The process of crystallization seems a microcosm of the universe.  Radiata, mollusca, feathers, flowers, ferns, mosses, palms, pines, grain-fields, leaves of cedar, chestnut, elm, acanthus:  these and multitudes of other objects are figured on your frosty window; on sixteen different panes I have counted sixteen patterns strikingly distinct, and it appeared like a show-case for the globe.  What can seem remoter relatives than the star, the starfish, the star-flower, and the starry snow-flake which clings this moment to your sleeve?—­yet some philosophers hold that one day their law of existence will be found precisely the same.  The connection with the primeval star, especially, seems far and fanciful enough, but there are yet unexplored affinities between light and crystallization:  some crystals have a tendency to grow toward the light, and others develop electricity and give out flashes of light during their formation.  Slight foundations for scientific fancies, indeed, but slight is all our knowledge.

More than a hundred different figures of snow-flakes, all regular and kaleidoscopic, have been drawn by Scoresby, Lowe, and Glaisher, and may be found pictured in the encyclopaedias and elsewhere, ranging from the simplest stellar shapes to the most complicated ramifications.  Professor Tyndall, in his delightful book on “The Glaciers of the Alps,” gives drawings of a few of these snow-blossoms, which he watched falling for hours, the whole air being filled with them, and drifts of several inches being accumulated while he watched.  “Let us imagine the eye gifted with microscopic power sufficient to enable it to see the molecules which composed these starry crystals; to observe the solid nucleus formed and floating in the air; to see it drawing towards it its allied atoms, and these arranging themselves as if they moved to music, and ended with rendering that music concrete.”  Thus do the Alpine winds, like Orpheus, build their walls by harmony.

In some of these frost-flowers the rare and delicate blossom of our wild Mitella diphylla is beautifully figured.  Snow-flakes have been also found in the form of regular hexagons and other plane figures, as well as in cylinders and spheres.  As a general rule, the intenser the cold the more perfect the formation, and the most perfect specimens are Arctic or Alpine in their locality.  In this climate the snow seldom falls when the mercury is much below zero; but the slightest atmospheric changes may alter the whole condition of the deposit, and decide whether it shall sparkle like Italian marble, or be dead-white like the statuary marble of Vermont,—­whether it shall be a fine powder which can sift through wherever dust can, or descend in large woolly masses, tossed like mouthfuls to the hungry South.