Above the snow-line all of the mountain tops that are not precipitous are heavily loaded with snow. Where the snow breaks off at the verge of a precipice and descends in avalanches a depth of more than a hundred feet is frequently revealed, but in the valleys and amphitheatres the snow has far greater thickness. Pinnacles and crests of rock, rising through the icy covering, indicate that the thickness of the névé must be many hundreds of feet. There are no evidences of former glaciation on the mountain crests which project above the névé fields. There are no polished and striated rock surfaces or glaciated domes to indicate that the mountains were ever covered by a general capping of ice, as has been postulated for similar mountains elsewhere. When the {183} glaciers had their greatest expansion the higher mountains were in about their present condition. The increase in the volume of the glaciers was felt almost entirely in their lower courses. CHARACTERISTICS OF ALPINE GLACIERS BELOW THE SNOW-LINE. The first feature that attracts attention on descending from the névé region to the more icy portion of the glaciers is the rapid melting everywhere taking place. Every day during the summer the murmur and roar of rills, brooks and rivers are to be heard in all of the ice-fields. The surface streams are usually short, on account of the crevasses which intercept them. They plunge into the gulfs, which are many times widened out by the flowing waters so as to form wells, or _moulins_, and join the general drainage beneath. The streams then flow either through caverns in the glaciers or in tunnels at the bottoms. While traversing the glacier one may frequently hear the subdued roar of rivers coursing along in the dark chambers beneath when no other indication of their existence appears at the surface. When these subglacial streams emerge, usually near the margin of the ice, they issue from archways forming the ends of tunnels, and perhaps flow for a mile or two in the sunlight before plunging into another tunnel to continue their way as before. The best example of a glacial river seen during our exploration was near the western border of the Lucia glacier. It is shown in the illustration forming plate 12, which is reproduced mechanically from a photograph. This Styx of the ice-world has been described on an earlier page. The lakes formed at the southern end of nearly every mountain spur projecting into the Malaspina glacier discharge through tunnels in the ice, which are similar in every way to those formed by the stream already mentioned.