But these organic remains are not the only evidence of the geological winter.  There are a number of phenomena indicating that during this period two vast caps of ice stretched from the Northern pole southward and from the Southern pole northward, extending in each case far toward the Equator,—­and that ice-fields, such as now spread over the Arctics, covered a great part of the Temperate Zones, while the line of perpetual ice and snow in the tropical mountain-ranges descended far below its present limits.  As the explanation of these facts has been drawn from the study of glacial action, I shall devote this and subsequent articles to some account of glaciers and of the phenomena connected with them.

The first essential condition for the formation of glaciers in mountain-ranges is the shape of their valleys.  Glaciers are by no means in proportion to the height and extent of mountains.  There are many mountain-chains as high or higher than the Alps, which can boast of but few and small glaciers, if, indeed, they have any.  In the Andes, the Rocky Mountains, the Pyrenees, the Caucasus, the few glaciers remaining from the great ice-period are insignificant in size.  The volcanic, cone-like shape of the Andes gives, indeed, but little chance for the formation of glaciers, though their summits are capped with snow.  The glaciers of the Rocky Mountains have been little explored, but it is known that they are by no means extensive.  In the Pyrenees there is but one great glacier, though the height of these mountains is such, that, were the shape of their valleys favorable to the accumulation of snow, they might present beautiful glaciers.  In the Tyrol, on the contrary, as well as in Norway and Sweden, we find glaciers almost as fine as those of Switzerland, in mountain-ranges much lower than either of the above-named chains.  But they are of diversified forms, and have valleys widening upward on the slope of long crests.  The glaciers on the Caucasus are very small in proportion to the height of the range; but on the northern side of the Himalaya there are large and beautiful ones, while the southern slope is almost destitute of them.  Spitzbergen and Greenland are famous for their extensive glaciers, coming down to the sea-shore, where huge masses of ice, many hundred feet in thickness, break off and float away into the ocean as icebergs.  At the Aletsch in Switzerland, where a little lake lies in a deep cup between the mountains, with the glacier coming down to its brink, we have these Arctic phenomena on a small scale; a miniature iceberg may often be seen to break off from the edge of the larger mass, and float out upon the surface of the water.  Icebergs were first traced back to their true origin by the nature of the land-ice of which they are always composed, and which is quite distinct in structure and consistency from the marine ice produced by frozen sea-water, and called “ice-flow” by the Arctic explorers, as well as from the pond or river ice, resulting from the simple congelation of fresh water.