Those who have read his startling investigations will remember that by sending a beam of sunlight through ice he brought to view the primitive crystalline forms to which it owes its solidity, and that he insisted that these star-shaped figures are always in the plane of crystallization. Without knowing what might be their origin, I had myself noticed these figures, and represented them in a diagram, part of which is reproduced in the annexed wood-cut. I had considered them to be compressed air-bubbles; and though I cannot, under my present circumstances, repeat the experiment of Dr. Tyndall upon glacier-ice, I conceive that the star-shaped figures represented upon Pl. VII. figs. 8 and 9, in my "Systeme Glaciaire," may refer to the same phenomenon as that observed by him in pond-ice. Yet while I make this concession, I still maintain, that besides these crystalline figures there exist compressed air-bubbles in the angular fragments of the glacier-ice, as shown in the above wood-cut; and that these bubbles are grouped in sets, trending in the same direction in one and the same fragment, and diverging under various angles in the different fragments. I have explained this fact concerning the position of the compressed air-bubbles, by assuming that ice, under various pressure, may take the appearance it presents in each fragment with every compressed air-bubble trending in the same direction, while their divergence in the different fragments is owing to a change in the respective position of the fragments resulting from the movement of the whole glacier. I have further assumed, that throughout the glacier the change of the snow and porous ice into compact ice is the result of successive freezing, alternating with melting, or at least with the resumption of a temperature of 32 deg. Fahrenheit in consequence of the infiltration of liquid water, to which the effects of pressure must be added, the importance of which in this connection no one could have anticipated prior to the experiments of Dr. Tyndall. Of course, if the interior temperature of the glacier never falls below 32 deg., the changes here alluded to could not take place. But if the _vacuous spaces_ observed by Dr. Tyndall are really identical with the spaces I have described as _extremely flattened air-bubbles_, I think the arrangement of these spaces as above described proves that it freezes in the interior of the glacier to the depth at which these crosswise fragments have been observed: that is, at a depth of two hundred feet. For, since the experiments of Dr. Tyndall show that the vacuous spaces are parallel to the surface of crystallization, and as no crystallization of water can take place unless the surrounding temperature fall below 32 deg., it follows that these vacuous spaces could not exist in such large continuous fragments, presenting throughout the fragments the same trend, if there had been no frost within the mass, affecting the whole of such a