Another difficulty is the following:--Mr. Thomson will have seen that I have recorded certain winter measurements made on the Mer de Glace, and that these measurements show not only that the ice moves at that period of the year, but that it exhibits those characteristics of motion from which its plasticity has been inferred; the velocity of the central portions of the glacier being in round numbers double the velocity of those near the sides. Had there been any necessity for it, this ratio might have been augmented by placing the side-stakes closer to the walls of the glacier. Considering the extreme coldness of the weather which preceded these measurements, it is a moderate estimate to set down the temperature of the ice in which my stakes were fixed at 5 deg. Cent. below zero. [Sidenote: REQUISITE PRESSURE CALCULATED.] Let us now endeavour to estimate the pressure existing at the portion of the glacier where these measurements were made. The height of the Montanvert above the sea-level is, according to Prof. Forbes, 6300 feet; that of the Col du Geant, which is the summit of the principal tributary of the Mer de Glace, is 11,146 feet: deducting the former from the latter, we find the height of the Col du Geant above the Montanvert to be 4846 feet. Now, according to Mr. Thomson's theory and his brother's experiments, the melting point of ice is lowered .0075 deg. Centigrade for every atmosphere of pressure; and one atmosphere being equivalent to the pressure of about thirty-three feet of water, we shall not be over the truth if we take the height of an equivalent column of glacier-ice, of a compactness the mean of those which it exhibits upon the Col du Geant and at the Montanvert respectively, at forty feet. The compactness of glacier ice is, of course, affected by the air-bubbles contained within it. [Sidenote: ACTUAL PRESSURE INSUFFICIENT.]