thermal wave outside (first maximum: day 154); in the lower part, the thermal wave of Station 4–“Grand Central” inside the cave (first maximum: day 214). The delay is of 60 days and the attenuation is about 1/20. Dots: experimental values; line: sinusoidal best fit; horizontal axis: days (Jan 1, 1996=0); vertical axis: °C.

Airflow Air circulation in a cave depends on the number of entrances and the general shape of the cave. Near the end of the 19th century Crammer (1899) showed that a cave with a single entrance may act either as a warm trap (if the entrance is in the lower part of the cave) or as a cold trap (if the entrance is in the upper part) where the outside air is “trapped”, respectively, in the summer or in the winter season (Figures 2 and 3). When there are two entrances at different altitudes, the difference in air density between the cave atmosphere and the corresponding outside air column results in airflow. The flow is directed from the lower to the higher entrance in winter and the reverse in summer (Figure 4). In general the driving force depends on temperature, warm air being less dense than cold air but when the outside temperature is close to the cave temperature relative humidity may play a role, since humid air is less dense than dry air. Where there are more than two entrances, the airflow pattern may be more complicated, being influenced also by the shape and the size of the passages involved.