International audienceFossil-bearing shale specimens that include sulfides are chemically reactive and sometimes also mechanically fragile. This decay is provoked by iron sulfate efflorescence resulting from the oxidation of sulfide compounds. The processes underlying these degradations are poorly known, thus impeding the elaboration of curative or preventive treatments. The present contribution aims to identify the origin of museum specimen alterations. It focuses on the Flouest collection housed at the Muséum national d’Histoire naturelle (MNHN, Paris, France) and originating from the Autun Basin (Saône-et-Loire, France; Permian). To evaluate the alteration of MNHN specimens, it appeared necessary to compare their composition with that of unaltered shale so as to identify chemical changes occurring during ageing. Therefore new material was collected in the Autun Basin, among others on the locality of Muse that corresponds to the same lithostratigraphic unit than that of the MNHN specimens. This work is divided in three parts. The two first, presented elsewhere, deal with the composition of the shale matrices and led to the conclusion that these matrices could not account for the large iron(II) sulfate efflorescence provoking damage on the MNHN specimens. The last part of this work, presented here, focus on artificial ageing experiments performed on new shale material. Most of the alterations observed on artificially aged samples correspond to dispersed crystals of calcium sulfate (gypsum). Similar crystals may be found on MNHN specimens, but they are relatively few and sporadically distributed. They are thus considered as damage of secondary importance with respect to iron sulfate efflorescence. These latter could be reproduced on three samples only (upon the 142 aged samples). They were also found on a coprolite and on a wood remain that had got severely damaged in ambient conditions within a few months after their excavation. On all these samples iron sulfate had grown on brownish layers consisting in crystals of framboidal pyrite (1 to 3 μm) and eventually sulfur (20 to 50 μm). These brownish layers are associated to thin maceral layers probably because of bacterial activity: during fossil diagenesis, bacteria need organic matter for their metabolism to produce hydrogen sulfide, a precursor of sedimentary pyrite formation. Most of the damaged specimens of the Flouest collection show as well a thin maceral layer nearby iron(II) sulfate efflorescence. On one of them, this layer is particularly thick. It corresponds to vitrinite and shows in some areas a brownish aspect. The topology of this surface (observed with scanning electron microscopy) shows numerous small holes (< 2 μm) and large holes (10–20 μm). This morphology is compatible with a former occurrence of isolated grains and aggregates of framboidal pyrite. These observations suggest that iron sulfate efflorescence was provoked by the oxidation of framboidal pyrite forming deposits on maceral layers. Because of their brownish aspect, they can easily ...