In 1899, Mons. Martel gave an account of the Glacière de Naye. In this paper, he abandons definitely fossil ice, salts and the capillary theory as possible causes of underground ice. He considers that there are four causes: 1, shape of the cavity; 2, free access of snow in winter; 3, high altitude; 4, evaporation due to wind currents. The last two causes he thinks are not necessarily always present. For instance he considers that, at the Creux-Percé, and at Chaux-les-Passavant, the ice is due especially to the sack or hour-glass shape of these hollows where the summer air cannot get in on account of its lightness. At the Glacière de Naye, which is a big windhole, situated at an altitude of 1750 to 1820 meters, Mons. Martel thinks that the ice is formed by the snow and cold of winter, but that its preservation is assured by the evaporation caused by the action of the windhole. Dr. Terlanday, in 1893, asserted that ice does not form in Szilize in winter, and that the ice first forms in the winter in the upper part of rock fissures and that in the spring, at the time of an increase of temperature, this fissure ice is brought to the melting point by the successive entering of heat into the earth and that it then arrives at the cave, where it aids the formation of icicles. This theory about fissure ice is probably in so far correct, that the ice in the upper parts of fissures, near the surface of the ground, melts before the ice in the lower parts of fissures. The drip would then naturally run into the cave and, as long as the temperature of the cave was low, help to form cave ice. Dr. Hans Lohmann, in 1895, published some valuable notes about several glacières. While considering the cold of winter as the main cause of the ice, he thought evaporation a secondary cause of cold. He says: "That the cold from evaporation bears its share in cooling a cave, will not be denied. * * * The air saturated with aqueous vapor makes one think of constant evaporation. The aqueous vapor spreads itself by diffusion throughout the entire cave, and if the outside air is driest, goes to that. Through this, more ice and water can always be vaporized, and to the warming elements there is furnished a cooling one. If dry winds get into the cave, then must evaporation be very lively and the chilling especially strong. Through this cause alone can be explained the remarkably low temperature of +6.3° in the new part of the Garischen Stollen, in contrast to the temperature of +7.9° in the old part. The strong draught in the last drew out through its suction the damp air of the new adit, so that there had to be a strong evaporation."