In the North of Charente department, on the watershed of the Bonnieure, the site of Les Marais in Puyréaux, is a cultual complex with ditched enclosures which spread on a little more than two hectares. The aerial and geophysic surveys put evidence of 17 circular monuments and 5 quadrangular, squared and rectangular ones, as well as numerous other tracks more or less identifiable, that are excavated or built structures. By accumulating datas from ancient discoveries, different surveys and the two programmed excavations – excavations yet very partial, of less than a tenth of the identified remains – the site appears frequented from early Bronze Age until gallo-roman period, without or a few chronological gaps, like sites as Antran, in Vienne or Perigny, in Charente-Maritime, occupied from the Chalcolithic to the gallo-roman period. In Les Marais, ditched enclosures seem organized compared to others, without overlap, indicating a sustainability, at least visual, between the old structures and the new ones. If the question of funeral vocation is asked for a lot of enclosures sites, here, it find a clear answer with the two graves excavated in the center of the double ditched enclosure. A first sepulchral quadrangular pit with rounded corners was dug between 1900 and 1740 BC to take in the corpse of an adult individual of undetermined sex. He was lying on his left side with flexed arms and legs, in a wooden container, covered and surrounded by limestone stones. Later, between 1890 and 1680 BC, a second grave was dug partially above and shifted towards the north-east related to the previous one. The officiants from then probably borrowed some stones out of the first grave for the layout of the second one. The corpse of a probable woman was lying down, head towards the west-north-west. Later again, between 1730-1720 or 1690 and 1510 BC, the grave was opened to bury a young adult, probably a male one, lying down in the opposite direction compared to the previous body, which one was yet reduced at one end of the pit. No artefact, excepted a bone pendant found under the skull of the woman, was preserved. The principal ditch of this monument, contemporaneous of the establishment of the first inhumation grave, was dug during early Bronze Age. His filling took place on the long term, until Late Bronze Age. Pottery vessels found in the different layers of the ditch filling prove it. Three filling phases clearly appear on the ditch stratigraphy. The first filling layer, without any artefact, is made of sediment from the erosion of the sandy and gravely walls. It is inside the second filling layer, a natural sedimentary input, where there are the first ceramic deposits, intentional or not but never of a complete vessel. Several potsherds of presentation potteries were found but it is the culinary and storage ceramic which is the most represented. Let us mention an outstanding artefact: a palet recut in a thick potsherd of a big vase that seems to be the oldest mention for this kind of object. The third and last filling phase of the external ditch provide us a ceramic material that is not chronologically homogeneous. Three stages of the Bronze Age were represented: Middle Bronze Age, by a fragment of a rounded bottom vase, Late Bronze Age Ib-IIa/BzD1b-Ha A1, by a fragment of a plate and finally the transition Late Bronze Age IIIa-IIIb/Ha B1-B2-3 by another plate fragment. Regarding the circular enclosures, some architectural transformations could have been enough to substantiate a vocation change of some of the monuments, going from funeral to cultual with the establishment of a stone crown at the top of the principal ditch. In view of the variety of the structures identified on the Marais site, various practices, perhaps not always funeral, presumably existed overtime on the site. We especially notice an original and presumably never yet viewed large subrectangular ditched enclosure (33m on 20 m) that invite us to make a connection with the big structures of Langgräbe type used during Late Bronze Age, although his function is unknown. On one of its small side, it presents two curved advances and one central interruption, device which we call « entrance with antenna. » In the absence of excavation this structure remains undatable.