The bottom of the Lachoong valley at Momay is broad, tolerably level, grassy, and covered with isolated mounds and ridges that point down the valley, and are the remains of glacial deposits. It dips suddenly below this, and some gneiss rocks that rise in its centre are remarkably _moutonneed_ or rounded, and have boulders perched on their summits. Though manifestly rounded and grooved by ancient glaciers, I failed to find scratches on these weather-worn rocks.* [I have repeatedly, and equally in vain, sought for scratchings on many of the most conspicuously moutonneed gneiss rocks of Switzerland. The retention of such markings depends on other circumstances than the mere hardness of the rock, or amount of aqueous action. What can be more astonishing than to see these most delicate scratches retained in all their sharpness on rocks clothed with seaweed and shells, and exposed at every tide, in the bays of western Scotland!] The Lachoong is here twelve or fifteen yards wide, and runs over a pebbly bed, cutting a shallow channel through the deposits, down to the subjacent rock, which is in some cases scooped out six or eight feet deep by its waters. I do not doubt that the flatness of the floor of the Momay valley is caused by the combined action of the streams that drained the three glaciers which met here; for the tendency of retiring glaciers is to level the floors of valleys, by giving an ever-shifting direction to the rivers which drain them, and which spread detritus in their course. Supposing these glaciers to have had no terminal moraines, they might still have forced immense beds of gravel into positions that would dam up lakes between the ice and the flanks of the valleys, and thus produce much terracing on the latter.* [We are still very ignorant of many details of ice action, and especially of the origin of many enormous deposits which are not true moraines. These, so conspicuous in the lofty Himalayan valleys, are not less so in those of the Swiss Alps: witness that broad valley in which Grindelwald village is situated, and which is covered to an immense depth with angular detritus, moulded into hills and valleys; also the whole broad open Upper Rhone valley, above the village of Munster, and below that of Obergestelen. The action of broad glaciers on gentle slopes is to raise their own beds by the accumulation of gravel which their lower surface carries and pushes forward. I have seen small glaciers thus raised 300 feet; leaving little doubt in my mind that the upper Himalayan valleys were thus choked with deposit 1000 feet thick, of which indeed the proofs remain along the flanks of the Yangma valley. The denuding and accumulating effects of ice thus give a contour to mountain valleys, and sculpture their flanks and floors far more rapidly than sea action, or the elements. After a very extensive experience of ice in the Antarctic ocean, and in mountainous countries, I cannot but conclude that very few of our