For a long time, landforms were studied according to the criteria exposed in the theory of the geographical cycle, at least for geomorphologists claiming a Davisian approach. In this context, particular importance was attached, concerning landforms, to the remains of " peneplains ". After 1950, Davisian conceptions were, either abruptly given up (Strahler), or severely criticized (Tricart, Hack, Chorley), or deeply modified (King, Baulig, Klein), in particular to make them compatible with situations where it is not possible to identify in the topography the remains of several geographic cycles. So the question becomes now to understand how far it is relevant and how it is possible for a geomorphologist to travel back in the past, a delicate and speculative succession of operations: as noted by Schumm (1979, 485), " The extrapolation of measured average rates of erosion and deposition to longer periods of time is misleading, in the sense that they do not reveal the natural complexity of landform development or the variability of existing landforms ".