POLAR GEOGRAPHY BROWN 351 



The existence of an Antarctic continent is still based on circum- 

 stantial evidence, and until more than some 5,000 miles of its coast 

 line, or only about 35 per cent of the total length, are known, direct 

 evidence of Antarctica will be lacking. It is not a little remarkable 

 that all the exploration of the twentieth century has merely modi- 

 fied the probable outline of that continent as it was predicted by 

 Sir John Murray in 1886. He had little but the reports of Ross, 

 d'Urville, Wilkes, a few sealers, and the Challenge^' to go on, and. 

 mainly on circumstantial evidence, he built his Antarctic continent. 

 The one considerable change in that map has been the curtailment of 

 the Weddell Sea and the removal of its southern extremity some 4° 

 north of Murray's position in latitude 82° South. But that south- 

 ward prolongation of the Weddell Sea and Atlantic Ocean at the 

 expense of Antarctica was based solely on Ross's mistaken sounding 

 of 4,000 fathoms, no bottom, in latitude 68° 48' South, longitude 12° 

 20' West. 



Most of the Antarctic " lands," and certainly nearly all those that 

 may be classed as key positions to the coast line of Antarctica, date 

 from last century, some of them from 100 years ago. Coats Land, 

 Wilhelm Land, and Oates Land are among the few exceptions. 

 Enderby Land, the one certain or nearly certain land in over 3,000 

 miles of hypothetical coast line, has never been seen or seriously 

 searched for since Biscoe found it in 1831. It should be the base 

 of an expedition that is prepared to work westward. Heavy ice 

 congestion so far found by all vessels that have tried to push south 

 between Enderby Land and Coats Land suggests that this stretch 

 of coast line will have to be put in by sledge journeys along the 

 edge of the ice cap. The western shores of the Weddell Sea are 

 another ice-girt region which no ship has been able to penetrate, 

 a region of dangerous ice pressure. Here, too, the advance must 

 be by land journey, but it should be relatively simple, since accessible 

 bases are known in Oscar Land and adjoining parts of Graham 

 Land. Lastly, there is the great gap south of the Pacific between 

 Charcot and Edward Lands, which leaves ample scope for an attack 

 from both ends. A minor problem in the outline of Antarctica for 

 an expedition based on Edward Land is the determination of the 

 eastern side of the Ross Sea and the elucidation of Amundsen's 

 sighting of land to the south of Edward Land, the appearance of 

 land which he called Carmen Land. 



But even more important than the discovery of the " missing " 

 stretches of the Antarctic coast line — a mere matter of descriptive 

 geography — is the explanation of the structure of the continent 

 and its former connections w^ith other lands of the Southern Hemi- 

 sphere. The problem is made more difficult of solution by the im- 

 mense covering of ice that completely hides the underlying rock in 



