328 PROBLEM. [CHAP.V. 



set to the westward against a westerly wind, and 

 with fresh squalls from the north a progress 

 south of only four miles. For this, I am at a loss 

 to account in any other manner than by sup- 

 posing the passage between the islands to have 

 been closely blocked up ; so that the southern ice, 

 acted upon by the floods which attend the dis- 

 ruption of the spacious lakes and rivers of the in- 

 terior, had been pressed forward with such force 

 as to resist even the vast body of ice bearing down 

 upon it from the north. The mast-head thermo- 

 meter, which throughout the winter had been in 

 a position eighty-seven feet above the sea now 

 varied so little from those on deck, seldom more 

 than 2° or 3° at the most, that it was taken down 

 and the registering of it discontinued ; and I may 

 take this opportunity of observing that there 

 were very few occasions which on trial were found 

 favourable for flying a kite with an appended 

 self-registering thermometer, during the former 

 portion of the winter when the floe was un- 

 broken, and that the operation was utterly im- 

 practicable in the latter part, when the ice was 

 ground into thousands of peaked and irregular 

 heaps, mounds, and barriers, which defied the 

 activity of the most alert. 



The weather continuing obscured, nothing 

 could be distinguished beyond a mile or two from 

 the ship, and on May L 2d there was no change ex- 



