It melts, and leaves nothing but water as a residuum. Snow, then, is frozen water,--water which existed in the atmosphere in the state of vapour, and which, to speak the language of physicists, has passed from the gaseous state into the liquid, and thence into the solid. If you doubt its identity with water, let a chemist analyse a portion of it for you: he will tell you that it is composed, like distilled water, of hydrogen and oxygen, in the proportion of two parts of the former to one part of the latter. The reader will, of course, understand that we abstract all foreign substances which may accidentally have got mixed up with it. [Illustration: FIG. 5.--A Snowy Landscape.] It was once a wide-spread opinion that snow is favourable to vegetation, on account of the salts which it contains. Analysis, however, gave a negative result; it demonstrated the absence of these salts. Recourse was then had to another hypothesis: it was supposed that the air contained in snow is richer in oxygen than the free air, and that to the action of this gas must be attributed its fertilising property. Another error! The truth really is, that snow maintains the soil which it covers at a perceptibly constant temperature, and that, when thawing, it mellows it by its aqueous infiltrations; so that if, before a fall of snow, the earth has experienced the action of a strong frost capable of killing injurious insects, all the chances will be in favour of a fertile year. * * * * * Snow forms crystals. To observe them clearly, you must examine the snow which falls in very cold and dry weather. It then appears to be a dust composed of little thin plates. Look at the small flake which has fallen on your coat-sleeve; it is isolated; hasten to examine it before it melts, or before other flakes become amalgamated with it. What a graceful star! (Fig. 6, _a_). It is formed of six regular rays. There are others which have only three, four, or five rays. But on inspecting these more closely, you see that many of these rays are broken or abortive, and that, when finally analysed, each star possesses the same number of rays. Why are there continually six rays? Why are there never more nor fewer than this number? One might suspect in nature a peculiar affection for the number _six_; as, for example, in the cells of the bees and the wasps, which form a regular hexagon (Fig. 6, _b_). Why, in the infinity of polygons, has the instinct of these insects only chosen one hexagon? What is the reason for this preference? [Illustration: _a b c_ FIG. 6.]