f the
prismatic structure, with limpid crockets and finials. The pipes of ice
which formed a network on the walls were in some cases so exquisitely
clear, that we could not be sure of their existence without touching
them; and in other cases a sheet 4 or 6 inches thick was found to be no
obstruction to our view of the rock on which it was formed. In one of
the domes we had only one candle, and the bearer of this after a time
contrived to let it fall, leaving us standing with our heads in perfect
darkness; while the indistinct light which strayed about our feet showed
faintly a circle of icicles, hanging from the lower part of the dome,
the fringe, as it were, of our rocky petticoats.

In one of the lower parts of the cave, where darkness prevailed, and
locomotion was only possible on the lowest reptile principles, M.
announced that she could see clear through the ice-floor, as if there
were nothing between her and the rock below. I ventured to doubt this,
for there was an air of immense thickness about the whole ice; and as
soon as A. and I had succeeded in grovelling across the intervening
space, and converged upon her, we found that the appearance she had
observed was due to a most perfect reflection of the roof, as shown by
the candles we carried, which may give some idea of the character of the
ice. We did not care to study this effect for any very prolonged time,
inasmuch as we were obliged meanwhile to stow away the length of our
legs on a part of the ice which was thinly covered with water,--one
result of its proximity to the arch communicating with the smallest pit.

It has been said that the whole ice-floor sloped slightly towards one
side of the cave, the slope becoming rather more steep near the edge.[52]
Clearly, ever so slight a slope would be sufficiently embarrassing, when
the surface was so perfectly smooth and slippery; and this added much to
the difficulty of walking in a bent attitude. On coming out of one of
the domes, I tried progression on all-fours--threes, rather, for the
candle occupied one hand,--and I cannot recommend that method, owing to
the impossibility of putting on the break. The pace ultimately acquired
is greater than is pleasant, and the roof is too near the floor to allow
of any successful attempt to bring things to an end by the reassumption
of a biped character.

We placed a thermometer in the line of greatest current, and another in
a still part of the cave. The memorandum is lo