Un gisement de surface exceptionnellement riche en bifaces a été découvert sur la colline de Moragne, en Charente-Maritime. Dans le but de définir leur typologie, leur pétrographie et leur attribution culturelle, 120 bifaces ont été échantillonnés. Ils sont issus des collections E. Dépré, D. Néraudeau et F. Champagne. L’analyse typologique a mis en évidence de nombreux bifaces amygdaloïdes, ovalaires et discoïdes, mais aussi des bifaces triangulaires et subtriangulaires, des ficrons, des bifaces lancéolés et micoquiens, des cordiformes et subcordiformes, des proto-limandes, un naviforme, des hachereaux, des nucléiformes, des abbevilliens, des chopping-tools, des bifaces partiels et divers. Le matériel lithique est principalement constitué de silex autochtones de couleur rouge à marron, représentant près de 90 % de la série. Ce silex provient des altérites tertiaires superficielles des dépôts crétacés sous-jacents. Certains des bifaces possèdent une ou plusieurs veines gréseuses et peuvent même parfois apparaître entièrement gréseux lorsqu’ils sont façonnés dans une veine épaisse. D’autres pièces ont la particularité d’être en quartz et représentent moins d’un dixième de l’échantillon total. Les origines géologique et géographique de ce matériau quartzeux ne sont pas clairement identifiées. L’absence d’indices stratigraphiques liée au mélange des couches superficielles n’a pas permis une datation précise de ces outils paléolithiques. Cependant, l’importante diversité typologique et l’étude de leur aspect ont révélé la présence d’industries paléolithiques très anciennes, qui ne sont que peu, voire pas encore, documentées en Charente-Maritime. D’autres semblent rattachés à des industries plus récentes comme le Moustérien de Tradition Acheuléenne (MTA) ou l’Acheuléen supérieur, témoins d’une transition avec le Paléolithique moyen. Homo heidelbergensis pourrait donc avoir occupé le site de Moragne au Paléolithique inférieur, ainsi qu’Homo neanderthalensis au début du Paléolithique moyen. A surface deposit with an exceptionally large number of bifaces was discovered on the Moragne hill, in Charente-Maritime, western France (fig. 1). The recent Prehistoric archaeology of the area has been widely studied, with the oldest human occupations dating to the Neolithic (Bouin, 1993, 1994; Kerdivel, 2009) and the Mesolithic (Favre, 2004; Michel, 2007, 2009). The Paleolithic has not been yet been the focus of scientific publication until now. 120 bifaces were sampled in order to define their typology, petrography and cultural attribution. 65 bifaces come from the E. Dépré collection, 46 from the D. Néraudeau collection and 9 from the F. Champagne collection. 95% of the bifaces were collected in a field on a hill and 5% at the bottom of the slope in a neighboring field (fig. 2). There are 97 classical bifaces in the sample, representing 81% of the total (fig. 3). The remaining 23 bifaces are considered non-classical. The typological analysis revealed lanceolate (n=7) and micoquian bifaces (fig. 4a ; n=3), ficrons (fig. 4b ; n=2), triangular bifaces (fig. 4c ; n=5 including 4 flat triangular bifaces) and subtriangular bifaces (fig. 5a-b ; n=10 including 2 flat subtriangular bifaces), cordiform (fig. 5c ; n=1) and subcordiform bifaces (n=3 including one elongated subcordiform biface), numerous amygdaloidal bifaces (fig. 6a ; n=35 including 20 short amygdaloidal bifaces), ovoid (fig. 6b ; n=14) and discoid bifaces (fig. 6c ; n= 14), proto-elliptical bifaces (n=3), but also Abbevillian bifaces (fig. 7a ; n=4), a naviform (fig. 7b), nucleiforms (fig. 7c ; n=2), cleavers (fig. 8 ; n=3), chopping-tools (n=4), and partial (n=6) and diverse bifaces (n=3). The lithic material consists mainly of native red to brown flints, representing nearly 90% of the series. The bifacial flint tools are diversified and occupy all the typological categories mentioned above, with the exception of the chopping-tools. This flint comes from the superficial alteration and silicification, during the Paleogene (Eocene-Oligocene), of the underlying Cretaceous calcareous and sandy deposits (Bourgueil et al., 1972 ; Moreau et al., 2019). Some of these flint specimens have one or more sandstone veins, and may even be entirely of sandstone when the biface is shaped in a thick vein. Other pieces have the particularity to be made of milky white quartz which seems of plutonic quartz-vein origin and represent less than one-tenth of the total sample. Quartz bifacial tools are represented by chopping-tools, amygdaloid bifaces and Abbevillian bifaces. The provenance of this material is not clearly identified, as there are no quartz veins intersecting the Cretaceous sediments in this area. Another source of quartz could be the quartz pebbles of the ancient alluvial terraces, some of which are very large, but these quartz pebbles are generally more translucent, neither white nor milky, and do not often correspond to the type used for the bifaces. Furthermore, the quartz-veins and the white quartz pebbles are not reported on the geological map of the area (Bourgueil et al., 1972). The lack of stratigraphic evidence due to the mixing of surface layers has presented difficulties for dating of these Paleolithic tools. Furthermore, the evolution of typological facies and production systems can lead to imprecise boundaries between various industries even when assemblages are stratigraphically in place (Vieillevigne et al., 2008). However, the important typological diversity and the study of their aspect revealed the presence of very old Paleolithic industries, which are little or not yet documented in Charente-Maritime. Bifaces are characteristic of the Early and Middle Paleolithic (Moncel, 1995; Bordes, 2000), periods dating respectively from 1.5 Ma to 250,000 years BP and from 250,000 to 35,000 years BP (Gagnepain, 2015; Tuffreau, 2004; Arzarello, 2015). Nonetheless, the presence of real bifaces in abundance is only found until the beginning of the Middle Paleolithic. It is yet possible to find bifaces sporadically in the rest of the Middle Paleolithic and in the Upper Paleolithic, but in small proportions and often with atypical shapes (Bordes, 2000). Since the Moragne site has a high concentration of bifaces, it is highly probable that the lithic industries present here are characteristic of the Lower Paleolithic and the early Middle Paleolithic. Indeed, there are bifaces characteristic of the Lower Paleolithic, particularly of the Early Acheulean, such as Abbevillian bifaces and proto-elliptical bifaces (Bordes, 2000). Most of the shaping of these tools is done with a hard hammer and the debitage methods are not very sophisticated, resulting in bifaces that can be very thick, with sinuous edges, cortical reserve, and weakly retouched. In addition, many notched flakes are associated with these old bifaces, which is typical of the Early Acheulean (Notter, 2015), although choppers are rare. The Middle Acheulean is also recognized at Moragne thanks to the presence of cleavers. The other bifaces of this industry are distinguished from the Early Acheulean by a more finely shaped, sharper edges and retouching. The presence of Levallois elements in small quantities associated with these bifaces is another marker of the Middle Acheulean (Bordes, 1970; Turq, 2000; Notter, 2015). Others seem to be associated with more recent industries, such as the Mousterian of Acheulean Tradition type A, indicated by flat triangular bifaces (Bordes, 2000). Smaller bifaces, strongly retouched and with symmetrical shapes, indicate a Late Acheulean or Mousterian Traditional Acheulean, witnessing a transition to the Middle Paleolithic (Jaubert, 1994). However, there are very few typical Levallois elements found, such as Levallois points, well geometric Levallois flakes, beautiful Levallois cores, and very sophisticated racloirs and scrapers. This under-representation of beautiful, characteristic Levallois tools could be explained by the preferential sampling of collectors, but this hypothesis is refuted by the abundance of bifaces, which are nevertheless highly sought after. The Moragne site thus attests to a proven human presence in the Early and Middle Paleolithic, extending the limits of the oldest known Mesolithic occupations of the site. Moreover, the high density of bifaces suggests that the site was occupied for a long time (Meignen et al., 2007). Homo heidelbergensis could therefore have occupied the Moragne site in the Lower Paleolithic, as well as Homo neanderthalensis in the early Middle Paleolithic, thereby making the Moragne site one of the witness prehistoric sites of the “neandertalization” process.