This is quite conceivable, but how is it that the flints left behind on the plateau remain on the original ancient surface? How is it conceivable that if (on the old theory) these plateaus were in Palæolithic days clothed with forest, the Palæolithic flints could even in a single instance remain undisturbed from Palæolithic times to the present day, when the forest in which they were made and the forest soil on which they reposed have entirely disappeared? If there were woods and forests On the heights, it would seem impossible that we should find, as we do, Palæolithic implements lying _in situ_ on the desert surface, around the actual manufactories where they were made. Yet if the constant rainfall and the vegetation of the Libyan desert area in Palæolithic days is all a myth (as it most probably is), how came the embedded palaeoliths, found by Gen. Pitt-Rivers, in the bed of diluvial detritus which is apparently _débris_ from the plateau brought down by the Palæolithic _wadi_ streams? Water erosion has certainly formed the Theban _wadis_. But this water erosion was probably not that which would be the result of perennial streams flowing down from wooded heights, but of torrents like those of to-day, which fill the _wadis_ once in three years or so after heavy rain, but repeated at much closer intervals. We may in fact suppose just so much difference in meteorological conditions as would make it possible for sudden rain-storms to occur over the desert at far more frequent intervals than at present. That would account for the detritus bed at the mouth of the _wadi_, and its embedded flints, and at the same time maintain the general probability of the idea that the desert plateaus were desert in Palæolithic days as now, and that early man only knapped his flints up there because he found the flint there. He himself lived on the slopes and nearer the marsh.