This source is the grand ice-cascade which descends in a succession of precipices from the plateau of the Col du Geant into the valley which the Glacier du Geant fills. We see from our present point of view that the bands _are confined to the portion of the glacier which has descended the cascade_. Fig. 37 represents the bands as seen from the Cleft-station above Trelaporte. [Illustration: DIRT-BANDS OF THE MER DE GLACE, AS SEEN FROM THE CLEFT STATION, TRELAPORTE. Fig. 37. _To face p. 369._] We are now however at such a height above the glacier and at such a distance from the base of the cascade, that we can form but an imperfect notion of the true contour of the surface. Let us therefore descend, and walk up the Glacier du Geant towards the cascade. At first our road is level, but we gradually find that at certain intervals we have to ascend slopes which follow each other in succession, each being separated from its neighbour by a space of comparatively level ice. The slopes increase in steepness as we ascend; they are steepest, moreover, on the right-hand side of the glacier, where it is bounded by that from the Periades, and at length we are unable to climb them without the aid of an axe. Soon afterwards the dislocation of the glacier becomes considerable; we are lost in the clefts and depressions of the ice, and are unable to obtain a view sufficiently commanding to subdue these local appearances and convey to us the general aspect. We have at all events satisfied ourselves as to the existence, on the upper portion of the glacier, of a succession of undulations which sweep transversely across it. The term "wrinkles," applied to them by Prof. Forbes, is highly suggestive of the appearance which they present. [Sidenote: SNOW-BANDS ON THE GLACIER DU GEANT.] From the Cleft-station bands of snow may also be seen partially crossing the glacier in correspondence with the undulations upon its surface. If the quantity deposited the winter previous be large, and the heat of summer not too great, these bands extend quite across the glacier. They were first observed by Professor Forbes in 1843. In his Fifth Letter is given an illustrative diagram, which, though erroneous as regards the position of the veined structure, is quite correct in limiting the snow-bands to the Glacier du Geant proper.