Glaciers ()f' Spitzbergen. 315 



ner, and having grounded in twenty-four fathoms of water, had 

 still an elevation of fifty feel above the water. Glaciers of this 

 kind are common in all the Arctic Regions, and it is their break- 

 ing up which gives rise to those mountains of soUd ice wliich 

 abound in these seas. 



As these glaciers, which are analogous to those of the Alps, 

 must be, like them, covered by fragments of rock which have 

 fallen from the adjacent mountains, it is to be supposed that 

 the islands of ice, which are detached from them, must sometimes 

 float away these blocks of rock on the surface of the ocean, and 

 give rise to phenomena of transport, whose examination might 

 furnish interesting terms of comparison for a part of the theory 

 of the phenomenon of erratic blocks. 



But, independently of these ordinary phenomena, we may 

 also ask if some great volcanic eruption, taking place near the 

 pole, may not have put in motion the polar ices loaded with 

 masses of rock, and thus suddenly caused a great dispersion of 

 erratic blocks ? The physical possibility of phenomenon of this 

 nature ought to give particular importance to all the rocks of 

 eruptive origin observable in the frozen zone. Spitzbergen is 

 not devoid of them. I have already mentioned the trap-rocks 

 observed by Professor Keilhau in Stans-Foreland. They occur 

 also in other places. 



To the east of Spitzbergen, says Pennant, there is a very low 

 island, almost opposite to the entrance of the Waygat. It is 

 remarkable from being merely a part of the basaltic chain ob- 

 servable in many places in the northern hemisphere. It is, he 

 adds, a kind of marble of the finest grain, of a deep black, and 

 shining like polished steel, never occurring in beds in the earth, 

 but standing upright in pillars with regular angles, com})osed 

 of a number of parts placed one above the other, with so much 

 exactness, that one would say they were formed by the hand of 

 a skilful architect. Here the pillars are from eighteen inches to 

 thirty inches in diameter, for the most part hexagonal, and fonn- 

 ing a superb pavement or floor of marble. 



According to all appearance, traps or basalts occur there ; 

 and, from some remarks of Captain Scoresby, we might be 

 tempted to ask further if MofFen Island and Low Island, situa- 



