All my own observations have tended more and more to make me believe that the cold of winter is the cause of the ice. Before elaborating my own views, however, I wish to take up seriatim the theories which have been formulated, principally in explanation of the belief that the ice was a summer product, and to give my reasons for my disbelief in them. _Glacial Period._--The first theory, perhaps, to touch on, is the one that the ice is a remnant of a glacial period. This seems to occur to many persons as a solution of the question when they first hear of glacières, and it has been several times propounded to me, and naturally enough, always by scientific men.[42] Still I do not think it has ever been held by anyone who had made a study of glacières. [42] See Part IV.: Hitchcock, page 284; Bonney, page 291; Dawkins, page 292. The theory is, indeed, untenable in regard to freezing caves, as it does not accord with the observed facts of the yearly disappearance of the ice in many caves and taluses. At Szilize every year the ice has disappeared pretty completely by November, and the cave is free; but in April or May the floor is again covered with ice, and columns and icicles have formed on the roof and sides. At La Genollière the cave is used by the people of the neighboring châlets, through the spring and early summer, to help in the operation of butter making; by the middle or the end of August it has entirely disappeared, but is found formed afresh the following spring. At the Rumney Talus, at the Cave of Decorah, at the Gorge of Ellenville, and at the Williamstown Snow Hole, I found no snow or ice.[43] Yet it is abundant in all these localities in the spring. Too many examples of the complete melting away of the ice every year can be cited, to permit any doubt: glacière caves are not connected with a glacial period. [43] See Part I.: Rumney, page 85; Decorah, page 88; Ellenville, page 91; Williamstown, page 98.