256 THE ANTARCTIC. 



and is, moreover, to be inferred from the height of 

 icebergs near the walls, which are uniform with it, and 

 are undoubtedly floating masses. How far out into the 

 sea this floating ice is swept it is impossible to tell ; but, 

 considering the enormous horizontal dimensions ot the 

 icebergs, this distance must often be very considerable. 

 This fact throws a brilliant light on the well-known 

 theory of the melting of the icebergs. Sir John Ross's 

 theory that the difference of temperature between the 

 intensely cold upper surface and the much warmer 

 lower surface of the icebergs causes extreme tension, 

 and ultimately in winter the severance of the berg, may 

 turn out to be a true solution of the problem. Another, 



-" * -.— T-— - .- ' ■" - ~"- * :-— ,3^. »...,,» -,.,^- rk _ ■■■ P 



Iceberg, after an original water colour of the Challenger Expedition. 



albeit merely occasional, cause of the severance of 

 especially enormous ice masses may be very violent 

 volcanic eruptions, and the waves generated by them ; 

 it is possible that some occurrences of this kind may be 

 responsible for the colossal ice-drifts of the years 1891 

 to 1896. 



The height of these floating walls of ice, and the 

 icebergs newly detached from them, affords the only 

 means at our disposal for an approximate estimate of 

 the thickness of the Antarctic inland ice. Croll, for 

 example, the eminent student of the glacial periods 

 assumed a maximum thickness of 120,000 to 130,000 

 feet, but sober observation by no means bears out this 



