424 THE CLIMATE OP THE PLEISTOCENE EPOCH. [Aug. I9OI. 



western quarter. To have given the latter country a milder climate 

 at that season, the prevalence of south-easterly winds from the 

 North Pacific would have been necessary, and these could only 

 have occurred when the isobaric lines ran south-east and north- 

 west. 



The present meteorological conditions of this region during 

 winter are represented statistically by an elliptical cyclone, lying, 

 to the south of Behring Strait, with its longest axis ranging east 

 and west, wedged in between the high-pressure systems then 

 prevalent over Siberia and North America respectively ^ (fig. 17, 

 p. 450), that of the latter being confluent with the anticyclone 

 of the North Pacific. Such a state of things, however, could 

 hardly have obtained during the existence of an ice-sheet over 

 North America. The centre of the North American anticyclone 

 may then have been farther north, or north-east, in winter (fig. 18, 

 p. 452), while the oceanic area now occupied by the Pacific anti- 

 cyclone at that season, having a temperature warmer than that over 

 the ice-sheet, would have been prevalently cyclonic. 



We have seen above that to produce westerly or south-westerly 

 winds over the Sahara, the anticyclonic system of the Northern 

 Atlantic, a part of which now lies statistically oft' the western coast 

 of North Africa, must have moved towards the south. I have 

 ventured to show, hypothetically (in fig. 18), how similarly a 

 displacement of the North Pacific anticyclone may have occurred 

 in winter, during the maximum glaciation of North America. 



The position of the low-pressure system of the North Pacific at 

 that season varies from day to day, in accordance with the movements 

 of the anticyclones of Siberia and of North America respectively. 

 When the centre of the Asiatic anticyclone advances towards Behring 

 Strait, and that of North America retires eastward, or conversely, 

 the cyclone moves with them eastward or westward as the case 

 may be. The alignment of the latter changes, moreover, with the 

 varying movements of the anticyclones ; sometimes it inclines to- 

 wards the north-west, at others towards the north-east ; occasion- 

 ally it even lies north and south. Fig. 8, p. 425, copied from the 

 Daily Meteorological Bulletins of the IJnited States War Depart- 

 ment, shows an example of the former kind. On February 13th, 

 1884, the centre of the high-pressure system of North America, 

 lying farther north and west than its statistical position in winter 

 (compare fig. 4, p. 414), the cyclone was in consequence situated 

 near the Asiatic coast, with a south-easterly and north-westerly 

 alignment, causing mild south-easterly winds over the Behring 

 Strait region. A temperature of 39° Fahr. was recorded on that 

 date at a point south of the Alaskan peninsula, and 4° Fahr. on the 

 eastern side of the Strait. Ten days later, on February 23rd, these 

 conditions were reversed, as shown in fig. 9, p. 427 : the centre of 



^ The rain now caused by the oceanic winds of this low-pressui-e system does 

 not reach Northern Siberia, but falls on the Pacific coasts of North America, 

 and in Kamtchatka ; a shifting of the centre of the cyclone, and a change in its 

 alignment, would, however, have carried the moisture farther north-westward. 



