THE NARRATIVE OF CHARLES WILKES. 407 
While in this state they rarely exhibit any signs of stratification, 
and some appear to be formed of a soft and porous ice; others are quite 
blue; others, again, show a green tint, and are of hard flinty ice. Large 
ice-islands are seen that retain their tabular tops nearly entire until 
they reach a low latitude, when their dissolution rapidly ensues ; while 
some have lost all resemblance to their original formation, and had 
evidently been overturned. The process of actually rending asunder 
was not witnessed by any of the vessels, although in the Flying-Fish, 
when during fogs they were in close proximity to large ice-islands, they 
inferred from the loud crashing, and the sudden splashing of the sea on 
her, that such occurrences had taken place. As the bergs gradually 
become worn by the abrasion of the sea, they in many cases form large 
overhanging shelves, about 2 or 3 feet above the water, extending out 
10 or 12 feet; the under part of this projecting mass exhibits the 
appearance of a collection of icicles hanging from it. The temperature 
of the water, when among the icebergs, was found below or about the 
freezing poiut. 
T have before spoken of the boulders embedded in the icebergs. All 
those that I had an opportunity of observing, apparently formed a part 
of the nucleus, and were surrounded by extremely compact ice, so that 
they appear to be connected with that portion of the ice that would be 
the last to dissolve, and these boulders would therefore, in all prob- 
ability be carried to the farthest extent of their range befure they were 
let loose or deposited. 
The ice-islands, on being detached from their original place of 
formation by some violent storm, are conveyed to the westward by the 
south-east winds which are prevalent here, and are found, the first 
season after their separation, about seventy miles north of the barrier. 
This was inferred from the observations of both the Vincennes and Por- 
poise, the greatest number having been found about that distance from 
the barrier. That these were recently detached is proven by their 
stratified appearance; while those at a greater distance had lost their 
primitive form, were much worn, and showed many more signs of decay. 
Near the extreme point of the barrier visited, in longitude 97° E., 
latitude 62° 30' S., and where it begins to trend to the westward, vast 
collections of these islands were encountered. From this point they 
must pass to the northward during the next season, partly influenced 
by the current, and partly scattered by the prevailing winds, until they 
reach the sixtieth degree of latitude, when they encounter the easterly 
and north-easterly streams that are known to prevail, which carry them 
rapidly to the north. 
Our data for their actual drift, though not altogether positive, are 
probably the best that can be had, and will go far towards ascertaining 
the velocity of their progress to lower latitudes ; our observations also 
furnish some estimate of the time in which they are formed. On our 
