The same writer mentions a similar example of summer ice in Auvergne.[151] There is a natural grotto in the basalt near Pont Gibaud, some miles to the north-west of Clermont, in which a small spring is found partly frozen during the greatest heats of summer, while the water is said to be warm in winter; probably, Scrope observes, only seeming to be warm by contrast with the external temperature.  The water is apparently frozen by means of the powerful evaporation produced by a current of very dry air proceeding from some long fissures or arched galleries which communicate with the cave.  In this case also the writer suggests that the air owes its dryness to the absorbent qualities of the lava through which it passes:  he repeats, too, the remark that the phenomenon is of common occurrence in caverns in volcanic districts.[152]

There is a remarkable instance of ice occurring under lava, near the Casa Inglese on Mount Etna, which it may be as well to mention, though the causes of its existence have probably nothing in common with the phenomena of ice-caves, or summer ice.  An account of it is to be found in Sir Charles Lyell’s ’Elements of Geology.’[153] It appears that the summer and autumn of 1828 were so hot, that the artificial ice-houses of Catania and the adjoining parts of Sicily failed.  Signer M. Gemmellaro had long believed that a small mass of perennial ice at the foot of the highest cone of Etna was only a part of a large and continuous glacier covered by a lava current, and from this he expected to derive an abundant supply of ice.  He procured a large body of workmen, and quarried into the ice; but though he thus proved the superposition of lava for several hundred yards, the ice was so hard, and the expense of quarrying consequently so great, that the works were abandoned.  This was on the south-east of the cone, not far from the Casa Inglese.  Sir Charles Lyell suggests that, probably, at the commencement of some eruption, a large mass of snow has been thickly covered with volcanic sand, showered upon it before the arrival of the lava itself.  This sand is a non-conductor of heat, and would therefore tend to preserve the snow from complete fusion when the hot lava-stream passed over it, and thus the existence of the underground glacier may be explained.  The peasants of the district are so well acquainted with the non-conducting properties of volcanic sand, that they secure an annual store of snow, for providing water in summer, by strewing a layer of sand a few inches thick upon a field of snow, thus effectually shutting out the heat of the sun.  It is curious that when De Saussure visited Chamouni for the first time, his attention was arrested by the sight of women sowing what seemed to be grain of some kind in the snow; but, on enquiring, he found that it was only black earth, which the inhabitants spread on the snow in spring, in order to make it disappear sooner.  He was told that snow thus treated would melt a fortnight or three weeks before the ordinary time for its disappearance in the valley; but it will be seen that this does not contradict the theory of the Sicilian peasants.[154]