834 MR G. W. TYRRELL ON 



radiolaria, may also be partly of Jurassic age.* Amongst the loose blocks strewn 

 on the surface, in the moraines, or occurring as boulders in the Late Mesozoic or 

 Tertiary conglomerates, ■ dark phyllitic slates, some of which are highly sericitic, 

 are frequent. 



In Dundee Island, close to Joinville Island, off the coast of Graham Land, a few 

 specimens have been got which are very similar to some South Georgian types. 

 These were collected by Captain Th. Robertson of Dundee during a sealing trip, and 

 have been described by Sir A. Geikie and Dr J. J. H. Teall.I They consist of 

 various types of granite, a tuff, and a red jasper pebble containing radiolaria, 

 amongst which, according to Dr Hinde, the genera Cenosphcera, Carpoxphtera, and 

 Cenellipsis are represented, along with Hexactinellid sponge spicules. These radio- 

 laria have also been identified in the radiolarian rocks of South Georgia. The tuff is 

 a greenish rock consisting largely of crystals and fragments of water-clear felspars 

 which are generally untwinned. There are also small lapilli of andesitic character, 

 and grains of pale augite and iron-ore. One crystal of brown hornblende was 

 also identified. The resemblance of this rock to the interbedded tuffs of South 

 Georgia is obvious. 



It may therefore be concluded that, while many of the rocks of South Georgia 

 appear to be similar to those found in other West Antarctic areas, their identity as 

 regards petrographic character and age cannot be accepted on the meagre evidence 

 at present available. Furthermore, pending the detailed examination of the rock- 

 specimens collected by the Filchner Expedition, it cannot be said that the igneous 

 rocks of South Georgia are of decided Andean affinities. The granito-dioritic series 

 discovered by Heim may prove to be Andean in type, but the characteristic andesitic 

 volcanic rocks have not yet been discovered, either in situ or as loose blocks. The 

 interbedded tuffs of South Georgia are distinctly of alkaline or " Atlantic " affinities, 

 if we may use a much-debated term, and differ entirely from the rocks of the 

 Andean suite. 



The question of the tectonic relations of South Georgia to South America on the 

 one hand, and Graham Land on the other, must of course be decided primarily on 

 tectonic evidence, but the petrographical character of the rocks may afford valuable 

 confirmatory evidence. Unfortunately, our knowledge of South Georgian petrography 

 is as yet too limited for the evidence of the rocks to be decisive on the problem of 

 the " Southern Antilles." The discovery of a granito-dioritic series in the south-east 

 of the island perhaps tilts the balance of evidence towards the interpretation favoured 

 by Suess ; although an alternative interpretation mentioned as possible by Professor 

 Gregory, that South Georgia may form part of Professor Schwarz's Flabellites 

 Land,j lH by no means out of court. 



* J. W. GREGORY, "The Geological Relations of South Georgia,* Geol. May., Feb. 1914, p. 64. 

 + Sir A. Geikie, " Notes on Rock Specimens from the Antarctic Regions" (with petrographic notes by Dr J. J. H. 

 TBAXL), Proc. Hoy. Soc. Edin., xxii, 1898, pp. 66-70. 

 J Loc. cit., p. 64. 



