late 2nd century and the early 1st century B.C., was probably followed by the blocking or alteration of some rooms, in a period that is difficult to date as the structures were in unbaked clay and the changes occurred very soon after the original construction. There is little doubt about the construction date, somewhere round about the year 100 B.C. (based on the ceramics and inscriptions on tiles and _opus_ _doliare_). The date of the villa’s destruction caused by a disastrous fire seems to date to within the 1st century A.D., given the absence of materials datable to the 2nd century A.D. However, only the full study of the finds from the last two campaigns will provide indications that are more precise. The Roman villa was constructed a few generations before the large villa delle Grotte, situated on the adjacent promontory and excavated in the 1960’s. That villa was built in the Augustan period, and probably used this one as its pars rustica, and for the storage of foodstuffs. The construction of the building at S. Giovanni also represents the end of the metallurgical phase in the bay of Portoferraio, yet to be identified below the late Republican levels. The waste products, which the builders must have found in abundance on the site, were in fact used as make-ups for the _coccio_ _pesto_ and _opus_ _signinum_ floors. With the discovery of burials on top of the villa collapse and evidence for the robbing of the villa structures this season’s excavations also confirmed that following the building’s destruction there was a later occupation phase, only previously suggested by a few scarce remains. This occupation may date to the late antique period when similar phenomena are attested at the villa delle Linguelle and villa delle Grotte on Elba.

The 2018 excavations on the site of San Giovanni, situated at the eastern end of the coastal plain of the same name, aimed to continue the investigation of the rural villa begun in 2012. The Roman villa is thought to have belonged to the _latifundium_ of the _gens_ _Valeria_. The building had two floors and a wine cellar with six large _dolia_ in which the wine was fermented. The villa also had a room for the storage of cider in amphorae. It was immediately clear that the villa’s structures were in a very good state of preservation: it was built in the late 2nd century B.C., mainly in unbaked clay, and destroyed in the 1st century A.D. by a fire that by “baking” the structures guaranteed their survival. The villa had two storeys and, the removal of the collapse of the upper floor revealed high-quality wall paintings with 1st Style Pompeian frescoes and moulded stucco. The previous excavations showed that the building situated in area 1had a single phase. However, this year’s excavations suggest there was a first construction phase, datable to between the late 2nd century and the early 1st century B.C. This was probably followed by interventions to close or alter rooms, which are difficult to date as they were pisé structures and the alterations took place very soon after the original construction. There is little doubt that the villa was built around the year 100 B.C. (based on the study of the ceramics and inscriptions found on tiles and the _opus_ _doliare_). The suggested date for its destruction, following a disastrous fire, is sometime within the 1st century A.D., given the lack of 2nd century A.D. materials. The last two excavation campaigns clearly showed the existence of a subsequent occupation phase, only concealed by a few remains found during previous seasons, attested by the presence of a burial on the villa’s collapse, by robbing activities and the presence of small hearths and forges. This occupation probably dates to the late antique period when, similar phenomena are documented in the contexts of the villa della Linguella and the villa delle Grotte on Elba. This building was erected some generations before the large villa delle Grotte (situated on the adjacent promontory and excavated in the 1960s). Of Augustan date and situated on the promontory above, it came to represent the _pars_ _rustica_, where wine and foodstuffs were stored. The construction of the building at S. Giovanni also represents the end of the metallurgical phase in the bay of Portoferraio, yet to be identified below the late Republican levels. The waste products, which the builders must have found in abundance on the site, were in fact used as make-ups for the _coccio_ _pesto_ and _opus_ _signinum_ floors.