242 
GREENLAND VOYAGE. 
columns that come in contact with it. They 
seemed to have a very slight attachment to the 
surface of ice upon which they were found, and 
little or none to one another. On another mass 
of ice, under similar circumstances, all the snow 
that remained on the surface was converted in¬ 
to small transparent bits of ice. These, accord¬ 
ing to their situation, varied from the size of a 
pea to that of a nutmeg, or even of a walnut. 
They were somewhat globular, but, being like the 
prisms bounded by a varying number of planes, 
they seemed to have derived their form partly from 
the shape and number of the contiguous pieces, 
and partly from a tendency to crystallization. 
Several very perfect figures were observed, which, 
had they been found detached, would have been 
considered as ice-crystals, and their formation the 
sole effect Of crystallization. These were in particu¬ 
lar dodecahedrons, cubes, rhomboids, prisms, and 
pyramids. This conversion of snow into transpa¬ 
rent pieces of ice, under a thawing temperature, 
may serve to account for the parallel lines of air- 
bubbles, that occur in most masses of fresh-water 
ice; also, for the resolution of ice into vertical 
prisms, when it is slowly dissolved in a proper po- • 
sition; for the formation of fields of fresh-water 
ice ; and for the manner in which the icebergs 
receive their enlargement. The enlargement of 
