We found the névé of the Marvine glacier differing greatly from the lower or icy portion previously traversed. Instead of ice with blue and white bands, as is common lower down, the {127} entire surface, and as far down in the crevasses as the eye could distinguish, was composed of compact snow, or snow changed to icy particles resembling hail and having in reality but few of the properties of ordinary snow: it might properly be called névé ice. Usually the thickness of the layers varied from ten to fifteen feet. Separating them were dark lines formed by dust blown over the surface of the glacier and buried by subsequent snow-storms, or by thin blue lines formed by the edges of sheets of ice and showing that the snow surface had been melted during bright sunny days and frozen again at night. The horizontal stratification so plainly marked in all the crevasses in the névé was almost entirely wanting, or at least was not conspicuous, in the lower portion of the glacier, where, instead, we found those narrow blue and white bands already mentioned, the origin of which has been so well described and explained by Tyndall. The center of the Marvine glacier, as in most similar ice-streams, is higher and less broken by crevasses than its borders. The crevasses at the side trend up stream, as is the case with marginal crevasses generally. In the present instance the courses of these rents could be plainly distinguished on each border of the glacier, when looking down upon it from neighboring slopes. The crevasses occur at quite regular intervals of approximately fifty feet, and diverge from the bank at angles of about 40°. In the banks of snow bordering the glacier similar crevasses diverge from the margin of the flowing glacier and trend down along its banks. The marginal crevasses and the crevasses in the bordering snow-fields, to which no special name has been given, fall nearly in line; but between the two there is a series of irregular cracks and broken snow, sharply defining the border of the moving névé. The origin of the marginal crevasses trending up stream was explained during the study of the glaciers of Switzerland. The following diagram and explanation illustrating their development are copied from Tyndall: