152 DISCOVERY REPORTS 



II. Islands without activity, but showing evidence of warmth. 



Leskov^. Apparently only a portion of a pre-existing crater. Probably not now active, 

 but with no glacier and little snow. 



Vindication. Thinly glaciated on higher slopes : extensive areas without snow. 



Montagu. Greater part of island and summit deeply glaciated, but with patches of 

 bare ground and indications of melting snow on and near peak at south-eastern 

 corner. 



III. Glaciated islands shoiving no signs of warmth or activity. 



Bristol. Cook. Thule. Very heavily glaciated throughout, the glacier at sea-level 

 reaching an average thickness of 200 ft. 



During our visit all the volcanoes were quiescent, and the conditions appear to sug- 

 gest that the volcanic activity is slowly abating: no eruptions were witnessed, nor was 

 any glow from the craters visible at night. We think it likely, however, that eruptions may 

 still occur from time to time on Zavodovski and Candlemas. This may also be true of 

 Bellingshausen; but Visokoi and Saunders appear to be in the solfatara phase, their 

 activity being limited to the emission of steam and vapour. 



Mode of formation. There is clearly a tendency on some of the islands for volcanic 

 activity to cease at one point and to break out at another. Thus the northern end of 

 Candlemas, now the only active part of the island, appears to be a more recent formation 

 than the high and heavily glaciated southern end. In the Southern Thule Group, Cook 

 and Thule Islands, with the submerged crater between them (see p. 179), are extinct, 

 the outlet for volcanic activity having shifted to Bellingshausen. In Montagu also, only 

 the small cone at the south-east corner, probably the most recently formed part, still 

 shows traces of warmth. Saunders is a less simple case. The high glaciated cone of Mt 

 Michael is still active ; but to the south-east there are craters formed of volcanic mud 

 or ash, seemingly extinct, but with most of the ground bare of snow and evidently still 

 warm. It may perhaps be surmised that the main crater was extinct for a period, during 

 which the south-eastern outlets were in operation, and that later it resumed its 

 activity. 



If these suppositions are correct the mode of formation of most of the islands is 

 tolerably evident. For Zavodovski, Visokoi and Bristol retain their original character 

 of simple volcanic cones, and Leskov is probably a fragment of what was once a crater. 

 The only island that presents difficulties is Vindication ; for it seems certain that it does 

 not represent the whole of a cone, and the soundings we have taken do not afford any clue 

 to the way in which it was formed. 



In its original form each island would appear to have consisted of a circular plateau 



of basalt, with a central cone formed of alternate layers of tuff and ash — a condition most 



nearly realized in Zavodovski. As a result of the prevailing westerly winds abrasion is 



most active on the western aspects of the islands, while lighter materials from volcanic 



1 This island was partly covered with mist during almost all the time it was under observation. 



