Chap, x.] WEATHERING OF ICEBERGS. 205 



circumstances is complete. Caves, cliffs, pinnacle-like outliers, 

 and a shore platform at the base of the cliffs, are formed in a 

 closely similar manner in each case. 



In order that a horizontal platform of any wide extent should 

 be formed beneath the water, it is necessary that the berg 

 should float at almost exactly the same level for a very long 

 period. I do not properly understand how this occurs. Each 

 time that a mass of ice falls from the undermined cliff in 

 order that the equilibrium should be maintained, it is neces- 

 sary that nine times that bulk of ice should be removed from 

 the base. 



No doubt portions of the platforms below water are con- 

 stantly being split off by the upward pressure and floating to 

 the surface as " calves." The formation of a large platform 

 under water must, however, depend on such a " calving " not 

 taking place, unless on sides of the berg other than that on 

 which the platform is formed. Nevertheless, by some means 

 or other, either by melting or calving, a very uniform wasting 

 of the berg below water must take place in order to form a 

 platform. It cannot be supposed that the amount of snow 

 which falls on the berg when set free can be sufficient to 

 balance the loss by the action of the sea. 



There must be a reason why the bergs which thus become 

 two-storied have their lower storey commonly, as in the example 

 figured here in the text, only at one of their ends. Probably a 

 certain amount of lower platform existed all round this berg 

 when it first rose, but this was cut away on all the sides where 

 it was narrow, by being undermined by the waves. The line 

 of the main upper cliff was thus soon reached on these sides, 

 and this cliff was then itself further undermined, so that, as 

 shown in the sketch, the old wash-line was obliterated, and 

 remained only at the base of that cliff which was protected by 

 the still remaining secondary platform. 



The greater undermining of bergs at one side may, no 

 doubt, be due to their taking up, from the shape of their parts 

 exposed above water and the relation of these parts in position 

 to the form of the parts below water, a particular direction with 

 regard to the wind, and maintaining this so that one particular 

 side is usually the windward one, and therefore most battered 

 by the waves. 



It seems far more difficult to explain how it occurs that ^(^J^^ 

 bergs suddenly rise to a considerable height further out of 

 water than that at which they have floated before. Such a >if^ ^u ^ 

 sudden rise must necessarily be supposed in order to account 

 for the two-storied form. 



In order that, in the case of the berg figured for instance, a 



