50 THE VOYAGE OF H.M.S. CHALLENGER. 



THE ANTARCTIC FLORA. 

 ORIGIN OF THE VEGETATION OF THE ISLANDS OF THE SOUTH INDIAN OCEAN. 



Many recent writers on phytogeography allude to the American element in Australasian 

 vegetation, and some have recognised the fact that the composition of the flora of the 

 coldest southern zone in which flowering plants grow is essentially the same in all parts ; 

 but nobody, except Engler, 1 has very fully discussed this subject. Sir Joseph Hooker 

 (Flora Tasmanise : Introductory Essay, pp. 89-91) follows Forster in designating this 

 vegetation " Antarctic ; " and he gives a list of species occurring in Australia and New 

 Zealand, or in one or both of these countries, and Kerguelen, Tristan da Cunha, or Fuegia. 

 Grisebach (Vegetation der Erde) defines an "Antarctic Forest Region" in South America, 

 but the vegetation of all the southern islands, including New Zealand, is treated of under 

 the general head of " Oceanic Islands." 



Engler (loc. infra cit.) divides the vegetation of the earth into kingdoms, and the 

 Antarctic flora is included in his "Old Oceanic Kinp-dom," which is subdivided into 

 several regions and provinces, and comprises New Zealand, except the northern island, the 

 greater part of Australia, the islands in the South Indian Ocean, South Africa, the Antarctic 

 forest region of South America (which does not embrace the Falklands and Juan Fernan- 

 dez), and the Tristan da Cunha, St Helena, and Ascension islands. We have not space to 

 discuss the merits of this plan, which we believe the author himself would considerably 

 modify with our data before him ; but we agree generally with Drude that insular floras 

 should be attached to the continental ones to which they exhibit the greatest affinity. 



Drude (Petermann's Mittheilungen : Ergiinzungskeft, No. lxxiv. p. 70) regards Fuegia, 



the Falklands, South Georgia, South Shetlands, Tristan da Cunha, Prince Edward, the 



Crozets, St Paul, Amsterdam, Macquarie, and Emerald Islands, as forming a distinct 



subregion, which he calls the " Antarctic Islands." Fagus he looks upon as properly 



belonging to this subregion and overlapping in other regions. Though this is a much 



more philosophical method of dealing with the subject than Grisebach's, it is not altogether 



satisfactory, because the proposed subregion is not correctly limited. It includes either 



too much or too little. As shown in Part II., pp. 143 and 2G1, the vegetation of the 



Tristan da Cunha group, and of Amsterdam and St Paul, consist largely of the same 



species, and is quite different in its main features (Phylica and Spartlna) from 



that of Kerguelen and the other islands in a higher latitude. Further, there is no good 



reason why Macquarie Island should be placed in a different subregion from Auckland and 



Campbell Islands. Perhaps it would be better to extend this subregion, both in America 



and New Zealand, in spite of the large endemic element with which the types common to 



1 Versuch einer Entwicklungsgescbichte der l'flanzcnwelt ; in which work he has tabulated and fully 

 analysed, among others, the floras of Australia and New Zealand. 



