with the mean temperature of the district. In winter, the external columns of air are as much heavier than the columns in the fissures as they are lighter in summer; and so cold currents of air blow from the cave into the fissures, though such currents are not of course colder than the external air. Thus the mean temperature of the cave is much lower than that of the rock in which it occurs; for the temperature of the currents varies from the mean temperature of the rock to the winter temperature of the external atmosphere. The descending columns of warmer air, in summer, must to some extent raise the temperature of the fissures above that which they would otherwise possess, that is, above the mean temperature of the place; but that may be considered to be counteracted by the corresponding lowering of the temperature of the fissures by the introduction of cold air from the cave in winter. By a similar reasoning, it will be seen that for some time after the spring change of direction in the currents takes place, the temperature of the cave will be less than would have been expected from a calculation founded on the true mean temperature of the rock through which the fissures pass. This, together with the fact of the porous nature of the rock in which most of the curious caves in the world occur, which allows a considerable amount of moisture to collect on all surfaces, and thereby induces a depression of temperature by evaporation, may be held to explain the presence of a greater amount of cold than might otherwise have been fairly reckoned upon in ice-caves. The idea of cold produced by evaporation Pictet took up warmly, believing that when promoted by rapid currents of air it would produce ice in the summer months; and he thus explained what he understood to be the phenomena of glacières. But it will have been seen, from the account of the caves I have visited, that the glacières are more or less in a state of thaw in the summer; and M. Thury's observations in the winter prove conclusively that they are then in a state of utter frost, so that the old belief with respect to the season at which the ice is formed may be supposed to have been exploded. The facts recorded by Mr. Scrope[191] would appear to depend upon the peculiar nature of rocks of volcanic formation; and I am inclined to think there is very little in common between such instances as he mentions and the large caves filled with ice which are to be found in the primary or secondary limestone.