In 1822, M.A.  Pictet, of Geneva, took up the question of natural glacieres, and read a paper before the Helvetic Society of Natural Sciences,[189] describing his visits to the caves of the Brezon and the Valley of Reposoir.  In order to explain the phenomena presented by those caves, M. Pictet adopted De Saussure’s theory of the principle of caves-froides, rendering it somewhat more precise, and extending it to meet the case of ice-caves.  It is well known that, in many parts of the world, cold currents are found to blow from the interstices of rocks; and these are utilised by neighbouring proprietors, who build sheds over the fissures, and so secure a cool place for keeping meat, &c.  Examples of such currents are met with near Rome (in the Monte Testaceo), at Lugano, Lucerne (the caves of Hergiswyl), and in various other districts.  It is found that the hotter the day, the stronger is the current of cold air; in winter the direction of the current is changed, and it blows into the rock instead of out from it.[190] De Saussure’s theory, as developed by M. Pictet, was no doubt satisfactory, so far as it was used to account for the phenomenon of ‘cold-caves,’ but it seems to be insufficient as an explanation of the existence of large masses of subterranean ice; of which, by the way, De Saussure must have been entirely ignorant, for he makes no allusion to such ice, and the temperatures of the coldest of his caves were considerably above the freezing point.

Pictet represents the case of a cave with cold currents of air to be much the same as that of a mine with a vertical shaft, ending in a horizontal gallery of which one extremity is in communication with the open air, at a point much lower, of course, than the upper extremity of the shaft.  The cave corresponds to the horizontal gallery, and the various fissures in the rock take the place of the vertical shaft, and communicate freely with the external air.  In summer, the columns of air contained in these fissures assume nearly the temperature of the rock in which they rest, that is to say, the mean temperature of the district, and therefore they are heavier than the corresponding external columns of air which terminate at the mouth of the cave; for the atmosphere in summer is very much above the mean temperature of the soil, or of the interior of the earth at moderate depths.  The consequence is, that the heavy cool air descends from the fissures, and streams out into the cave, appearing as a cold current; and the hotter the day is—­that is, the lighter the columns of external air—­the more violent will be the disturbance of equilibrium, and therefore the more palpable the cold current.  Naturally, in this last case, the air which enters by the upper orifices of the fissures is more heated, to begin with, than on cooler days; but external heat so very slightly affects the deeper parts of the fissures, that the columns of air thus introduced are speedily impressed with the mean temperature of the district.  In