This page has been validated.
Chap. VIII.]
DISRUPTION OF ICEBERGS.
225
1841
some must occasionally break away. But a little reflection soon furnished an explanation: in summer the temperature of the atmosphere and of the ocean seldom differ more than three or four degrees, the air being generally the colder, but never more than eight or ten degrees: it is therefore probably of rare occurrence that any great disruption should occur at that season of the year, the whole mass being then of so uniform a temperature. But in the winter, when the air is probably forty or fifty degrees below zero, and the sea from twenty-eight to thirty degrees above, the unequal expansion of those parts of the mass exposed to so great a difference of temperature could not fail to produce the separation of large portions. These, impelled by the prevailing southerly winds, drift to the north as soon as the winter breaks up, and are met with abundantly in the lower latitudes, where they rapidly melt away and break in pieces. We have often in the arctic regions witnessed the astonishing effects of a sudden change of temperature during the winter season, causing great rents and fissures of many miles extent; especially on the fresh-water lakes of those regions, where the ice being perfectly transparent, affords better means of observing the effects produced: a fall of thirty or forty degrees of the thermometer immediately occasions large cracks, traversing the whole extent of the lake in every variety of direction, and attended with frequent, loud explosions; some of the cracks opening in places several inches by the contraction of the upper surface in
