INTRODUCTORY ESSAY. xv 



"peculiar is its vegetation, and the smaller its proportion of species to genera. In the case of very 

 isolated islands, moreover, the generic types are often those of very distant countries, and not of the 

 nearest land. Thus the St. Helena and Ascension forms are not so characteristic of tropical Africa 

 as of the Cape of Good Hope. Those of Kerguelen's Land are Antarctic American, not African nor 

 Indian. The Sandwich Islands contain many North-west American and some New Zealand forms. 

 Japan presents us with many genera and species unknown except to the eastward of the Rocky 

 Mountains, in North America* So too American, Abyssinian, and even South African genera and 

 species are found in Madeira and the Canary Islands ; and Fuegian ones in Tristan d'Acunha. 



22. There is a strict analogy in this respect between the Floras of islands and those of lofty 

 mountain- ranges, no doubt in both cases owing to the same causes. Thus, as Japan contains 

 various peculiar N.E. American species which are not found in N.W. America nor elsewhere on the 

 globe, and the Canaries and Azores possess American genera not found in Europe nor Africa, so the 

 lofty mountains of Borneo contain Tasmanian and Himalayan representatives ; the Himalayas con- 

 tain Andean, Rocky Mountain, and Japanese genera and species ; and the alps of Victoria and Tas- 

 mania contain assemblages of New Zealand, Fuegian, Andean, and European genera and species. 

 We cannot account for any of these cases of distribution between islands and mountains except by 

 assuming that the species and genera common to these distant localities have found their way across 

 the intervening spaces under conditions which no longer exist. 



23. There is much to be observed in the condition and distribution of the introduced or natu- 

 rabzed plants of a country, which may be applied to the study of the origin of its indigenous vegeta- 

 tion. The greater proportion of these are the annual and other weeds of cultivated land, and plants 

 which attach themselves to nitrogenous soils; naturalized perennials, shrubs, and trees occur con- 

 secutively in rapidly diminishing proportions. I can find no decided relation between complexity 

 of structure and proneness to migrate, nor much between facilities for transport or power of endur- 

 ance or vitality in the seed, and extent of distribution by artificial means. I shall return to this 

 subject (which I have elsewhere discussed at length with reference to the Galapagos Archipelagof) 

 when -fef ating of the naturalized plants of Australia. 



24. I venture to anticipate that a study of the vegetation of islands with reference to the 

 pecubarities of their generic types on the one hand, and of their geological condition (whether as 

 rising or sinking) on the other, may, in the present state of our knowledge, advance the subjects of 

 distribution and variation considerably. The incompleteness of the collections at my command 

 from the Polynesian islands, has frustrated my attempts to illustrate this branch of inquiry by 

 extending my researches from the Australian Flora over that of the Pacific. I may however 

 indicate as a general result, that I find the sinking islands, those (so determined by Darwin's able 

 investigations) characterized as atolls, or as having barrier reefs, to contain comparatively fewer 

 species and fewer peculiar generic types than those which are rising. Thus, commencing from the 

 east coast of Africa, I find in the Indian Ocean the following islands marked in Darwin's chart J as 

 bounded with fringing reefs or active volcanos, and hence rising : — The Seychelles, Madagascar, 

 Mauritius, Bourbon, Ceylon, the Andamans, Nicobar, and Sumatra ; the vegetation of all which is 

 characterized by great diversity and much peculiarity of generic type : whereas those marked as 



* Whilst these sheets are passing through the press, I have been informed by Professor Asa Gray that the 

 Flora of Japan and N.E. Asia is much more closely allied to that of the Northern United States than to that of 

 America west of the Rocky Mountains. 



f Linn. Trans, xx. 235. t See his works on volcanic islands and on coral reefs. 



