FORMATION OF ICEBERGS. 
241 
tcnioon, wo stood to the soutli-castward, until we 
were stopped by a chain of floes, through which 
wc could not discover a passage. In the even¬ 
ing the weather moderated, and the fog increased 
in density to that degree, that we had no safety 
but in having the ships moored to the ice: the 
three ships were therefore made fast to a floe. 
The influence of a temperature some degrees 
above the freezing point, with the action of wet¬ 
ting fogs and a little rain, had formed lakes upon 
the surface of the sheet of ice to which we moor¬ 
ed, so large, that, while the fog continued, we 
could not see across them. Where the snow was 
not wholly dissolved, there was another effect of 
this state of the weather, deserving attention, in¬ 
asmuch as it casts a considerable light upon the 
mode of formation of icebergs, and of the enlarge¬ 
ment of ice-fields. The upper stratum of the 
floe, which had originally consisted of loose light 
snow, was now much reduced in thickness, and 
formed, by the infiltration of the dissolving sur¬ 
face, into vertical needles, and irregular prisms of 
transparent ice. These prisms, upon a sheet of 
ice, formerly examined, that was in a similar 
state, were five or six inches in length, and seem¬ 
ed to have given form one to another, the num¬ 
ber of sides iu each, like what occurs in pillars of 
basalt, being equal to the number of prisms or 
Q 
