Supposing a river, instead of the glacier, to sweep through this valley; _its_ point of maximum motion would not always remain central, but would deviate towards that side of the valley to which the river turned its convex boundary. Indeed the positions of towns along the banks of a navigable river are mainly determined by this circumstance. They are, in most cases, situate on the convex sides of the bends, where the rush of the water prevents silting up. Can it be then that the ice exhibits a similar deportment? that the same principle which regulates the distribution of people along the banks of the Thames is also acting with silent energy amid the glaciers of the Alps? If this be the case, the position of the point of maximum motion ought, of course, to shift with the bending of the glacier. Opposite the Ponts, for example, the point ought to be on the Glacier du Geant, and westward of the centre of the trunk-stream; while, higher up, we ought to have another change to the eastern side, in accordance with the change of flexure. On the 25th of July a line was set out across the glacier, one of its fixed termini being a mark upon the first of the three Ponts. The motion of this line, measured on a subsequent day, and reduced to its daily rate, was found to be as follows:-- Fourth Line.--Daily Motion. No. of stake. Inches. East 1 moved 6-1/2 2 " 8 3 " 12-1/2 4 " 15-1/4 5 " 15-1/2 6 " 18-3/4 7 " 18-1/4 8 " 18-3/4 9 " 19-1/2 10 moved 21 11 " 20-1/2 12 " 23-1/4 13 " 23-1/4 14 " 21 15 " 22-1/4 16 " 17-1/4 17 " 15 West. This line, like the third, was set out and numbered from the eastern side of the glacier, the theodolite occupying a position on the heights of the Echelets. A moment's inspection of the table reveals a fact different from that observed on the third line; _there_ the most easterly stake moved with more than twice the velocity of the most westerly one; _here_, on the contrary, the most westerly stake moves with more than twice the velocity of the most easterly one. To enable me to compare the motion of the eastern and western halves of the glacier with greater strictness, my able and laborious companion undertook the task of measuring with a surveyor's chain the line just referred to; noting the pickets which had been fixed along the line, and the other remarkable objects which it intersected. The difficulty of thus directing a chain over crevasses and ridges can hardly be appreciated except by those who have tried it. Nevertheless, the task was accomplished, and the width of the Mer de Glace, at this portion of its course, was found to be 863 yards, or almost exactly half a mile.