When M. de Cossigny visited the cave, there were thirteen or fourteen columns of ice, from 6 to 8 feet high, and he was in consequence inclined to doubt the accuracy of the statement of M. Billerez, that in his time (1711) there were three columns only, from 15 to 20 feet high. But my own observation of the shape of the columns suggested that the largest of all was probably an amalgamation of several others; so that it is not unreasonable to suppose that after the Duc de Lévi removed the large columns seen by M. Billerez, a number of smaller columns were formed on the old site, and that these had not become large enough to amalgamate in 1743. Not satisfied with these visits of August and October, M. de Cossigny visited the cave in April 1745. He found the temperature at 5 A.M. to be exactly at the freezing point, and at noon it had risen 1°. From this he concluded that the stories of the greater cold in the cave during the summer, as compared with the winter, were false. In 1769, M. Prévost, of Geneva, visited the cave, as a young man; and in 1789, he wrote an account of his visit in the _Journal de Genève_ (March), which was afterwards inserted as an additional chapter in his book on Heat.[186] He believed that one or two hundred _toises_ was the utmost that could be allowed for the height of the hill in which the glacière lies,--a sufficiently vague approximation. He rejected the idea of salt as the cause of ice, and came to the conclusion that the cave was in fact nothing more than a good natural ice-house, being protected by dense trees, and a thick roof of rock, while its opening towards the north sheltered it from all warm winds. He accounted for the original presence of ice as follows:--In the winter, stalactites form at the edges of various fissures in the roof, and snow is drifted on to the floor of the cave by the north winds down the entrance-slope. When the warmer weather comes, the stalactites fall by their own weight, and, lying in the drifted and congealed snow, form nuclei round which the snow is still further congealed, and the water which results from the partial thaw of portions of the snow is also converted into ice. Thus, a larger collection of ice forms in winter than the heat of summer can destroy; and if none of it were removed, it might, in the course of years, almost fill the cave. At the time of his visit (August), M. Prévost found only one column, from 6 to 8 feet high.