Freezing gorges, however, show, also, certain degrees of kinship to freezing caverns and taluses, principally in the protection afforded to the ice against external destructive influences. The ice is almost always found in positions where it receives little, if any, of the direct rays of the sun and, also, where it is scarcely, if at all, exposed to any winds. The sides of the fissures and surrounding trees generally afford the necessary protection. Some of the forms which the ice assumes in gorges, such as long pendent icicles, are also more characteristic of underground than of overground ice. The freezing troughs or basins found in Siberia are evidently closely related to gorges, and the fact that the ice is found in less sheltered places may be explained by the high northerly latitudes of these troughs, in general between fifty-seven and sixty degrees. _The Winter's Cold Theory._--The places where ice is found underground differ in one important respect from gullies and troughs, and that is, in the fact that above the ice there is rock or soil, which, in true caves, takes the form of a roof. This causes some important distinctions between overground and underground perennial ice. It means that the ice is formed directly in the caves, and that it is genuine subterranean ice, and not, except perhaps near the entrance, solidified snow. The roof, while not admitting the winter snows, is, however, a protection against warm summer rains, and, of course, entirely cuts off radiation from the sky. If, therefore, it keeps out some cold, it also acts as a protector against heat. That the cold of winter is the source of the cold which produces the ice which forms underground, and that it is through its influence, with the assistance of certain secondary causes, that some caves are converted into what are practically natural ice houses, seems to me the true explanation of the phenomenon of subterranean ice, not only since it is the simple and obvious explanation, but also because all the facts, so far as I have myself observed, are in accord with this theory.[56] [56] Among those who have written or said that the cold of winter plays a more or less important part in the formation of subterranean ice may be mentioned: Poissenot, 1586; Gollut, 1592; DeBoz, 1726; Nagel, 1747; Cossigny, 1750; Jars, 1774; Hacquet, 1778; Girod-Chantrans, 1783; Hablizl, 1788; Prévost, 1789; Townson, 1797; Humboldt, 1814; Dearborn, 1822; Deluc, 1822; Dewey, 1822; Lee, 1824; Reich, 1834; Hayden, 1843; Guyot, 1856; Rogers, 1856; Petruzzi, 1857; Smyth, 1858; Hager, 1861; Thury, 1861; Browne, 1865; Raymond, 1869; Krenner, 1874; Ritchie, 1879; Benedict, 1881; Schwalbe, 1881; Fugger, 1883; Trouillet, 1885; Girardot, 1886; Russell, 1890; Martel, 1892; Krauss, 1894; Lohmann, 1895; Balch, 1896; Cvijic, 1896; Butler, 1898; Kovarik, 1898; Cranmer, 1899.