SCIENTIFIC RESULTS 115 



ity of the Grand Bank durinjj: summer and northward on the Lab- 

 rador shelf, as over other bodies of cold water, is usually much 

 warmer at an altitude of a few hundred feet than it is close to the 

 sui'face of the sea. An iceberg which rises a hundred feet or more 

 will form an obstruction to the wind, tending to stir up the air to a 

 still greater altitude. Around many bergs, therefore, one might 

 expect the air to be much warmer than in the immediate neighbor- 

 hood, and fact sustains theory, for a surprisingly warm blast of air 

 has often been experienced by the ice-patrol ship when passing close 

 by and just to the leeward of an iceberg. On the other hand, under 

 diiferent atmospheric conditions, cold air is experienced in the 

 vicinity of the ice. 



Erosion, washing by the waves and ocean swell is most effective 

 of all destructive agents for North Atlantic bergs. The erosion 

 ]u-ocesses are always at work from the time the ice drifts any appre- 

 ciable distance out from the i)rotection of harbors and bays. As the 

 seas continuallv wash back and forth thev find a crevasse, an irregu- 

 lar shape, or a slight depression, all of which they enlarge. The 

 ceaseless surging of the swell, pouring back and forth through the 

 small channels, soon erodes valleys, which in turn grow larger and 

 larger, until they develoji into the main features of the berg. The 

 inner slopes retreat until only two thin side walls remain, and eventu- 

 ally the valley bottom deepens and disappears below the waves, thus 

 giving the impression of two separate pinnacles or towers of ice. 

 Finally in the later stages of wasting, far south in the Atlantic, 

 calving and fracture outstrip erosion with the falling of the fragile 

 side walls. Calving is then in turn superseded by melting agencies 

 which turn the last ])age in our iceberg's history. It is interesting 

 to note that the attacks of erosion are always directed at the middle 

 of tlie mass of ice that is above water. If the sea sculptures too 

 much away from one side, that rcLnon through loss of weight is 

 lifted out of reach of the waves, which then concentrate on the other 

 side, the midsection always receiving the brunt of the attack. 



During early season around the Grand Bank, February to ]March, 

 icebergs are subjected to a severe washing by waves which some- 

 times dash completely over them, causing a great deal of Avastage. 

 The long fK'ean swell which makes U]) during a gale also sets up 

 stresses and strains in the above-water portion, causing much rend- 

 ing, cracking, and fracture. But only in the small-sized bergs, even 

 in such gales, can one detect a ])ei-cei)tible roll and sway. The ordi- 

 nary run of bergs remain as immovable as the Rock of Gibraltar. 

 Calving begins when the meltinix and erosive processes have set up 

 strains that exceed the structural strenirth of the ice. Prominences 

 and overhanging ledges ^-alve away, sliding down from th? steepest 

 parts of the berg's sides and slopes. Unequal detachments around 

 the edges interfere Avith equilibrium and occasionally initiates calv- 

 ing on a major scale. Tlv.' berg begins to roll slowly and deeplv to 

 and fi'o. and when some Ijidging jji-ominenc^^ swings far away from 

 till' perpendicular thousands of tons of ice rupture to fall, avalanche- 

 like, down to the sea. In the case of many tons of ice. the effect 

 is very curious; it seems to fall much more sIoavIv than really is the 

 case. " Stability is, of course, seriously disturbed, and the berg may 

 again suffer one or more successive calvings of its irregular parts 



