DISCOVERY REPORTS 



any of the other islands, for though material of similar appearance is to be seen on 

 Montagu and Bellingshausen, it is of small extent and there is no indication that it was 

 extruded from a separate outlet. Because of the evident way in which the centre of 

 volcanic activity has shifted from one point to another in certain of the islands (e.g. 

 Candlemas and Southern Thule), it may perhaps be supposed that Mt Michael— no 

 doubt the original seat of activity— was for a time quiescent and that it was during this 

 period that the south-eastern hills were formed. 



In his atlas Bellingshausen gives a sketch of Saunders Island seen from the north 

 (Fig. 14, cf. PI. XXI, fig. i), in which the bare hills behind Nattriss Point culminate in a 

 lofty cone, evidently far higher than at the present day. If this sketch accurately 

 represents the conditions in 1820 it is clear that the hills have undergone considerable 

 alteration, due either to renewed volcanic activity, or to erosion. 



Fig. 14. Saunders I., distant about 4 miles. (After Bellingshausen.) 



Cook mentions 'two hills, seen over the northern point, which might be two islands, 

 and adds, "these only were clear of snow and seemed to be covered with green turf". 

 This statement must refer to the hills at the south-eastern end, but at the time of our 

 visit they showed no traces of vegetation. In Cordelia Bay, on low ground where the 

 glacier ends and the mud hills begin, small patches of green were to be seen, and at this 

 point there is a large rookery of Ringed penguins. A little further east, where a valley 

 gives greater extent to the foreshore, some Elephant seal had hauled up, and here a 

 number of giant petrels were gathered. Round the eastern end of the bay is a steeply 

 shelving beach of black sand where parties of Ringed penguins, with a few King and 

 Gentoo, were walking, and on which some seal (apparently Weddell) were lying. Other 

 colonies of penguins are to be found north of Rocky Point on the west coast, and on 

 Harper and Nattriss Points. 



We made an attempt to land in Cordelia Bay, but were prevented by the heavy swell 

 and strong undertow on the beach. The only recorded landing is by Larsen in 1908. 



Montagu Island 

 Lat. 58° 25I' S ; long. 26° 21I' W 



(Plates XXII and XXIII) 



Montagu is the largest island in the South Sandwich Group. It is roughly quad- 

 rangular in outline, and is rather less than 6\ miles in both length and breadth, with a 

 circumference of about 24 miles. In the south-east there is a promontory ending in 

 Allen Point, ^ and standing on it is a conical hill 1660 ft. in height, which forms a con- 

 spicuous feature from many points of view. To the west of the hill is Phyllis Bay," 



'^ H. T. Allen, member of the Discovery Committee. ^ Phyllis V. Horton. 



