CHAP. V.] DISUNION OF FLOE. 2^7 



part of this was, that the opening did not con- 

 tinue along the crack to the edge of the floe, 

 but took place only in a particular spot, as if the 

 ice had been scooped out from the interval so 

 created. At 10 h p. m. it closed a little, and imme- 

 diately a mound of ice was raised at the western 

 termination, adjacent to the starboard quarter ; 

 this had probably eased the pressure from the 

 ship, since little more was felt on board than a 

 few squeezes and an occasional concussion. 

 Meanwhile, the body of the ice outside the 

 ramparts, which had been for some time at rest, 

 began to be again disturbed. February 19th 

 arrived, and we looked with some anxiety to 

 the approach of the same hours, during which, 

 on the preceding morning, we had suffered so 

 much annoyance, fully expecting a repetition ; 

 but, fortunately the chief pressure fell on the new- 

 ly-opened crack, extending however occasionally 

 as far as the ship, and ending by heeling her over 

 to starboard. In fact, when the sun rose, it was 

 found she had forged about eight inches from 

 the bank of the dock on the larboard side, where 

 bay or young ice had closed up the interval. 

 About 10 h a. m. another crack close ahead, 

 or rather on the starboard bow, opened, and en- 

 larged others near it, which in their turns pro- 

 duced fresh fractures, thus gradually separating 

 the floe into its original constituent parts. In 



q 2 



