§ 17. This cloud of evaporation, however, does not always rise. It sometimes rests in absolute stillness, low laid in the hollows of the hills, their peaks emergent from it. Fig. 92 shows this condition of it, seen from a distance, among the Cenis hills. I do not know what gives it this disposition to rest in the ravines, nor whether there is a greater chill in the hollows, or a real action of gravity on the particles of cloud. In general, the position seems to depend on the temperature. Thus, in Chamouni, the crests of La Côte and Taconay continually appear in stormy weather as in Plate 36, Vol. IV., in which I intended to represent rising drift-cloud, made dense between the crests by the chill from the glaciers. But in the condition shown in Fig. 92, on a comparatively open sweep of hill-side, the thermometer would certainly indicate a higher temperature in the sheltered valley than on the exposed peaks; yet the cloud still subsides into the valleys like folds of a garment; and, more than this, sometimes conditions of morning cloud, dependent, I believe, chiefly on dew evaporation, form first on the _tops_ of the soft hills of wooded Switzerland, and droop down in rent fringes, and separate tongues, clinging close to all the hill-sides, and giving them exactly the appearance of being covered with white fringed cloth, falling over them in torn or divided folds. It always looks like a true action of gravity. How far it is, in reality, the indication of the power of the rising sun causing evaporation, first on the hill-top, and then in separate streams, by its divided light on the ravines, I cannot tell. The subject is, as the reader perceives, always inextricably complicated by these three necessities--that to get a cloud in any given spot, you must have moisture to form the material of it, heat to develop it, and cold[4] to show it; and the adverse causes inducing the moisture, the evaporation, and the visibility are continually interchanged in presence and in power. And thus, also, the phenomena which properly belong to a certain elevation are confused, among hills at least, with those which in plains would have been lower or higher. I have been led unavoidably in this chapter to speak of some conditions of the rain-cloud; nor can we finally understand the forms even of the cumulus, without considering those into which it descends or diffuses itself. Which, however, being, I think, a little more interesting than our work hitherto, we will leave this chapter to its dulness, and begin another. FOOTNOTES: [1] One of the great difficulties in doing this is to distinguish the portions of cloud outline which really slope upwards from those which only appear to do so, being in reality horizontal, and thrown into apparent inclination by perspective.