Dôme extreme moisture may succeed almost absolute dryness in a few instants, in a clear sky and without any change of wind.
The parts of the atmosphere included within the same limits of temperature or humidity therefore rarely form concentric or parallel layers. They constitute regions interlacing zones which the clouds, thick or scattered in groups, often mark to our eyes, showing us those which are saturated with the vapor of water. The distribution of these zones in space depends chiefly on the heat action of the sun, and upon preceding and present movements of the atmosphere. Now, if a certain increase of heat is adequate to melt a piece of ice and to transform the water of the fusion into vapor, in like manner a corresponding cooling may suffice to cause the vapor to return to the state of a liquid and then to that of ice. The processes in the atmosphere are not different, and all showers, results of the more or less extensive condensation of the vapor of water, may arise from the cooling of that vapor or of the water which it produces.
A certain volume of atmospheric air is capable of holding in suspension a quantity of water proportioned to the elevation of its temperature. But, for each determined temperature, there is a maximum which can not be exceeded without the excess of vapor returning to the liquid state. If, therefore, an atmospheric region is saturated with vapor, and its temperature falls, that region will give rain. Immense and superabundant causes for the cooling necessary to provoke rain exist in such an atmosphere as we have described. The cooling may take effect in three principal ways: first, by the radiation of different regions between one another and toward interplanetary space, the temperature of which is extremely low, as has been indicated by measurements made in high balloon ascensions; second, by the expansion which air rising in the atmosphere undergoes in being rarefied; and, third, by the mingling of masses of warm or moist air with cold or dry.
Cooling by mixture is the sufficient cause in the majority of cases; and this may be effected from above, by descent of the air from the upper regions; from below, by ascent, with the assistance of rising currents created by solar radiations; or, finally, in any and every direction under the influence of the winds and the general movements of the atmosphere. Furthermore, the cooling need not be very great in order to provoke rain under certain conditions of temperature and humidity of frequent occurrence.
Rain clouds very frequently descend a little below the altitude of the Puy-de-Dôme. It is, therefore, not difficult, in order to determine the degree of cooling necessary for the formation of rain, to take advantage of observations that have been made there. The