January 25, 1906] 



NA TURE 



299 



projection eastward of Ross Island and the peaks 

 called by Ross the Parry Mountains, which were 

 all regarded as part of the mainland, suggested that 

 behind Ross's ice-sheet the mountains of Victoria 

 Land trend to the east. Captain Scott tells us that the 

 Parrv Mountains do not exist; but a group of islands, 

 White Island, Black Island, and Minna Bluff, occur 

 in almost the same relation to Mounts Erebus and 

 Terror as Ross marked his Parry Mountains. 

 Behind this archipelago there is a great bight, the 

 land first trending somewhat westward, and bending 

 to the east after about 8i° S. Thence, so far as 

 Captain Scott could see, the land has an average 

 trend to S.S.E. from Mount Wharton to the most 

 distant southern peak observed beyond Mount Long- 

 staff. Captain Scott concludes that the mountains 

 continue in the same direction to Graham's Land. 



slope southward to the Pole and across it northward 

 to the Atlantic." This view is fully supported by 

 Captain Scott's opinion that, according to his view 

 of the course of the main mountain chain, " the 

 geographical pole would be situated 200 miles more 

 from it, and on the high ice-plateau which must 

 continue behind " (vol. ii. p. 427). 



The lands, problematical and proved, to the 

 south of the Pacific probably belong to one of those 

 island festoons, which are still so characteristic, and 

 apparently once occurred along all the Pacific coasts. 

 The only objection to placing the main coastline of 

 the South Pacific south-west of Ross's ice-sheet, in- 

 stead of along a line north-eastward through King 

 Edward Land, is the ice-barrier, on Captain Scott's 

 theorv of its formation. If it be land ice, and be 

 flowing rapidly northward, a mile in three years, it 



it 1 



1 the "Ross's Ice-Sheet," sho» 



; the snowy 



There is nothing a priori improbable in the con- 

 nection of Victoria Land and Graham's Land along 

 this line ; for coasts of the Pacific type are character- 

 istically straight for long distances, and have broad 

 open curves rather than sudden bends. 



This does not affect the essential part of my sugges- 

 tion, made in Nature, April 25, 1901, " that we may 

 expect the greatest elevations on the Antarctic Lands 

 will lie along the Graham's Land-Victoria Land line, 

 and will be near the sea. To the south of the main 

 mountain range there may be an undulating ice- 

 covered region descending slowly across the Pole to 

 the shore of the Weddell Sea. The main ice drainage 

 would then be not from the Pole radially in all direc- 

 tions ; the ice-shed would run along the Pacific Shore 

 with a short steep northern face and a long gradual 



NO. 189I, VOL. 73] 



he surface. From " The Voyage of the Discovery" 



must be fed from snow-fields among mountains to 

 the south, and is probably confined between high 

 lands to east and west. 



It is here that we feel most the need of more 

 precise information regarding the ice of Ross's ice- 

 sheet, as Ferrar proposes it should be re-named. That 

 this ice is land ice flowing out to sea has been the 

 generally accepted explanation of the facts described 

 by Ross. The difficulty presented by Ross's ice- 

 sfieet, if it be advancing northward along its whole 

 face at anything like the rate of the ice movement 

 round Minna Bluff, is that its surface appears to be 

 practically level, so far as it was followed by Royds 

 to the south-east and by Scott to the south. Hence its 

 rapid movement cannot be due to flow down a slope 

 as in the case of ordinary glacier ice. The best 



