Abstract: Obsidian was used as a raw material for the lithic industry during prehistoric times. This material has specific properties allowing to identify its geological provenance. The rarity of obsidian sources and the interest of early men for obsidian make it one remarquable tracer of ancient exchanges and human contacts over large distances. We show the relative advantages, for chemical characterisation, of ion beam analyses by PIXE -Particle Induced X-ray Emission-, which allows non-destructive analysis and of Inductively Coupled Plasma, either as ICP-AES -Atomic Emission Spectrometry- and ICP-MS -Mass Spectrometry-, which need small samplings but allow the measurement of a large spectrum of elements. In an Andine area of South America (Colombia, Ecuador) where the knowledge on obsidian circulation was very limited, we brought new data on sources and started a PIXE and ICP study of several collections of prehispanic artefacts. ICP measurements allowed a more detailed characterisation of obsidian sources in this area. Our results constitute the most complete data base to date for this region. In the Near-East (Syria, Turkey), we brought informations on five néolithic sites of the middle Euphrate valley (10 400-4 500 years BC). Our ICP results confirm the pre-eminence of Cappadocia as an obsidian source during this period and the emergency of Eastern Anatolia contribution. They also show the difficulty of discriminating between some of these sources, when their chemical composition is the only parameter considered. Due to the difficulty of discrimination between sources, we proposed the use of a double characterisation -chemical composition/fission track dating- of all obsidian artefacts in Colombia and Ecuador, and for specific cases in the Near-East. ; L'obsidienne a été utilisée comme matière première de l'industrie lithique au cours de périodes préhistoriques. Ce matériau possède des propriétés spécifiques qui caractérisent son lieu d'origine. La rareté de ses sources et l'intérêt dont l'obsidienne a fait l'objet ...
No Comments.