i 3 4 SPITSBERGEN chap, ix 



snow that they seemed like sky-carpets cast upon the earth. 

 We decided to attempt the ascent. Lightly we set forth 

 over the wide snowfield, imagining that it would resemble 

 the others we had crossed. At first all went well ; the snow 

 was pretty good, and we sank only calf-deep into it. After 

 half-an-hour we began to go in knee-deep. Then followed 

 an area of mere flour and pie-crust — a surface too weak 

 to bear our weight, but strong enough to resist the forward 

 pressure of the knee. Advance became absurdly slow, and 

 half the passage was not accomplished. Sometimes the ice- 

 axe, plunged up to its head, reached the hard ice below, 

 but often soundings failed to reveal any firm bottom. We 

 erroneously supposed that this was because the surface 

 snow was deeper than the length of the axe. The winter 

 snow was still present in overwhelming quantity. If the 

 snowfield we expected to cross to Agardh Bay were to 

 prove in this condition, a week would be needed to tra- 

 verse a few miles of it. The reflection did not add to our 

 happiness. At last, when I sank in deeper than usual, an 

 ominous tinkle was heard below. It was caused by icicles 

 falling into a crevasse. This was more than we bargained 

 for. We had not brought a rope. Careful inspection now re- 

 vealed a maze of crevasses ahead, deeply snow-buried, whilst 

 away to the right were the tops of a quantity of seracs. 



We did not yet realise that there were crevasses all round 

 us, and that we had been walking over a whole series of 

 them for a couple of hours. Every time that the axe failed 

 to touch bottom it was because we were over a crevasse, 

 and not because the winter snow was too deep. This I dis- 

 covered many days later, when looking down upon the 

 glacier from the top of Mount Lusitania. All the fresh 

 snow was then melted away from this area, and the net- 

 work of crevasses was revealed in naked complexity. The 

 schrunds were wider than the ice-walls dividing them, and I 



