IS4 DISCOVERY REPORTS 



Other islands, including all the largest, which are entirely different. In these the "ice- 

 cap" exists only as a fringing or foreland glacier, through which protrude hills and 

 mountain ranges with steep escarpments, and surrounded by a bergschrund. It is 

 clear that this type of glaciation is a combination of ice-cap and mountain glacier, for 

 denudation is proceeding by two quite different methods. In the foreland, by plucking 

 and by erosion, all irregularities have been smoothed away, whereas farther back the ice 

 is cutting into the hill-ranges and in so doing tends always to sharpen the existing relief. 



Geology. It was thought by Suess, and is maintained by many geologists, that the 

 chain formed by the Burdwood Bank, South Georgia, the South Sandwich Islands and 

 Graham Land (sometimes known collectively as the South Antillean Arc) represents 

 an extension of the folds of the Andes, curving eastwards into the Atlantic. Gregory, 

 on the other hand, considers that so far as South Georgia is concerned this view is un- 

 tenable and from the evidence afforded both by its igneous and sedimentary formations 

 he^ concludes that South Georgia is " a fragment of an ancient South Atlantic Land, and 

 the south-western corner of the Flabellites-\and of Schwarz", in no way related to the 

 Andean system. Tyrrell,- however, has confirmed Nordenskjold's view that the 

 eruptive rocks of Graham Land and the South Shetlands are completely identical 

 with those of the Andes, and Holtedahl,^ more recently, has contested Gregory's 

 views on the relationships of the rocks of South Georgia. The echo-soundings 

 that we have taken during the last two years show conclusively that most of the sub- 

 merged portions of the arc claimed by Suess can still be traced in the relief of the 

 sea-bottom. 



In these circumstances the geological structure of the South Sandwich Islands 

 should be of some interest ; it is a matter for regret that our lack of knowledge pre- 

 cluded field observations of real value, and that owing to the difficulty of making 

 landings rock samples were obtained only on Thule Island. The specimens from 

 this island, collected by Mr T. J. Hart, have, however, kindly been examined by 

 Dr Tyrrell and his report is published as an appendix to this paper (p. 191). From 

 microscopic examination and chemical analyses Dr Tyrrell has reached the important 

 conclusion that the South Sandwich rocks are typical Andean lavas — a fact which goes 

 far to support Suess's theory. 



So far as we were able to observe, the islands themselves are all composed of eruptive 

 rocks; but Freezeland Peak, situated to the west of Bristol Island, differs conspicuously 

 from anything seen elsewhere. From a geological point of view this islet appears to be 

 of considerable interest, for if our view is correct, it is here alone that an exposure of 

 the underlying sedimentary rocks is to be found. To effect a landing on Freezeland 

 Peak would be extremely difficult, but it might not be impossible in exceptionally good 

 weather. 



^ Gregory, "The Geological History of the Atlantic Ocean", Quart. Journ.Geol. Soc, lxxxv, p.cxi(i929). 

 See also Nature, cxxvi, p. 837 (1930). 



- Tyrrell, Trans. Roy. Soc. Edinburgh, Liii, p. 78 (1921). 

 ^ Holtedahl, loc. cit. supra, pp. 104-11. 



