for ascertaining the Deviation of the Magnetic Needle. 367 



" In order to follow up the observations on this phenomenon 

 on some other day, I sent a boat to fix a flag upon the ice, by 

 way of marking the spot, but the margin was so broken up 

 that it was impracticable to land upon it; a light buoy was 

 therefore moored for the same purpose, though with little 

 chance of retaining its station, on account of the depth of wa- 

 ter. During the remainder of the night, when the wind and 

 weather obliged us to keep more to the northward, the com- 

 passes were not thus influenced *. 



" The weather cleared up in the morning of the 2d. We 

 found that a strip of ice about half a mile in width had been 

 lately separated from the fixed ice, but this to our impatience 

 appeared like a drop of water in the ocean; considerable 

 streams and patches were also drifting along the margin du- 

 ring the day, and we were employed in breaking through them 

 in order to make fast to the floe, the weather being unfavour- 

 able for keeping under way. In the evening we secured the 

 ships to the ice, being in twenty-three fathoms, at the distance 

 of two miles to the westward of Tern Island. For several hours 

 in the course of this day there was something in the atmo- 

 sphere which distorted objects into very curious shapes. The 

 principal feature in this phenomenon was a constant waving 

 tremulous motion near the horizon, causing the whole body of 

 ice to appear at times as if turning round, and making one al- 

 most giddy to look steadfastly at it. The distant land was 

 sometimes flattened down so as to appear like a single thick 

 black line upon the horizon ; then again it would assume a 

 shape of this kind — 



when its real ouriine, when not thus distorted, was this, — 



" The trenuilous appearance is, in a greater or less degree, 

 a very common phenomenon in the Polar seas. Such, indeed, 

 is the frctjuent occurrence of extraordinary and variable ter- 

 restrial refraction, and the consequent uncertainty with respect 

 to tlie dip of the horizon, that observations made by the hori- 



• The s[)ots near whicli this local attraction was found, arc designated 

 on the chart by this mark, 0. 



zon 



