278 THE BORDERLAND OF SCIENCE. 



So far, indeed, as the geographical evidence extends, 

 it seems probable that there exists within the Antarctic 

 circle an elevated region bearing somewhat the same 

 relation to the great promontories terminated by Cape 

 Horn and the Cape of Grood Hope, as well as to the 

 relatively elevated region indicated by the islands 

 south and south-east of Australia, which the Hindoo 

 Koosh bears to the great mountain-ranges of Asia. 

 We seem to have in the Antarctic high lands the great 

 central elevation whence three great lines of elevation 

 extend. That the great mountain range which forms 

 the backbone of South America is continued under 

 water, rising again in the South Shetland Isles and 

 Graham's Land, would indeed seem altogether pro- 

 bable; and it may be remarked as a coincidence of 

 some importance that the mountains seen by Eoss on 

 the other side of the Antarctic Circle Mounts Sabine, 

 Crozier, Erebus, and Eoss lie in a chain tending in 

 the same direction. But although we might thus be 

 led to regard the Antarctic regions as forming a great 

 central region of elevation, it by no means follows that 

 this region is of the nature of a table land. 



Meteorological considerations have been urged by 

 Maury for the theory of Antarctic lands in large masses, 

 'relieved by high mountains and lofty peaks.' He 

 considers that it is to such mountains (performing the 



stress laid by Maury on the observed relation seems to me, indeed, as 

 unwarranted as that laid by Humboldt on the fact that the great 

 southerly projections of the land lie nearly in the same longitude as the 

 great northerly projections. 



