The causes of the movements of air in these windholes are exceedingly simple. The movements of air depend on the fact that in summer the air in the tube becomes colder from contact with the rocks and, therefore, heavier than the air outside, and by gravity the heavy inside air displaces the lighter outside air and comes rushing out at the lower opening. This leaves a vacuum, which is filled by the warmer air dropping into the tube from above. In winter on the contrary, the air within the tube is warmed by contact with the rocks and becomes lighter than the air outside. It, therefore, rises and streams out from the upper opening, and the vacuum is filled by the heavy cold air pushing in at the lower opening. [Illustration: Fig. 11. Vertical Section of a Windhole.] G. F. Parrot's[27] explanation is so satisfactory that I give it with one or two changes. He considers the air movements an ordinary statical phenomenon of the air, in caves which have two openings at different altitudes. Let E G D represent the section of such a cave with the openings A and B. Let us think that there are over C and B two vertical air columns and from B to C a horizontal air column B C; then the two air columns over B and C are at all times of the year equal in weight. Not so the air columns A C and A E G D B, because their temperatures are different. Assume the temperature in the cave G is +12° the whole year round. If in summer the air column A C is at a temperature of +25°, then the heavy air in the cave G pours out through A and is replaced by air flowing in through B. If in winter the air column A C is at a temperature of -1°, then the air pours with equal inverse velocity at A into the cave, and out at B. The velocity of the current in both cases depends on the difference of temperature within and without. [27] _Grundriss der Physik der Erde und Geologie_, 1815, pages 92-99. The foregoing explanation makes it evident that the movements of air in these windholes do not depend on the presence of ice. In many of those I have examined myself there was no ice visible, and from the temperature of the air current, there could not have been any ice within the mountain. Still, there are numerous cases where ice is found in windholes among boulders, and a few cases where windholes exist in connection with apparently static glacière caves. Undoubtedly the great majority of windholes do not contain ice in summer, or, indeed, at any time of the year, and, as far as I can see, windholes, according to their temperatures, belong rather to the class of normal caves than to that of glacières.