Besides the streams of new snow, there were occasional avalanches of a different character, caused by the breaking away of portions of the cliffs of old snow, accumulated, perhaps, during several winters. These start from the summits of precipices, and are caused by the slow downward creep of the snow-fields above. The snow-cliffs are always crevassed and broken in much the same manner as are the ends of glaciers which enter the sea, and occasionally large masses, containing thousands of cubic yards, break away and are precipitated down the slopes with a suddenness that is always startling. Usually the first announcement of these avalanches is a report like that of a cannon, followed by a rumbling roar as the descending mass ploughs its way along. The avalanches formed by old snow are quite different from those caused by the descent of the new surface snow, but are frequently accompanied by surface streams in case there has been a recent storm. The paths ploughed out by the avalanches are frequently sheathed with glassy ice, formed by the freezing of water produced by the melting of snow on account of the heat produced by the friction of the moving mass. A third variety of avalanches, due to falling stones, has already been noticed. The floor of my snow-chamber was the surface of the old snow on which we had pitched our tents at the time we first reached that camping place. On this hard surface, and forming the walls of the cell, there were thirty inches of clear white snow, the upper limit of which was marked by a blue layer of ice about a quarter of an inch thick. This indicated the thickness of snow that fell during the first storm. Its surface had been melted and softened during the days of sunshine that followed its fall, and had frozen into clear ice. Above the blue band which encircled the upper portion of my chamber was the soft, pure white snow of the second storm. The stratification of snow which I had seen fall rendered it evident that my interpretation of the stratification observed in the sides of crevasses was correct. The snow when it fell was soft and white, and composed of very fine crystals; but under the influence of the air and sunshine it changed its texture and became icy and granular, and then resembled the névé snow so common in high mountains.