6 On the Columnar Crystallization of Ice. 
siderable portions of the walks in which these occurred were, 
sufficiently fur our purpose, protected by the shade of the 
contiguous shrubs and trees. 
Of the theory of these beautiful crystallizations, I have 
but little to submit. An idea which was intuitively sug- 
gested by observing the upraised surface of gravel and earth, 
and the interior stratum of columnar crystals, was of growth 
or expansion upward. It was hardly possible, indeed, to con- 
ceive of any other mode of progress. For in one region to 
which my observation was directed—the side of the terrace- 
walk—there was an alinost continuous range of these forma- 
tions, extending for, perhaps, 15 or 20 yards along the walk, 
and of a varying breadth of half-a-yard to a yard, in which the 
thin superficies of earth and gravel had been disrupted and 
elevated by means of the hidden stratum of diversified and 
beautiful crystallizations. . 
The surface of the ground, as it seemed, had, by draining 
and evaporation, been dried probably to the depth of two or 
three tenths of an inch; whilst the materials beneath were 
in the condition of a somewhat compressed sponge, having 
the pores pretty well, not entirely, filled with water. The 
frost of the night acting upon this wetted portion, had frozen 
it to no inconsiderable depth ; and the expansion of the water 
on freezing was, apparently, a chief cause of the phenomena. 
But whilst this might account for the general fact of the 
disruption of the gravelly surface, it could hardly account 
for the formation and exudation of a vertically-arranged bed 
of crystals. Was it that the resistance above being the least, 
the crystallization assumed this direction, and was thrust 
progressively upward as the consolidation and expansion ad- 
vanced below? Or was it that, whilst this least resistance 
might facilitate an upward growth of the columnar fascicule, 
that certain electric relations betwixt the earth and the air 
might dispose to this direction of enlargement and form of 
crystalline development ? 
To the impression that the latter principle had to do with 
the origination of the phenomena described, I much incline ; 
for it is hardly to be doubted, but that, if after a long period 
of damp weather, the electricities of the air and earth should 
