According to the researches of Dr. J. W. Spencer, one great sheet of water covered most of the great lake region about the close of the ice age; and the lower strands of these inland seas are known to be connected with old marine shore lines. The probable reason why so few sea-shells collected on the glacial drift during such times was because of so much marine life having been exterminated in the high northern latitudes during the frigid age. Therefore, the sea, in the short period of northern submergence, left but few traces on the glacial drift it once flowed. Thus it will be seen that, if the ocean waters were attracted northward through the preponderance of northern ice-sheets, they not only assisted in melting the northern ice, but also served to greatly reduce the waters in the Cape Horn channel, and so largely prevented the independent circulation of the southern ocean, thus furthering a mild climate in the southern hemisphere until the prevailing winds, after the northern ice-sheets were melted, were able to move more of the ocean waters southward than they could move northward, owing to the ocean currents setting southward being less obstructed than the lesser currents setting northward. This tendency of the ocean waters to move southward I have before explained in the preceding pages. But I will say in addition that, on further consideration, it seems that one of the main causes of the waters of the augmented northern oceans moving southward so soon after the melting of the ice from the northern lands was on account of so much water being attracted southward to the great low sea-level east of Cape Horn. This vast low sea-level remained a great area of attraction for the northern seas until so much northern water was moved into the southern ocean as to reduce the seas of the northern hemisphere and augment the southern ocean sufficiently to enlarge the Cape Horn channel, thus causing the extinction of the vast low sea-level that furnished such great attraction for the waters of the more northern latitudes. If the earth-movement hypothesis, so wholly rejected by Professor Geikie, fails to explain the cause or causes of a northern ice age, it seems to be still more inadequate for explaining the occurrence of ice periods extending over both hemispheres. For it is not probable that portions of continents and large islands rose above the snow-line in both temperate zones during the same period of time, and then again obtained their present level with the occurrence of a mild era. Those who maintain that the continents of North America and Europe rose to great elevations during the ice age, in order to prove their assertions, point to the fiords which indent the eastern and western coasts of North America, and also to the fiords of Norway, as having been eroded by streams of ice that flowed along the bottom of such gorges when they were above the sea.