In 1963, N. Tabaste described disconnected fossil material from the Saharan Cenomanian site of Gara Tabroumit, and especially three sets of isolated scales that she all referred to the genus Lepidotes. Although the rhomboid shape and overall morphology of one set of scales allow regarding them as ganoid scales, the external aspect of the ganoine is not as regularly smooth and shiny as it is in most lepidosteoid scales commonly referred to ginglymodian (“semionotiform”) fishes. A photonic microscopic study of scale ground sections definitively excludes the allocation of these scales to the genus Lepidotes, or to closely related genera, because of an interrupted ganoine layer, and, more generally to ginglymodians: i) the bony basal plate is deprived of Wil- liamson's canaliculi, which characterized holostean scales; ii) the basal plate is well vascularised and overlaid by a thick layer of dentine; iii) there is an elasmodine layer between dentine and the basal plate. These various histomorphologic characters allow us to consider these ridged scales as belonging to a polypteriform fish, probably Bawitius. The histological specificities of the two other scale series allow their allocation to holostean taxa, respectively to “Lepidotes”pankowskii and to Obaichthys africanus. A new generic name, Adrianaichthys, is coined for “Lepidotes” pankowskii.