Underwater research began at Torre S. Sabina in 1972 and has continued intermittently over the years. The sea-bed produced and continues to produce ceramic material which varies in both provenance and manufacture, but above all in chronology: from Mycenean pottery to Late Roman C from the late antique period. The archaic materials, dating to between the end of the 7th century B.C. and the first half of the 5th century B.C., were all imports and seemed to be the result of dumping from ships. This material included Corinthian and Laconian pottery, Ionic cups, Graeco-oriental type wares, Attic Black-figure ware and Corinthian-Corcyran A and B amphorae. A headless statuette of a Rhodian type Kore was also found. Amongst the Hellenistic and Roman material was relief pottery, Gnathian ware, black glaze, coarse, table and cooking wares, Italian sigillata, Arretine, Italic and north Italic wares, ARS and African cooking wares and lamps. With the exception for the few examples of early Greco-Italic amphorae from the end of the 4th-first half of the 3rd century B.C. the amphorae were mainly local, late Republican, productions for oil and wine transport. Amongst the finds there was also a bone medallion, an anthropomorphic vase in the form of Priapus, glass, a hopper grindstone and oscilla. Reports had been made in the past regarding the presence of at least two wrecks in the northern area of the bay, calculated on the basis of the number of finds, their variety and the presence of burnt pieces of wood in various points on the sea-bed. However, it may be suggested that the presence of such a substantial assemblage of ceramics and amphorae material cannot be attributed to dumping activities in the port and above all, that the finds and wooden structures may have belonged to a single vessel. Furthermore, as a boat may have been wrecked during the course of the 2nd century B.C., a date provided by the Lamboglia 2 amphorae, the Megaran pottery also found could have been part of the same cargo. In fact, it cannot be excluded that this refined pottery, having arrived at Brindisi from the Aegean and mainland Greece, travelled on to another destination. As part of a scenario of commercial redistribution it could have been loaded together with a cargo of basic necessities (oil and wine) from the Salento contained within locally produced amphorae. It cannot be excluded that this mixed cargo was from an Adriatic port, possibly on the opposite coast, where finds of these ceramic types are common. What is certain is that this ship never reached its destination as it was wrecked, perhaps due to a fire on board, as suggested by the traces of burning on the wood and pottery fragments. Other wrecks are known within the bay: one of these (Torre S. Sabina 1), identified in the 1970s along the south side of the largest inlet, came to light again in 1989 after its traces were lost due to silting over. A second ship (Torre S. Sabina 3) was found orientated towards the north immediately adjacent to the slope of the cliff. The C14 analysis undertaken showed both wrecks dated to within the 4th century A.D. Lastly, it should be noted that in 1998 another wreck emerged: a cylindrical iron object sticking out from the sea-bed was connected to large pieces of wood, thought to belong to the prow of a medieval ship due to the presence of glazed wares and majolica datable to between the 16th-17th century.

A number of wrecks lie on the seabed in the bay, one of which was excavated in 2007. Situated at 2.50 m in depth, a few metres from the shore, it was beached at a right angle to the shore and thus seems to represent an interesting marker of the variations in the sea level. The excavated portion, circa half the boat, showed the external planking, assembled using mortice and tenon joints, to which numerous planks of the internal frame and the robust stepped keel were fixed. The wreck was of great interest due to the presence of elements of the bridge, props, beams and planks, which are only preserved in exceptional cases (only one other case is known for antiquity, in France). Originally the boat must have been over 20 m long and been of a reasonable tonnage. On the basis of the pottery finds, including an intact African amphora and others of the same typology, fragmented but largely reconstructable, the wreck can be dated to the beginning of the 4th century A.D. The 2009 campaign continued the investigations at the foot of the western bank of rock, at the base of the submerged rocks, to a depth of 5 m, in the so-called area B. Here there was a deposit of material partially recovered between the 1970s and 80s (circa eleven thousand pieces), extremely homogeneous both in provenance and fabrication and above all chronology: from Mycenean pottery (LH III) to Late Roman C, from late antique to medieval pottery. The excavation investigated a limited but intact and densely stratified section of the submerged deposit. At least two layers were identified within the deposit, which although they were heavily disturbed by stormy seas, seemed to be in situ. The state of preservation and conditions of the deposition indicate that US 2 was a late Republican cargo, constituted by amphorae and perhaps bricks/tiles (saleable ballast?) of local production and containers, mostly for wine, from the Aegean area. These travelled with table and kitchen wares of oriental production, according to a commercial model of redistribution which had its hub in the great port of Brindisi. Furthermore, it is also necessary to mention three peculiarities which further support the identification of the deposit as the remains of a cargo: the overturned position of the finds; the high incidence in US 1 of cobbles that may be allocthonous, the presence noted in previous campaigns, of timber remains with traces of burning in the cobble layer. All of these elements suggest an upside down stratigraphy, that is an overturned cargo, with the ballast (the cobbles) which, originally spread on the bottom of the hold, now covering the cargo, and with the few remains of the hull, wrecked following a fire on board, overlying or mixed with the presumed ballast. This deposit is separated by a thin diaphragm US 3 from the layer below, different in composition and dating of the materials. US 4 is identifiable, for analogous reasons, with a much earlier cargo, of late archaic date. Therefore, over the centuries a number of boats came to be wrecked on this heavily exposed stretch of coast, and here several cargoes or parts of cargoes precipitated and dispersed along the rocky scarp. At the same point, probably caused by the effects of the sea’s movements, traces of the port’s unloading activity had come together, attested by materials that were isolated and mixed (“intrusions”) both in production and chronology. Lastly, fragments of impasto pottery dating to the middle Bronze Age were found at the base of US 4. These probably related to a prehistoric settlement that developed on this bay and that immediately to the north.

Preserved on the seabed in the bay are the remains of several cargoes, stratified deposits and wrecks (5), one of which (T. S. Sabina 1) was excavated in 2007. This was revealed to be unique for the presence of elements of the upperworks, the chronology (beginning of the 4th century A.D.) and for the remains of a North African cargo (cf. 2009 report) Very close to the latter lay the remains of another boat (T.S. Sabina 5), discovered in 1998, at a very shallow depth (1.70 m). Several metal artefacts from this boat were recovered along the sides of the bay by metal detector: seven helmets, four of which stacked and stuck together by concretions, a canon ball, various iron nails etc. The operations of sorbonatura intercepted a structure with juxtaposed timber and metal parts, 2.74 m long, circa 40 cm maximum height, orientated east-west. The timber beam constituting the central part had a quadrangular section (c. 15 cm sides). The lower face adhered to a metal element with an irregular form. At present it is not possible to confirm the initial attribution of the structure to a post-medieval wreck, nor propose other hypotheses regarding its identification or function, also due to the lack of associated material. Timber samples were taken for C14 dating by the CE.DA.D (Centre for dating and diagnostics) at the University of Salento. During the 2010 investigations continued at the foot of the western bank of rock at the foot of the submerged reef/cliff (area B). This is the site of the stratigraphic sequence made up of several dispersed cargoes or partial cargoes mixed with clear traces of dockside unloading activities. A trench was laid out across the north-south channel/paleo-riverbed, in which the depth and stratigraphy of the deposit is conserved – below about two metres of sandy sediment – in greater quantity and quality with respect to that encountered at the base of the scarp. The aim was to define the sequence, documenting any other presences beyond the two overlying cargoes (T. S. Sabina 4, late Republican and T. S. Sabina 3, late Archaic) and check the incidence of materials from the unloading activities. During the excavation a small column drum was recovered and a quarry block was exposed. Layer US 1 was again exposed, with local and Aegean amphorae (Rhodian and Cnidian), _opercula_, coarse and fine ware pottery, clearly relating to a late Republican cargo. Lastly, a survey was undertaken along the western ridge of the trench towards the north, as far as the Tower and at a depth of 6-7 m. It was seen that a substantial removal of sediment had revealed, among the stones at the base of the rock bank, a large quantity of pottery fragments, often concentrated in small pockets. This vast area of dispersion showed the usual predominance of local Adriatic productions (Greco-Italic, transitional forms between Greco-Italic and Lamboglia 2, Lamboglia 2), in line with the presence and composition of the late Republican cargo from the Torre Santa Sabina 4 wreck.

I fondali della baia conservano resti riferibili a vari carichi, depositi stratificati e relitti (5), uno dei quali (T. S. Sabina 1) è stato oggetto di un saggio di scavo nel 2007, rivelandosi un eccezionale unicum per la presenza di elementi dell’opera morta, per la cronologia (inizi IV secolo d.C.) e per i resti del carico nordafricano (v. precedenti relazioni). Nel corso delle prospezioni effettuate nel 1996 dalla Soprintendenza per i Beni Archeologici della Puglia durante le operazioni di copertura di quel relitto, fu possibile riscontrare una nuova concentrazione di frammenti lignei in una zona attigua alla fiancata nord dell’imbarcazione (50-70 m a nord), adiacente al declivio della scogliera. Gli autori del rinvenimento osservarono la presenza di fasciame assemblato a mortase e tenoni. Accanto ai resti, che apparivano però smembrati in vari punti, vennero collocati anche i numerosi frammenti lignei sparsi nei dintorni e presumibilmente pertinenti anche l\'altro relitto; per tutti si decise di adottare lo stesso sistema di protezione, con l’impiego di altri sei pannelli di CLS. La campagna di scavo subacqueo 2011 ha interessato i resti lignei attribuiti a questo relitto (Torre S. Sabina 2), rimuovendo la copertura di protezione, riposizionata alla fine dell’intervento. Nell’area Nord si è verificata la presenza, al di sotto di un sottile strato di sedimento e detriti, di scarsi e minuti frammenti di tavole di fasciame, gravemente danneggiati ed erratici; ne sono stati recuperati alcuni a testimonianza del pesante stato di degrado. Nell’area Sud sono venuti in luce vari frammenti lignei, sempre erratici, oltre ad un grosso elemento ligneo presumibilmente non lavorato, che insiste su due tavole frammentarie, una delle quali ha visibili in sezione ed in faccia laterale il foro di alloggio di caviglie lignee. Sembra che tali porzioni di quello che si è sempre supposto un relitto a se stante siano state quasi completamente disgregate dall’azione del moto ondoso sotto i lastroni non ben poggianti su una superficie continua, a causa della maggiore intensità dell’energia ambientale in questo tratto. La datazione al C14 dei campioni potrà accertare la distinzione tra questi resti, ormai quasi nulli, e il relitto Torre S. Sabina 1. Si è condotta, parallelamente alle operazioni di scavo, una ricognizione lungo il costone W, nell’area B, dettata dall’esigenza di controllare lo stato dei depositi ai piedi della scarpata dopo alcune forti mareggiate estive; si sono individuati un corno in piombo pieno, con chiodi in ferro infissi obliquamente nel corpo, forse interpretabile come uno scandaglio, ed un ducato argenteo della Serenissima con Leone di Venezia e S. Giustina, attribuibile al periodo di reggenza del Doge Pasquale Cicogna, regnante tra il 1585 e 1595, e la parte superiore di un’anfora corinzia B ellenistica. Un’ulteriore ricognizione ha interessato il tratto settentrionale del costone E della baia, dove si è rinvenuta un’olla biansata integra.

The campaign’s main objective was to intercept a better preserved, “tidier” column of stratigraphy. The aim was an attempt to understand whether, as well as the two deposits interpreted as cargo remains (late Republican the upper one - TSS 4/SR 4, late archaic the lower - TSS 3/SR 3) that had smashed against the rocks, there were other wrecks that could be linked to the other groups of materials that do not seem to relate simply to unloading activities in the port. Therefore, a sector in correspondence with the canyon cutting the seabed in the bay was investigated. Here, continuing from and adjacent to the trench dug in 1978 where a concentration of intact or complete amphorae was uncovered, the sediments seemed to be thicker. The deposit resulting from the disturbed upper cargo (US 1 and the underlying 10-SAS I, 11-SAS II, 13, 15 – SAS III) appeared in all three of this season’s trenches (but to a greater extent in III) below the surface layer constituted by rock slippage. It was made up of allocthonous small rounded pebbles, identified as ballast, and by a substantial concentration of materials, probably _in situ_, smashed under the rocky collapse but partially reconstructable: amphora fragments (mainly produced in the Salento), coarse pottery (table and cooking wares) and various fine wares: Megarian bowls, some complete, _lagynoi_, black gloss cups, fragments of _sovradipinta_ pottery, thin walled ware, Eastern sigillata and colour-coated ware, as well as lamps, glass etc. A timber element found in the central area of SAS III at a depth of about 5 m, appeared particularly interesting. It had a mortise and tenon joint, and seemed, although heavily deteriorated, to be a piece of the boat’s longitudinal structures, perhaps a part of the keel or the bow or stern, which adds to those already recovered in the same area between 1972 and 1983. However, no evidence of other deposits clearly identifiable as cargo remains was found. The only exception was a small nucleus of late-antique finds (LR 2 amphora and African cooking ware), found in SAS II, in the interface between US 1 and the surface layer, which can be linked to the coeval materials recovered between 1972 and 1983. This evidence may represent the remains of a deposit that has largely disappeared (another cargo?), datable to the 5th-6th century, that overlay the much earlier US 1. Lastly, at the base of US 11 – SAS II, at 5.10 m below sea level, a concentration of lithic industry was identified. Datable to the Neolithic period (5th millennium B.C.), together with other elements it may indicate the remains of a work area, now submerged by material washed down from a higher level, or more probably due to a rise in the sea level that submerged the paleo-shoreline in the Neolithic and Eneolithic periods. Moreover, this evidence is a reflection of that found on other important sites on the coast of the Salento during research undertaken with the aim of reconstructing the ancient coastal landscape.