Chap. X.] STRUCTURE OF ICEBERGS. 



203 



exposed part is of lighter, less compact ice, and often further 

 lightened by excavation of caves, and presence of crevasses. 



So large a proportion of the bergs being required to be 

 immersed, in order that the bergs broken off from the parent 

 ice masses should float in stable equiHbrium, with their surfaces 

 originally uppermost maintained still in that position, it is 

 necessary that the pieces thus breaking off, supposing their 

 upper surfaces to be square, should be at least as wide as they 

 are thick. If this were not the case, if the density of the ice 

 masses were uniform, the bergs would necessarily topple imme- 

 diately they broke free, and this fact would be shown liy their 

 stratification being vertical to their plane of flotation. This, 

 however, seems never, as far as I could judge from the bergs 

 I saw, to occur. Tilting only takes place after bergs have been 

 long weathered. The bergs seem nearly always to be of large 

 area in proportion to their thickness, and to maintain their 

 original balance for very long periods. No doubt the much 

 greater density of the ice composing the lower portions of the 

 bergs tends to keep them in their original position. 



The waves, partly no doubt because of the water at the very 

 surface being warmed by the sun, and partly no doubt by heat 

 resulting from their motion, cut a wash-line all round the bergs, 

 which appears as a concave groove-like channel with a polished 

 inner surface, just at the water-level. 



When bergs rise to a higher level, or tilt, these wash-lines 

 remain marked on them, as straight polished streaks, visible 

 from a great distance, giving evidence of the former lines of 

 flotation of the bergs. Sometimes, several ancient wash-lines 

 are visible on one berg, and where the cliff surfaces on which 

 they are scored are protected at their base from the waves by 

 secondary cliffs or projections, they may remain intact for very 

 long periods. 



The wash-lines being hollowed out at the bases of the cliff's, 

 these latter soon overhang, and large masses split off along the 

 lines of joint and cleavage, and fall. The masses evidently 

 split off tolerably evenly from the whole height of the cliffs, 

 for these are nearly always, when thus still w^ater-worn at their 

 bases, perpendicular, and on our firing a shot at a berg cliff", 

 the ice split off in this manner from the whole height of the 

 cliff". 



When there are crevasses at the level of the wash-line 

 leading into the ice from it, the wash of the waves hollows 

 out caverns, w^hich resemble in general form caves cut in the 

 same manner by waves on coast-lines, and have their mouths 

 wide at the levels of the wash-lines. 



The presence of caves is a proof that a berg has floated at 



