ANTARCTIC ADVENTURE AND RESEARCH 



thick, to test its strength. The ice undulated Hke rub- 

 ber, but did not break. 



Later ice crystals in the sea develop as vertical plates 

 having a somewhat triangular form. Often there is a 

 layer of square plates about two inches thick above a 

 layer of more or less vertical plates which is some 

 eight or ten inches in thickness. It is these wedges of 

 vertical plates which give rise to the fibrous appearance 

 of new sea-ice, and this is partly due to the layers of 

 air entangled between the crystals. A very important 

 result of this structure is that sea-ice tends rather 

 readily to break into cakes along these vertical planes of 

 weakness. On the under side of older ice there is a 

 great tendency for melting to occur irregularly. Holes 

 one or two inches deep penetrate the ice, primarily along 

 these planes of weakness, where the ice is particularly 

 saline. 



David and Priestley recorded the history of the sea- 

 ice in MacMurdo Sound off Cape Royds in 1908. At 

 the end of February Pancake Ice developed from 

 countless ice crystals which had given the sea the ap- 

 pearance of paraffin wax. * 'Little by little these crystals 

 felted themselves together into small cakes, which 

 jostled by the gentle breezes continually collided along 

 their growing edges. The latter being very flexible 

 became gradually turned up to form raised rims, and 

 the whole pancake became slightly saucer shaped." 

 Often a large pancake would be formed from a lot of 

 little pancakes jammed together. 



Very curious structures called '*ice flowers" develop 

 during a sudden fall of temperature, when the sea-ice 



164 



