I will not pretend to say that I fully understand this passage, but, taking it and the former one together, I think it is clear that the water which is supposed to gorge the capillaries of the glacier is assumed to be essential to its motion. Indeed, an extreme degree of sensitiveness has been ascribed to the glacier as regards the changes of temperature by which the capillaries are affected. In three succeeding days, for example, Professor Forbes found the diurnal summer motion of a point upon the Mer de Glace to increase from 15.2 to 17.5 inches a day; a result which he says he is "persuaded" to be due to the increasing heat of the weather at the time. If, then, the glacier capillaries can be gorged so quickly as this experiment would indicate, it is fair to assume that they are emptied with corresponding speed when the supply is cut away. [Sidenote: TEMPERATURE AT CHAMOUNI; WINTER 1859.] The extraordinary coldness of the weather previous to the Christmas of 1859 is in the recollection of everybody: this lowness of temperature also extended to the Mer de Glace and its environs. I had last summer left with Auguste Balmat and the Abbe Vueillet thermometers with which observations were made daily during the cold weather referred to. I take the following from Balmat's register. Minimum Date. temperature Centigrade. December 16 -15 deg. " 17 -20 " 18 -16-1/2 " 19 -9 " 20 -13 " 21 -20-1/2 " 22 -4-1/4 December 23 -4-1/2 deg. " 24 -6-1/2 " 25 -2 " 26 +2 " 27 -3 " 28 -10-1/2 " 29 -6 The temperature at the Montanvert during the above period may be assumed as generally some degrees lower, so that for a considerable period, previous to my winter observations, the portion of the Mer de Glace near the Montanvert had been exposed to a very low temperature. I reached the place after the weather had become warm, but during my stay there the maximum temperature did not exceed -4-1/2 deg. C. Considering therefore the long drain to which the glacier had been subjected previous to the 29th of December, it is not unreasonable to infer that the capillary supply assumed by Professor Forbes must by that time have been exhausted. Notwithstanding this, the motion of the glacier at the Montanvert amounted at the end of December to half its maximum summer motion. [Sidenote: BALMAT'S MEASUREMENTS.]