38 DR. NILS EKHOLM ON THE METEOROLOGICAL [Feb. 1902, 
at the beginning of the hot season. This season is short, but 
relatively warm with frequent rainfall. In Southern Greenland, 
on the contrary, a great deal of snow falls during the whole winter,. 
owing to frequent cyclones passing across or south of that area 
during that season. Moreover, this cyclonic state of weather 
persists in summer as well: the sky is generally cloudy, and the 
effect of the summer sun is counteracted thereby, and by the 
cold and damp winds coming in from the ice-filled sea surrounding: 
the Greenland coast. Of course, the height above sea-level and 
the thick ice-sheet already covering that land also contribute to 
maintain its state of glaciation; and it is possible that the present 
climatic conditions of this region are such that, if the ice-sheet 
were now removed, it would not return. It is remarkable, more- 
over, that the northernmost part of Greenland and the surrounding 
islands are not completely ice-covered, despite the much lower mean 
annual temperature, a fact which is explained by the insignificance 
of the snowfall there. 
Such instances might be multiplied, and from all of them the 
conclusion may be drawn that a region where a permanent 
anticyclone prevails during the winter, cannot be 
covered with a permanent ice-sheet, however low the 
winter-temperature may fall, unless possibly the summer is not 
only cold but accompanied by an abundant snowfall, so that more 
snow falls in that season than is thawed away. This latter state 
of weather might perhaps be found somewhere in the unexplored 
Polar regions—the South Polar cap for instance—but certainly not 
in any known country. 
The most favourable climate for land-glaciation is to be found 
in the great Southern Ocean. In Cape Horn and Kerguelen Island 
the glaciers come down to the sea, and the snow-line is comparatively 
very low, so that it probably coincides nearly with the isotherm of 
32° Fahr., despite the high latitude (49° to 55°S8.). Moreover, 
the cyclonic state of weather prevails there all the year round, and 
the summer is relatively very cool, 
If we now consider the area of glaciation in North America and 
Europe during the Great Ice-Age, we find that it nearly coin- 
cides with the area now crossed by the most regularly 
frequented tracks of storms or cyclones. More strictly 
speaking, those regions are the site of moving or temporary cyclones 
and anticyclones, which alternate. Regions occupied by stationary 
anticyclones during the greater part of the year have little or no 
permanent ice-sheet, even though the mean annual temperature be 
much below the freezing-point. Such are several great plains and 
highlands in Asia and America. 
We find also that the ditference between the mean annual tem- 
peratures at the same latitude in Europe and North America probably 
was nearly the same during the Great Ice-Age as it is now, for the 
southern limit of the ice-sheet coincides approximately with the mean 
annual isotherm of 55° Fahr. in both continents (or 50° Fahr., if 
