We have now to consider those movements of the water which depend upon the fact that at ordinary temperatures the sea yields to the air a continued and large supply of vapour, a contribution which is made in lessened proportion by water in all stages of coldness, and even by ice when it is exposed to dry air. This evaporation of the sea water is proportional to the temperature and to the dryness of the air where it rests upon the ocean. It probably amounts on the average to somewhere about three feet per annum; in regions favourably situated for the process, as on the west coast of northern Africa, it may be three or four times as much, while in the cold and humid air about the poles it may be as little as one foot. When contributed to the air, the water enters on the state of vapour, in which state it tends to diffuse itself freely through the atmosphere by virtue of the motion which is developed in particles when in the vaporous or gaseous state.

The greater part of the water evaporated from the seas probably finds its way as rain at once back into the deep, yet a considerable portion is borne away horizontally until it encounters the land. The precipitation of the water from the air is primarily due to the cooling to which it is subjected as it rises in the atmosphere. Over the sea the ascent is accomplished by the simple diffusion of the vapour or by the uprise through the aërial shaft, such as that near the equator or over the centres of the whirling storms. It is when the air strikes the slopes of the land that we find it brought into a condition which most decidedly tends to precipitate its moisture. Lifted upward, the air as it ascends the slopes is brought into cooler and more rarefied conditions. Losing temperature and expanding, it parts with its water for the same reason that it does in the ascending current in the equatorial belt or in the chimneys of the whirl storms. A general consequence of this is that wherever moisture-laden winds from the sea impinge upon a continent they lay down a considerable part of the water which they contain.

If all the lands were of the same height, the rain would generally come in largest proportion upon their coastal belt, or those portions of the shore-line districts over which the sea winds swept. But as these winds vary in the amount of the watery vapour which they contain, and as the surface of the land is very irregular, the rainfall is the most variable feature in the climatal conditions of our sphere. Near the coasts it ranges from two or three inches in arid regions—such as the western part of the Sahara and portions of the coast regions of Chili and Peru—to eight hundred inches about the head waters of the Brahmapootra River in northern India, where the high mountains are swept over by the moisture-laden airs from the neighbouring sea. Here and there detached mountainous masses produce a singular local increase in the amount of the rainfall. Thus in the lake district in northwestern England the rainfall on the seaward side of mountains, not over four thousand feet high, is very much greater than it is on the other slope, less than a score of miles away. These local variations are common all over the world, though they are but little observed.

In general, the central parts of continents are likely to receive much less rainfall than their peripheral portions. Thus the central districts of North America, Asia, and Australia—three out of the five continental masses—have what we may call interior deserts. Africa has one such, though it is north of the centre, and extends to the shores of the Mediterranean and the Atlantic. The only continent without this central nearly rainless field is South America, where the sole characteristic arid district is situated on the western slope of the Cordilleran range. In this case the peculiarity is due to the fact that the strong westerly setting winds which sweep over the country encounter no high mountains until they strike the Andean chain. They journey up a long and rather gradual slope, where the precipitation is gradually induced, the process being completed when they strike the mountain wall. Passing over its summit, they appear as dry winds on the Pacific coast.

Even while the winds frequently blow in from the sea, as along the western coast of the Americas, they may come over water which is prevailingly colder than the land. This is characteristically the case on the western faces of the American continent, where the sea is cooled by the currents setting toward the equator from high latitudes. Such cool sea air encountering the warm land has its temperature raised, and therefore does not tend to lay down its burden of moisture, but seeks to take up more. On this account the rainfall in countries placed under such conditions is commonly small.

By no means all the moisture which comes upon the earth from the atmosphere descends in the form of rain or snow. A variable, large, though yet undetermined amount falls in the form of dew. Dew is a precipitation of moisture which has not entered the peculiar state which we term fog or cloud, but has remained invisible in the air. It is brought to the earth through the radiation of heat which continually takes place, but which is most effective during the darkened half of the day, when the action is not counterbalanced by the sun's rays. While the sun is high and the air is warm there is a constant absorption of moisture in large part from the ground or from the neighbouring water areas, probably in some part from those suspended stores of water, the clouds, if such there be in the neighbourhood. We can readily notice how clouds drifting in from the sea often melt into the dry air which they encounter. Late in the afternoon, even before the sun has sunk, the radiation of heat from the earth, which has been going on all the while, but has been less considerable than the incurrent of temperature, in a way overtakes that influx. The air next the surface becomes cooled from its contact with the refrigerating earth, and parts with its moisture, forming a coating of water over everything it touches. At the same time the moisture escaping from the warmed under earth likewise drops back upon its cooled surface almost as soon as it has escaped. The thin sheet of water precipitated by this method is quickly returned to the air when it becomes warmed by the morning sunshine, but during the night quantities of it are absorbed by the plants; very often, indeed, with the lowlier vegetation it trickles down the leaves and enters the earth about the base of the stem, so that the roots may appropriate it. Our maize, or Indian corn, affords an excellent example of a plant which, having developed in a land of droughts, is well contrived, through its capacities for gathering dew, to protect itself against arid conditions. In an ordinary dew-making night the leaves of a single stem may gather as much as half a pint of water, which flows down their surfaces to the roots. So efficient is this dew supply, this nocturnal cloudless rain, that on the western coast of South America and elsewhere, where the ordinary supply of moisture is almost wanting, many important plants are able to obtain from it much of the water which they need. The effect is particularly striking along seashores, where the air, although it may not have the humidity necessary for the formation of rain, still contains enough to form dew.

It is interesting to note that the quantity of dew which falls upon an area is generally proportioned to the amount of living vegetation which it bears. The surfaces of leaves are very efficient agents of radiation, and the tangle which they make offers an amount of heat-radiating area many times as great as that afforded by a surface of bared earth. Moreover, the ground itself can not well cool down to the point where it will wring the moisture out of the air, while the thin membranes of the plants readily become so cooled. Thus vegetation by its own structure provides itself with means whereby it may be in a measure independent of the accidental rainfall. We should also note the fact that the dewfall is a concomitant of cloudless skies. The quantity which is precipitated in a cloudy night is very small, and this for the reason that when the heavens are covered the heat from the earth can not readily fly off into space. Under these conditions the temperature of the air rarely descends low enough to favour the precipitation of dew.

Having noted the process by which in the rain circuit the water leaves the sea and the conditions of distribution when it returns to the earth, we may now trace in more detail the steps in this great round. First, we should take note of the fact that the water after it enters the air may come back to the surface of the earth in either of two ways—directly in the manner of dewfall, or in a longer circuit which leads it through the state of clouds. As yet we are not very well informed as to the law of the cloud-making, but certain features in this picturesque and most important process have been tolerably well ascertained.

Rising upward from the sea, the vapour of water commonly remains transparent and invisible until it attains a considerable height above the surface, where the cooling tends to make it assume again the visible state of cloud particles. The formation of these cloud particles is now believed to depend on the fact that the air is full of small dust motes, exceedingly small bits of matter derived from the many actions which tend to bring comminuted solid matter into the air, as, for instance, the combustion of meteoric stones, which are greatly heated by friction in their swift course through the air, the ejections of volcanoes, the smoke of forest and other fires, etc. These tiny bits, floating in the air, because of their solid nature radiate their heat, cool the air which lies against them, and thereby precipitate the water in the manner of dew, exactly as do the leaves and other structures on the surface of the earth. In fact, dew formation is essentially like cloud formation, except that in the one case the water is gathered on fixed bodies, and in the other on floating objects. Each little dust raft with its cargo of condensed water tends, of course, to fall downward toward the earth's surface, and, except for the winds which may blow upward, does so fall, though with exceeding slowness. Its rate of descent may be only a few feet a day. It was falling before it took on the load of water; it will fall a little more rapidly with the added burden, but even in a still air it might be months or years before it would come to the ground. The reason for this slow descent may not at first sight be plain, though a little consideration will make it so.