THE PROBLEMS OF ANTARCTIC PLANT LIFE. 17 



present loci for the scanty Hora of Antarctica can only have become such when glaciation 

 had for some time diminished. 



It is difficult to believe that any species, unless possibly a lichen or two, can be a 

 survivor of an older Antarctic flora. 



At the period of severest glaciation the sub-antarctic islands were heavily glaciated 

 of that we have proof in many cases but probably not to such an extent as to 

 exterminate any pre-existing flora, only greatly to diminish it, though there are indica- 

 tions that in South Georgia and Macquarie Island the flora was wiped out. 



Turning now from the true Antarctic regions to the austral or sub-antarctic regions, 

 consisting mainly of the many islands that gird the Antarctic seas, it must be said that 

 it is here that the most fruitful botanical collections of future expeditions will probably 

 be made. This ring of circum- polar islands includes the following : Fuegia, the Falk- 

 lands, South Georgia, South Sandwich group, Tristan da Cunha with Gough Island, 

 Bouvet Island, Prince Edward and Marion Islands, the Crozets including Possession 

 Island, Kerguelen, Macdonald and Heard Islands, St Paul and Amsterdam, Campbell 

 and Auckland Islands, Macquarie Island, and Dougherty 1 or Keates Island, with a few 

 others whose existence is somewhat hypothetical. In passing it may be as well to note 

 that I have included all these islands in the general category of sub-antarctic merely for 

 the sake of convenience in this place, and do not intend to imply that on botanical 

 grounds they can be grouped in the same domain : for a discussion of the classification 

 of these islands reference should be made toDr Skottsberg's paper (loc. cit.). Of these 

 islands Fuegia and adjoining Patagonia, as well as the Falklands, have been well studied 

 by various expeditious, including the recent most fruitful one led by Dr Skottsberg ; 

 South Georgia has been recently re-explored by the Swedish Antarctic Expedition ; 

 Tristan da Cunha has hardly been exhausted despite the visit of the Challenger ; 

 Gough Island, on which 1 had the privilege of being the first botanist to land, would 

 well repay a visit ; Prince Edward and Marion Islands, the Crozets, Kerguelen, 

 Macdonald and Heard Islands, are far from well known, except perhaps Kerguelen ; 

 St Paul and Amsterdam Islands are better known, and the New Zealand group, includ- 

 ing Campbell, Auckland, and Macquarie Islands, and the Antipodes, have lately 

 received more attention. But all would be worth the attention of a careful explorer, 

 especially as regards the lower forms of plant life. Bouvet and Dougherty Islands 

 are altogether unknown from a botanical or almost any other standpoint. Bouvet 

 Island, according to the Vctldivias reports, is entirely covered with ice, and is 

 devoid of vegetation : moreover it offers no landing-place. On the other hand, 

 previous voyagers have given the island a slightly better reputation, Bouvet (1739) 

 and Lindsay (1808) both reporting trees and shrubs (? tussock grass), and Morrell 

 (1823) speaking of small spots of vegetation. Whatever may be the case it well 



1 The Discovery reported that this island does not exist in its formerly assigned position, and Captain J. K. Davis 

 of the Nimrod cast grave doubts on the existence of Emerald Island and the Royal Company Islands as well as 

 Dougherty Island. 



VOL. Ill 3 



