Résumé
The geological context of the discoveries made by Boucher de Perthes at Moulin Quignon in 1863-1864 was examined through the testimonies left by himself, by his detractors, then by geologists and archaeologists who later worked on the alluvial deposits of the Somme at Abbeville. This critical approach has been completed by specific studies, through a restricted excavation performed at Carpentier Quarrya locality known since the end of the 19th century as a sequence analogous to Moulin Quignon, and by a series of new analyses (microvertebrate, malacofauna, dating). The main results of these studies confirm the fluvial nature of the 'archaeological' layers at Moulin Quignon that correspond to a high terrace of the Somme River, about 40 m above the current valley incision at Abbeville. Dating results provide an age of 584 +/- 48 ka for the fluvial levels at Carpentier Quarry, in good agreement with the available biostratigraphic data, inferred from malacofaunas collected at both Moulin Quignon (Boucher de Perthes collection) and Carpentier Quarry, and from mammals fauna recovered since the end of the 19th century from the white marl at Carpentier Quarry which allow the allocation of these alluvial deposits to an interglacial stage of the second half of the Cromerian, the marine isotopic stage (MIS) 15. From these data and taking into account the identification of the human remains of Moulin Quignon as an anatomically modern man, the contemporaneity of these remains with the alluvial deposits of a high terrace of the Somme River is virtually impossible and supports the hypothesis of a hoax. Besides, it is possible that at least some of the lithic material collected at the time of the discovery by Boucher de Perthes could derive from fluvial levels dated of approximately 600,000 years. They would be then the oldest evidence of a human presence in the Somme Valley. (C) 2016 Elsevier Masson SAS. All rights reserved.