If, then, the pressure of forty feet of ice lower the melting point .0075 deg. Centigrade, it follows that the pressure of a column 4846 feet high will lower it nine-tenths of a degree Centigrade. Supposing, then, the _unimpeded thrust of the whole glacier, from the Col du Geant downwards_, to be exerted on the ice at the Montanvert; or, in other words, supposing the bed of the glacier to be absolutely smooth and every trace of friction abolished, the utmost the pressure thus obtained could perform would be to lower the melting point of the Montanvert ice by the quantity above mentioned. Taking into account the actual state of things, the friction of the glacier against its sides and bed, the opposition which the three tributaries encounter in the neck of the valley at Trelaporte, the resistance encountered in the sinuous valley through which it passes; and finally, bearing in mind the comparatively short length of the glacier, which has to bear the thrust, and oppose the latter by its own friction merely;--I think it will appear evident that the ice at the Montanvert cannot possibly have its melting point lowered by pressure more than a small fraction of a degree. The ice in which my stakes were fixed being -5 deg. Centigrade, according to Mr. Thomson's calculation and his brother's experiments, it would require 667 atmospheres of pressure to liquefy it; in other words, it would require the unimpeded pressure of a column of glacier-ice 26,680 feet high. Did Mont Blanc rise to two and a half times its present height above the Montanvert, and were the latter place connected with the summit of the mountain by a continuous glacier with its bed absolutely smooth, the pressure at the Montanvert would be rather under that necessary to liquefy the ice on which my winter observations were made. [Sidenote: MEASUREMENTS APPLY TO SURFACE.] If it be urged that, though the temperature near the surface may be several degrees below the freezing point, the great body of the glacier does not share this temperature, but is, in all probability, near to 32 deg., my reply is simple. I did not measure the motion of the ice in the body of the glacier; nobody ever did; my measurements refer to the ice at and near the surface, and it is this ice which showed the plastic deportment which the measurements reveal. Such, then, are some of the considerations which prevent me from accepting the theory of Mr. Thomson, and I trust they will acquit me of all desire, to make his theory co-operate with my views. I am, however, far from considering his deduction the less important because of its failing to account for the phenomena of glacier motion. THE PRESSURE-THEORY OF GLACIER-MOTION. (22.) [Sidenote: POSSIBLE MOULDING OF ICE.]