Hypogea 2023

200 St. Euphemia’s cave inscriptions: ancient navigation, beliefs and devotion (Vieste, Italy) cated near the rock context, confirmed the presence of a freshwater spring, active until recent times (fig. 5). The presence of wells and freshwater storage facilities can be related to the need for water supply for seafarers: especially in ancient times, the route was chosen not only for its length, but also for its safety in relation to winds and weather conditions and the presence of intermediate stopovers useful for logistical or commercial needs or to prevent any unforeseen situations; one of the important factors was precisely the possibility of stopping during the voyage to refuel with fresh water, the so-called aquata (Leone and Turchiano, 2017). Refuelling points were also always indicated in portolans, along with conspicuous points, warnings of dangers (shoals, shallow waters, etc.), advice on advantageous ports, landings, and anchorages in relation to winds, the direction to follow, favourable and unfavourable winds for certain routes. Archaeological investigations have also recovered elements on the dimension of the sacred, a further characteristic of the frequentation of this islet over time. A survey and a complete mapping of the vast repertoire of inscriptions covering the walls of the two cavities were carried out, testifying to the cult function of the settlement, its Christianisation during the Middle Ages, and at last its use in modern and contemporary times. In a general perspective, archaeological investigations have begun to shed light on a vast rock system, whose traces remain clearly legible all over the rocky face of the islet, and of which the two cavities constitute only a part. The numerous traces of other niches and structures support the hypothesis that a small community of people engaged in productive activities and cult management lived here. In this sense, the conspicuous quantities of ceramic materials (amphorae of African and eastern origin, tableware and kitchenware, common pottery, etc.) from the Late Antique period (IV-VI AD) can be interpreted. These findings confirm the frequentation of the islet in Late Antiquity as well when the conversion of the pagan sanctuary into a Christian sanctuary can be dated and allow us to reconstruct the dense network of Mediterranean trade in which Uria was still fully integrated. Finally, two burials also document the funerary destination, which seems to have developed from the postclassical age, as suggested by the C14 dating of some bone fragments (V-VI and XI-XII centuries AD) and the Late Antique pottery recovered inside. Shrine inscriptions Research activities also involved the survey and analysis of the inscriptions found in the two cavities. The smaller one is a small chamber, on the walls of which there are some graffiti inscriptions, mainly from the modern period. The room is closed externally by a modern wall and is internally equipped with an arcosolium and several small niches. The main cavity, on the other hand, is a large, very irregularly shaped room with a floor of slabs made in recent times (fig. 6). The cavity has undergone various modifications over the centuries, and only traces of a rather regular parallelepiped room and some flaps of a massive, moulded lintel remain from the earliest phase in the front, near the present entrance. On the other hand, the shape of the rear part of the hypogeum, which is much more irregular and characterised by two large niches on the short sides, is in all probability later, perhaps late antique and medieval. Lastly, the transformation into a dwelling and warehouse which has severely compromised the legibility of the oldest inscriptions dates to modern and contemporary times. A dense array of inscriptions stretches along the walls of the cavity, reflecting the long chronological span of frequentation, from Roman times to the 20th century, and testifying to sacred use in ancient and medieval times, but also to later uses after the construction of the 19th-century lighthouse. The oldest inscriptions are located at the front of the main cave. They consist of at least four texts (the poor state of preservation makes it impossible to tell if there were more), datable between the 2nd and 1st centuries BC, in Greek and Latin, in which the name of Venus Sosandra recurs (fig. 7). These exceptional epigraphs constitute only a small part of an extraordinary heritage of around two hundred engravings traced on the walls of the main cave. These are simple and small signs, the most common being crosses, in a repertoire of shapes and sizes Fig. 5 – Inspection of the well and freshwater spring (photo M. Vitelli).

201 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa ranging from very large symbols to tiny, evanescent engravings found almost everywhere on the walls. Other more elaborate inscriptions give the names of presbyters and other individuals who frequented the shrine and testify to the transformation of the site into a Christian shrine during the centuries of the Middle Ages. Extraordinary, both for its monumentality and its historical importance, is the inscription engraved inside a niche, recalling the passage, in the year 1002, of the Venetian doge Pietro II Orseolo, who stopped by the island with hundred ships on the route that would bring him to the rescue of Bari besieged by the Saracens (Beatillo, 1637). The last two lines, reporting the victorious outcome of the battle, were added by a different hand later (fig. 8). Other visitors engraved their names, and fortunately also the year, between the 17th and 19th centuries, but it was mainly the lighthouse keepers stationing on the island after the construction of the lighthouse in 1867 who left a mark of their stay in the cave. It is conceivable that it is precisely the presence of this vast heritage of inscriptions that may have encouraged them to draw their names and to create inscriptions that were also very meticulous in both outline and decoration, giving rise to an original practice, which, however, is at the same time the cause of the loss of the oldest inscriptions. The presence of some intact or more probably abraded epigraphic frames is a clear indication, at least in some places, of the continuous reworking of the walls and the erasure of previous testimonies, if not of a precise intention of damnatio memoriae. The cult of Venus Sosandra The sea has always been perceived as a complex, unknown, and dangerous reality, which, in imagination and in real experience, has been imbued with multiple meanings and values. Throughout the centuries of antiquity, a sacred dimension linked to navigation has therefore developed, as witnessed also by the number of cults and deities linked to sea travel, from Zeus to Poseidon, Apollo, Hera, Artemis, Venus, Isis, Serapis and the Dioscuri, each associated with very precise aspects of the marine dimension (Lamboley, 2000). Venus, for instance, always had a positive effect, with her ability to calm the waters and ensure a peaceful voyage for sailors. As is well known, shrines to Aphrodite-Venus, the celebrated euploia deity, protector of seafarers, dotted the Mediterranean coastlines, particularly at strategic points along cabotage routes, landing places, river mouths, headlands (Sandberg 1954). Examples include the sanctuary of Aphrodite Ericina in Sicily, the extramural sanctuary of St. Venera in Paestum, the sanctuary of San Giovanni in Venere at the mouth of the Sangro on the Adriatic coast, or Populonia. The association between such sanctuaries and the practice of so-called sacred prostitution is Fig. 6 – Three-dimensional survey of the main cave (G. De Felice).

202 St. Euphemia’s cave inscriptions: ancient navigation, beliefs and devotion (Vieste, Italy) also well known; once again, Populonia yields a possible trace of a hierodula (i.e., a sacred slave), Filica, whose speaking name (‘loving’, ‘delightful’, ‘amiable well disposed’) is graffitied on a black-painted cup (De Tommaso and Patera, 2002). In the case of Vieste, the devotion to Venus Sosandra (the saviour of men), attested by the inscriptions on the main cavity of the islet of St. Euphemia, is to be seen in relation to other attestations: Catullus seems to allude to the presence of the cult of the goddess when, in Carme 36, he mentions Uria (... Nunc o caeruleo creata ponto, quae sanctum Idalium Uriosque apertos quaeque Ancona Cnidumque harundinosam...) among the localities connected to the veneration of the goddess. In this regard, mention should also be made of the discovery of a pre-Roman cippus with a dedication in Messapian language to Zeus, Demeter and Aphrodite and a draped statuette of the goddess, which occurred in front of the islet of in 1930, which has unfortunately been lost. The sea of the sacred Other coastal shrines are attested in the Otranto Channel and, more generally, in the Adriatic. The coastal landscape of the lower Adriatic, in particular, is, in fact, marked by the presence of small bays, promontories, cliffs and islets, crucial places of navigation, also connoted as places of worship. These are sometimes true ‘coastal maritime sanctuaries’, which not only represented mere geographical references or emergency shelters, but also symbolised the interaction between needs related to sea travel, religious beliefs, superstition, devotion and, in some cases, economic practices (Lamboley, 2000; Fenet, 2005). In the maritime space of the lower Adriatic, the religious and devotional dimension connected to navigation and, in some cases, to the economic sphere, have taken on peculiar forms that define an entirely original ‘sacred geography’ and landscape of the sacred. Emblematic cases are, along the coast of Salento, the promontory of Roca with the sanctuary of Grotta Poesia, and Punta Matarico with the cave of San Cristoforo, gravitating on the inlet of Torre dell’Orso; or again, Punta Ristola, with Grotta Porcinara, in connection with the Leuca landing place (Pagliara, 1971- 73; 1987; 1991; Auriemma, 2001; 2004). The bay of Grammata, with its open-air sanctuary along the Karaburun peninsula on the Albanian side, also falls within this category, as it is an isolated place located at a ‘strategic’ passage for navigation, the Strait of Otranto, also characterised in Antiquity by a sinister reputation for danger (Volpe et al., 2013; Leone and Turchiano, 2017). Many of these ‘sanctuaries’ belong to the so-called euploia sanctuaries, identifiable with places frequented by sailors who left traces of their devotion, beliefs and hopes in the face of the Fig. 7 – Two of the dedicatory inscriptions to Venus Sosandra (photo G. Volpe).

203 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa dangers of the sea through inscriptions, engraved directly on the rocks, expressing either vows before departure, or thanks after the crossing, or through other manifestations. A typical characteristic of these sanctuaries, and of the context of St. Euphemia, in which traces of worship seem to be confined to the underground environments, is the absence of more structured constructions or other cultic devices. The architectural and monumental aspect, in fact, was not a relevant element for these sites, which, particularly in the Adriatic, are characterised by their integration into the natural landscape. Priority factors underlying the location and fame of such coastal shrines were their geographical position, their maritime visibility, their role as topographical markers and their function as landing places for navigation. The archaeological research activities on the islet of St. Euphemia have revealed a context of great importance not only for reconstructing the articulation and transformations of the rock sanctuary that took place over the centuries, but also for acquiring valuable information on the general organisation of the structures and devices connected to navigation. When it appeared on the horizon to sailors, the islet must not only have represented a temporary landing place, an anchorage point where one could pause while waiting for weather conditions to improve and replenish the water supply, but was also a stopover on the cabotage route and the longer-distance routes that wound their way along this stretch of the Adriatic coast, both as a point of nautical orientation and at the same time as a place of devotion, the recipient of the prayers of sailors, to whom it must evidently have been well known. Acknowledgements Images by Giuliano De Felice (Uniba), Maria Potenza (Uniba), Marco Vitelli (ASSO ETS) and Giuliano Volpe (Uniba). Bibliography AA.VV., 1998, Uria Garganica e la grotta di Venere sull’isola del Faro di Vieste (3rd century BC), 1998, Atti del Convegno Internazionale di Studi (Vieste, 17-18 October 1987), Foggia. Auriemma R., 2001, Gli approdi minori del Salento Adriatico: il contributo della ricerca archeologica subacquea, in Zaccaria C. (ed), Strutture portuali e rotte marittime nell’Adriatico di età romana, Antichità altoadriatiche XLVI: pp. 415-429. Fig. 8 – The inscription of Doge Pietro II Orseolo (1002 AD). Other inscriptions graffitied on the sides can be seen (photo G. Volpe).

204 St. Euphemia’s cave inscriptions: ancient navigation, beliefs and devotion (Vieste, Italy) Auriemma R., 2004, Salentum a salo. Porti, approdi, merci e scambi lungo la costa adriatica del Salento, Congedo, Galatina, 360 pages. Beatillo A., 1637, Historia di Bari Principal Città della Puglia nel Regno di Napoli, Napoli. D’Ercole C., 2015, Les commerces dans l’Italie adriatique (VIe - IVe s. a.C.): notes pour un bilan, in Marion Y., Tassaux F. (eds), AdriAtlas et l’histoire de l’espace adriatique du VIe s.a.C. au VIIIe s.p.C., Actes du colloque international de Rome (4-6 November 2013), Bordeaux: pp. 403-418. De Felice G., Leone D., Turchiano M., Volpe G., 2020a, Un santuario marittimo lungo le rotte adriatiche. La grotta di Venere Sosandra a Vieste, L’archeologo subacqueo XXVI: pp. 16-24. De Felice G., Leone D., Turchiano M., Volpe G., 2020b, Il mare e il sacro. Landing places of men and gods. La Grotta di Venere Sosandra a Vieste, in Martinelli N., Carlone G. (eds), Vieste. Il faro di Sant’Eufemia, Mario Adda Editore, Bari: pp. 40-50. De Tommaso G., Patera A. (eds.), 2002, Il mare in una stanza. Un pavimento musivo dall’acropoli di Populonia, Museo archeologico del territorio di Populonia, Piombino. Fenet A., 2005, Sanctuaries marins du canal d’Otrante, in Deniaux E. (éd.), Le canal d’Otrante et la Méditerranée antique et médiévale, Colloque organisé à l’Université de Paris X-Nanterre (Nanterre, 20-21 November 2000), Bari: pp. 39-49. Lamboley J.-L., 2000, Les cultes de l’Adriatique méridionale à l’époque républicaine, in Delplace C., Tassaux F. (éd.), Les cultes polytheists dans l’Adriatique romaine, Bordeaux: pp. 133-141. Leone D., Turchiano M., 2017, Liburna 1. Underwater archaeology in Albania. Ports, landings and maritime routes, Bari. Lippolis E., 1984, Testimonianze di età romana nel territorio garganico, in La ricerca archeologica nel territorio garganico: pp. 171-196 (= in Uria garganica, 183-201). Mazzei M., Volpe G., 1998, La documentazione archeologica di Vieste: la città e il territorio, in Uria garganica: pp. 119-158 (= Studi Bitontini, 63, 1997, 5-28). Pagliara C., 1971-73, La Grotta Porcinara al Capo di S. Maria di Leuca. I. Le iscrizioni, AnnLecce 6: pp. 5-67. Pagliara C., 1987, La Grotta Poesia di Roca (Melendugno-Lecce). Preliminary notes, AnnPisa 17.2, s. 3: pp. 267-328. Pagliara C., 1991, Santuari costieri, in I Messapi, Atti del Trentesimo Convegno di Studi sulla Magna Grecia (Taranto-Lecce, 4-9 October 1990), Taranto: pp. 503-526. Petrone M., 1984, Note di storia antica garganica e viestana, Vieste. Ruggieri V., 1989, Una nota sulla topografia medievale di Vieste, Vetera Christianorum 26, 361-369. Russi A. 1971, Uria Garganica and a new funerary inscription, in Terza miscellanea greca e romana, Rome: pp. 211-223. Russi A., 1992, Sul complesso epigrafico rupestre della “Grotta di Venere Sosandra” nell’isolotto del faro di Vieste (Foggia), in Uggeri G. (ed.), L’età annibalica e la Puglia, Atti del Convegno (Mesagne, 24-26 March 1988), Fasano: pp. 103-110. Russi A., 1998a, Navi, marinai e dèi in epigrafi greche e latine scoperte in due grotte del Gargano nord-orientale, in Laudizi G., Marangio C. (a cura di), Porti, approdi e linee di rotta nel Mediterraneo antico, Atti del seminario di studi, (Lecce, 29-30 novembre 1996), Lecce: pp. 104-135. Russi A., 1998b, Gli dei e il mare: le grotte del Gargano, in R. Cassano, R. Lorusso Romito, M. Milella (eds.), Andar per mare. Puglia e Mediterraneo tra mito e storia, Bari: pp. 97-102. Sandberg N., 1954, Euploia. Études épigraphiques, Acta Universitatis Gotoburgenis, Göteborgs Högskolas Arsskrift LV, 8, Goteborg. Volpe G., Leone D., Turchiano M., 2013, Il progetto Liburna: Ricerche archeologiche subacquee in Albania (campagne 2007- 2010), AS Atene LXXXIX, s. III, 11 (2011): pp. 251-286.

205 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa 1   Doctor of Architecture, Professor of Geoecology, NUACA; President of the Armenian Speleological Center 2   Sumerologist, Linguist, Byurakan Observatory, [email protected] 3   PhD, Lecturer at Brusov State University, the Armenian Speleological Centre, [email protected] 4   Archeologist and Lecturer, Institute of Ethnography and Archaeology and Yerevan State University, [email protected] *  Reference author: Samvel M. Shahinyan - [email protected] Use of natural caves for ritual purposes as a forerunner of the formation of architecture Samvel M. Shahinyan1,*, Armen Davtyan2 , Smbat Davtyan3 , Boris Gasparyan4 Abstract At the dawn of the civilization, in the processes of transition from totems to the formation and establishment of regulated naturalistic religions, special natural geomorphological and landscape locations were selected for worshipping rituals For example, the Platanus Forest in Shikahogh Reserve, Portakar in Sisian, megalithic monuments – Vishaps in high and visible places of the landscape (Bobokhyan et al., 2019), sacred Eridu (Lloyd, 1984) reed pond or basin in Sumer, Mesopotamia (Schmidt, 2001; Mellaart, 1982; Lloyd, 1984) etc., including caves. Later on, in conditions of continuous development of civilization, drastic changes in public requirements occurred. Buildings, and later on temples were built in the sites of rituals, with explicit canonic structures and appearances Such as the Eridu sacred basin (Meller and Avetisyan, 2011; Mellaart, 1982; Lloyd, 1984) where reeds were growing and there were fish swimming. Ziggurats were constantly built and rebuilt on this basin for almost a millennium. Another typical example are the Urartian sun temples in the form of a cube (Shahinyan, 2004, 2011b; Shahinyan et al., 1985), and three columns in each of the two rows in front of it, with thick anchors and an equal-winged cross-shaped plan in the basis. Ayanis (Turkey), Aghdzk, Erebuni, Karmir Blur (Armenia). (Haykazun, 2005; Meller and Avetisyan, 2011). Parallel to architectural structures, cavesanctuaries built by the pantheistic principle were continuously preserved almost until the present day. They have been in use from time to time; however, the natural appearance was not distorted and no construction interventions were made. The caves formed as a result of denudation processes of water in the lithomarges of the hypogenic-clastokarst layers in Areni-1 and those of the basalt layers in “Red Cave” of Geghamavan and Byurakan (Khechoyan et al., 2014); (Gasparyan et al., 2003); (Shahinyan et al., 2022); (Gasparyan et al., 2014) have been considered sacred for many centuries. Areni-1[5, 16] was a place of worship of grape and wine, while the caves of Yeghvard and Byurakan were water worship places [11]. Rituals were held in the caves. However, speleological and archeological research has revealed that no changes in the natural appearance or other changes caused by human intervention were noticed for centuries, other than rock falls. This phenomenon is clearly expressed in the caves of Areni-1 and Yeghvard, Gegamavan-1 which have been sacred places even in the High Middle Ages (tab. 1: figs. 1, 2; tab. 3). Keywords: natural cave sanctuaries, underground temple buildings, temple architecture. The problem and the goal of our research Caves such as the one with a water source located on the river bank, next to the Sun Temple/Tomb characteristic of the Urartian times of Byurakan, or Jrovank in Vayk, located at the end of the raven of one of the tributaries of Arpa river, have not been altered by man and have been preserved in their natural form. And the rock wall of the spring source in the deepest part of Geghard monastery has not changed, although the remaining three walls of the hall were cut and treated in High Middle ages. This is not the case with Areni-1 cave (Shahinyan, 2004; Gasparyan and Sargsyan, 2003), where there is no spring though taking into account the geological environment of cave formation. i.e., the limestone conglomerates, we can argue that there were water flows from the cave from time to time.; however, for many millennia, in the conditions of changing times of different social formations and confessions, rituals did take place, using ritual items, since there are lots of artefacts and a winepress found during the excavations, with no change whatsoever in the appearance of the cave. This approach has been persisting for over 6000 years. Even with the many examples of rock-cut churches and sacred structures survived in the territory of the Republic of Armenia and neighboring countries, all of them cannot compete with Areni-1, even in terms of longevity; for example, Geghard in Armenia, Hochants in Artsakh, Karaftu in Iran, The Bayazet rock cut temple in Turkey. Etc. Our research was aimed at identification of ritual objects, such as winepress and wine amphoras, in naturally occurred and not made by human hands caves that turned into sanctuaries as well as versions of functional separation of space for rituals.

206 Use of natural caves for ritual purposes as a forerunner of the formation of architecture Archaeological excavations, research and the latest discoveries (over the recent 30 years) uncover new realities of the period from the Middle Stone Age to Copper and Bronze ages, which are related to the origin of religions on the pantheistic basis, and ritual ceremonies. Rituals and ritual functions in the cave required functional spaces for the class of priests and the religious society (TTa). Taking into account the geological environment of cave formation. i.e., the limestone conglomerates, we can argue that there were water flows from the cave from time to time. In this regard, it is difficult to explain why the Areni-1 cave, which has been a sacred place for centuries and millenniums, has never undergone any architectural intervention. No structures of the period of the Kingdom of Ararat (Urartu), nor that of the Christian period was discovered, while we are well aware, and the bibliographic sources have documented that the first Christian churches were built in the place of old temples or in their foundations. Moreover, the ritual functions in the naturally formed caves have aroused the need of not only separation of the functional spaces, but also of particular items and structures for conducting the rituals (Gasparyan and Arimura, 2014). For example Areni Cave, winepress, wine amphoras with ideabearing images and magical (tab. 2). For centuries and even more, those rituals and ritual items have been processed, standardized, and received a special style and shape. The form and the design of ritual and worship items have crystalized also in the form of symbol-bearing signs (symbols) or idea-bearing images composition of symbols For example, Tree of Life, Khachkars. see: (Shahinyan, 2012a; Arqepiskopos, 2016) and ended up in the form of compositions with idea-bearing signs: B. Gasparyan, Pottery with dragons and with the symbol of the sacred Sun (tab. 2: fig. 1, 2, 3). We are concerned with one question with many whys: a number of ritual artefacts of different and confirmed periods found in the archaeological layers of the above-mentioned caves indicate their use for religious purposes for millennia. Why there are no facilities and architecture characteristic of the sacred structures of religions alternating each other, such as in Geghard, where, in the spring hall, we see both construction and architectural intervention? Written and archaeological information on the permanence and the longevity of sacred sites Developments and variations of emergence and architecture of sacred temples have stemmed from various ancient pantheistic faiths, in accordance with the dogmas and rituals of those religions (Sargsyan, 2022; Toramanyan, 1984) (tab. 5). However, the change in belief did change the geomorphological landscape or the place of rituals, in our case - the Tab. 1: Fig. 1 – The Areni-1 cave. (The floor plan S. Davtyan; Graphical drawings A. Badasyan). Tab. 1: Fig. 2 – The Frontal view of Areni-1 Cave.

207 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa cave. In other words, the temple of the new faith was built on the foundations of the old one or on the old one: for example such as Mother See of Holy Echmiadzin, under which a pagan temple with ritual items was discovered. This very fact was recorded since the 5th century in the chronicle works of the Armenian historians, such as Khorenatsi and Agathangelos (Khorenatsi, 1913, 1990; Agathangelos, 1985). The cave sanctuaries are especially durable in this regard. Even with no reliable information about the pagan period of Geghard (Ayrivank-is located north of the village of Garni in Kotayk marz of Armenia, on the slope of Goght tributary of Azat river) due to the lack of relevant archaeological research, we can nonetheless state with confidence that it has been a pre-Christian sacred complex for a very long Tab. 2: Figs. 1, 2, 3 – The worship of the Tree of Life and the Sun has always been an integral part of Armenian culture. Among the characters in mythology are the two dragons who guard the Tree of Life, the Sun, the Seed, and the cube-shaped crust, the Earth, the Sky, from Chaos (Shahinyan, 2010). Tab. 2: Fig. 4 – Sacred amulet with the image of the foot of a typical representative rodent of Areni-1 cave biotope. Fig. 1 Fig. 3 Fig. 2 Fig. 4

208 Use of natural caves for ritual purposes as a forerunner of the formation of architecture time. We learn from written sources that Armenian Catholicoses have lived and worked here since 301: Grigor Lusavorich (Gregory the Illuminator), who founded the first church (probably in a cave) in the place of a sacred pagan spring, then Nerses I Partev, Sahak A Partev, Hovhannes D Ovayetsi, Hovhannes E Draskhanakerttsi and Catholicos Stepanos of Aghvank (Albania). In other words, St. Geghard Monastery has been continuously operating during the entire period of Christianity, more than 1700 years. However, we have chronicled information about the buildings and architectural interventions only starting from AD 7th – 9th centuries. This means, after the Christianization of the pagan sanctuary, for hundreds of years, the rituals were performed in a natural cave environment. There are many examples of similar longevity throughout more than 15 thousand years of the hisTab. 3 – Geghamavan - 1. In the top row, the entrance to the cave, on the right is the Persian inscription, which testifies to the fact that the spring is sacred. In the bottom row: The part of the canyon where you can go down into the cave. On the right side. At the beginning of the trail leading to the cave, a petroglyph carved into the basalt.

209 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa Tab. 4 – Comparison between Sumerian and Armenian words and their meanings.

210 Use of natural caves for ritual purposes as a forerunner of the formation of architecture tory of human civilization. The fact that Gobekli Tepe is a result of civilization, can be easily proven calculating the natural resources of the particular biotope, which was necessary for human existence and get convinced that the number of nomads and gatherers could not possibly provide the opportunity to create a large army of masons who could create 27 temples even for 5 centuries. One of them is Enki Temple in Eridu. Archaeological excavations have shown that the temple, which has been built and rebuilt for millennia, was originally just a reedy lake full of fish. It was sanctified and became a place of worship. This is evidenced by the many bones remains of Getatsatsan (Carp) fish found during excavations. As a result of his studies, Yu. Sargsyan (Sargsyan, 2022), based on the geological and geomorphological features of the terrain, on the research of F, Safar, S. Lloyd and other archaeologists, as well as the legacy that has come down to us from the Sumerian folklore culture, draws three interesting and fundamental conclusions. The first conclusion is that the lake was fed by underground springs and not terrestrial waters. The second conclusion is that the lake has remained intact underneath the foundations of ziggurats built and rebuilt dozens of times on the lake. And thirdly, the Enki Temple was the only one, where, due to the specificity of the geological structure, the Holy Font unlike all the known ziggurats of Mesopotamia, was constantly existed under the ziggurat: In our opinion, it is an interesting and also scientific hypothesis that should be put into circulation. In other words, in the case of the Sacred Pond, we are dealing with an underground or cave pool. It is worth noting that the Sumerian word “avazan” is used in Armenian with the same sound and meaning to this day (“font”, “pond”, “basin”) (tab. 4); moreover, the babtismal font, avazan, is the same word “avazan” used in Sumerian (Shahinyan, 2011a) (tab. 4). Tab. 5 -– The plan of the Zvartnots and baptismal font (Avazan). The floor plan (by Toramanyan, 1984; graphical drawings by A. Badasyan). 1 – River and lake sedimentary rocks, boulders, cobbles, sands, clays (4 meters). 2 – Lake sediments. Sandstones and clays (6 meters). 3 – Clays, sandstones (36 meters). 4 – Basalt permeable (28 meters). 5 – Sandstones (7 meters). 6 – Basalt deep fissured aqueduct (42 meters) (S. Shahinyan).

211 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa Results of research, analyses and conclusions Research of Areni-1 cave began in 1979. The scientific exploration began with the efforts of the Speleological Expedition of the Geographical Society of Armenia. The task of the expedition was to describe the morphological and geological features, the faun and the archaeological perspectives of the cave. Based on the research results, the Scientific Council of Geographical Society of the Arm SSR AS proposed in 1982 to start archaeological research. The ethnography and archaeology expeditions of the RA NAS started the excavations only in late 1990s. Excavations became more effective with international scientific expeditions getting involved. The excavations were participated by archaeologists from Armenia. U.S., UK, Germany, Japan, Israel, The Netherlands, etc. Institute of Ethnography and Archeology of NAS 2010-2021. The discovered and classified archaeological material made it possible to get an idea of the numerous different, sometimes, radical, changes in the social life of the indigenous ethnos over the course of 6200 years. At the very beginning of the archaeological work, the separation of the archaeological layers at the entrance of the cave with a large opening, caused significant difficulties. The reason was that in different eras, the people who used the cave for different ritual and other purposes made holes for amphoras and for other purposes and thus disturbed the sequence of layers of archaeological significance. It should be stated to the credit of archaeologists that they accomplished the task assigned to them with honor and were able to restore both the chronology and the stratigraphy of the cave culture with the help pf the chronology of each finds. The solution of this problem was greatly supported by the excavations of the large hall deep in the cave, where the stratigraphy was comparatively undistorted unlike the that at the entrance. As a result, we have now the opportunity to get an idea on g’inedzon (“libation”) ritual as close to the reality as possible (see the plan of the cave and the explanation in tab. 1: fig. 1). Based on the findings presented in this article and the opinions within the many publications on the basis of excavations, we have made the following radical conclusions: • Areni-1, Geghamavan, Jrovank, Geghardavank and other sanctuaries for millennia have been sanctuaries, regardless of religious dogmas and the type of faith (see the inscription in Persian on the rock wall in Geghamavan). • Caves from which springs flowed out were sanctified. • Pantheistic and pre-Christian religious beliefs have not only been preserved in folklore works (eposes, myths, fairy tales); but also, rituals have long functioned in parallel to the adopted major religion. Summary From a purely architectural point of view, the sacred caves can be divided into three groups. 1. The cave has preserved in its natural appearance as a temple, with the allocated sites and sacred items being allocated for functions making the architectural component. 2. With the cave being partially built, the worshippers of the underground waters and underground deities earnestly and continuously professed their faith until they were adapted to the new faith, incorporating old rites (like blessing of water and blessing of grape in the calendar of the Armenian Church) into the new religion. 3. The temple was built on a cave, sacred spring or a pond, with the former sanctuary in its natural initial shape becoming the integral functional component of the temple (Enki in Eridu - the ziggurat dedicated to the God the Creator Hye). Bibliography Agathangelos, 1985, History of Armenia, YSU, Yerevan, 240 pages. Arqepiskopos S. J., 2016, Jampordutyun I Metsn Hayastan (A trip to Great Armenia), Yerevan (Armenia), Dall LTD, 697 pages. (in Armenian). Bobokhyan A., Gilibert A., Hnila P. (eds), 2019, Vishap Between Fairy Tale and Reality, in Fundamental Armenology № 1 (11) 2020, Publishing House of the Institute of Archeology and Ethnography, Yerevan, 651 pages. Gasparyan B., Arimura M., 2014, Stone Age of Armenia, Institute of Archaeology and Ethnography of the National Academy of Sciences of the Republic of Armenia, Gfoeller Fund of America Corporation, Armenian Branch, Center for Cultural Resource Studies, Kanazawa University, 369 pages. Gasparyan B., Sargsyan G., 2003, A newly discovered group of petroglyphs in the “Red Cave” of Geghamavan. The ancient culture of Armenia in Proceedings of the conference dedicated to the memory of Telemak Khachatryan, 2003, pp. 5-12 (in Armenian). Haykazun A., 2005, Simbolism in Armenian Architecture. ALAS, Yerevan, 177 pages. 106 Tabel. Khechoyan A., Gasparyan B., 2014, Rock- Painting in the RA (pp. 315-337), in Gasparyan B., Makoto A. (eds) Stone Age of Armenia, Institute of Archaeology and Ethnography of the National Academy of Sciences of the Republic of Armenia, Gfoeller Found of America Corporation, Armenian Branch, Center for Cultural Resource Studies, Kanazawa University, 369 pages.

212 Use of natural caves for ritual purposes as a forerunner of the formation of architecture Khorenatsi M., 1913, History of Armenia, Tbilisi 1913, book III, 165 pages. Khorenatsi M., 1990, History of Armenia, Hayastan, Yerevan, 324 pages. Lloyd S., 1984, Archaeology of Mesopotamia. Moscow: Nauka, 279 pages. Mellaart J., 1982, Ancient civilizations of the Middle East (translated from English by Ye.V Antonova, reface by N.Ya. Merpert.), M. Nauka, Chief Editorial Board of Oriental Literature, 152 pages. , “In the wake of the disappeared cultures of the East” series. Meller H., Avetisyan P., 2011, Archäologie in Armenien, Landesmuseum für Vorgeschitchte, Helle, pp. 72-110. Sargsyan Y., 2022, Tayni Araratskikh gor. Yerevan, Secrets of Ararat mountains, Lusakn, 136 pages. (in Russian). Schmidt K., 2001, Göbekli Tepe, Southeastern Turkey. A Preliminary Report on the 1995-1999 Excavations, Paléorient 26.1, pp. 45-54. Shahinyan S., 2004, Caves of Armenia. Vol. A. Zangak, Yerevan, 186 pages. Shahinyan S., 2011a, Global Climate Changes and Pantheistic Spiritual Perceptions, in Armenia Highland Before and After the Last Wurm Glaciation, http://www.inqua2011.ch Shahinyan S.M., 2012a, Khachkar. The sacred system of cosmic creation, Yerevan: YSYAC, 120 pages. Shahinyan S., 2012b, About one peculiarity of rehabilitation of Ancient Sacred Structures of peoples of Indo-European family and medieval Armenian Churches, Proceedings of the 4th International Conference on contemporary problems in Architecture and Construction. Collection of articles, Czestochowa, Poland, pp. 707-712. Shahinyan S., Balyan S., Davtyan S., 1985, Caves of Artashat, Ararat and Yeghegnadzor regions of Arm SSR, report of the speleological detachment, expedition of the GS of RA AS, 1984-85. Toramanyan T., 1984, The temples Zvardnoc and Gagikashen, printing in Sovetakan Grokh, Yerevan, Armenia, 142 pages. (in Armenian).

213 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa 1   Università degli Studi “G. D’Annunzio” di Chieti, Dipartimento di Lettere, Arti e Scienze Sociali (DiLASS), Architect, Research Fellow* - mobile +39 3478707575 - [email protected] Physical evidence of dedication rites in rock churches of Basilicata and Apulia (Italy). Some case studies Sabrina Centonze1 Abstract This article analyzes the physical evidence of the rite of dedication, collecting some case studies, coming from a large investigation on walls, columns, altars and vaults of the rock churches of Basilicata and Apulia, and compares them to other regions and foreign countries cases. We’ll see the colour, the shape and the materials of the surviving dedication crosses, the former presence of relics on the altars, inscriptions and other evidence of dedication, but also some clear deconsecration signs.The article aims to bring to the attention this rite, to give a correct reading of its rare physical evidence in the rupestrian context, where they are often misunderstood, and tries to define a recognition method, to understand and read better the stratifications of a church. Keywords: dedication rite, consecration, rock churches, altars, relics, dedication crosses, painted crosses, graffito crosses, epigraphs, inscriptions. Purification and blessing gestures The dedication rite of a church consists in a series of purification and blessing gestures, including prayers and litanies, undertaken by a Bishop in order to exorcise and hallow any place that was built or carved to become a place of worship (tab. 1). This applies to all newly built or existing places, which went under restoration, new wall decoration, or - even more - profanation. The ancient ceremony of dedication was handed down to us evolving over the centuries. Even the term “dedication” is an evolution, preferred to “consecration”, as it has been recognized as the most ancient term used. * Research Fellow of the European Research Council AdvGrant “Graff-IT Project” - Writing on the Margins: Graffiti in Italy (7th to 16th centuries) - for Diagnostic, Geo-Archaeological Study, Mapping and Monitoring of Hypogea with graffiti. This paper aims to spread and update some previous case studies (Centonze, 2022a) and falls within the activities of the ERC AdG project “Graff-IT”, coordinated by professor Carlo Tedeschi at the University “Gabriele D’Annunzio” of Chieti- Pescara. The Project was funded by the European Research Council (ERC) under the Horizon 2020 Framework Program for Research and Innovation (GA no. 101020613). If in the Old Testament the altar was the only part of the church involved in the ceremony2 , from the Early Middle Ages onwards various liturgical texts set the rules for dedicating the whole place. One of the oldest, concerning the description of the abecedarium phase (tab. C) already appeared in the 8th century (Treffort, 2008, p. 225), and the first official text, the Pontifical of Guillaume Durand, bishop of Mende, was written in 13th century by (Durand de Mende, 1280/1999; Andrieu, 1940); the rite was fixed and printed in the 1485 Liber Pontificalis, then the 1596 Pontificale Romanum extended it to the whole Catholic Church. The current ceremony follows the Ordo dedicationis Ecclesiae et Altaris of the 1977 Pontificale Romanum3 , of which we use the 1994 version. After the construction, the church was prepared for the rite by painting or carving twelve dedication 2 In the Old Testament we see Noah building the first altar and then sacrificing burnt offerings on it (Genesis 8:20); Abraham miraculously found a ram caught in a bush and offered it as a burnt offering on the altar, instead of his son Isaac (Genesis 22:9-13). Jacob erected a stone stele on the place where God had spoken to him, then made a drink offering and poured oil on it (Genesis 35:14-15). In the Exodus God himself instructed Moses how to build a wooden Tabernacle and altar, to overlay them with gold and how to prepare the chrism and the fragrant incense (Exodus 25-27). Solomon built Jerusalem’s Temple in seven years, with a golden altar with four fronts (1 Kings 6-7; 2 Chronicles 3); he also instituted the first Feast of Consecration (2 Chronicles 7:8). 3 The latest Pontifical was elaborated according to the prescriptions of Paul VI after the Second Vatican Council.

214 Physical evidence of dedication rites in rock churches of Basilicata and Apulia (Italy) crosses on the columns or on the walls, at the height of 2.30 m, so that the Bishop used - and still uses - a ladder to anoint them with the chrism, a consecrated anointing oil (tab. 1L), but, first of all, the church was entirely hallowed, outside and inside its walls (tab. 1A, 1F, 1G). The Altar of Footprints To better understand the profound and sacred meaning that makes the rite of dedication a necessity, let’s recall one of the oldest “tangible” signs of dedication in natural caves, coming from an hagiographic memory. According to the mid-8th century Liber de apparitione Sancti Michaelis in Monte Gargano, the legend of the foundation of Saint Michael’s Sanctuary in Monte Sant’Angelo, of Longobard origin (text e traslation in Trotta, 2012: pp. 63-68,159-165), when the Bishop of Siponto arrived on the site of the apparition, the Archangel informed him that the cave had already been consecrated by his presence. This source underlines a key theme of the rite of dedication: the need to neutralize the presence of negative forces in the cave by an exorcism performed by the Bishop, a Papal messenger. But the case of this Sanctuary is unique, because the Archangel appears as a Divine messenger. An anomimous 11th century Irish Homily, discovered in 2007 (Leabhar Breac, XVI Homily, composed by for 29 September dies festus in Picard, 2007: p. 141; Trotta, 2008: pp.161-174; text e traslation in Trotta, 2012: p. 170), reporting the same foundation episode, provides us new information, integrating what we already knew on the origin and conformation of the Sanctuary: first of all the Archangel’s fight with a dragon and his far and wide flight inside the Gargano cave in the form of a beautiful bird («in forma avis»; Centonze, 2022b), Tab. 1 – The dedication rite according to Pontificale secundum ritum sacrosancte Romane ecclesie (Castellano,1520). A) The Bishop and the clergy leave the church, except for a deacon, and begin to bless the water mixed with salt. In the meantime, inside the church the deacon lights 12 candels placed in front of the 12 dedication crosses previously made on columns and walls. The Bishop soaks a branch of hyssop in the holy water mixed with salt and with it he sprinkles the external walls, turning around the church three times, followed by the clergy. B) After any round of sprinkling, the Bishop knocks three times on the door with his Pastoral and at the third round the deacon opens it. After signing a cross on the threshold with the holy Pastoral, the Bishop enters the church among prayers and sacred litanies, symbolically taking possession of the building. C) Abecedarium: a Saint Andrew’s cross is drawn on the floor with the ash where the Bishop marks the Latin and Greek alphabet with the Pastoral. D) The water is mixed again with salt, adding wine and ash. E) On the altar mensa are marked four crosses at the corners of with this holy mixture. F) The mensa is sprinkled with the hyssop, turning around it seven times. G) Then the interior of the church is sprinkled from bottom to top with the hyssop and the Bishop throws the remaining holy water below the altar. H) A shrine is prepared inside the altar for relics, anointing with chrism four crosses inside the corners and putting in four incense grains; then relics enter the church to be placed in the shrine of the altar and then the closing stone is sealed along the edges. I) The altar is dedicated, by anointing five points (middle and corners) with crosses, first with oil and then with chrism. Incense is burned above the same points. L) With a ladder the Bishop reaches one by one the 12 dedication crosses and anoint them with chrism. M) The altar is “confirmed at the front” by anointing its frontal dedication cross with chrism. N) A white altar cloth is laid and the ceremony of the dedication Mass begins.

215 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa aimed at delimiting its sacred border, indeed consecrating the place. He finally imprints on the rock the marks of his claws «that still remain there»4 . The old path that the pilgrims took to the Sacred Grotto passed through the place recognized as the exact point of the apparition: it’s the so-called Altar of the Footprints (tab. 2A), that today is not immediately visible, because it’s now incorporated into the museal itinerary (tab. 2B). We’re speaking of a preeminent nucleus for the Gargano’s cult: a consecrated altar that received the contact with the Saint, it’s itself a relic. Altars and relics In all other cases, the altar was (and is) consecrated by the Bishop with the deposition of relics (tab. 1H,1I) and by anointing the frontal dedication cross (tab. 1M). This cross still survives in some rupestrian examples. In Matera we can see it carved in relief in the monolithic altar of Santa Maria al Vitisciulo (tab. 2C), or fresco painted in San Falcione (tab. 2D) and in Sant’Eustachio, where the cross was also repainted during a rededication (tab 2F). In some cases we can imagine the former presence of relics in an hemispherical housing, “the shrine”, usually excavated in the upper part of the altar, that by now is empty: to give a few examples, this can be seen in Mottola, on the altars of Santa Margherita (tab. 2E) and of San Nicola di Lamaderchia, in San Simine in Pantaleo in Massafra, or again in Matera, in Santa Maria al Vitisciulo (tab. 2C). External and internal inscriptions One of the most representative places of the rupestrian area, that testifies its foundation and dedication through an inscription, is the church of Santi Andrea e Procopio in Monopoli. Here we see an incised cross pattée in the middle of the lunette, and above it an engraved 11th century epigraph (tab. 3A), which mentions the client, the magister and the Archbishop Pietro who officiated the rite5 . Also the anonymous church improperly known as Sant’Andrea, near the Masseria Irene, between Matera and Montescaglioso, has an inscription engraved next to the entrance (tab. 3B) and another text is red painted on an arcosolium. Unfortunately they are both compromised and not readable. Other churches and chapels only show a sacralizing cross on the front lintel (tab. 7B,C): probably several of them have lost their inscriptions, so today only the 4 The manuscripts refers to the signs imprinted in the rock «as if it was soft wax». 5 «Hoc templum fabricare fecerunt Johannes, Alfanus, abbas Petrus, Paulus in onore sancti Andree Apostoli et sancti Procopii martyris per manus Joannis diaconis atque magistri et dedicatum est per manus domini Petri archiepiscopo secondo die intrante mense nobember. Hoc scripta fieri fecit Iaquitnus presbyter, filius suprascripti magistri per manus Rodelberti presbiteri». deeply carved ones survive. It’s very likely that these external inscriptions were painted too and the red pigment has gone for the exposition to the elements. In fact, some of the inscriptions carved inside the places of worship partially maintain the red colour, as we see in the right niche of Cappuccino Vecchio in Matera6 (tab. 3C) and in the apse of the anonymous church at Vallone Tufara of Montescaglioso (tab. 3D), which opens with a crux and closes with EP (abbreviation of Episcopus) and the monogram of the Bishop. Other internal dedication inscriptions are simply painted with a brush or a finger. It’s the case of the right altar rail of San Nicola di Lamaderchia in Mottola, where a red inscription dedicates the chapel to Saint John the Baptist (tab. 3E). Roberto Caprara brought to attention the case of San Giuliano al Bradano in Matera (also identified as San Possidonio by Pelosi and Lionetti, 2021: p. 445): an 80 cm cursive inscription, painted in red on the apse of the church (tab. 3F,G). The reading provided by Caprara «(crux) Ego magister et presbiter indignus»7 (Caprara, 2016, revised in Caprara, 2017) must be amended into «(crux) Ego Magiper presbiter indignus»8 . As for the dating, by calling the writing a “Gothic”, Caprara attributed it to the 14th century. On the contrary, as confirmed by Carlo Tedeschi, the writing is a Beneventan minuscule, dating back to the 11th -12th century. Other investigations (Centonze, 2022a: p. 123) have rebuilt a more complex situation: the red inscription was painted to overwrite the faded trace of a previous red Latin cross9 that was in the apse; in the upper zone there’s a cloud of other red crosses, among which a Jerusalem cross stands out, then at the right follow some potent crosses, a Papal cross (with three crossbars) and a Menorah sign. Moreover on the left there’s another inscription in which the text is partially missing, but the word INTRANTE seems to be repeated; the last line incipit is «(crux) Ego», as we saw in the apse. All these elements attest that the bishop Magiper was in charge of rededicating this church. The Papal cross seems to recall the Pope’s delegation to the dedication, and for this reason the Bishop humbly considers himself indignus, unworthy. Let’s also pay attention to the Menorah, painted in a “V” shape, as It’s described in the Bible, representing the flame that had to burn permanently on the altar10. 6 The text, at the moment is illegible, but in the past someone has read «(crux) Mihi Johannis» (BMR 1916: p. 24). Today it is not possible to confirm or deny the news. A similar imprint is in the first symmetrical niche of the left aisle, but this is even less readable. 7 In Caprara’s reading «(crux) Ego magister et presbiter indignus», was the anonimous signature of the magister, the architect of the church. 8 For this new reading I have to thank Carlo Tedeschi, Full Professor in Palaeography at the “G. D’Annunzio” University of ChietiPescara, and Principal Investigator of the ERC-2020-AdV Graff-IT Project. 9 The old faded cross is decorated at the base with some red and ocra yellow swirls. 10 In the etymology alta res or alta ara, the altar has the meaning of “sacrificial pyre” on which the priests burned incense and where a flame must always burn by divine disposition (Lev 6).

216 Physical evidence of dedication rites in rock churches of Basilicata and Apulia (Italy) Tab. 2 – Altars and relics. A) Altar of Footprints, Sanctuary of Saint Michael the Archangel, Monte Sant’Angelo, Foggia (photo S. Centonze). B) North-south section of Monte Sant’Angelo’s Sanctuary: we see the 12 m natural elevation of the original floor in the Sacred Grotto, the Angevin changes (forepart and walking surface), and The Altar of Footprints that is hidden below (graphic edit from Trotta M., 2012, p. 298). C) Monolithic altar of Santa Maria del Vitisciulo, Matera, with an embossed dedication cross on the front (photo S. Centonze). D) San Falcione’s altar in Matera: the remains of the red Latin dedication cross inscribed in a red frame. In the right quadrants we can hardly read a C at the top and an A at the bottom, preceded by another red sign that stands for K, so we can reconstruct the acronym IC XC and NI KA, «Jesus Christ conquers» in Greek (Antros Archive). E) Altar of Santa Margherita in Mottola: an empty relic chamber carved in the calcarenite monolith (photo S. Centonze). F) Sant’Eustachio’s altar in Matera, with two layered fresco crosses (dedication and rededication) (Antros Archive).

217 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa Tab. 3 – Inscriptions. A) Santi Andrea e Procopio’s church in Monopoli: a central cross pattée and an engraved 11th century dedication epigraph mentioning the client, the magister and Archbishop Pietro who officiated the rite (photo R. Latorre). B) The anonymous church near Masseria Irene, between Matera and Montescaglioso, has an engraved dedication inscription next to the entrance, anticipated by a cross (Antros Archive). C) Dedication inscription in the right niche of Cappuccino Vecchio in Matera, engraved, painted red and anticipated by a Latin potent cross (photo S. Centonze). D) Dedication inscription in the apse of the anonymous church in Vallone Tufara, Montescaglioso: it’s engraved and painted red, with a Latin potent cross at the beginning and the Bishop’s monogram at the end (Antros Archive) E) Red inscription on the right altar rail of San Nicola in Lamaderchia in Mottola, dedicating the chapel to Saint John the Baptist (photo S. Centonze). F) San Giuliano al Bardano, Matera. Sacralizing inscriptions painted red (with a brush or a finger) at the left of the apse (Antros Archive). G) Apse of San Giuliano al Bardano, Matera. Dedication inscription painted red with a brush or a finger, with a Jerusalem cross and several potent crosses, including a Papal cross and a “V” Menorah ( (Antros Archive).

218 Physical evidence of dedication rites in rock churches of Basilicata and Apulia (Italy) Surviving dedication crosses on vertical surfaces: shape, colour, materials and location The shape, the dimensions, the colours and the materials of dedication crosses are always very variable. According to the realization technique and to the surface, we can find them carved in low or high relief on the rock, incised on rock, plaster or previous frescoes, directly painted on the bare rock or fresco painted11. For the shape we can speak of large recurrence of potent crosses and cross crosslets typology, but also the Latin and Greek crosses had their part in dedication. Another recurrent element is the rota, a circle or a frame in which the cross was inscribed. Then we have to mention the compass drawn crosses, like crosses pattée, four or six petals flowers and more rarely eight petals flowers. In the first centuries of the Middle Ages, red and simple shapes were the most frequent kind, the red colour has always been preferred, alluding to Christ’s sacrifical blood, shed on the Cross. In postmedieval and later examples more complex shapes appeared, as interlaced crosses, floral elements and polychrome decorations. Red painted and polychrome crosses An important testimony of red cross dedication is in Ognissanti’s church, located on the Gargano promontory12. Although the church is currently used as a shelter for tools and animals, here we still see some intact attestations of the rupestrian dedication in the internal walls, where we can record ten red crosses. The peculiarity lies in the central hole of the carved and painted Latin crosses (tab. 4A), which was used to house the wooden arm that held the candle lit in front of the cross durind the rite (tab. 1L); here also appear some Latin cross crosslets painted on the calcarenite (tab. 4B). The case of San Nicola di Lamaderchia in Mottola, Taranto, is one of the most interesting and best preserved. The church has an unusual typology of red painted dedication crosses (also described in Maglio, 2021): a Latin cross with four stylized nails, arranged in an X (crux decussata or Saint Andrew’s cross) at the crossbar connection point, alluding to the four nails of the Crucifixion13. Inside the church there are twelve red crosses: six of them are painted along the intrados of the barrel vault of the central apse (tab. 4C), alter11 Stucco and marble crosses should be mentioned, they appear almost exclusively in subdival cases or in late remodeled rock churches of the Baroque period. 12 Along the Scala Santa, a mountain pilgrimage pathway that leads from the Gulf of Manfredonia to the Sanctuary of Saint Michael in Monte Sant’Angelo. Along the way, near Jazzo Ognissanti the sacred attestations carved in the rock (simple and inscribed crosses) intensify. 13 As we see in several medieval representations in that period (and further) the nails were thought to be four, two for the hands and two for the feet. nating crosses with nails and crosses of various kinds (Latin, Greek or Calvary crosses); three are inside the right impost of the central arch leading to the presbytery and another (faded) Greek cross is on the left on the counterfacade. Two more Latin crosses with four nails are at the two ends of the dedication epigraph of Saint John’s altar, seen before (tab. 3E). The upper right of San Nicola’s church facade also features four red crosses, painted in the in the basin of a collapsed arcosolium tomb (tab. 4D): a large potent Latin cross stands in the middle, two eight-pointed Latin crosses are in the lower quadrants and a powered Latin cross, inscribed in a circle (a tipical in rota dedication cross), is in the lower left quadrant. Even these external crosses are provided with four nails. It’s an added value for Cappuccino Vecchio being the only rupestrian church in Matera to preserve a certain number of red dedication crosses. The red potent crosses are painted on the bare calcarenite on the intrados of the right parabolic arch (tab. 4E): a potent Greek cross is clearly visible on the keystone, and there are traces of at least three other crosses placed at regular distance, following the curve of the arch downwards. Other red potent Latin crosses are on the intrados of the two arches that lead to the presbytery, placed at the arch’s impost: the most visible ones are in the left arch (tab. 4F). These testimonies of the dedication adds to the epigraph engraved and painted in red in the first niche of the right aisle, we saw before (tab. 3C). Four red crosses inscribed within clypei have been reported (Sammarco, 2000: pp. 45-51) in the apse of Sant’Elia’s crypt in Patù, Lecce. Other two red crosses are located on the west wall, on either side of a fresco frame, and on the south end wall (unspecified number), painted at a regular distance of 1 m from each other (tab. 6B). It’s a clear, rare case of a rock church that still keeps its red dedication crosses. A different peculiar case is Madonna delle Rose’s church in Massafra, where there are at least three coeval floral variants of painted dedication crosses in the same place (tab. 6C,D,E): they’re not in their best conditions, but we have a white cross pattée contrasting with an outlined red rota and with red and white petals, there’s another red cross pattée on a white background (worn, encrusted and not very legible), and a green and white free anchored cross (without a rota), outlined in red, with a distinct floral character. Incised, engraved and sculpted crosses Graffiti has always been used to represent crosses, in every historical period, both in rock and subdival areas14. Graffito dedication crosses are always a large number in the rupestrian areas: incised on the bare rock, scratched on plaster and frescoes, or engraved, 14 According to the paleographic literature also red inscriptions and crosses, directly painted on rock or plaster, can be classified as graffiti, as well as the incised or scratched ones (Favreau, 1997: p.31).

219 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa Tab. 4 – Red dedication crosses. A) Ognissanti’s church, Monte Sant’Angelo, Foggia: one of the carved and red painted Latin crosses, with a central carved hole used to house the wooden arm that held a dedication candle (photo R. Rinaldi). B) Ognissanti’s church, Monte Sant’Angelo, Foggia: a dedication Latin cross crosslet, painted red on the bare calcarenite (photo R. Rinaldi). C) San Nicola di Lamaderchia, Mottola, Taranto: some of the red dedication crosses painted on the presbytery vault in a unusual typology of a potent cross with four nails, arranged in a X (crux decussata) at the crossbar connection, alluding to the four nails of the Crucifixion (photo S. Centonze). D) San Nicola di Lamaderchia, Mottola, Taranto: the basin of a collapsed arcosolium tomb, in the upper right of the facade, features four red crosses of various types, painted in red and ocra yellow and provided with four nails (photo S. Centonze). E) Cappuccino Vecchio, Matera: at the keystone of the right parabolic access is clearly visible a potent Greek cross and there are still traces of at least three other crosses following the curve downwards (photo S. Centonze). F) Cappuccino Vecchio, Matera: the most preserved red potent Latin cross at the intrados of the left access to the presbytery, placed at the arch impost (photo S. Centonze).

220 Physical evidence of dedication rites in rock churches of Basilicata and Apulia (Italy) they are often misunderstood and confused among the layered devotional signs To identify the dedication marks, we should isolate the signs drawn with more care and proportion. Other important data is the recurrence of one or more types in the same place, the similarity of the “hand” who traced them, the localization and their modular repetition at a regular distance. As we’re going to see, the Pontifical’s fixed heigh of 2.30 m, for placing the crosses, is hardly respected in rock churches, due to the lower height of the liturgical space. On the walls of the aniconic lower crypt of Madonna della Scala’s Sanctuary, Massafra, Taranto, both in the narthex and main body, we find several Greek and Latin crosses of dedication, deeply engraved in the calcarenite at regular distance between them and at the height of the arches’ impost (tab. 5C). Also in Madonna delle Tre Porte, in Matera, some deep, simple Greek crosses are placed on the arches between the naves (tab. 5B): they are likely to be the oldest dedication crosses, for to the precision of the sign and their position at the arch impost (at the midpoint of the intrados and on the front). Sometimes dedication crosses are integral part of the church decoration, as it happens in the Candelora, in Massafra, where an inscribed in rota dedication cross pattée becomes the bas-relief of the capitals (tab. 5D)15. With the destruction of the facade in the 19th century, the apse was almost completely lost, but in the upper tympanum we can still identify other three dedication crosses pattée inscribed in rota, two symmetrical and one in the middle (tab. 5E). Partial evidence, rededications and desacration signs Due to the perishability of the attestations (both materials and surface) and to all the changes occurred to our rock churches, we often find less than 12 crosses of the same typology. The other missing exemplars may have been covered up by later plaster or frescoes, or they may have disappeared for vandalism. Moreover, in some cases we clearly recognize the intention of desecrating the church. A typical case of partial dedication evidence is in Madonna with Child’s chapel, in Santa Maria della Palomba’s Sanctuary in Matera, where we find six dedication crosses in the shape of eight petals flowers with gray pistils, which form an eight rays chrismon16, four on the bottom and one on each side (tab. 5A), all in good conditions, painted in red and ocra yellow. We 15 Here Father Abatangelo noticed a certain number of patent crosses in rota: they were so well defined on capitals and archivolts, that for the first time he revealed doubts about the purely decorative function of these crosses, in favor of the sacral one. He did not arrive at the real solution, but he depicted one of them on his book cover (tab. 5D). 16 The style of these fresco crosses is contemporary with the other frescoes made between the second half of the 17th century and the first thirty-five years of the 18th century. should assume this crosses belong to a complete 17- 18th century rededication, of which we lost the second half of the 12 crosses. In Madonna delle Tre Porte’s church our attention goes to the fresco red cross on the central pylon (tab. 6A): it’s a Greek cross inscribed in a triple square frame alternating black, white and red, at its base there’s a floral swirl decoration. This cross suggests a 15th-16th century rededication, contemporary to the frescoes of the nearby apses. This church is also known as “Madonna delle Croci”, for the large number of crosses scratched on its pylons and walls: the church went under various planimetric changes and decoration, so the cross crosslets and the layered crosses at mediumlow height belong to subsequent rededications. In Santa Croce’s church, Massafra, we can recognize some partial cross pattée: they are compass drawn in a four petals flower and inscribed in a rota on a thin fresco plaster, on which they are red painted. We can find their traces in the apse (tab. 6F), on the pylons and on a previous 13th -14th century fresco (tab. 6G), thanks to which we can connect all these crosses to a rededication of the place of worship, occured starting from the 14th century (Caprara, 2006: p.16). In a final stage the crosses have been disfigured in order to desecrate the church. Before we saw another immediate, tangible sign of deconsecration of the altars, in the carved empty container, which lost its relics (tab. 2E). Comparisons with subdival cases: exceptions and other evidence of dedication Comparisons between subdival and rock cases (an extract in tab. 8) show almost a perfect match in the evidence of dedication, confirming that architecture by subtraction is the mirror of the built one also in the liturgical sphere. At the same time they underline some important exceptions occuring in rupestrian churches. We have just seen the derogation in placing the crosses at the fixed heigh of 2.30 m, as the rupestrian spaces had a remarkable lower height than the built ones. Other exceptions were due to the natural conformation of a rock church, which usually had only one free front, and the rest of the walls, vault and floor were part of the ground. It results that the Bishop could not go around the place of worship and hallow it, therefore we should assume that the different evidence of dedication - we are going to analize - are related to variations of the rite. The abecedarium sign (tab. 1C, 8A) was an ephemeral phase that left no traces, even if sometimes graffiti fix its memory on the walls17 (tab. 8B). We can record cases of medieval crypts and subdival churches where the floor is sacred by one or more in rota crosses (tab. 8F, even the continuous Cosmatesque inlays alluded 17 At the moment we can record some alphabetic series alluding to abecedarium only in crypts and subdival churches.

221 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa Tab. 5 – Main localization of internal dedication crosses. A) Madonna with Child’s chapel, Santa Maria della Palomba, Matera. Six dedication crosses are painted in red and ocra yellow near the vault impost, in the shape of an eight petals flower with gray pistils, which form an eight rays chrismon (photo S. Centonze). B) Madonna delle Tre Porte’s church, Matera. Simple Greek crosses, deeply carved on the arches impost between the naves: due to the precision of the sign and to their position, they are likely to be the oldest dedication crosses in this church (photo S. Centonze). C) Lower crypt of the Madonna della Scala Sanctuary, Massafra, Taranto. Simple Greek and Latin dedication crosses deeply engraved in the calcarenite at regular distance, placed at the impost height (photo G. Mastrangelo). D) Cover of a book depicting the bas-relief of a capital of the Candelora’s church in Massafra, Taranto, with an inscribed in rota dedication cross pattée (Abatangelo,1966). E) Half collapsed apse of the Candelora’s church, Massafra, Taranto: in the upper tympanum archivolt we still can find three inscribed in rota dedication crosses pattée (photo G. Vezoli).

222 Physical evidence of dedication rites in rock churches of Basilicata and Apulia (Italy) Tab. 6 – Dedication crosses of different shapes and colours. A) Madonna delle Tre Porte’s church, Matera. The isolated red cross painted on the central pylon, suggesting a rededication. Below several scratched and layered rededication cross crosslets (photo S. Centonze). B) Sant’Elia’s crypt, Patù, Lecce. A potent red cross inscribed in a rota, one of several (and misunderstood) dedication crosses of the rupestrian church (Patù Centopietre Archive). C, D, E) Madonna delle Rose, Massafra, Taranto. Three different floral shapes for these polychrome dedication cross (photo A. Conforti). F) Santa Croce’s church Massafra, Taranto. This dedication cross in the apse has been deliberately disfigured in order to desecrate the church. It was originally an in rota inscribed cross pattée (formed by a four petal flower) scratched with a compass on the calcarenite and then painted (photo S. Centonze). G) Santa Croce’s church Massafra, Taranto. In the same church we can find a similar in rota cross pattée, scratched with a compass on a previous older fresco and then painted. In this way this cross testifies the rededication of the place (photo S. Centonze).

223 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa Tab. 7 – Exceptions: sacralizing crosses on vaults, archivolts and upper rock extrados. A) Calvary cross carved in the rock bank above San Basilio Magno, Castellaneta, Taranto (photo A. Gregucci). B) Cross pattée inscribed in a rota on the archivolt of the rock chapel of San Basilio Magno in Castellaneta, Taranto (photo P. Manigrasso). C) High relief Latin cross on the archivolt entrance of Madonna del Giglio in Matera (photo S. Centonze). D) The complex rib vault of Santa Eugenia in Matera, with a bas-relief Latin cross pointing east (photo S. Centonze). E) Madonna dell’Attarico, Marina di Andrano, Lecce. A rare case of a still red painted cross pattée on the vault: a typical cross pattée inscribed in a rota used the dedication rite that reconnects the unpainted ones sculpted on the vault, clarifying their utility in the dedication of a church (photo A. Romano). F) Cross pattée inscribed in a rota on the altar vault of Santa Maria de Balneolo, Matera (photo F. Foschino). G) In rota Cross pattée carved on the altar vault of San Gregorio in Mottola, Taranto (photo S. Centonze). H) Lily cross pattée in rota carved on the vault of San Donato, Convicinio di Sant’Antonio, Matera (Antros Archive).

224 Physical evidence of dedication rites in rock churches of Basilicata and Apulia (Italy) Tab. 8 – Comparisons: subdival and rock dedication/rededication evidence. A) Abecedarium during the rededication ceremony of Westminster Abbey in London in 1910 (newliturgicalmovement.org). B) Bartoccini Tomb in Tarquinia, Viterbo. Graphic restitution of a dedication graffito on a fresco: an abecedarium next to a Latin cross crosslet (Tedeschi C., 2012: p. 39). C) Matera’s Cathedral: an avellana rededication cross on a previous in rota dedication cross pattée (photo S. Centonze). D) A red rededication cross pattée on a fresco of the Basilica of SS. Martyrs in Cimitile, Benevento (photo S. Centonze). E) One of the rededication cross pattée on a previous fresco of Santa Maria Maggiore in Tuscania, Viterbo (photo S. Centonze). F) Floor of the crypt of Isina’s Cathedral, Matera: a large polychrome rota with a six petals flower used as a cross to sacralize the ground flooring (photo S. Centonze). G) Lily cross with four heart-shaped petals on the vault of the Eremo di Poggio Conte, Ischia di Castro, Viterbo (photo S. Centonze). H) Crosses and other dedication symbols painted in red on the plaster of San Giovanni d’Antro, Pulfero, Udine: lily crosses, branches and crowns recalling the hyssop (photo J. Ariton). I) In rota six petals flower painted as dedication cross in San Giovanni d’Antro, Pulfero, Udine (photo F. Serino).

225 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa to a form of floor protection), on the contrary, the floor of a rock church was the same rock, had no finish with other flooring materials, so these kind of attestations did not reach us. Nevertheless in the Middle Ages, some variants appear on the vaults, along the nave or upon the presbytery, expecially above the altar, where a cross (inscribed in a rota or free) is sculpted in relief. This means there was a particular attention in sacralizing ceilings and vaults, in order to isolate them with the exorcism of the dedication cross, the holy Christ’s cross. All the crosses marked on the entrance and those hat were engraved on the extrados of some rupestrian churches, appear useful to this purpose (tab. 7A, B, C). Matera adorns the vaults with a plenty of these rotae: some lily crosses pattée are in Sant’Antonio Abate and San Donato (tab. 7H), three different variants (a cross pattée, a potent cross and a bordoned one) are sculpted in Madonna della Croce, a potent cross is in Sant’Agnese, two potent crosses are in Cappuccino Vecchio (on the altars), while in San Pietro e Paolo’s crypt there’s a barely detectable red painted cross. A smaller cross pattée inscribed in a circle is carved on the altar vault of the church identified as Santa Maria de Balneolo (Foschino, 2017, or Sant’Eustachio ai Pedali in Pelosi and Lionetti, 2021: p. 205), and three crosses pattée in bas-relief are on the hemispherical domes of Madonna delle Virtù’s church. Other isolated sculpted crosses protect the presbyterial area from the vault, as the cross pattée in Oratorio dell’Ofra (Centonze, 2020: p. 121) and as the Greek cross in Santa Lucia alle Malve; on the rare complex ribbed vault in Santa Eugenia’s church (tab. 7D) we find a bas-relief Latin cross pointing the east raising sun on a specific day. There are other examples of sculpted vault cross in Mottola, among which we should mention an irregular bordoned cross in Madonna delle Sette Lampade’s church and a cross pattée on San Gregorio’s altar (tab. 7G). Moving to Ischia di Castro, Viterbo, we find a large lily cross pattée with four heart-shaped petals18, sculpted in relief on the vestibule of the Eremo di Poggio Conte (tab. 8G): thanks to this polychrome exemplar with the petals painted in red and blue, we can assume that some of these vault crosses where painted in the past. The presence of a red cross on the vault of Madonna dell’Attarico in Marina di Andrano, Lecce, has led a wrong attribution to the Templar knights, but it’s clear this shape is a classic dedication cross pattée inscribed in a rota (tab. 7E). This last is a fundamental specimen, because it manages to reconnect all the unpainted sculpted examples we have just seen, clarifying the role of the vault crosses in sacralizing these rock churches: we have no written sources, but we can imagine that all the aspersions, lustrations and anointations during the dedication rite could reach to touch the vaults, thanks to the low height of the places. Vault crosses phenomenon is limited to the Middle Ages period, then - probably due to a change of mentality - these specimens slowly disappeared and the vaults became smooth. 18 Similar to the Cistercian crosses schemes appearing in Fossanova Abbey, Priverno, Latina. Bibliography Abatangelo L., 1966, Chiese-cripte e affreschi italo-bizantini di Massafra, Vol.1, Cressati, Taranto. Andrieu M., 1940, Le pontifical romain au Moyen-Age, III, Le Pontifical de Guillaume Durand, Studi e testi, 88, Città del Vaticano. Caprara R., 2006, La Chiesa rupestre di Santa Croce a Massafra, Kikau, Massafra. Caprara R., 2016, Iscrizioni e graffiti da chiese rupestri otrantine ed altri siti pugliesi, in Archeogruppo 7, Bollettino dell’Archeogruppo “E. Iacovelli”, pp. 11-48, Massafra. Caprara R., 2017, L’inedita iscrizione di San Giuliano al Bradano, in Rivista Mathera 1, pp. 17-21, Antezza Tipografi, Antros, Matera. Castellano A., 1520, Pontificale secundum ritum sacrosancte Romane ecclesie cum multis additionibus pportunis ex apostolica bibliotheca sumptis & alias non impressis: quarum breuis index post epistolam, Giunta, Venezia. BMR, 1916, Anonimo Manoscritto n. 959, Notizie sulle chiese di Matera e sugli oggetti d’arte in esse contenuti raccolte dal compianto Senatore Conte Giuseppe Gattini (Copiato del manoscritto esistente nella biblioteca di casa Gattini), Biblioteca del Museo Nazionale Domenico Ridola, Matera. Centonze S., 2020, L’esaltazione della croce e del Tabernacolo nei graffiti della cappella di contrada Ofra a Matera, in Rivista Mathera 11, pp. 121-124, Antros, Matera. Centonze S., 2022a, “Ego indignus dedicavi” La consacrazione delle chiese rupestri. Rituali di purificazione e di dedicazione in rupe, in Rivista Mathera 19, pp. 106-127, Antros, Matera. Centonze S., 2022b, Dal cielo alla piuma: l’incarnazione di San Michele Arcangelo, inviato divino in forma avis. Attestazioni graffite e dipinte di memoria agiografica, dal “VIII Convegno Nazionale “Le presenze Longobarde nelle Regioni d’Italia. Archeologia, Diritto, Economia, Costume nei Principati Longobardi del Sud”, Massafra 29-31 ottobre 2021, in Rivista Mathera 20, pp. 171- 182, Antros, Matera. Durand de Mende G., 1280/1999, Manuale per comprendere il significato simbolico delle cattedrali e delle chiese, trad. di Rosanna Campagnari, Edizioni Arkeios, Roma. Favreau R., 1997, Epigraphie médiévale, Turnhout. Foschino F., 2017, Santa Maria de Balneolo a Matera. Una proposta di identificazione, in Rivista Mathera 2, pp. 44-51, Antros, Matera.

226 Physical evidence of dedication rites in rock churches of Basilicata and Apulia (Italy) Maglio S. A. N., 2021, L’Ordine degli Ospedalieri di San Giovanni di Gerusalemme a Mottola e la chiesa rupestre di San Nicola di Lamaderchia, KDP, Grendel Edizioni, Mottola. Pelosi M., Lionetti G., 2021, Riflessi storici e toponomastici di Matera: la Santa Visita di Mons. Fabrizio Antinori e altri fonti inedite dal medioevo ai nostri giorni, Antros, Matera. Picard J. M., 2007, La diffusion du culte de saint Michel en Irlande médiévale, in Culto e santuari. Sammarco M., 2000, Gli insediamenti rupestri nel Capo di Leuca, Congedo, Galatina. Tedeschi C., 2012, Graffiti templari. Scritture e simboli medievali in una tomba etrusca di Tarquinia, Viella, Roma. Treffort C., 2008, Une consécration «à la lettre». Place, rôle et autorité des textes inscrits dans la sacralisation de l’église, in Mises en scène et mémoires de la consécration d’église au Moyen Âge, études réunies par Didier Méhu, (Collection d’études médiévales de Nice, 7), pp. 219-251, Brepols, Turnhout. Trotta M., 2012, Il santuario di San Michele sul Gargano dal tardoantico all’altomedioevo, Adda, Bari. Online sources newliturgicalmovement.org, https://www.newliturgicalmovement.org/2016/06/the-consecration-of-westminster.html, consulted on 30/06/2023.

227 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa 1   Ilia State University, 3/5 K. Cholokashvili Ave, Tbilisi 0162, Georgia 2 Georgian National Museum, 3 Rustaveli Ave, Tbilisi 0105, Georgia Nodar Bakhtadze: phone + 995 32 2321582 - [email protected] New considerations on the Uplistsikhe rock-cut ensemble (Georgia) Nodar Bakhtadze1,2 Abstract In Georgia, in the river Mtkvari valley, artificial cavities complexes, dated back to the 11th century BC up to the 4th century AD, are found. The crowning touch of this kind of building is the great cave ensemble of Uplistsikhe. However, despite its almost 150-year study, the artistic and planning solution of this site and questions of its functional differentiation were challenged by Georgian scholars of different generations. Researchers expressed different, frequently diametrically opposed views on the purpose of the entire complex as well as individual cave facilities. This trend continues until now. Most researchers consider this site as a fortified city, and the caves forming it as part of the fortification or palace halls, household areas, theatre, and so on in use over the centuries. We think, that in some areas of Eastern Georgia, similarly to many regions of the Near East and the Mediterranean basin a tendency of the gradual assignment as burial and sacred functions to the complexes of the man-made cave settlements by the population is already attested in the Late Bronze-Early Iron Ages. This tradition must have played a decisive role in the emergence of cave templar centers of the cult of the ancestors and other religions merging with it on the basis of rock settlements. The origin of the Uplistsikhe temple ensemble should be related to the same event. Throughout the Hellenistic period, large and small temples in Georgia were often carved into the rock. This tradition lasted until the Late Antiquity and was finally reflected in a large part of Christian monasteries. Keywords: Georgia, rock-cut, Uplistsikhe, ensemble, temple. Introduction Uplistsikhe is an well-known, great rock-hewn ensemble in eastern Georgia, some 90 kilometers west of the city Tbilisi (Figs. 1, 2). The complex was cut in a flat, straight, but slightly inclining mountain. The river Mtkvari itself closed access to the town from the south, but it was possible to get into the complex through a 3 m tunnel that functioned as a water supply. The tunnel was closed by a metal gate in case of invasion. Two Fig. 1 – Uplistsikhe ensemble (photo N. Bakhtadze).

228 New considerations on the Uplistsikhe rock-cut ensemble (Georgia) out of the three roads approaching the ensemble were not protected, but the third, coming from the north, was cut in the rock, with up to 10 m high walls. The Uplistsikhe complex can tentatively be divided into three parts: south (lower), middle (central) and north (upper) covering an area of approximately 8 hectares. The middle part is the largest, it contains a bulk of the Uplistsikhe rock-cut structures. Among them, a ceremonial hall is the most notable. A pillared hall with two adjacent rooms is one of the most important structures in the ensemble (Fig. 3). Its ceiling was supported by two pillars. A stone bench, probably, served as a ruler’s seat. Most of the caves lack any decorations, although some larger structures have coffered tunnel-vaulted ceilings, with the stone carved in imitation of logs (Fig. 4). Some of the larger structures also have niches in the back or sides, which may have been used for ceremonial purposes. The decoration of the halls reflects both quite rich wooden interiors and Roman lacunas. The facade of the large ceremonial hall of the southern part is decorated by a Romantype arch with pediment. Approximately two decades ago, the view prevailed in Georgian and foreign scholarly circles according to which the origin of the Uplistsikhe ensemble was believed to be the earliest and the only pre-Christian rockcut monument in Georgia. Despite its almost 150-year study, the artistic and planning solution of this site and questions of its functional differentiation were challenged by Georgian scholars of different generations. Researchers expressed different, frequently diametrically opposed views on the purpose of the entire complex as well as individual cave facilities. A part of the researchers considers this site as a fortress-stronghold Fig. 2 – Plan of the Uplistsikhe ensemble (photo N. Bakhtadze). Fig. 3 – The Plan and interior of the rock-cut hall of Uplistsikhe (drawing B. Gabekhadze).

229 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa or fortified city, and the caves forming it, respectively, as the fortification or secular purpose components (palace halls, household areas, theatre, and so on) (Beridze, 1974: 15; Amiranashvili, 1963: 81-85; Mepisashvili, Zinzadse, 1986: 41-42). Some scientists think Uplistsikhe is a complex of cave temples (Sanikidze, 2002). Several years ago, an absolutely original view was advanced, according to which the cave facilities of Uplistsikhe form a unity of rock-cut burials (Kipiani, 2002). The phenomenon of “sudden” creation of a plot rock-cut complex of Uplistsikhe on the Georgia territory, was explained by a famous German orientalist of the 19th20th centuries C. Lehmann-Haupt as follows (LehmannHaupt, 1938: 245-246). He supported the hypothesis of the Georgian tribes’ migration to the South Caucasus from the territory of Corduene (North Mesopotamia, south of Lake Van) in the 1 th millennium BC and believed that the tradition of carving caves was “imported” by this population from there. During the 20th century, this version of Uplistsikh’s origin was shared by many well-known Georgian scientists (Amiranashvili, 1963; Khakhutaishvili, 1964 and others). We decided to verify this issue due to the following circumstances. The architectural masterpieces of all civilizations of the world, including the artistically planned cave complexes, are mostly the result of long-term development of one or another style in a place. Of course, certain ethnic group, when migrating to other areas, could introduce a completely new style of buildings, but it was unlikely that migrants would create such a high-class architectural ensemble on the first attempt, and that too the only one. This critical view was reinforced by the landscape archaeological works conducted by us in a number of provinces of Eastern Georgia, during which we identified a number of cave complexes with quite archaic plans, completely undiscovered and unexplored before (Bakhtadze 2007: 62-79). It turned out that directly in the province of Shida Kartli, adjacent areas of Uplistsikhe, distant from each other to some extent, several cave monuments were created in the Late Bronze - Early Iron periods. The age of their origin slightly predates the early stages of the Uplistsikhe construction. These ordinary rock sites, following definite reconstruction work, functioned actively in the Hellenistic period too. The results of the archaeological research of the Shiomgvime Cave complex The caves of Shiomgvime are located in the closest neighborhood of a large, famous Christian monastery. Hiding places and necropolises carved into the rock of pre-Christian times in many regions of the Eastern Byzantine world (Syria-Palestine, Anatolia, etc.) were very intensively used for monastery needs (Hirschfeld, 1992: 54-55; Gilli, 2017: 450 and others) So, the exclusion of similar phenomenon would not be legitimate in Shiomgvime either. It sould also taken into account that the layout of these caves does not at all correspond to the principles of rock monastery architetecture of the Middle Ages: there are no churches, prayer niches, refectoies so characteristic of Christian cave monasteries. On the other hand, a significant part of the caves, due to their rather complex spatial and planning solution, cannot be considered monk’s living cells or shelters. To carry out archaeological work from cave monuments in the vicinity of Shiomgvime, we chose a group of caves where the probability of finding tangible material from the most ancient phases of life was high (Fig. 5 A). Our expectation was justified: in several of these caves, under very poor cultural layers of the developed and late Middle Ages, in the recesses of rocky floor, we found a certain amount of tangible materual, fragments of pottery from various periods of pre-Christian times (Bakhtadze, Kipiani, 2000). For example, fragments of red-polished yellowish dishes are similar to pottery of the Ellenistic Hera found within the territory of Shida Kartli (Gagoshidze, 1981: 45-48) (Fig. 5 B). The material of this series certainly indicates that this complex of caves should have been created no later than the turn of the 2th-1th millennium BC, and its earlier intensive reconstruction, expansion and use for various purposes took place in the early ancient, Hellenistic and subsequent periods. The hard-to-reach premises included in this complex are Fig. 4 – The interiors of the rock-cut halls of Uplistsikhe (drawing B. Gabekhadze).

230 New considerations on the Uplistsikhe rock-cut ensemble (Georgia) their creation for residential purposes is less likely. It is quite obvious that in their person we are dealing with a set of rock-cut tombs of a simple type, and the material found here is the remains of grave goods. The results of the archaeological research of the Grakali cave complex Grakali caves are carved in 4 tiers on the left slope of the Kura River gorge (Bakhtadze, 2002: 24-26). Up to 30 caves caves have been preserved, the vast majority of which are inaccessible today and we explored them using climbing equipment (Fig. 6). The caves are carved in easily processed calcareous rocks, so it was not difficult for their creators to obtain shapes that were close to geometric, and the rooms practically have a rectangular layout. In most premises, a trace of carving with a tetrahedral chisel has been preserved. Most of the premises included in the complex turned out to be very small in size - their length and width do not exceed 2-2.5 cm, and their height varies between 1.4-1.7 m. The ceilings are all flat. Most of the premises are isolated from each other and have very narrow and low (approximately 1.2-1.5 m high) openings and a high threshold protruded to the surface of the earth. Even though burial shelve-beds and niches in the caves are not confirmed, based on the miniature volumes, low height and shapes of the entrances of these rooms, we think we are dealing with a system of rock-cut tombs of a simple type. The burial chambers have more or less close parallels both from the Uplistsikhe ensemble and the rock tombs of the ancient period in the countries of the Middle East. With repeated use of these cave complexes, the original cultural layers were destroyed. In return, the archaeological study of the directly front gentle slopes of the caves brought greater clarity to the establishment of the chronological framework in the origin and functioning of the complexes. In the material found here with the help of exploration tools, fragments of ceramic products of a well-known type from the archaeological sites of Shida Kartli in the first half of the 1th millennium BC were revealed (Lordkipanidze, 1989: 129-144). It is noteworthy that two shapeless pottery shards of this type of pottery were also found on the floor of one inaccessible small cave in the upper register of the Grakli caves. The results of the archaeological research of the Kaspi rock-cut necropolis In the nearby areas of Uplistsikhe, we also found a number of rock-cut monuments of the late Antique Period. The political and economic decline of the country and the consequences of the orientation of Iberian ruling strata towards Iran Parthian and Sassanian states (2th-3th cc AD) must be reflected in the architecture of the Kaspi rock-cut necropolis (Fig. 7), where, on the one hand, decline of the tradition of monumental rock architecture, as well as the Zoroastrian custom of burying nobles is in evidence. In Georgian written sources, Kaspi is mentioned among the most ancient cities of Iberia (Bagrationi, 1959: 364, 366). The construction of Kaspi, like Uplistsikhe attributed to the legendary Ethnarck - Uplos. This is what led us to the idea to look for monuments of the ancient period carved into the rock and in the vicinity of the supposed localization of Kaspi settleFig. 5 – A) a group of Shiomgvime cave complex; B) pottery from this caves. Fig. 6 – Grakali cave complex (drawing B. Gabekhadze).

231 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa ment. The study yielded an important result: in the area of the alleged localization of Kaspi settlement, we discovered a very interesting complex carved into the rock (Bakhtadze 2007: 66-73). The cave complex is carved into a cape-shaped massif of a hillock. They are badly damaged, the front parts of all of them collapsed, of the majority only minor fragments remain. Today, 4 of them have been preserved in somewhat complete form, and the remains of up to 10 caves make it possible to approximately reconstruct them. An architectural analysis of these monuments has led us to the conclusion that before damage and later alteration, each of these structures was based on a similar planning principle, which we can learn from the example of relatively well-preserved caves. The first three caves in one cape are carved side by side, often with a four-sided metal tool (Fig. 8). There is no visible trace of wall wear. The plan of the caves approaches a round-angled square, and the ceilings are vaultedlancet (dimensions No. 1 - 2.5x2.3m, maximum height 1.96m, No. 2: 2.4x2.3m, maximum height 1.6m). In the western and northern walls of both rooms, one roundangled shelf-bed (length 1.6-1.75m, width about 55- 0.6m, height 0.65-0.7m) is carved, from the floor the shelf rises by 0.2- 0.5m. There were open vestibules in front of the caves, from where through low and narrow doors (height 0.65-0.9 m, width 0.5 m) one could get into the main rooms. The entrances are decorated with simply carved decorative frames. Cave No. 4 unlike the first three, does not have a vestibule, and the ceiling is 1.8 m high here. Opposite the entrance and in the right walls niche-shelves (1.75-1.8 x 0.6- 0.9 m, height 0.6-0.8 m) are carved, out of which the sides and bottom of one of them are separated from the space of the room by thin partitions 15cm long. It is obvious that the attributes of the described caves, and especially the shelf-niches, were typical of the badly damaged caves preserved around them. Here we note that in the sheer rocks adjacent to the caves, such shelves are carved openly, without rooms. The shape and dimensions of the shelves are suitable only for the dead to rest there in an extended position. Low entrances suitable for one-time use and miniature dimensions of the rooms are also quite suitable for this Fig. 7 – Kaspi rock-cut necropolis (photo N. Bakhtadze). Fig. 8 – Plans and sections of Kaspi necropolis tombs (drawing N. Bakhtadze).

232 New considerations on the Uplistsikhe rock-cut ensemble (Georgia) function. Thus, it can be convincingly said that this complex of caves cannot be anything other than a collection of tombs-mausoleums of a very mature type, a necropolis carved into the rock. In these tombs, the bodies rested emphatically on different sides, perpendicular to each other, while in the walls opposite the chambers, shelves could be freely carved along the east-west axis. Thus, here there is a clearly non-Christian order of burials of the dead. In a number of details, this monument is related to the samples created in the bowels of the Zoroastrian religions of the ancient Iranian world. Fire worship forbade the burial of the dead in the ground (the doctrine of Zoroastrianism, along with fire, recognized the holiness of the earth), therefore, even the kings and nobles of Media, creating rock tombs, did not insult the earth with corpses, since the rock was not considered earth (Boyce, 2001). The Kaspian necropolis, given its size, almost indisputably points to the widespread in the territory of Shida Kartli during a not so short period of fire worship or some kindred religion based on this doctrine. Features of the architectural details of the Kaspian tombs should indicate the origin of this monument in the late Antic era. The functioning of the necropolis in this era is also confirmed by small red-colored pottery shards from late antique times, which we found in the recesses of the floor in the tomb No. 3, which may be accidentally preserved remains of the grave goods. We also think that such dating of the Kaspian tombs should indicate that, like some regions of the Middle East, in Shida Kartli, the ensemble of mausoleumstemples of Uplistsikhe, highly artistically carved in the early antique and Hellenistic style, was replaced by cave tombs created in the late Antique era by relatively modest architectural methods. Conclusion Finally, we can conclude that in some areas of Eastern Georgia, similarly to many regions of the Near East and the Mediterranean basin a tendency of the gradual assignment as burial and sacred functions to the complexes of the man-made cave settlements by the population is already attested in the Late Bronze-Early Iron Ages. This tradition must have played a decisive role in the emergence of cave templar centers of the cult of the ancestors and other religions merging with it based on rock settlements. We think that the origin of the Uplistsikhe temple ensemble should be related to the same event. In the process of the establishment and development of an ancient state in Eastern Georgia, in the wake of the political, economic and cultural developments in the Iberian kingdom, and in conformity with concrete cultic procedures, the scale of buildings of this type and architectural structures as well as the aesthetic aspect of their execution altered. Brilliantly designed and executed on the Classical-Hellenistic traditions Uplistsikhe constructions – accord well with the period of the ascendancy of the Iberian kingdom. It seems that throughout the Hellenistic period, large and small temples in east Georgia were often carved into the rock. This tradition lasted until the Late Antiquity and was finally reflected in a large part of Christian monasteries. Bibliography Amiranashvili S., 1963, History of the Georgian art, Moscow, Isskustvo, pp. 81-85, [in Russian]. Bagrationi V., 1959, Description of the Kingdom of Georgia. Kartlis Tskhovreba, Vol. IV, Sabchota Saqartvelo, pp. 364-366 [in Georgian]. Bakhtadze N., 2000, Genesis of rock-cut architecture and first stages of its development in Georgia. Caucasian Messenger, 6, Tbilisi. pp. 22-30 [in Georgian]. Bakhtadze N., 2007, The genesis and paths of development of rock-cut architecture in Georgia. Tbilisi, georgian national museum, pp. 62-79 [in Georgian]. Bakhtadze N., Kipiani G., 2000, The results of the archaeological exploration of the caves in the vicinity of Shio-Mgvime monastery. Tbilisi, Proceedings of IV session of the Mtskheta institute of the Georguan Academy of sciences, Tbilisi, pp. 11-15 [in Georgian]. Beridze V., 1974, Old Georgian architecture, Tbilisi, Khelovneba, p. 15, [in Georgian]. Boyce, M., 2001, Zoroastrians: their religious beliefs and practices. London, Routledge, 252 pages. Gagoshidze I., 1981, Samadlo, 2. Tbilisi, Mecniereba, pp. 45-48 [in Russian]. Gilli E., 2017, From columbaria to dovecotes: two thousand years of use of cave dwellings in Ağirnas (Kayseri, Turkey). Proceedings of II International Congress of Speleology in Artificial Cavities, HYPOGEA 2017, Cappadocia, pp. 446-453. Hirschfeld Y., 1992, The Judean desert monasteries in the Byzantine period. New Haven and London, Yale University Press, pp. 54, 55. Khakhutaishvili D., 1964, Uplistsikhe, I. The results of the archaeological research of 1957-1963, Tbilisi, Mecniereba, 236 pages [in Georgian]. Kipiani G., 2002, Uplistsikhe. Tbilisi, Logosi, 172 pages. Lehmann-Haupt C., 1938, Introductory lecture on the histori and culture of the Khalds. Prceedings of Tbilisi State University, VI. Tbilisi, 1938, pp. 245-267 [in Russian]. Lordkipanidze O., 1989, Heritage of ancient Georgia. Tbilisi, Mecniereba, pp. 129-144 [in Russian]. Mepisashvili R., Zinzadse W., 1986, Georgien. Wehrbauten und Kirchen. Leipzig, VEB E.A. Seemann Verlag, pp. 41-42. Sanikidze T., 2002, Uplistsikhe. An essay on the history of Georgian architecture. Tbilisi, Khelovneba, 196 pages [in Georgian].

233 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa 1 Dipartimento di Ingegneria, Università degli Studi della Campania “Luigi Vanvitelli”, Aversa, Italy 2   Ministero della Pubblica Istruzione, Italy 3   ISPRA – Istituto Superiore per la Protezione e la Ricerca Ambientale, Rome, Italy 4   Federazione Speleologica Campana, Catasto Cavità Artificiali *  Reference author: [email protected] Artificial cavities under worship places: case studies from the province of Caserta (Italy) Emilia Damiano1 , Francesco Fabozzi2 , Maria Assunta Fabozzi1 , Paolo Maria Guarino3 , Ivana Guidone4 , Erika Molitierno1 , Lucio Olivares1 , Arcangelo Pellegrino2 , Marco Vigliotti1 , Daniela Ruberti1,* Abstract In northern Campania (Southern Italy), the historic center of many towns is characterized by the widespread presence of cavities in the subsoil, excavated over the centuries for quarrying tuff blocks for buildings, along with cathedrals, churches and chapels. A singular feature of these places of worship is, in fact, the presence of a wide and frequently connected network of underground cavities and tunnels, which were used for hydraulic, religious or connecting purposes. Very often the above cavity network is unknown, abandoned or even buried, thus representing a risk for their susceptibility to sinkholes. Such elements are important as cultural heritage of inestimable value and, then, as attractors for tourism; for this reason the multidisciplinary study conducted on two places of worship in the Caserta area is illustrated herein: the Cloister of Sant’Agostino, in Caserta (XVI century CE), and the Monumental Complex of San Francesco delle Monache, in Aversa (XIII century CE). A geological characterization of the subsoil was performed at first. A laser scanner survey of the accessible cavities and the external churchyard was carried out. The resulting 3D model of the underground sector allowed a clear understanding of the room size, their location, the levels and the path of the corridors. In order to understand the extension and layout of the crypts, Electrical Resistivity Tomography (ERT) surveys were undertaken in the surrounding areas. The analysis of the ERT measurements revealed some anomalies that could be ascribed to unknown structures (crypts). The results allowed us to improve the knowledge of the study sites and provide useful tools for the planning of future targeted investigations, providing indispensable support both in the management and mitigation of geological risks in urban areas and in the sustainable reuse of hypogea. Keywords: artificial cavities, worship places, northern Campania, multidisciplinary characterization, risk management and mitigation. Introduction In northern Campania (southern Italy), the historic centre of many towns is characterized by the widespread presence of underground cavities. These underground environments testify to the multi-centennial exploitation of local geo-resources through the extraction of tuff rock, widely used in ancient, medieval, and modern age constructions. From ancient age structures to the seventeenth-century churches and palaces, passing through medieval convents and basilicas, Campanian tuffs have been the primary material for the development of inhabited centres and related infrastructures. Among the latter we can also include the same underground environments that responded to various social needs; citing the most attested uses, it should be remembered the conservation of agricultural products, the burial of the dead, the cult of saints and the general hydraulic needs of the inhabited areas such as transport and storage of water, both rain and spring (Ebanista 2007, 2011). The artificial cavities, therefore, now recognized as an “underground heritage”, represent a precious object of study and an irreplaceable testimony of the historical and deep-rooted relationship between Man and his territory. In relation to the worship and funerary needs of the Campanian settlements, the local communities resorted not only to the natural cavities present along the limestone ridge of the Campanian Apennines, but also to the excavation of specific artificial hypogea found in almost all the areas of northern Campania where weak rocks abound. Of about 1130 natural caves surveyed by the Speleological Federation of Campania, 83 were manmodified and exploited by the medieval Christian community; of the 375 artificial cavities, on the other hand, about twenty have religious and burial uses (Meneghini, 2010; Guidone and Izzo, 2015). In connection with these worship structures, it is not uncommon to find cisterns and water pipes pertaining to the convent communities and the various monasteries formed in the modern age. These underground environments, in addition to guarding any archaeological finds, are themselves architectural assets of considerable cultural and scientific value. The protection and enhancement of the architectural, artistic and land-

234 Artificial cavities under worship places: case studies from the province of Caserta (Italy) scape heritage cannot therefore be separated from a precise and systematic knowledge of the underground environments of anthropic origin. Furthermore, these empty spaces, while representing a cultural heritage of inestimable value, are still little known, abandoned or even buried, to the point of representing a risk for their susceptibility to sinkholes. On this basis and considering the importance of such elements as attractors for tourism, the multidisciplinary study conducted on two places of worship in the Caserta area is illustrated herein: the Cloister of Sant’Agostino, in Caserta (XVI century CE), and the Monumental Complex of San Francesco delle Monache, in Aversa (XIII century CE). Study area The study area, in north-western part of Campania Region (Italy; fig. 1) corresponds to the northern and north-eastern part of the Campania Plain, a broad, complex graben closely controlled by NE-SW, NW-SE and E-W normal fault activity, established in Late Pliocene (Ippolito et al., 1973) or the Early Pleistocene (Cinque et al., 1987; 2000) along the Tyrrhenian side of the Apennine Mountains. The sedimentary evolution of this graben was conditioned by the fluvial and marine processes and the volcanic activity of the Campi Flegrei, Somma-Vesuvius and Roccamonfina volcanoes (Amorosi et al., 2012; Vitale and Ciarcia, 2018; Ruberti et al., 2022). The considered sector is characterized by a flat morphology between 95 and 20 m a.s.l. In this area the subsoil is formed by the succession of different units composed of volcanoclastic deposits, in particular related to the Campania Grey Tuff (CGT; 39 ky B.P; De Vivo et al. 2001) and Neapolitan Yellow Tuff (NYT; 15 ky B.P.; Deino et al., 2004) pyroclastic eruptions from the Campi Flegrei volcanic district (Ruberti et al., 2020). The CGT deposits, in particular, were settled on the whole Campania Plain, giving rise to a thick (up to 40 m thick), laterally continuous, volcanoclastic unit, characterized by different lithofacies mostly derived from the different mineralogic composition. The good mechanical characteristics of the tuff lithofacies justify the presence of numerous quarries and/ or cavities, according to the availability of adequate thicknesses of coherent lithofacies (Vigliotti and Ruberti, 2018). Methods Geological and geotechnical characterization To assess the states of stress induced by the presence of cavities in the surrounding soils and the potential Fig. 1 – Location map of the study sites: (a) Schematic geologic map of the study area (modified from Sacchi et al., 2014); air view of (b) Cloister of Sant’Agostino, in Caserta (14.331657° E, 41.072397° N) and (c) Monumental Complex of San Francesco delle Monache, in Aversa (14.206782° E, 40.973504° N).

235 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa risks associated with them, comprehensive investigations were carried out at the two sites. Boreholes, penetrometer tests (DL030 and DPSH types) and geophysical investigations (MASW) were used to reconstruct the stratigraphic sequences and the mechanical characterization of the subsoils. In detail, the empirical correlations available on literature for cohesionless soils, based on the results of standard penetrometer tests, were adopted to determine parameters regarding the state properties and the strength characteristics of the pyroclastic soils (Skempton, 1986; Hatanaka and Uchida, 1996) whereas the results of geophysical investigations were used to evaluate their stiffness. Regarding the tuff formations, due to the lack of direct measurements, typical values reported in the literature for pyroclastic weak rocks (Aversa and Evangelista, 1998; Ceroni et al., 2004) were adopted. At the present stage of study, it was not possible to examine in detail the structural characteristics of the tuff formation (i.e. possible presence of joints or discontinuities) and therefore the latter was assumed to be continuous. Electrical Resistivity Tomography In order to understand the extension and layout of the hypogea, Electrical Resistivity Tomography (ERT) surveys were undertaken in the surrounding areas of the churches. Geoelectric surveys represent a modern methodology of non-invasive geophysical investigation and are based on the detection of electrical resistivity of the various types of investigated soil (Carrara et al., 1992) This type of prospecting is very useful to detect the presence of cavities and to localize the critical points for exploratory surveys, avoiding to perforate in insignificant areas (Evangelista et al., 2017). The test provides a qualitative representation of the electrostratigraphy of the subsoil that gives us an idea of the distribution of real resistivity in the subsoil and allows us highlighting local anomalies determined by “too high” or “too low” values resistivity with respect to adjacent values. In both case studies, Sant’Agostino in Caserta and Piazza Municipio in Aversa, to understand the extension and layout of the crypts, Electrical Resistivity Tomography (ERT) surveys were undertaken in the surrounding areas of the church. The data processing has allowed the geometry and characteristics of geological bodies to be defined, as well as indications on the presence of voids in the subsoil. However, geophysical surveys, as indirect surveys, do not replace direct surveys such as continuous core drilling and/or open excavations, which should be carried out in order to confirm and calibrate the results obtained from geophysical tests, especially in the presence of recognized or hypothesized hypogea. Laser scanner A laser scanner survey of the cave networks extending beneath the Cloister of S. Agostino in Caserta and the Municipio square on the opposite side of the Monumental Complex of S. Francesco delle Monache in Aversa was performed, using a Leica BLK360 laser scanner. Leica BLK360 works, as all laser scanner systems, basing on the emission and subsequent receive of laser beams sent towards the object to be detected, measuring the return time of the laser beam and, according to the time-of-flight TOF principle, measuring distance and angle of the reflection point. The result of the scan consists of a point cloud, defined by their coordinates and by other parameters, i.e. the reflectivity coefficient k, useful for providing information aimed to discriminate different materials for porosity, compactness, etc. The instrument used is equipped with a photographic camera (the images are used to “texturize” the three-dimensional obtained model) and with a thermal camera. In this work, it was possible to carry out the survey without target positioning, thanks to the redundant number of scans and the high precision of the used instrumentation. Indeed, a number of 12 setups in the case of the Cloister of S. Agostino and 19 setups in the case of Municipio square were performed. Firstly, the results of the setups were used to reconstruct, through the software partner Register 360, a unique point cloud and the related 3D model of each cave network, allowing the analysis of morphological and geological features. The performing of setups up to the surface has allowed a detailed reconstruction, by means orthographic and perpendicular projections, of relationships amidst underground cavities and buildings. Lastly, the point clouds have been exported to other cad software to enable i) the comparison with the surveys performed with other methodologies, ii) the processing of sections useful for stability analyses. Results and discussion The Cloister of Sant’Agostino The Sant’Agostino Complex is located in the historic center of the city of Caserta and was built in 1441 by the Augustinian Friars. It includes a sixteenthcentury cloister with adjoining convent, the Church of Sant’Agostino which houses seventeenth-century paintings, the Museum of Contemporary Art and the Museum of Traditions, both declared of regional interest. This structure has recently been the subject of a restoration project which required investigations inside the cloister, in whose subsoil a cavity, with an adjoining cistern, is known and accessible (fig. 2). First of all, the stratigraphic structure of the subsoil was reconstructed, integrating bibliographic data deduced from geological investigations carried out in situ and in the surroundings. The first 4-5 m from the ground level are characterized by loose pyroclastic deposits, deriving from eruptions subsequent to the emplacement of the CGT, which instead characterizes the underlying levels with the zeolitized yellow facies. The contact between the two units is marked by a thin

236 Artificial cavities under worship places: case studies from the province of Caserta (Italy) layer of “cappellaccio”, representing the upper portion of the lithoid tuff formation (CGT) altered in subaerial environment. This stratigraphic sequence is also confirmed by the analysis of the MASW and DL030 performed inside the cloister. They display a very loose granular soil in the first 5 m; at the base of this unit a sudden increase in measurement parameters occurs, indicating a transition to the CGT formation. On the base of such results the geotechnical model was derived and the subsoil was schematized as a single layer of cohesionless pyroclastic soil 4.5m thick directly lying on the CGT formation. The parameters used for the rock and soil are shown in Table 1 in terms of unit volume weight (γ), Poisson coefficient (ν), effective cohesion (c’), friction angle (φ’) and the uniaxial compression (σc ) strength of the lithoid tuff. It has to be noted that the indicated value of cohesion Fig. 2 – Plan view of the Cloister of Sant’Agostino and location of shooting points of parts of the hypogeum; a) type of access: stair with three ramps; b) view of the northeastern corner of the outer wall of the cistern; c) view of the southwest corner with vertical access point.

237 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa for the Post-CGT soil (15kPa) is not a true cohesion but it takes into the account for the beneficial effect of the unsaturated state on the shear strength of the soil (Damiano and Olivares, 2010). The survey of the cavity with laserscanner techniques has highlighted the extension of a pentagonal shaped cave, connected by means vertical shafts to the courtyard above (fig. 4). Given the complexity of the monumental complex, the presence of other voids in the area was suspected. The electrical tomography investigations for the Sant’Agostino Cloister were performed to understand the real extension of the cavity body and to confirm or not the presence of other voids (fig. 5). The performed tomography confirmed the presence of the known and accessible cavity, the access shaft, the presence of the stairwell to the north, corresponding to high resistivity values. In the initial hypothesis, the presence of underground voids was also suspected in the northern part of the cloister courtyard, and the results obtained would seem to confirm the existence of an inaccessible empty space. In fact, in section BB’ (fig. 5) there are strong contrasts of electrical resistivity starting from a depth of about 2 m, rather superficial, considering that in the study area the tuff notoriously characterized by high resistivity values is present at a depth of 4 meters approximately, it is excluded that these resistivities can be attributed to the tuffaceous formation. All this suggests a posFig. 3 – Stratigraphical sequence and results of penetrometer test used for soil characterization at the Sant’Agostino Cloister. Legend: a) pedogenic soil; b) post-GCT deposits; c) GCT “Cappellaccio”; d) CGT. Depth (m) Soil γ (kN/m3 ) ν c’ (kPa) φ’ (°) σc (kPa) E (MPa) 0 - 4.5 Post-CGT 15 0.3 15 35° - 170 > 4.5 CGT 14 0.3 120 35° 800 620 Tab. 1 – Cloister of Sant’Agostino: mechanical properties of the soils. Fig. 4 – Sant’Agostino Cloister: relationship between the courtyard and the cave below: 1) access; 2) shafts; 3) cave.

238 Artificial cavities under worship places: case studies from the province of Caserta (Italy) Fig. 5 – Plan view of Cloister of St. Agostino and the map of the cavity, the cistern and the vertical shafts below the courtyard. Red lines indicate the ERT profiles traces. On top and left the resistivity profiles; the black arrows indicate the intersection of the two profiles. For explanation refer to text.

239 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa sible existence of an underground void, also confirmed by the profile AA’ which intersects the previous one, reporting anomalous resistivity values. In fact, the north stairwell has a part of a walled-up side wall in its intact part, as if to signify a sealing of a possible connection to the inaccessible void. However, the accessible cavity does not allow reaching and inspecting this part of the subsoil, to identify which it would be advisable to carry out a coring. It is worth specifying that geoelectric investigations being indirect investigations do not replace direct investigations such as continuous core drilling and open pit excavations, which should always be performed in order to confirm and calibrate the results obtained from the electrical tomography. The Monumental Complex of San Francesco delle Monache The Monumental Complex of San Francesco delle Monache originated between 1219 and 1234. It was a true monastic citadel measuring approximately 1 hectare. The cloister and the convent that make it up have undergone many alterations over the centuries and the entire citadel was dismembered especially during the Fascist period. In particular, what was originally the garden of the ecclesial complex was transformed into a square and connected directly to the railway station, cutting through the Franciscan insula (Pellegrino, 2020). Below the current complex there are several voids: one used as a burial area, one as a cellar and finally a water cistern. Other voids are known on the southern edge of the original garden, while there is no information on the subsoil of the current Piazza Municipio, characterized by frequent deformations of the road pavement. In order to verify the subsoil conditions in the area of the former courtyard of the monumental complex, which could be affected by hypogea, an investigation was conducted both directly on the known and accessible hypogea and indirectly on the subsoil through geophysical surveys. Again, the stratigraphic structure has been reconstructed using literature data and borehole stratigraphies from neighboring areas. Here a stratigraphical sequence very similar to the one of the cloister of S.Agostino was found. Below a pedogenized soil layer about 2m thick, pyroclastic sands of different relative densities lie above the soft rock belonging to the CGT formation which is altered in its most superficial part giving rise to the so-called Cappellaccio layer, a slightly cemented sand. The analysis and interpretation of drillings and penetrometer tests (DPSH type) allowed the reconstruction of the mechanical characteristics of the lithotypes reported in Table 2 (fig. 6). It is worth noting that to perform a detailed analysis of the state of stress of the cavity it is necessary to investigate the characteristics of the rock formation through laboratory tests on samples appositely collected at the investigated site. The stratigraphic and mechanical characteristics of the investigated subsoil, so similar to those found at the Caserta site, suggest that local erosion phenomena of the loose pyroclastic soils due to rainwater infiltration or to leakages from the water and sewer networks, associated with the presence of cavity, may, also in this case, pose a severe risk of instability of the subsoil especially in correspondence of unlined wells. Given the extension and complexity of the entire monumental complex, the search for other underground hypogea is very complicated. The hypogeum on the southern edge of the square was surveyed using a laser scanner technique (fig. 7) which revealed that it extends below the building that represented the edge of the original garden. Over time, the cavity has been used for the production and conservation of wine and was probably connected to a more complex underground system which extended toFig. 6 – Stratigraphical sequence and results of penetrometer test used for soil characterization at the San Francesco Complex. Legend: a) pedogenic soil; b) post-GCT deposits; c) GCT “Cappellaccio”; d) CGT. Depth (m) Soil γ (kN/m3 ) ν c’ (kPa) φ’ (°) σc (kPa) E (MPa) 0 - 4.6 Post-GCT 15 0.3 0 35° - 170 4.6 – 6.2 CGT Capp 14.5 0.3 15 37° - 420 > 6.2 CGT 14 0.3 300 37° 1100 800 Tab. 2 – Monumental Complex of San Francesco delle Monache: mechanical properties of the soils.

240 Artificial cavities under worship places: case studies from the province of Caserta (Italy) wards the historic center of the city of Aversa. The sepulchral area and the cistern instead extend only below the current cloister of San Francesco, to the current state of knowledge. The electrical tomography was performed starting from the south side of the original garden, diagonally up to the north border (fig. 8). The results of the investigation show the presence of an excavation in the tuffaceous material of approximately 10 m wide and 5 m deep, filled with material showing signs of reworking. There is no documentary information about the presence of tanks in the area used as a garden, nor does the single profile show us the 3D extension of this hypothetical excavation. Furthermore, what is noteworthy is the high resistivity value recognized right on the border with the current cloister and which would Fig. 7 – Monumental Complex of San Francesco delle Monache: (a) cellar, (b) cistern, (c) burial places (or crypta); hypogeum on the southern edge of Municipio square (d-e) warehouse; f) shaft once communicating with the house above; g) laser scanner image: perspective view of the cavity and the building above.

241 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa suggest the presence of another hypogeum in this portion of the former monumental complex. Also in this case, however, further investigations are needed to understand the 3D development of the void. Conclusions The results allowed us to improve the knowledge of the study sites and provide useful tools for the planning of future targeted investigations. Of particular importance was the verification of the reliability and potential of the geophysical investigations which, in addition to providing an overall characterization of the site of interest, return rather detailed information especially in the presence of hypogea, in order to be able to guide other types of investigations that integrate the results achieved. Above all, this study underlines how integrated research between applied disciplines (in this case, geology, geotechnics, speleology, cultural heritage) can provide indispensable support both for the management and mitigaFig. 8 – (a) Aereal view of the Monumental Complex of San Francesco delle Monache and Municipio Square and the location of ERT profile trace (in red); (b) ERT resistivity profile; dotted lines indicate the beginning of each segment. For explanation refer to text.

242 Artificial cavities under worship places: case studies from the province of Caserta (Italy) tion of geological risks in urban areas and for the sustainable reuse of hypogea thus contributing to enhance the cultural and tourist promotion of a territory. Bibliography Amorosi A., Pacifico A., Rossi V., Ruberti D., 2012, Late quaternary incision and deposition in an active volcanic setting: The Volturno valley fill, southern Italy. Sedimentary Geology, 282, 307–320, doi:10.1016/j.sedgeo.2012.10.003. Aversa S., Evangelista A., 1998, Mechanical behaviour of a volcanic tuff: yield, strength and “destructuration” effects. Rock Mechanics and Rock Engineering, 31(1), 25-42. Carrara E., Rapolla A., Roberti N., 1992, Le indagini geofisiche per lo studio del sottosuolo: metodi geoelettrici e sismici. Ed. Liquori. Ceroni F., Pece M., Manfredi G., Marcari G., Voto S., 2004, Analisi e caratterizzazione meccanica di murature di tufo. Atti del 15° Congresso C.T.E. Bari, 4-5-6 novembre 2004. Cinque A., Alinaghi H.H., Laureti L. & Russo F., 1987, Osservazioni preliminari sull’evoluzione geomorfologica della Piana del Sarno (Campania, Appennino meridionale). Geogr. Fis. Dinam. Quat., 10, 161-174. Cinque A., Ascione A. & Caiazzo C., 2000, Distribuzione spaziotemporale e caratterizzazione della fagliazione quaternaria in Appennino meridionale. In: F. Galadini, C. Meletti & A. Melez (a cura di): Le ricerche del GNDT nel campo della pericolosità sismica(1996-1999), CNR-GNDT Spec. Publ., 203-218, Roma. Damiano E., Olivares L., 2010, The role of infiltration processes in steep slope stability of pyroclastic granular soils: laboratory and numerical investigation. Natural Hazards, 52(2):329-350. De Vivo B., Rolandi G., Gans P.B., Calvert A., Bohrson W.A., Spera F.J. and Belkin H.E., 2001, New constraints on the pyroclastic eruptive history of the Campanian volcanic Plain (Italy). Mineralogy and Petrology 73: 47-65. Deino A.L., Orsi G., de Vita S., Piochi M., 2004, The age of the Neapolitan Yellow Tuff caldera-forming eruption (Campi Flegrei caldera-Italy) assessed by 40Ar/39Ar dating method. Journal Volcanol. Geotherm. Res. 133:157-170. Ebanista C., 2007, L’utilizzo cultuale delle grotte campane nel medioevo, in Campania Speleologica, Atti I Convegno Regionale di Speleologia, Oliveto Citra, 127-150. Ebanista C., 2011, Abitati e luoghi di culto rupestri in Campania e Molise. In: E. Menesto (ed.), Le aree rupestri dell’Italia centro·meridionale, nell’ambito delle civiltà italiche: conoscenza, salvaguardia, tutela. Atti IV Convegno Internazionale sulla civiltà rupestre, Savelletri di Fasano, 39-94. Evangelista L., de Silva F., d’Onofrio A., Di Fiore V., Silvestri F., Scotto di Santolo A., Cavuoto G., Punzo M., Tarallo D., 2017, Application of ERT and GPR geophysical testing to the subsoil characterization of cultural heritage sites in Napoli (Italy). Measurement, 104, 326–335. Guidone I., Izzo F., 2015, L’importanza del Catasto delle Cavità Artificiali in Regione Campania, in Atti del XXII Congresso Nazionale di Speleologia Euro Speleo Forum 2015 “Condividere i dati”, 30 maggio-2 giugno 2015, Pertosa-Auletta (SA), pp. 639-644. Hatanaka M., Uchida A.,1996, Empirical Correlation Between Penetration Resistance and Internal Friction Angle of Sandy Soils. Soils and Foundations, 36 (4), 1-9 https://doi.org/10.3208/sandf.36.4_1. Ippolito F., Ortolani F., Russo M., 1973, Struttura marginale tirrenica dell’Appennino campano: reinterpretazioni di dati di antiche ricerche di idrocarburi. Mem. Soc. Geol. Ital., 12, 227-250. Meneghini M., 2010, Il Catasto Nazionale delle Cavità Artificiali della Società Speleologica Italiana. In: Le Cavità Artificiali dalla riscoperta alla valorizzazione turistica. Corso Naz. III livello, 2-4/12/2011 Urbino, 105-113. Pellegrino A., 2020, Aversa, San Francesco le Moniche nel Medioevo. Quorum edizioni, 128 pp. Ruberti D., Buffardi C., Sacchi M. & Vigliotti M., 2022, The late Pleistocene-Holocene changing morphology of the Volturno delta and coast (northern Campania, Italy): Geological architecture and human influence. Quaternary Int. 625, 14-28. https://doi. org/10.1016/j.quaint.2022.03.023. Ruberti D., Vigliotti M., Rolandi R., Di Lascio M., 2020, Effect of paleomorphology on facies distribution of the campania ignimbrite in the northern campania plain, southern Italy. In Vesuvius, Campi Flegrei, and Campanian Volcanism; De Vivo, B., Belkin, H.E., Rolandi, G., Eds.; Elsevier: Amsterdam, The Netherlands; pp. 207–229, doi:10.1016/B978-0-12-816454-9.00009-2. Sacchi M., Molisso F., Pacifico A., Vigliotti M., Sabbarese C., Ruberti D., 2014, Late-Holocene to recent evolution of Lake Patria, South Italy: an example of a coastal lagoon within a Mediterranean delta system. Glob. Planet. Change 117, 9-27. Skempton A.W., 1986, Standard Penetration Tests Procedures & the Effects in Sands of Overburden Pressure, Relative Density, Particle Size, Ageing and Overconsolidation. Géotechnique. 36(3): 425-447. Vigliotti M., Ruberti D., 2018, Campania Grey Tuff and anthropogenic tuff cavites in the Southern metropolitan area of Caserta (Southern Italy). Alpine and Mediterranean Quaternary, 31), 247 - 251 (Quaternary: Past, Present, Future - AIQUA Conference, Florence, 13-14/06/2018, ISSN (print): 2279-7327, ISSN (online): 2279-7335. Vitale S, Ciarcia S., 2018, Tectono-stratigraphic setting of the Campania region (southern Italy). J Maps 14(2):9–21. doi. org/ 10. 1080/ 17445 647. 2018. 142

243 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa 1   Centro Studi Basilica di San Gavino di Torres, Porto Torres, Italy 2 Gruppo Speleo Ambientale Sassari, Italy *  Reference author: Giuseppe Piras - [email protected] The hypogeum of San Gavino a mare in Porto Torres (Sassari, Sardinia, Italy): preliminary epigraphic, glyptographic and speleological investigations Giuseppe Piras1,*, Pier Paolo Dore2 Abstract The study analyses the hypogeum complex of San Gavino a mare in Porto Torres (Sardinia, Italy) and the homonymous small church adjacent to it, a site locally known to have been the burial place of the three Christian martyrs Gavino, Proto and Gianuario, immediately after their decapitation that probably took place in 303 AD under the emperors Diocletian and Maximian. Geological, and speleological features of the hypogeic complex are surveyed. In addition, a preliminary investigation is made of the epigraphic evidence and graffiti found not only in the hypogeum’s rock walls, but also in the internal walls of the room behind the presbytery area of the church 1. Keyword: epigraphy, glyptography, San Gavino a mare, Balai. Introduction The toponym Balai is currently used to indicate the long strip of coastline in the north-eastern sector of the municipal territory of the city of Porto Torres (Sassari, Sardinia, Italy), which extends for just over 2 km and includes, as natural and monumental elements that distinguish its landscape, the beach of the same name, some natural sea caves and two small churches, closely linked to the cult of the Christian martyrs Gavino, Proto and Gianuario. It is precisely the site where one of these places of worship stands, the church today dedicated to S. Gavino a mare (also commonly known as ‘di Balai vicino’ to distinguish it from another one, located 2.2 km to the east and known by the appellative ‘di Balai lontano’), that is the subject of the present study and to which the first documentary evidence of the place name of this locality is connected. Today’s name Balai would, in fact, derive from the ancient form Balagay or Baragai; the first variant of the toponym is mentioned in the Inventio corporum sanctorum martyrum Gavini, Prothi et Ianuarii, the drafting of which has been variously assigned by scholars to a period between the first half of the 11th and the 14th century, which constitutes the lectio IX of an Officium of the three martyrs contained in a volume printed in Venice on 30 May 1497 (Piras, 2019: 57, footnote 79), while the second variant is related to the surname de Baragai mentioned in card 299 (datable between 1154 and 1170) of the Condaghe di San Pietro di Silki (Piras, 2019: 74, footnote 86). Specifically, it is a surname that can be classified in the category of Herkunftsnamen, i.e., surnames formed directly from toponyms or ethnic names, which suggests that a small settlement with the name Balagay or Baragai may have existed in Late Antiquity. An important reference for tracing the origin of the toponym Balagay is certainly its use in Arab sources of the 10th-13th centuries and its association in Catalan, French and High-Provençal languages with the word used to indicate the broom plant (Piras, 2019: 74-80, footnote 86). Overview of geology and speleology The coastal area concerned by the study consists of a limestone cliff that belongs to the territory of Porto Torres, an inhabited centre located on the northwestern coast of Sardinia and overlooking the Gulf of Asinara. Geologically, the Turritana plain was formed by the intense fracturing that occurred during the opening of the Sardinian-Corsican basin, when shallow-water sediments filled the great rift known as the ‘Fossa Sarda’ (Sardinian Trench). On the northwestern slopes of this trench lies the territory of Porto Torres, filled by a sequence of shallow-water marlycarbonate sediments, subsequently affected by nu1 Although conceived as a single unit, this article is divided into five paragraphs: §§ 1 and 3-5 are by Giuseppe Piras, § 2 by Pier Paolo Dore. The English texts are translated by Dr. Giovanni Ferrandu, to whom the authors extend their sincere thanks.

244 The hypogeum of San Gavino a mare in Porto Torres (Sardinia, Italy): preliminary investigations merous secondary tectonic processes. Two main lithologies have been distinguished in the sector studied, representing different sedimentation environments. In the western part, up to the beach of Scoglio Lungo, outcrops are mainly biocalcarenites composed mainly of algal nodules, very porous, where numerous small karst manifestations are present. To the east, along the promontory, the composition of the sedimentary rock changes, becoming richer in clay and turning into marl, often arenaceous, where karst phenomena are mainly set into fractures (Funedda, 2000). From a geomorphological point of view, the survey area consists of a karst plateau without watercourses, set on the Miocene sedimentary formation, bordered towards the sea by a very jagged cliff coast. The area also features riversides zones, crevasses and karst-erosive forms filled and camouflaged by Pleistocene formations consisting of red sands and Holocene colluvial deposits. The Turritan karst had an ancient genesis, testified by the presence of Plio-Pleistocene deposits found within some cavities. During this period there were important eustatic variations, linked to Pliocene glacial cycles, which generated the different levels in the karst system; these eustatic variations are evident along the coastline, where it is still possible to recognise traces of ancient grooves flying (Martini, 2000). In the caves of the upper level, we often find continental, or beach deposits related to the Tyrrhenian period that testify to an earlier karstification. In the slab furrows it is also often possible to observe holes of lithodomous organisms that colonised the coasts of Sardinia during the last interglacial. The high escarpments and morphological data have allowed interesting palaeogeographical reconstructions, first of all, the erosive escarpment, corresponding to the old Tyrrhenian grooves flying located at about 5.5 metres from the present sea level, in some cases, this groove is very evident; another less pronounced groove between 8 and 9 metres has also been identified, which could be due to a previous interglacial; interspersed by the erosive plain at around 6/7 metres. On the other hand, the present-day thrust furrow is very pronounced, indicative of a longer eustatic stasis, which has deepened further at submarine karst resurgences, due to the action of seawater/freshwater mixing. The groundwater network was reconstructed not only by direct exploration of the cavities, but also by sampling from the various boreholes carried out by the Turritan municipality (Dore et al.; 2020). The karstification of the Turritan coast had previously been studied and made known by an excellent publication published in 1983 by Mucedda and Cossu of the Gruppo Speleologico Sassarese. The publication described all 17 cavities found from the Scoglio Lungo to the Torre di Abbacurrente (Mucedda 1984, 1996). A further study by the writer then led to the discovery of a further three cavities, bringing the number to a total of 20 caves. The karstic processes that have acted in Fig. 1 – Front view of the church with one of the entrances to the hypogeum (photo G. Piras, P.P. Dore).

245 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa the formation of these cavities are manifold; most of the caves were formed in the low-solubility carbonate marls, settling exclusively on fractures. Coastal cavities are therefore to be considered originally mesocarstic and deepened by epigenetic pseudocarst processes, which through the action of seawater and the solid components transported in it, such as sand and pebbles, caused these cavities to be shaped by coastal erosion. A basic aspect of this Turritan sector is the presence of a deep groove of present-day flow, also very visible in some caves, formed mainly by hyperkarst. In carbonate areas, in fact, submarine resurgences create a layer of freshwater that, due to its density, floats above seawater; the anionic exchanges between these waters make this horizon highly aggressive to calcium carbonate salts, creating the splendid beating furrows visible almost everywhere in the area. As for the hypogeum of San Gavino a mare, it is evident that its genesis is in an erosive esplanade at about 5 m above sea level (fig. 1), although traces are scarce, the fracturing of the main room indicates a possible genesis from a natural cavity that was then completely excavated by hand. The hypogeum currently consists of three distinct rooms. The first room is the most external to the church and is currently closed by two large metal gates (“Chamber C”); it consists of a room with two small extensions and a small corridor that currently communicates with a small window in the second room (“Chamber B”), a smaller room in which there is also an old fireplace, probably a remnant from a period when hypogea were used by fishermen. Chamber C is irregularly shaped and has two deep niches; the main room is about 5 metres in length and from it a narrow corridor runs in the North direction leading to a second external entrance, which is also closed, and to a small window with a grille that connects Chamber B. Quadrangular in shape and smaller in size (2.5x3.0m), this room is used today as a hallway by the monks and is connected to the third and final Chamber A. The latter is certainly the most important room as, in fact, it is directly connected to the church, but is also the room where the statues of the saints are brought for adoration for about a month. This chamber measuring 7.6 x 5.5 m has an average height of 2.5 m and features large side seats, a small altar, and a large votive niche, as well as being the one with the most inscriptions (fig. 2). The church The church of San Gavino a mare or ‘di Balai vicino’ lays its foundations on a cliff overlooking the sea, near the beach of Balai (Tab. 1). It was erected in this Tab. 1 – Cartographic surveys of the survey site (graphics P.P. Dore, E. Dallocchio, G. Piras).

246 The hypogeum of San Gavino a mare in Porto Torres (Sardinia, Italy): preliminary investigations place because, according to hagiographic tradition, the three martyrs Gavino, Proto and Gianuario were buried in the hypogea adjacent to it, excavated in the rocky ridge, immediately after their beheading, which probably took place in 303 A.D. during the persecution of Christians by the emperors Diocletian and Maximian, by order of the praeses provinciae Barbarus. According to the tale of the Passio sanctorum martyrum Gavini, Proti et Ianuarii, a document containing the stories of the three martyrs (an edition can be found in Zichi, 20135 ), datable to a period between the second half of the 11th and the beginning of the 12th century, Proto, a presbyter of the Christian community of the colony of Turris Libisonis (today Porto Torres) and Gianuario, his deacon, preached the word of Christ despite the persecutory edict promulgated by the emperors. For this reason, the governor Barbaro had them imprisoned and tortured, entrusting them to the soldier Gavino, who, fascinated by their faith, converted, freed the two and went to turn himself in to the governor. He was thus beheaded on 25 October and two days later, on 27 October, Proto and Gianuario were also decapitated. The small church, originally dedicated to Sanctu Gavinu scapichatu (or San Gavino Descabeçado), i.e. ‘Saint Gavino decapitated’ (Piras, 2019: 36-37 and 71-72, footnote 84) as this site was most probably also the locus martyrii of the three martyrs (on this proposal cf. Piras, 2019: 57-60, footnote 79), it has a single barrel-vaulted nave supported by doubleaux arches (on the architectural structure see, among others, Spanu, 2000: 123; Piras, 2013: 25). It faces north due to the conformation of the rock on which it was built. It was built using ashlars of local organogenic limestone, except for the outer face of the end wall facing east, made of trachyte slabs, and the outer face of the back wall, made of well-squared ashlars larger than the rest of the masonry. Mistakenly, based on an indication by canon Giovanni Spano (Spano, 1856: 147, note 1), the structure we see today has been assigned in the past to the mid19th century. (e.g., Masala, 1988: 258, footnote 43), while in reality its oldest depiction so far found is included in a triptych preserved in the church of S. Giovanni in Bonorva (Sassari, Sardinia, Italy), a painting datable between the second half of the 17th and the beginning of the 18th century (Piras, 2019: 36-37 and 70-71, footnote 83). The existence of a room of worship or a martyr’s memorial chronologically prior to the present building is confirmed by the aedicule that came to light below the altar during archaeological investigations carried out in the church in 1980 (Mastino and Vismara, 1994: 100; Manconi, 2001: 40). Fig. 2 – Photo of the main room where the martyrs are placed (photo G. Piras, P.P. Dore).

247 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa The hypogeum Chamber A From the west side of the worship hall, it is possible to access the hypogeum that, according to hagiographic tradition, would have housed the bodies of the three martyrs of Turris Libisonis (Piras, 2019: 57-60, note 79). Three rooms of the hypogean complex, originally communicating with each other, have been preserved and have undergone continuous tampering over time, as shown by the comparison between the current situation and the one described in the 19th century by Vittorio Angius and Giovanni Spano (Angius, 1847: 652; Spano, 1856: 123-125, 138-144 and 145-147). Of the three hypogean chambers, inserted in the context of the eastern necropolis of Turris Libisonis or perhaps of that pertaining to the small settlement of Balagay (or Baragai), to be identified as the locus depositionis of the remains of the three martyrs is the one directly connected to the small church by a door. The chamber, with an irregular trapezoidal plan, can be interpreted as a funerary hypogeum of the Roman period (cf., among others, Mastino and Vismara, 1994: 100; Poli, 1997: 201; Spanu, 2000: 125; Manconi, 2001: 40), and presents along two sides of the wall a seat spared in the rock bench, a base of a rectangular structure made of trachyte and limestone blocks (perhaps an ancient altar) on the south side and a niche on the west side. Inside this apsidiole, which was probably excavated for liturgical purposes, considering the existence of an altar built in front of it until recent times (Spano, 1856: 147), traces of painted plaster have been found: more precisely, two evident fragments of red plaster (fig. 3) on top of a Latin cross carved in relief in the rock and another small ochre-coloured fragment preserved in the apsidal basin. As Spano already pointed out in 1856, what remains of the painted plaster film that must have covered the walls of the chamber can also be seen in other places: on the west wall, above an aedicule, a large section of yellow ochre can be seen with a brown band running underneath, and a circular plaster fragment in black and red can be seen at the bottom of the north wall (Piras, 2019: 61-62, footnote 80). In the centre of the chamber, with a structural function, was a column with an Ionic capital placed in an inverted position (Botteri, 1978: 110), elements removed in the 1980s to make the room more accessible to worshippers and now unfortunately missing. In the rock walls, quadrangular holes can be seen in which wooden beams were inserted in the past to support shelves. The central role played by the hypogean complex in the popular devotion towards the three martyrs of Turris Libisonis and its frequentation over the centuries are confirmed by the graphic testimonies left by the worshippers who went to the locus depositionis of Gavino, Proto and Gianuario to pay homage to them, to pray, to implore intercession for requests for support or to thank them for the assistance they had been granted. Past the door connecting with the cult room, on the north wall of the chamber, 36.5 cm from the northeast corner and 100.5 cm above the seat, is scratched a signum (h 8 cm) in which one could recognise a chrismon with a particular shape (Piras, 2019: 63, footnote 81). Continuing along the wall, towards the west, about 130 cm from the christogram, four signs are visible that can perhaps be interpreted as mason’s marks (fig. 4): the first has a ‘herringbone’ shape (a vertical rod of 11.5 cm with two arrows at the upper and lower extremities), the second (5 cm from the first) is a vertical segment (h 13 cm) with two hooks at the top and intersected at the bottom by a St. Andrew’s cross (its shape is similar to a late 15th-century mason’s mark found in the basilica of San Gavino in Porto Torres and attributable to the work of the master craftsman Zazius; cf. Piras, 2016: 34-36), the third (10 cm from the second) is a cross on Golgotha (h. 7 cm), the latter being represented with a triangle (the sign was already in Piras, 2005: 410; Piras 2019: 62, footnote 81), the fourth is a segment with two opposite arrows at the extremities. At 10 cm from the cross on Golgotha is another sign, possibly related to a devotional context: a lowercase letter h, approximately 19 cm long (the measurement is partial because part of it is concealed by a film of plaster that must have covered the walls of the hypogeum), which we also find on the west wall inside the niche, associated with other letters. Above these five signs, an inscription was painted on the rock with a red pigment in a later period than the Fig. 3 – Fragments of red plaster on the west wall of chamber A (photo G. Piras, P.P. Dore).

248 The hypogeum of San Gavino a mare in Porto Torres (Sardinia, Italy): preliminary investigations graffiti. It is lacunose because it’s under the plaster layer too, but four capital letters can still be identified (h 7 cm; total length 23 cm) of which the last three can be transcribed as follows: DEO. In the centre of the north wall, two, possibly three, inscriptions with Greek letters were also found, inscriptions that are now missing at the two lateral ends and the lower margin, as they are also covered by plaster. They currently occupy an area of 18.5 cm in height and 65 cm in length, the characters were engraved into the rock and the grooves appear to still retain traces of rubrication. The largest letter measures 5 cm in height, the smallest 2 cm. The inscriptions are unpublished and still being studied by the writer. Moving towards the north-west corner, 50 cm from the inscriptions, other evidence related to the religious sphere is engraved: a stylised palmette (h 10 cm), a Christian symbol linked to martyrdom, and, on either side, two silhouettes of sandal insoles (fig. 5), better known by the name of ‘pilgrim’s footprints’ (Piras, 2005: 410, footnote 125; Piras, 2019: 62, footnote 81), belonging to a typological variant similar in characteristics to the one found in the artificial caves of Santorkaria, near the village of Laño (Province of Alava, in the Basque Country), hypogeum that took on the function of hermitages in the early and late Middle Ages (about a redefinition of the symbolic meaning and diffusion of the ‘pilgrim’s footprints’, a specific category of graffiti not limited to Sardinia but extended in the medieval and post-medieval age to a large part of the West Europe, see Piras, 2012: 94-95 and 92-93, footnote 190 for the reference to the Santorkaria complex). At 25 cm from the palmette, it’s a hooked cross (h 5 cm), a religious symbol (fig. 6), as a Latin cross (h 11 cm) scratched on the west wall. The Latin cross has a curious descending oblique tract grafted into the lower half of the lower arm, a detail found with the same characteristics in the monogrammatic cross scratched, not far away, on the wall. In the staurogram (h 7 cm), the tract is grafted at the point where the horizontal arm of the cross joins the semicircular lobe pointing to the right (fig. 7). Further on, 37 cm from the apsidiole and 126 cm from the rock bench below, a vertical rod (h 11 cm) is scratched on the wall, cut at the bottom by an oblique descending line slightly curved at its end. Inside the apsidiole, in addition to the above-mentioned cross on Golgotha (41.5 cm high), excisa (i.e. carved in relief) in the rock, there is a cross on the right side and a few characters below, including a lower-case h, perhaps forming a titulus that, at the current state of research, cannot yet be attributed to a specific alphabet, let alone be chronologically framed. On the left side of the niche two other signs, similar to each other, one of which (length 3 cm) is reminiscent in shape of a trident or, vaguely, of the lower part of the schematic petroglyphs representing inverted anthropomorphs found in some domus de janas from the Neolithic period in Sardinia, while the other (length 2.7 cm), compared to the first, lacks the central vertical stroke. Chamber B A breach in the south wall of chamber A leads to another room, the dimensions of which are 3.5 x 3.4 metres. The room is the result of the construction in ancient times of the wall separating it from the church on the east side, while in modern times a small wall with a small window was erected to separate it from the rest of the hypogeum (extended towards the south and in direct communication with the outside on the east side) to protect the sacred place. Although the inspection of the inner walls of the chamber is not easy, by virtue of the reduced space in which to move, it was possible to identify the presence of a stylised palmette (h 6.5 cm) and a cross on Golgotha (h 11.5 cm), side by side, in the inner right jamb of the opening, at a height of 105 cm from the floor level. Continuing along the north wall, numerous mason’s marks engraved deeply into the rock are visible. Among the various shapes can be recognised: a square with two internal diagonals (three census specimens); a triangle with the vertex pointing downwards, cut internally by an ascending oblique line and provided with a cross; two contiguous squares crossed by two diagonals with opFig. 4 – Four signs that can perhaps be interpreted as mason’s marks and the lowercase h in the north wall (photo G. Piras, P.P. Dore). Fig. 5 – The palmette with the two ‘pilgrim’s footprints’ on either side (photo G. Piras, P.P. Dore).

249 Fourth IC of Speleology in Artificial Cavities Hypogea 2023 - Genoa posite and converging courses; a hooked cross (more regular than the one in chamber A); a backwards F; an ‘arched’ signum (a mason’s mark very common in Europe from the early Middle Ages onwards); a monogram in relief formed by the letters F and M. Other signs can be found near the north-eastern corner of the wall, which is difficult to reach; among these, the letters NA (or HA) in an epigraphic Gothic script can be identified. On the south wall, at 107 cm from the ground, another lapidary mark (h 5.1 cm) that proposes a geometric pattern depicting a cross on Golgotha (the mount is a triangle) terminating at the top with a right-angled triangle facing left. The sacellum The back wall of the worship hall divides the presbyterial area from a rectangular, barrel-vaulted room made of limestone blocks (which can be accessed through a door), the function of which has not yet been clarified, perhaps a Roman cistern transformed into a sacellum in the early Middle Ages (Mastino and Vismara, 1994: 100; Manconi, 2001: 40), although underneath the thick layer of plaster that covered, until recent times, the ashlars of the masonry, there do not appear to be any traces of hydraulic mortar (Piras, 2019: 80-81, footnote 88). There are numerous epigraphic testimonies on the walls of the room, many of which are signatures left in the 20th century. The room was frequented, perhaps already in ancient times, as show the two Greek letters rho and xi with a horizontal straight tilde above the latter character, engraved deeply in the ashlar (w. 42 cm; h. 38 cm) inserted in the north-west corner, under the cornice that runs along the wall of the room. The two letters (h of the rho 15.2 cm; h of the xi 17.7 cm), the first capital (or, more likely, lower case, but with a bow high above the line of text) and the second lower case, both rather coarsely engraved, could be interpreted as the identity mark of a stone mason, expressed either as an abbreviation by truncation (indicated by the tilde) or, if ρ and ξ are considered numerals, as a ‘utility sign’ indicating the number 160. In addition to the acronym ρξ on the west wall, other graffiti were found on the remaining walls of the room, below the plaster film: among those worth mentioning, on the north wall, in an ashlar (30x63 cm; fourth row from the bottom, fifth ashlar from Fig. 6 – 3D processing of the graffito depicting the hooked cross (photo G. Piras, P.P. Dore). Fig. 7 – The staurogram on the west wall (photo G. Piras, P.P. Dore).