its pole at a point located somewhere north of Hudson Bay. One effect of this oblique rotation is to establish in northern latitudes between 50 and GO degrees two areas of low barometer and one of high barometer, a disposition which is in strong contrast to the condition in the southern hemisphere, there being a zone of high pressure along the parallel of 35 degrees, south latitude, with decreasing pressure thence toward the Antarctic. These lows and highs differ from those which are familiar as features of daily weather maps, in that they are nearly stationary. The well-known migrant centers converge toward and run into the great fixed centers. The two permanent lows are situated one across the North Atlantic, from Hudson Bay to Scandinavia, the other in the North Pacific, from Japan to southern Alaska. They are centers of inflowing ascending air currents, and are, therefore, characterized by great precipitation. The region of maximum glaciation at the present time lies between them; one conspicuous development occurring in Greenland in the northwest quarter of the Atlantic low, another lying in Alaska in the northeast quarter of the Pacific low. By an analysis of the winds, it is shown that both Greenland and Alaska lie to leeward of the prevailing currents where they pass ashore. They are not necessarily the provinces of maximum precipitation, rain and snow both considered, but they are areas of copious snowfall, with low annual mean temperatures.
The northern high lies nearly midway between the two lows over Siberia. In contrast to them, it is a center of descending outwardflowing currents, marked by slight precipitation, and it is not now, nor was it in Pleistocene time, a scene of glacial development.
The centers of Pleistocene glaciation were so arranged with reference to the glacial regions of to-day that they would be determined by the oblique circulation and distribution of areas of low pressure, if existing conditions were intensified. An adequate occasion of intensification is found in the thermal transparency of the atmosphere, resulting from depletion of carbon dioxide, and thus the localization of Pleistocene ice sheets is explained in a manner consistent with the major hypothesis of the cause of glaciation.
Chamberlin's hypothesis is framed on an atmospheric basis, but the efficiency of the agencies which it postulates depends upon geographic conditions, upon distribution of land and sea and average heights of continents. The geography of the earth in the closing epochs of the Paleozoic era is known only in its broadest outlines, and they are but vaguely traced. With such imperfect data it is impracticable to explain satisfactorily the extraordinary phenomena of glaciation at that date in intimate association with the development of coal beds and extending within the tropics. Nevertheless, to carry out his purpose of developing a working hypothesis, the author feels obliged to arrange
