In Canada and the United States we experience the same difficulty as in Europe when we attempt to distinguish between glacial formations of submarine and those of supra-marine origin. In the New World, as in Scotland and England, marine shells of this era have rarely been traced higher than 500 feet above the sea, and 700 feet seems to be the maximum to which at present they are known to ascend. In the same countries, erratic blocks have travelled from north to south, following the same direction as the glacial furrows and striae imprinted almost everywhere on the solid rocks underlying the drift. Their direction rarely deviates more than fifteen degrees east or west of the meridian, so that we can scarcely doubt, in spite of the general dearth of marine shells, that icebergs floating in the sea and often running aground on its rocky bottom were the instruments by which most of the blocks were conveyed to southern latitudes. There are, nevertheless, in the United States, as in Europe, several groups of mountains which have acted as independent centres for the dispersion of erratics, as, for example, the White Mountains, latitude 44 degrees north, the highest of which, Mount Washington, rises to about 6300 feet above the sea; and according to Professor Hitchcock some of the loftiest of the hills of Massachusetts once sent down their glaciers into the surrounding lower country. GREAT SOUTHERN EXTENSION OF TRAINS OF ERRATIC BLOCKS IN BERKSHIRE, MASSACHUSETTS, U.S., LATITUDE 42 DEGREES NORTH. Having treated so fully in this volume of the events of the glacial period, I am unwilling to conclude without laying before the reader the evidence displayed in North America of ice-action in latitudes farther south by about ten degrees than any seen on an equal scale in Europe. This extension southwards of glacial phenomena in regions where there are no snow-covered mountains like the Alps to explain the exception, nor any hills of more than moderate elevation, constitutes a feature of the western as compared to the eastern side of the Atlantic, and must be taken into account when we speculate on the causes of the refrigeration of the northern hemisphere during the Pleistocene period. [Illustration: Figure 50. Erratic Blocks In Berkshire, Massachusetts] (FIGURE 50. MAP SHOWING THE RELATIVE POSITION AND DIRECTION OF SEVEN TRAINS OF ERRATIC BLOCKS IN BERKSHIRE, MASSACHUSETTS, AND IN PART OF THE STATE OF NEW YORK. Distance in a straight line, between the mountain ranges A and C, about eight miles.