468 SUBANTARCTIC ISLANDS OF NEW ZEALAND. [Systematic Botany. 



reached the islands by the circuitous route of the South Georgia - Kerguelen group, 

 and have probably been conveyed by the agency of birds. 



As to the manner in which the colonisation from New Zealand has been effected, 

 I think that the recognised agents of plant-dispersal across the ocean — the wind, 

 ocean-currents, and the birds — are ample, even on the assumption, which I consider 

 most probable, that the islands have not been united with New Zealand since the 

 early part of the Tertiary period. All naturalists agree that it is exceedingly im- 

 probable that the Azores, Bermuda, St. Helena, the Galapagos, and the Hawaiian 

 Islands have ever been connected with any of the continental land-masses. Yet all 

 these islands support a numerous fauna and flora, the ancestors of which must have 

 been carried across the ocean for distances far greater than that separating the 

 southern islands from New Zealand. 



Having shown how closely the flora is associated with that of New Zealand, 

 and having described its points of resemblance and divergence, I ought perhaps to 

 close this inquiry. To proceed much further would involve a consideration of the 

 origin of the New Zealand flora itself, a matter far too extensive to touch in this 

 memoir. There is, however, one branch of the subject on which a few words may 

 be expected. I allude to the presence in the three land-masses of South America, 

 New Zealand, and Australia of a considerable number of identical genera, constitut- 

 ing what Sir J. D. Hooker has called " antarctic types." To explain this fact. Hooker 

 suggested that these were the remains of a flora that was once spread over a larger 

 and more continuous tract of land than now exists. This supposition was accepted 

 and extended by Hutton ; and more lately Forbes, Woodward, and others have 

 advocated a land connection including the Antarctic Continent as well as the three 

 land-masses mentioned above. Wallace (" Island Life," pp. 488-91) considers that 

 the plants in question probably reached Australia and New Zealand by way of the 

 Antarctic Continent, but does not think that an actual land connection is necessarily 

 implied thereby, the distances from Cape Horn to Graham Land and from the Ant- 

 arctic Continent to Macquarie Island being considerably less than those which must 

 have been traversed by plants and land-animals in other parts of the world. The 

 recent proofs that have been obtained of the former existence of a warmer climate 

 in the Antarctic Continent, and the presence of a varied vegetation thereon, con- 

 stitute a powerful argument in favour of a probable intercommunication between 

 the three southern floras. It appears in every way likely that the ancestors of our 

 species of Colohanthus, Donatia, Phyllachne, Fagus, and many others, may have 

 arrived by the source indicated above, probably in very early Tertiary times. It is 

 possible that the Antarctic Continent may have been connected with South America 

 (see the remarks on page 461 of this memoir), but I regard it as exceedingly im- 

 probable that the broad and immensely deep ocean separating Antarctica from the 

 southern islands of New Zealand has been completely bridged by a land connection 

 during Tertiary times, although the breadth of the ocean may have been less. 



The assumption so often made of a previous land connection between the 

 southern islands and New Zealand is in a somewhat different position. With the 

 exception of Macquarie Island, which is said to be separated from the rest by a broad 

 oceanic depression, they stand on a comparatively shallow plateau, so that no great 

 amount of elevation would be required to unite them with New Zealand and with 

 one another. We have seen that the presence of certain South American (or ant- 



