Let us apply these facts to the explanation of the clink heard in my experiments. On sending a sunbeam through ice, liquid cavities are suddenly formed at various points within the mass, and these cavities are completely cut off from atmospheric contact. But the water formed by the melting ice is less in volume than the ice which produces it; the water of a cavity is not able to fill it, hence a vacuous space must be formed in the cell. I have no doubt that, for a time, the strong cohesion between the walls of the cell and the drop within it augments the volume of the latter a little, so as to compel it to fill the cell; but as the quantity of liquid becomes greater the shrinking force augments, until finally the particles snap asunder like a broken spring. At the same moment a lustrous spot appears, which is a vacuum, and simultaneously with the appearance of this vacuum the clink was always heard. Multitudes of such little explosions must be heard upon a glacier when the strong summer sun shines upon it, the aggregate of which must, I think, contribute to produce the "crepitation" noticed by M. Agassiz, and to which I have already referred. [Sidenote: FIGURES IN ICE; VACUOUS SPOTS.] In Plate VI. of the Atlas which accompanies the 'Systeme Glaciaire' of M. Agassiz, I notice drawings of figures like those I have described, which he has observed in glacier-ice, and which were doubtless produced by direct solar radiation. I have often myself observed figures of exquisite beauty formed in the ice on the surface of glacier-pools by the morning sun. In some cases the spaces between the leaves of the liquid flowers melt partially away, and leave the central spot surrounded by a crimped border; sometimes these spaces wholly disappear, and the entire space bounded by the lines drawn from point to point of the leaves becomes liquid, thus forming perfect hexagons. The crimped borders exhibit different degrees of serration, from the full leaves themselves to a gentle undulating line, which latter sometimes merges into a perfect circle. In the ice of glaciers, I have seen the internal liquefaction ramify itself like sprigs of myrtle; in the same ice, and particularly towards the extremities of the glacier, disks innumerable are also formed, consisting of flat round liquid spaces, a bright spot being usually associated with each. These spots have been hitherto mistaken for air-bubbles; but both they and the lustrous disks at the centres of the flowers are vacuous. I proved them to be so by plunging the ice containing them into hot water, and watching what occurred when the walls of the cells were dissolved, and a liquid connexion established between them and the atmosphere. In all cases they totally collapsed, and no trace of air rose to the surface of the warm water.