354 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [June 9,
they do now from the glaciers of Greenland. We have plenty of
evidence of iceberg action during the glacial period.
I believe I have shown that glaciation depends chiefly on a cold summer, but partly also on an abundant snow-fall. I have now to show that a period of cold summers, caused as I have explained, must be also one of snowy winters; so that the two conditions favourable to glaciation will occur together.
During the mild winter of the glaciated hemisphere, there is a hot summer in the opposite one. Increase of temperature promotes increase of evaporation in a much greater ratio than that of the increase of temperature; and increased evaporation in the summer hemisphere will produce increased snow-fall in the winter one. We know that at present the vapour raised in one hemisphere is to a great extent precipitated in the other; for, were it not so, the southern hemisphere, by reason of its greater extent of ocean surface, would have a rainier climate than the northern; and such does not appear to be the case on the whole. Besides, during a glacial period, the atmospheric circulation between the two hemispheres, at the time of the earth's minimum distance from the sun (which on my theory was in the winter of the glaciated hemisphere), must be more active than ever it is now; for when the earth, at either solstice, was nearer the sun than is ever the case now, and the difference of temperature between the two hemispheres consequently at its greatest possible amount, this would produce a very active circulation of atmospheric currents between the two hemispheres, which would involve the deposition as rain or snow in the winter hemisphere of a great part of the moisture evaporated in the summer one.
Some interesting collateral observations remain to be made. I believe that the foregoing remarks furnish the explanation of a very remarkable fact in physical geography. Not many coasts in the world are cut up into fiords; and all, or nearly all, that are so, are western coasts in high latitudes. The fiord formation is found in North-western Europe, including Norway, the west of Scotland, and the west of Ireland, in North America from Vancouver's Island northwards, and in South America from the island of Chiloe southwards. From Vancouver's Island to Chiloe is an immense stretch of nearly straight coast-line; but at those limits its character changes quite abruptly. The transition from straight to indented coast-lines coincides pretty nearly with that from dry to moist climates; and the change from the dry climate of Chile to the moist one of western Patagonia is accompanied, as we might expect, by a great depression in the snow-line on the Andes. (See tabular statement above.) It is now generally believed that the prevalence of lakes in high latitudes is, in some way, a result of glacial action; it can scarcely be doubted that this is equally true of fiords; and the coasts I have mentioned are those on which glacial action must necessarily be the most energetic, because west coasts, in high latitudes, are exposed to west winds (Maury's "countertrades"), which deposit on the mountains in snow the moisture they have taken up from the sea.
Geologists appear to be now agreed that the carboniferous climate
