THE SWEDISH ANTARCTIC EXPEDITION. 471 



long- continued, and bring the wtirniest weather. It is these differences 

 thatbringoutanotheroharacteristic of the climate — its great variability. 

 The variations from day to day are in winter time greater, as in most 

 other regions of the world, and, as far as our short experience goes, 

 it is quite probable that this also holds good from year to year. 

 Though the mean temperature of these two winters is nearly the same, 

 the difference between the two is very great. In the first year the 

 southwest storms were absolutely overwhelming, alternating wnth 

 periods of calm, warmer weather. In the second year the calm periods 

 were generally colder, and, at least during the first part of the winter, 

 very common and long continued, and to this cause during the second 

 half were due long periods of warm northerly Avinds. 



In the closest connection with that state of the weather stand the ice 

 conditions. After the winter of 1902, with its southwest storms, came 

 a summer that was not only the coldest hitherto known in an}- region 

 of the world, but also, and that to our l)ad luck, marked by an accu- 

 nudation of ice such as never has been seen in that region. It was in 

 the l)attle with that ice that the Antarctic was lost, but it nmy be said 

 that we on the station had really no reason for uneasiness, as never, 

 except perhaps for two or three days, was the sea in our neighborhood 

 so free from ice as to render the arrival of the ship probable. On the 

 other hand, even in the middle of the winter, the north and northwest 

 winds would cause large openings in the ice, and after the strong gales 

 in August to October, 1903, the second summer started with an almost 

 clear sea, and probably this year is of the same type as the summer of 

 1893-1, when Larsen made his well-known voyage south. 



During our stay 1 brought together rich material from investiga- 

 tions on the ice, both sea ice and land ice, and especially that typical 

 Antarctic ice cap of Snow Hill — its temperature compared with that of 

 the sea ice and the soil, its movement, surface structure, and stratifica- 

 tion. Interesting is the great accumulation of snow during the sum- 

 mer 1902-3, which is important for explaining the formation of such 

 ice caps and their great extension in those regions; a few 3-ears such 

 as this would Cover the whole region with snow. 



Because of this accumulation the land ice forms at all seasons an easy 

 traveling road, and only where there are large crevasses it might be 

 difficult to pass in summer time. On the contrary, just as in the north, 

 so also here the sea ice js during the summer to a great extent covered 

 l)y water, making the traveling very difficult. But even if this had 

 not been the case we could not have used the first summer for distant 

 sledge traveling, as we had then to wait for the return of the Antcwctic, 

 and later to provide ourselves with the supplies necessar}' for another 

 wintering. All sledge work, therefore, was mainly during the two 

 springs. Its results have been the survey of the coast with its outly- 

 ing islands from the end of Louis Philippe Land to our southernmost 



S5I 1!)03 31 



