At the place where the three welded tributaries of the Mer de Glace squeeze themselves through the strait of Trelaporte, the bands undergo a considerable modification in shape. Near their origin they sweep across the Glacier du Geant in gentle curves, with their convexities directed downwards; but at Trelaporte these curves, the chords of which a short time previous measured a thousand yards in length, have to squeeze themselves through a space of four hundred and ninety-five yards wide; and as might be expected, they are here suddenly sharpened. The apex of each being thrust forward, they take the form of sharp hyperbolas, and preserve this character throughout the entire length of the Mer de Glace. I would now conduct the reader to a point from which a good general view of the ice cascade of the Geant is attainable. From the old moraine near the lake of the Tacul we observe the ice, as it descends the fall, to be broken into a succession of precipices. It would appear as if the glacier had its back periodically broken at the summit of the fall, and formed a series of vast chasms separated from each other by cliffy ridges of corresponding size. These, as they approach the bottom of the fall, become more and more toned down by the action of sun and air, and at some distance below the base of the cascade they are subdued so as to form the transverse undulations already described. These undulations are more and more reduced as the glacier descends; and long before the Tacul is attained, every sensible trace of them has disappeared. The terraces of the ice-fall are referred to by Professor Forbes in his Thirteenth Letter, where he thus describes them:--"The ice-falls succeed one another at regulated intervals, which appear to correspond to the renewal of each summer's activity in those realms of almost perpetual frost, when a swifter motion occasions a more rapid and wholesale projection of the mass over the steep, thus forming curvilinear terraces like vast stairs, which appear afterwards by consolidation to form the remarkable protuberant wrinkles on the surface of the Glacier du Geant." [Sidenote: FORBES'S EXPLANATION.] With regard to the cause of the distribution of the dirt in bands, Professor Forbes writes thus in his Third Letter:--"I at length assured myself that it was entirely owing to the structure of the ice, which retains the dirt diffused by avalanches and the weather on those parts which are most porous, whilst the compacter portion is washed clean by the rain, so that those bands are nothing more than visible traces of the direction of the internal icy structure." Professor Forbes's theory, at that time, was that the glacier is composed throughout of a series of alternate segments of hard and porous ice, in the latter of which the dirt found a lodgment. I do not know whether he now retains his first opinion; but in his Fifteenth Letter he speaks of accounting for "the less compact structure of the ice beneath the dirt-band."