From the 6th-26th September 2009 investigations recommenced on the Middle Paleolithic deposit in the Riparo l’Oscurusciuto in the ravine of Ginosa. During the first few days the excavation of US9, squares E-F/11-12, was completed, with the definitive exploration of hearth US 43. On the basis of results from previous campaigns, in an excavated area of circa 10 m2, US 9 produced a complex of small circular and ellipsoid hearths concentrated in two distinct areas, each about half a square metre in size. The succession of a number of structures, some overlying and displaced, was seen. Below US 9 another interesting group of hearths came to light, situated at the top of the layer and within the underlying US 11. As in the preceding phase, the hearths were in small, mainly circular, hollows with a maximum depth of 5 cm. A particular concentration of these hearths was uncovered in an area of 2 m2 (q F-G/11), during excavation 19 were identified, partially cut by the successive excavations to a maximum depth of 20 cm. The average diameter of the hearths varied between 40 and 50 cm. The fill of these features was constantly of a blackish-brown colour; in some cases plates of concretionary light coloured sediment was present at the top, probably mixed with ash. Burnt bones were often present in the fill, partially concentrated at the base of the hollows. In the circa 10 m2 also investigated for this US other isolated hearths in hollows were also found. Of particular interest the presence of a pair of small hearths, very slightly overlapping which attests the fact that they were built and used in two successive phases. The diameter of these two structures (25 cm), clearly smaller than the others, suggests they had a different use. A preliminary examination of the faunal remains showed that mainly aurochs were hunted in an environment characterised by extensive areas of meadowland with trees. The lithic industry, still abundant, showed no marked differences to what had been seen in the overlying layers. The Levallois method was recurrent, mainly unipolar, producing flat, elongated or triangular artefacts. The presence was confirmed, although minor, of a different debitage method, unipolar direct percussion characteristic of the exploitation of cores by volume (and not by surfaces), anticipating the production typical of the Upper Paleolithic and was aimed at the production of flakes. Among the retouched implements together with the numerous scrapers, were a number of very regular points and two graters. As the excavation intercepted numerous burnt areas many of the lithics had been damaged by fire. It was possible to recognise the provenance from a single core for some. A fragment of burnt bone from the bottom of US 1 gave a C14 date of 38.500±900 BP.

The investigation of the middle Palaeolithic shelter of Oscurusciuto at Ginosa (TA) continued between the 5th and 23rd September 2011. During the first part of the campaign, the excavation of US 11 was completed. In the previous campaign numerous hearths in shallow circular pits (diam. 20-50 cm, depth 3-6 cm) had been uncovered in this layer. Within the excavation area, these features were concentrated in a band, orientated north-east/south-west leaving a corner of the shelter, circa two metres wide, completely free. This corner was probably used by the Neanderthals as a sleeping or living area. This season a pair of small hearths in partially overlapping pits (diam. circa 20 cm), and a larger hearth (diam. circa 40 cm) were uncovered in an area below the southern sector of the excavation, itself characterised by the presence of a concentration of hearths. The hearths in US 11 were often overlapping and it was not possible to work out which of the individual structures in the sediment (average thickness 8cm) had been in use at any one time, due to the lack of evidence for the intermediate occupation levels. The exception was the base layer, constituted by 16 hearths in shallow pits of circa 20 cm in diameter (including the pair during this campaign), which characterised the lower part of the stratum. All of these hearths were smaller than those in the upper levels of US 11 and 9, and probably belonged to a single occupation phase. Micro-stratigraphic analyses are being undertaken which may indicate whether differences in dimensions is related to differences in the use of the pits. In the base part of US 11 the presence of tephra in the sediment gradually increased and the number of lithic and bone finds decreased. Phase two of the excavation involved the removal of the underlying US 13, partially exposed in the previous year, with a matrix rich in tephra and small calcarenite fragments. Nine hearths in shallow pits (diam. 20-30 cm) were uncovered, together with a larger patch of burning whose real shape and extension are unknown, as a part of it had eroded away. The hearths, mainly single, occupied the same area as the overlying structures, that is none were present in the north-western corner of the shelter and a corridor along the back wall. Only three hearths were overlapping (83a, 83b, 84). Finds within this layer were sporadic and often in small concentrations. The bones, heavily fragmented, and the lithic industry continued to be characterised by the dominance of the Levallois technique with flakes parallel to the plane of percussion. No particular changes in the methods of the production chain were noted during the excavation. This stratigraphic unit covered a compact layer of volcanic ash circa 60 cm thick (US 14) sealing an ancient ground surface that was rich in materials. The hearths found in US 13 were therefore pertinent to the first sporadic occupations of the shelter, successive to the substantial ash fall. The analyses of the tephra will probably provide data about the attribution, and therefore on the chronology, of the volcanic eruption. The surface of US 14 was irregular and characterised by frequent depressions/excavations perhaps the result of erosive action. These irregular holes, in part still to be excavated, were filled by US 13 and contained few lithic and bone finds. The upper part of the layer of volcanic ash contained occasional finds, which had penetrated into the sediment, and will be investigated during the next campaign.

In September 2012, a new excavation campaign investigated the middle Paleolithic deposit of the Riparo l'Oscurusciuto at Ginosa (TA). The aim was to excavate the thick stratum of tephra (about 60 cm) identified during the first investigations in 1998. In fact, in this shelter the hill erosion has damaged the entire deposit, so that in the first trench part of the stratigraphy was identified simply by removing the top soil. The analysis of the tephra (undertaken by Roberto Sulpizio at Bari University), attributed the volcanic ash to the Tufo Verde di Monte Epomeo (Ischia), dated to around 55.000 years BP (pers. com.). This established a chronology for the upper part of the stratigraphy, which develops between this date and the 43.000 years BP cal (14C on collagen, AMS – Beta 181165: 38,500 ± 900 BP) obtained at the base of the latest level (US 1). This result is of great importance for the possible paleo-environmental correlations with other Apulian deposits, some of which present tephra layers that have yet to be identified. In this campaign, the tephra (US 14) in quadrants D/G – 13/16 was completely excavated (about 11 m2). As noted in last year’s campaign, the surface of the tephra stratum is characterised by small irregular depressions probably caused by erosion, filled by US 13, but containing almost no material. The upper part of the volcanic ashes attested the first occupation, which occurred during the final phase of ash deposition. In fact, lithic finds (few in number but significant) and occasional bone fragments were found down to a depth of 10-15 cm. The material was concentrated in small areas and preliminary studies have identified a number of joining lithic elements. All of the upper part of the excavated tephra was sieved with a 2 mm mesh. The few identifiable bones were attributable to Bos primigenius. Within the tephra there were numerous small areas with inclusions of compact light brown coarse sand. Also present were sporadic small jasper cobbles, probably from the summit of the shelter. The paleo-surface (US 15) emerged at the base, sealed by the volcanic ash. This was a horizontal surface with an abundance of lithic material, bone and stones, partially structured to form a semicircular perimeter about 2 m in diameter, facing the north wall of the shelter. The structured stones, varying between 15 and 30 cm in size, were arranged in small groups of 3-4 elements. There were a few bone and lithic finds on the surface enclosed by this structure, including a large fragment of aurochs’ mandible. On the contrary, an abundance of finds was present among the stones, along the edge of the structure itself. Outside the structure there were numerous finds including large fragments from the diaphese of long bones, aurochs mandible and jaw fragments and lithic industry. Other stones were present on the exterior, seemingly not in any alignment, partially covered by an articulated thick layer of concretionary vegetal remains. The next campaign aims to excavate another sector of the tephra and expose a new area of paleo-surface of about 8 m2.

Nella prima fase dei lavori della quindicesima campagna, lo scavo è proseguito nello strato di tefra (US 14 – Epomeo verde di Ischia - potenza 60 cm – ca 55mila anni BP ) in due aree di 4 mq l'una. Nei quadrati D-E-F-G/11 è stata asportata la parte superiore di US 14 (taglio 1) per uno spessore di 10-12 cm dove è presente materiale sporadico, soprattutto litico. Sono stati riconosciuti alcuni rimontaggi. Il taglio 2 di US 14 è stato indagato in D-E-F-G/11-12 per uno spessore di circa 10 cm. Le ceneri vulcaniche hanno restituito rarissimi reperti litici. Nel quadrante F12 III è stato isolato un rizolite con sviluppo verticale per tutto lo spessore del tefra. Questo reperto, del diametro medio di 2 cm, è risultato in continuità con l'esteso insieme di rizoliti a contatto e in prossimità con la sottostante paleosuperficie. Nella successiva fase di lavoro è stata messa in luce, evidenziandone i dettagli, la paleosuperficie sigillata dal tefra. Nella nuova area sono emersi altri due interessanti insiemi di pietre, uno in E12 III/E13 IV, l'altro in F11 II/F12 I. Questi due raggruppamenti formano, con una pietra di maggiori dimensioni parzialmente immersa nella sezione di scavo, un allineamento E/W situato tra la struttura a semicerchio evidenziata lo scorso anno e la parete W del riparo. Questo allineamento separa un'area più depressa, d’angolo fra le due pareti N/W del riparo e con scarso materiale antropico, da una zona ricca di frammenti di diafisi di ossa lunghe di grandi ungulati, litica e alcune pietre apparentemente non strutturate. L'abbondanza di frammenti di diafisi riguarda l'intera porzione Sud della paleosuperficie (US 15). Evidenti anche frammenti di mandibole e di mascellari di Uro, la specie nettamente dominante in questa fase. L'area della paleosuperficie finora indagata sembra quindi essere interessata da due strutture adiacenti. Una, individuata lo scorso anno, è costituita da un semicerchio di pietre addossato alla parete Nord. In corrispondenza delle pietre la paleosuperficie presenta un evidente leggero rilievo. La seconda struttura, sopra descritta, ha in comune con la prima la scarsità di materiale nella parte interna e un rilievo nella zona perimetrale. All'esterno delle due strutture sono abbondanti i resti ossei, in gran parte costituiti da frammenti di ossa lunghe, mandibole e mascellari. L’industria litica finora raccolta sembra per la maggior parte riconducibile al sistema di produzione Levallois unipolare, analogamente a quanto attestato nella parte superiore della serie, successiva al tefra. Nella terza fase di lavoro è iniziato il prelievo delle ossa e della litica presenti sulla superficie. Le ricerche sono state condotte su concessione ministeriale in accordo con la Soprintendenza per i Beni Archeologici della Puglia. Alla campagna di scavo hanno partecipato oltre agli scriventi: Francesco Boschin, Francesco Colopi, Jacopo Crezzini, Elena D'Itria, Paolo Gambassini, Olivia Lisi, Giulia Marciani, Anna Pizzarelli, Giulio Poggi, Marco Serradimigni, Vincenzo Spagnolo, Gianna Tinacci. Un sentito ringraziamento all'amico Piero Di Canio di Ginosa e alla sua famiglia per l'indispensabile appoggio logistico e all'Amministrazione Comunale per il contributo alle spese sostenute.

The excavation of the middle Palaeolithic deposit of the Riparo l'Oscurusciuto in the ravine at Ginosa (Taranto) continued in October 2014. During this campaign, the excavation of the settlement occupation surface (US 15) began. This surface was sealed by a layer of tephra (US 14) attributed to the green tufa from the eruption of Monte Epomeo on Ischia (55ky BP). The area was dug in 25 cm2 squares to a depth of 2/3 cm over an area of c. 12m2. The aurochs continued to be predominant among the bone remains: a first analysis estimates a minimum of five individuals of different ages. The analysis of the lithic industry requires further study in the laboratory in order to identify whether there are variations in the management of the raw materials and the running of the production line within a predominantly unipolar Levallois production common in all levels excavated to date. In this phase of the work traces of a small hearth emerged, probably created in a small pit, with ashes at the summit. The structure was situated inside structure “A”, identified during the 2012 campaign, and formed by a semicircle of stones with a radius of c. 2 m, abutting the north wall of the shelter. More interesting evidence relating to structure “B”, identified in 2013, came to light in this first examination of the surface. This structure is constituted by a row of three groups of stones that seem to enclose the space in the north-west corner of the shelter, characterised by the scarce presence of occupation evidence. Within one of these groups of stones, two of which in a vertical position, there was a small pit 25 cm wide and 10 cm deep (the excavation of what was probably a second fill may redefine the depth). In a hypothetical reconstruction of the structure, the pit is interpreted as housing for one of the timber elements abutting the shelter wall. All of the stones visible on the surface, some of which probably part of structures and all kept _in_ _situ_, were measured and their lithology analysed. The stones were mainly calcarenite blocks with a few smaller limestone elements. The latter were only present within the groups that delimit the structures. The analysis of the area will be completed following the excavation of the remaining 6 m2 during the next season. The present study of the spatial distribution of the materials on the occupation surface will provide information about the different uses made of the spaces.

This campaign saw the completion of the excavation of occupation surface US 15. The stones forming the structures were not removed as their bases were embedded in US 16. Following the removal of lithic and faunal material from the paleo-surface in 2013, visible after the removal of the tefra, in October 2014, excavation began of the first 2-3 mm of sediment of the occupation surface. The aim was to obtain a complete picture of the finds present in the shelter’s final phase of use, prior to the deposit of the thick layer of volcanic ash (US 14) relating to the Tufo Verde del Monte Epomeo di Ischia (55ky BP). Therefore, excavation of the remaining 6 m2 and recovery of materials continued, by 25 cm2 squares. In quadrants G11, E-F 12, E-F 13, the sediment was heavily concretionary. In these areas, it was difficult to extract the faunal remains. Most of the materials were found in a horizontal position. As seen in previous campaigns, the material inside the structures was sporadic. The expected hollow, like that documented last year between the stones in F 11-12, did not emerge in relation to the group of stones in E12 (structure B). The uroch was the predominant species among the identifiable, though heavily fragmented bones. The lithic industry, like that in the layers above, can be attributed to the Levallois production system and was mainly constituted by small flakes and debris, indicative of flint working on site. Future study will check whether the scarcity of medium/large elements and tools is due to their removal from the encampment at the time of its abandonment or not. At the beginning of the campaign, the north baulk was seen to have been disturbed by recent animal burrows. Given the instability of the upper part of this portion of sediment abutting a wall of the shelter, excavations began with the upper layer (US4) in quadrants C11 IV and C12 IV. The abundant material presented numerous vertical or sloped elements. The sediment was friable with the exception of a small concretionary patch against the wall in quadrant C11 IV. Small calcarenite fragments were frequent, which had probably broken away from the walls.

Nel settembre 2016 è proseguito lo scavo del testimone Nord, iniziato negli ultimi giorni della campagna dello scorso anno. Le indagini in questo lembo di stratigrafia addossato ad una delle pareti del riparo hanno tre obiettivi principali: a) recuperare i dati delle US 4 e 5 di questo testimone, in pericolo di crollo per la presenza di un complesso di tane; b) integrare i dati spaziali delle unità stratigrafiche con focolari: US 9, 11, 13; c) ampliare l'esplorazione della superficie di abitato US 15, mettendo in luce i rapporti tra le due strutture delimitate da pietre e la parete Nord del riparo stesso. Le ricerche in questa campagna hanno interessato l'unità stratigrafica 4 nei quadrati C11-12-13 per un'ampiezza complessiva di circa 2 mq. Questa piccola area è risultata particolarmente ricca di reperti litici e ossei, spesso rinvenuti in addensamenti. Lo scavo è stato effettuato in tre tagli artificiali di 8-10 cm di spessore. Il sedimento, compatto ma friabile, ha facilitato il recupero dei materiali, ad eccezione di una porzione nel quadrato C11, concrezionata contro parete. All'inizio dello scavo il sistema di tane (US 90) era visibile sia sulla parete del testimone (US 4 e 5), sia, molto parzialmente, sulla sommità di US 4 (questa unità costituiva il tetto del testimone stesso, conservato a partire dalla campagna del 2003). Nel corso dello scavo, nei tagli 2 e 3 di US 4 il complesso di tane è stato messo in luce per una maggiore estensione: una galleria principale si sviluppa da Est ad Ovest, con uno sbocco nell'angolo Nord-Ovest del riparo. In questo punto, le escavazioni da parte di roditori hanno probabilmente interessato anche unità più profonde (US 9, 11). La galleria principale, nel quadrato C11 e in parte in C12, si sviluppa addossata alla roccia. Rami secondari proseguono all'esterno sulla parete del testimone. Queste tane risultavano beanti con riempimento alla base sciolto ben distinguibile dal sedimento di US 4. Nei quadrati C12-13-14 è stata messa in luce la sommità di US 5, della quale è stata scavata una piccola parte in C12 III. Tra i resti faunistici di US 4, sono abbondanti i frammenti di diafisi di ossa lunghe: è stata riconosciuta la presenza di _Bos primigenius_ , _Dama dama_ e _Sus scrofa_. L'industria litica non presenta sostanziali novità: il sistema di produzione Levallois ricorrente è sempre dominante sui sistemi volumetrico o addizionale. Solo si osserva, rispetto ai livelli sottostanti, una maggiore frequenza della modalità unipolare convergente, accanto a quella a stacchi paralleli. Sono rappresentate tutte le fasi della catena operativa, dalla inizializzazione della messa in forma del blocco all'abbandono dei nuclei. Quanto ai prodotti ritoccati, sempre non particolarmente numerosi, i raschiatoi soprattutto lunghi prevalgono sulle punte e sui denticolati. Si segnala il rinvenimento, sporadico nelle campagne precedenti, di alcuni ciottoli che sembrano utilizzati come percussori/ritoccatoi. Alla campagna di scavo, effettuata dal 4 al 23 settembre, hanno partecipato oltre agli scriventi: Francesco Boschin, Serena Ciullo, Jacopo Crezzini, Lucia Dallafior, Loris Di Vozzo, Melania Farnese, Alessandra Macco, Giulia Marciani, Vincenzo Spagnolo. Un forte ringraziamento all'amico Piero Di Canio di Ginosa e alla sua famiglia per l'indispensabile appoggio logistico e all'Amministrazione Comunale per il contributo alle spese sostenute.

The excavation investigated the patches of US 4 (spit III) in quadrants C11-12, C14-15, the removal of which revealed the top of units 5 and 7. The sediment was not concretionary and the materials in quadrants C11 and C12 were easily recovered, while the concretionary terrain in quadrants C14-C15 made excavation more difficult. Once this first phase was completed, research took place in US 5 and 6, constituted by several alternating levels of ash and charcoal relating to a hearth c. 2.5 m in diameter, situated against the wall in the north-west corner of the shelter. The hearth was housed in a depression at the top of US 7 (US negative 35). In quadrants C10 and C11 the hearth was almost completely destroyed by a system of animal burrows (US 90), already exposed during previous campaigns. This complex also heavily cut the deposit in quadrant C12 (and to a limited extent in C13) with two overlying tunnels linked by several vertical tracts. US 5 was excavated in two spits 10 cm thick, while US 6, only identified in a limited area, was removed in a single spit. This revealed the negative US 35, forming the surface of the depression containing the hearth. In the area not disturbed by US 90, the edge of the depression was identified against the north wall of the shelter. Subsequently, US 7 was removed in quadrants C11-12-13-14-15 down to the top of US 8. US 7 contained fewer materials compared to the layers above and was compressed by the burrow system (US 90) in quadrants C11-12. As the excavation continued, this complex became larger and was revealed to be a single large tunnel extending west to east, which at the top of US 8, occupied quadrants C10-11-12 and in part C13. In order to leave the north bulk as stable as possible at the end of the campaign, the excavations halted at the top of US 9 in quadrant C12. In this area US 8 was poor in materials. Most of the faunal remains recovered during the course of the excavations seemed to belong to _Bos_ _primigenius_. Rare fragments of deer bone were also present. There was no new evidence about the lithic production (recurrent dominant Levallois). The working took place, at least in part, on site, as shown by numerous elements of the first reduction cores, exhausted cores or ones that were abandoned because damaged, by hammerstones and a number of chips. The artefacts transformed by retouching included points that were intact at the apex, and flake scrapers, some denticulated.

This season excavation continued of the north baulk in which quadrants C11-C15 extend, which were investigated during the previous three campaigns. There was no sign of any infraction in the area, while during the course of the year, rodent, reptile, and vegetal activity in the baulk sediments seemed to have intensified. Initially, the excavation concentrated on US 8 characterised by a scarce quantity of lithics and faunal material. The looser sediment in quadrants C12-C13 facilitated the recovery of the finds, while part of quadrant 14 and in quadrant 15 the excavation was harder due to the sediment concretion. The removal of US 8 revealed US 9, which was removed in two spits each 5 cm thick. In addition to containing a greater number of finds, US 9 was distinguished by the presence of hearths. In particular, contexts US 92 (fig. 2) and 94 were identified in quadrant C14. Both were hearths in depressions, constituted by levels of ash and charcoal. In the same quadrant US 93 was identified as a circumscribed concentration of charcoally material. Like some of the layers above it, in this area of the excavation US 9 also presented a system of animal burrows (US 90), constituted by very loose soil that was easily recognisable. The sediment disturbed by the rodent activity (and further modified by geckos, lizards and roots) was present in quadrants C11, C12 and, marginally, in C13. Most of the faunal remains in US 9 were from aurochs, although some elements from horse and deer were identified. The commonest fragments were those from the diaphyses of long bones (fractured to recover the marrow) and teeth. In some areas, especially in the vicinity of the shelter’s rock face, the finds were often in a vertical position or obliquely immersed in the sediment, attesting possible disturbance of the original deposition. Further spatial analyses will clarify whether the Neanderthal groups occupying the shelter could have caused a random accumulation of waste materials. The removal of US 9 in quadrants C12-15 exposed the top of US 11. US 9 was not excavated in quadrant C11, heavily disturbed by the burrow system US 90, in order to maintain the stability of the baulk. Lastly, in quadrant C14 two more hearths were excavated (US 10 and US 95), both at the top of US 11 and considered in phase with it. Both hearths were in small pits about 25-30 cm in diameter, and made up of a layer of charcoal remains covered by a crust of hardened ash. The top of another hearth emerged at the base of pit US 95, denominated US 96. Small blocks of sediment for micro-stratigraphic analyses were removed from the excavated hearths (US 10, 92, 93, 94 and 95). The natural radioactivity of the site was measured in correspondence with US 26, 20-21 and 11 in order to calibrate the OSL dating carried out on the sediments sampled the previous year. At the end of the excavation the state of the north baulk was the following: US 90 in C11 I-II and C12 II; US 9/1 in C11; top of US 11 in C12, C13, C14, C15; top of US 96 in C14 IV. There were no fundamental differences with regard to the modes of production/transformation of the lithic artefacts. The cores and products indicate the presence of the integral Levallois system, above all flakes parallel to the plane of percussion alongside supplementary systems, all aimed at obtaining generally elongated products. Neither was any change seen in the relationship between the retouched objects: scrapers usually on a laminar support, followed by points. Among the latter, there were morphologies similar to tools for shaping wood. As in previous campaigns several cobblestones were found that may represent hammerstones or retouching tools. Two cores were also found.

This season, the excavation continued of the North baulk in quadrants C11-C15 (Fig. 1). This baulk was investigated during the four previous campaigns and this season involved the last patches of US 9 in quadrants C11-C12 and US 11 across the entire excavation area. US 11 was 10-15 cm thick and rich in faunal and lithic materials. However, it also presented layers where the materials decreased, attesting the fact that it was a palimpsest made up of different occupation phases. As known from previous years, quadrants C14 and 15 presented a concretionary and strong sediment; on the contrary, the recovery of the finds became easier as the excavation moved towards the N-W corner of the shelter (quadrants C10-11). US 97 and 96, 98 and 99 (Fig. 2) were identified respectively in quadrants C13 and C14, all relating to hearths within small pits, which terminated the row of hearths along the north wall of the shelter found in US 11 during previous campaigns. Of interest was an area of charcoally material, perhaps the result of the cleaning of a hearth, in quadrant C14. The majority of the faunal remains found in US 11 are attributable to wild oxen, although elements belonging to horse and deer were also present. Diaphyses from long bones (fractured for marrow recovery) and teeth were predominant. The lithic industry was the same as that found in the rest of the context. The raw materials for flaking were local and seemed to be present in all phases of the production line. There were few re-touched elements. In the vicinity of the shelter’s rock wall, the finds were often in a vertical or heavily slanted position, attesting the possibility that they had been disturbed. Further analysis of the spatial distribution will evaluate the possibility of a random accumulation of discarded material in this area of the site that was created by the groups of Neanderthals who used the cave. The underlying US 13 was exposed in the areas where US 11 was completely removed (via the excavation of two artificial spits). This was characterised by a different matrix (a fraction of volcanic sand, associated with the presence of tefra in US 14, was added to the sediment originating from the decomposition of the local calcarenite). US 11 also presented the series of burrows and tunnels US 90 (Fig. 3), caused by the actions of roots, rodents and reptiles. The disturbed part of the deposit was extensive in quadrants C10-11 (almost covering their entirety). It decreased in quadrants C12-13 and finally disappeared in the eastern part of the deposit (C14-15). In quadrant C11, the base of the pocket of disturbed deposit was reached; it did not appear to have cut the deepest part of layer 11. Small blocks of sediment were taken from hearths US 96, 97, 98 and 99 for micro-stratigraphic analysis. Sediment samples were also taken from the entire sequence in the cave. The sediment analyses are part of a project financed by the National Geographic Society “The key role of cave and shelter clastic successions in defining the replacement of Neandertals by Modern Human”. In addition to the excavations, “sediment traps” were positioned along the cave walls in order to measure the natural deposition of calcarenite on the site during the course of the year. This will make it possible to evaluate the rate of sediment accumulation caused by the decomposition of the rock face and estimate the time it took for some of the levels in the archaeological deposit to form. At the end of the season, the stratigraphy in the North baulk was as follows: US 90 in C10 and locally in C11 and C12. The top of US 13 in C11, C12, and C13 I; US 11/2 in C13 II, C14 I-II and C15; US 11/3 in C13 III-IV and C14 III-IV (Fig. 4).