?82 



NA TURE 



[January 6, 1910 



floating, and, though fully exposed to the gales of 

 this stormy sea, they have not been broken up into ice- 

 bergs ; moreover, solution by sea water appears to be 

 verv slow, as Prof. David describes the floating end of 

 the Nordenskjold Glacier as dwindling away from the 

 failure of fresh supplies of ice from the snow-fields by 



which it was formerly fed. The Great Ice Barrier 

 appears to be of quite a different origin. Shackleton 

 records in his diary (vol. i., p. 2^3) that "The Bar- 

 rier surface is still as level as a billiard table, with no 

 sign of any undulation or rise." i\nd with his keen 

 geographical insight Shackleton discriminated (e.g. 

 NO. 2097, '^'OL. 82] 



p. 305) between the Barrier ice and the glacier ice, 

 which in places is thrust into it from the land. 



It would be difficult to explain the striking uni- 

 formity in level of the Barrier if it were land ice flow- 

 ing northward, the conclusion adopted by the staff of 

 the Discovery expedition. This view was rejected in 

 N.vruRE in the re- 

 views, both of Cap- 

 tain Scott's narra- 

 ;tive and of the 

 V o 1 u m e on the 

 geological results of 

 the Discovery ex- 

 pedition, in favour 

 of the origin of the 

 ice " by the accu- 

 mulation of layers 

 pi snow upon the 

 surface more 

 quickly than the 

 ice was dissolved 

 by the sea be- 

 neath " (Nature, 

 1906, vol. Ixxiii., 

 p. 300). This ex- 

 planation was then 

 iattended by the 

 difficultv that the 

 snowfall was said 

 to be too low. The 

 evidence, however, 

 of the photographs 

 brought back by 

 the Discovery 

 seemed so convinc- 

 ing that the writer 

 concluded that the 

 snowfall would be 

 ,found much larger 

 than was reported. 

 The new observa- 

 tions show that 

 the snowfall at the 

 winter quarters was 

 equivalent to a 

 rainfall of gj.inches 

 a year, and that on 

 the Barrier the 

 average for the 

 past four years has 

 been 7A inches. 

 This amount of 

 snow seems ade- 

 quate for the forma- 

 tion of the Barrier 

 ice from snowfall, 

 and this explana- 

 tion of its origin is 

 now fully confirmed 

 by Prof. David. 



1" h e meteoro- 

 logical results are 

 still indefinite, but 

 further light on the 

 hvnothetical South 

 Polar anticyclone 

 may be expected 

 from the full re- 

 cords. The narra- 

 tive shows that there was a less constant blizzard from 

 the south in the southernmost district than was inferred 

 from the cable report. There was a north wind, for 

 example, behind the party during its ascent of the 

 Beardmore Glacier. Nevertheless, the direction of the 

 ice ridges on the polar plateau shows that its pre- 



