191. Let us now apply our law to the explanation of a difficulty. Turning to the careful measurements executed by M. Agassiz on the glacier of the Unteraar, we notice in the discussion of these measurements a section of the "Système glaciaire" devoted to the "Migrations of the Centre." It is here shown that the middle of the Unteraar glacier is not always the point of swiftest motion. This fact has hitherto remained without explanation; but a glance at the Unteraar valley, or at the map of the valley, shows the enigma to be an illustration of the law which we have just established on the Mer de Glace. § 26. _Motion of Axis of Mer de Glace._ 192. We have now measured the rate of motion of five different lines across the trunk of the Mer de Glace. Do they all move alike? No. Like a river, a glacier at different places moves at different rates. Comparing together the points of maximum motion of all five lines, we have this result: MOTION OF MER DE GLACE. At Trélaporte 20 inches a day. At _les Ponts_ 23 " " Above the Montanvert 26 " " At the Montanvert 34 " " Below the Montanvert 33[C] " " [C] This is probably under the mark. I think it likely that the swiftest motion of this portion of the Mer de Glace in 1857 amounted to a yard in twenty-four hours. 193. There is thus an increase of rapidity as we descend the glacier from Trélaporte to the Montanvert; the maximum, motion at the Montanvert being fourteen inches a day greater than at Trélaporte. § 27. _Motion of Tributary Glaciers._ 194. So much for the trunk glacier; let us now investigate the branches, permitting, as we have hitherto done, reflection on known facts to precede our attempts to discover unknown ones. 195. As we stood upon our "cleft station," whence we had so capital a view of the Mer de Glace, we were struck by the fact that some of the tributaries of the glacier were wider than the glacier itself. Supposing water to be substituted for the ice, how do you suppose it would behave? You would doubtless conclude that the motion down the broad and slightly-inclined valleys of the Géant and the Léchaud would be comparatively slow, but that the water would force itself with increased rapidity through the "narrows" of Trélaporte. Let us test this notion as applied to the ice. 196. Planting our theodolite in the shadow of Mont Tacul, and choosing a suitable point at the opposite side of the Glacier du Géant, we fix on July 29 a series of ten stakes across the glacier. The motion of this line in twenty-four hours was as follows:-- MOTION OF GLACIER DU GÉANT. Sixth Line: H H' upon Sketch. Stake 1 2 3 4 5 6 7 8 9 10 Inches 11 10 12 13 12 13 11 10 9 5 197. Our conjecture is fully verified. The maximum motion here is seven inches a day less than that of the Mer de Glace at Trélaporte (192).